Download Page 630 The Lancet

Volume 370, Issue 9588, Pages 629-712 (25 August 2007-31 August 2007)
1.
Humane and compassionate elder care as a human right
Page 629
The Lancet
2.
Can caring for patients be taught?
Page 630
The Lancet
3.
Preventing chronic disease: a forthcoming initiative
Page 630
The Lancet
4.
The Kampala Forum on working together to overcome the workforce crisis—a call
for papers
Pages 631-632
Lincoln Chen, Francis Omaswa, Sigrun Mogedal, Anders Nordstrom, Suwit
Wibulpolprasert and Richard Horton
5.
Calcium and vitamin D for osteoporotic fracture risk
Pages 632-634
Jean-Yves Reginster
6.
Acute myocardial infarction and diabetes
Pages 634-635
Lionel H Opie
7.
Catecholamine treatment for shock—equally good or bad?
Pages 636-637
Mervyn Singer
8.
Speaking out about human rights and health in West Papua
Pages 637-639
Susan Rees and Derrick Silove
9.
Where have all the bees gone?
Page 639
Jessica Hamzelou
10.
Clinical update: obstetric anaesthesia
Pages 640-642
Allan M Cyna and Jodie Dodd
11.
Experts disagree over NICE's approach for assessing drugs
Pages 643-644
Richard Hoey
12.
Therapy for autistic children causes outcry in France
Pages 645-646
Laura Spinney
13.
Much to be done to improve the mental health of young people
Pages 647-648
Michael Rutter
14.
A cultural history of impotence
Page 648
George Rousseau
15.
May, 2007
Page 649
16.
Dennis Carlson
Page 649
Faith McLellan
17.
John R Hogness
Page 650
Ivan Oransky
18.
Folic acid for stroke prevention
Page 651
Vincenzo Sepe, Gabriella Adamo, Maria Grazia Giuliano, Carmelo Libetta and
Antonio Dal Canton
19.
Folic acid for stroke prevention – Authors' reply
Page 651
Xiaobin Wang, Xianhui Qin, Hakan Demirtas and Xiping Xu
20.
Recurrent anaphylaxis to synthetic folic acid
Page 652
Julie Smith, Marianne Empson and Clare Wall
21.
DFID's health strategy
Pages 652-653
Gorik Ooms and Nathan Ford
22.
DFID's health strategy
Page 653
Adrienne Germain
23.
Misleading confidence intervals
Page 653
Martin Bland
24.
Misleading confidence intervals – Authors' reply
Page 653
Francis Ka Leung Chan and Bing Yee Suen
25.
Minimisation of immunosuppression in transplantation
Page 654
RY Calne and KO Lee
26.
Minimisation of immunosuppression in transplantation – Authors' reply
Page 654
Mohamed H Sayegh and Giuseppe Remuzzi
27.
Human resources in developing countries
Pages 654-655
Per Ashorn
28.
Human resources in developing countries – Author's reply
Page 655
Claudio F Lanata
29.
MMC, Britain's aid deficit, and overseas health-care shortages
Pages 655-656
Colin Brown and Anjum Khan
30.
Quality of life in children with cerebral palsy
Page 656
Jon Sparkes and David Hall
31.
Department of Error
Page 656
32.
Use of calcium or calcium in combination with vitamin D supplementation to
prevent fractures and bone loss in people aged 50 years and older: a meta-analysis
Pages 657-666
Benjamin MP Tang, Guy D Eslick, Caryl Nowson, Caroline Smith and Alan
Bensoussan
33.
Incidence of new-onset diabetes and impaired fasting glucose in patients with
recent myocardial infarction and the effect of clinical and lifestyle risk factors
Pages 667-675
Dariush Mozaffarian, RosaMaria Marfisi, Giacomo Levantesi, Maria G Silletta,
Luigi Tavazzi, Gianni Tognoni, Franco Valagussa and Roberto Marchioli
34.
Norepinephrine plus dobutamine versus epinephrine alone for management of
septic shock: a randomised trial
Pages 676-684
Djillali Annane, Philippe Vignon, Alain Renault, Pierre-Edouard Bollaert, Claire
Charpentier, Claude Martin, Gilles Troché, Jean-Damien Ricard, Gérard
Nitenberg, Laurent Papazian, et al.
35.
Polycystic ovary syndrome
Pages 685-697
Robert J Norman, Didier Dewailly, Richard S Legro and Theresa E Hickey
36.
Violations of human rights: health practitioners as witnesses
Pages 698-704
James Orbinski, Chris Beyrer and Sonal Singh
37.
Inconvenient truths about effective clinical teaching
Pages 705-711
Brendan M Reilly
38.
Mysterious falls and a nasal voice
Page 712
Avijit Bhandari and Firdaus Adenwalla
Editorial
Last week, in a courtroom in Louisiana, the first criminal
trial to arise from deaths that occurred during Hurricane
Katrina, which struck New Orleans 2 years ago, got
underway. Mabel and Salvador Mangano, who owned
St Rita’s nursing home in St Bernard Parish, have been
charged with the deaths of 35 of their residents, all of
whom drowned, some in wheelchairs, others in their
beds, when the home was rapidly flooded. They have
also been charged with cruelty to the 24 residents who
survived after family members and aides evacuated
them using two nearby boats. The couple, who have
pleaded not guilty, refused an offer of two buses to
evacuate the home the day before the storm. They chose
instead to “shelter in place”, with catastrophic results.
If convicted, each of them could be sentenced to more
than 400 years in prison.
The plight of the residents of St Rita’s was, it is to be
hoped, singular. But even under ordinary circumstances,
there is little question that elderly people living in
nursing or care homes or in hospitals are often badly
treated. A UK parliamentary report released last week—
The human rights of older people in health care—describes,
in distressing detail through several personal narratives,
the all-too-common failures of their care and the dismal
environments in which institutionalised elderly people
often live. The problems the report documents include
premature discharge from hospitals, discriminatory
treatment (through covert rationing of services), physical
neglect leading to malnutrition, dehydration, and bed
sores, inappropriate medication, physical abuse, sexual
assault, lack of privacy, dignity, and confidentiality, and
negative attitudes towards older people.
None of these have a place in anyone’s care—indeed,
some are criminal offences—but a course correction
in elder care is especially urgent because of the sheer
numbers of people involved. In a population that is living
longer than ever before, by 2050 there will be twice as
many people aged 80 and older than there are today in the
UK. In the USA, more than 1·6 million people were living
in nursing homes in 1999; this number will continue to
rise. Attention to the attitudes toward, and care of, elderly
people will obviously become increasingly crucial.
The UK report acknowledges that much good care is
provided, and often by poorly paid and undervalued
staff. Nonetheless, abuse and neglect are apparently
www.thelancet.com Vol 370 August 25, 2007
widespread. The government committee focused on the
problem through a human-rights lens, with particular
reference to the European Convention on Human Rights
and the UK’s Human Rights Act of 1998, which came
into effect in 2000. Human-rights principles affirm
common-law rights to humane and equitable care
that is free from discrimination. In addition, by statute
or through litigation, human rights have evolved
to include the right to life, freedom from degrading
treatment, dignity, and respect for private life, including
physical and psychological integrity. Human rights
impose positive obligations on public authorities,
beyond the negative obligation not to infringe human
rights. This report suggests that, 7 years after the Act’s
implementation, health-care policymakers in the UK
have paid little more than lip service to those rights and
responsibilities, and that service providers themselves
have little understanding of the Act itself or how its
principles should be applied in practice. Therefore the
report suggests several approaches to filling these gaps,
ranging from the broad adoption of a human-rights
framework by health and justice authorities, to staff
training and whistleblower protection.
Will employment of the Human Rights Act really
work as a means for a cultural shift in elder care? We
would like to think so. But, as the report underscores,
this recommendation will work only if the principles of
human rights and their attendant responsibilities are
embraced by every segment of those involved in caring
for older people, from ministerial level officers and
policymakers, to local trusts and individual providers,
right down to those doing the most menial of chores
in hospitals and care homes, whose role in supporting
quality care is often overlooked.
Getting every person involved in care for elderly
people to consider their work as a positive contribution
to a civilised, human-rights based society will require
an enormous amount of continuing education and
training. Elderly people themselves and their surrogates or
advocates will have to be relentless in framing long-term
care in a human-rights context. Cultural change cannot
come about without the participation of all those affected
by the issues at hand. Given modern medicine’s success at
expanding life expectancy, that is a group that, more likely
than not, will eventually include most of us. ■ The Lancet
Getty Images
Humane and compassionate elder care as a human right
For the UK parliamentary report
see http://www.publications.
parliament.uk/pa/jt200607/
jtselect/jtrights/156/156i.pdf
629
Editorial
Science Photo Library
Can caring for patients be taught?
See Viewpoint page 705
For Francis Peabody’s essay on
the care of the patient see
JAMA 1927; 88: 876–82
For more on the Royal College
of Physicians’ description of
professionalism see Doctors in
society: medical professionalism in
a changing world. London: Royal
College of Physicians, 2005
Almost two decades have passed since Stephen Covey
brought the Seven habits of highly effective people to
the world’s attention. In his more recent 8th Habit,
his advice is to “find your voice and inspire others to
find theirs”. In this week’s Lancet, Brendan Reilly has
done just that when he describes the eight habits of
exemplary clinical teachers. Can clinical teachers talk
the TALK (Think out loud, Activate the learner, Listen
smart, Keep it simple) and walk the WALK (Wear
gloves, Adapt, Link learning to caring, Kindle kindness)
necessary to make them effective examples to the next
generation of doctors?
Reilly admits evidence of his theory is lacking but his
instincts are shared by many, especially in his point
about teaching caring. Care is not a “thing”—a product
provided by clinicians and received by patients. Rather,
as poignantly articulated by Francis Peabody in 1927:
“One of the essential qualities of the clinician is interest
in humanity, for the secret of the care of the patient is in
caring for the patient”.
Reilly comments, “the medical profession as a group has
not tried to discover who the careless clinicians are or how
they got that way”, but maybe it is too much to expect that
caring clinicians were there in the first place. A selection
system that over-emphasises academic achievements and
technical, disease-orientated training might not deliver
the attitudes and skills that Reilly so passionately values.
To make care more than a manufactured product
there also needs to be compassion—the ability to feel
for someone in trouble. Although compassion is often
cited as one of the core values of professionalism, there
remains a continuous and inconclusive debate about
whether compassion is innate or can be taught. So given
the vogue of acronyms in medical training and caring,
we propose another one: CARE—Compassion, Attention,
Respect, and Empathy. Although central to care, if the
ability to feel compassion is missing, or indeed even if
it is a trait that cannot be learned, paying attention to
patients, respecting them, and being empathic towards
them certainly can be. ■ The Lancet
Preventing chronic disease: a forthcoming initiative
WHO/Chris De Bode
On Dec 4, 2007, The Lancet will publish a series of papers
to signal a new effort to reduce the global burden of
chronic disease. A Lancet Chronic Disease Action Group,
led by Robert Beaglehole, together with an international
team of clinical and public health scientists, has spent a
year assembling evidence to advance the goal set out in
The Lancet’s 2005 series and WHO’s Preventing chronic
disease: a vital investment. The goal set was ambitious: to
reduce chronic disease death rates by an additional 2%
annually. If achieved, this target would avert 36 million
deaths and add 500 million years of life by 2015.
This collaboration has been undertaken with the
support of the UK’s Department of Health, the Public
Health Agency of Canada, and WHO. Indeed, the
knowledge gathered in the series aims to underpin
WHO’s own non-communicable disease action plan,
adopted by the World Health Assembly in May, 2007.
The purpose of the Lancet’s report is to provide
evidence for a core group of interventions for chronic
disease control in low-income and middle-income
countries.
630
The Dec 4 launch of The Lancet series will be marked by
a 1-day conference in London to present its findings. We
invite all those interested in the prevention and treatment
of chronic diseases in low-income and middle-income
settings to join us. We will use the findings of the report to
trigger a debate about how its recommendations can be
translated into practice at country level. Further details of
this event will follow later in the year.
Tackling chronic disease presents one of the most
complicated challenges facing the international community. In addition to making sure that services for heart
disease, stroke, and cancer treatment, among other
conditions, are available, an array of direct risk factors
(hypertension, tobacco, cholesterol, obesity, and physical
inactivity) and indirect influences (eg, urbanisation and
transport) somehow have to be incorporated into a
total strategy. Not all of these problems can be solved
at once. But there is enough reliable science available to
start. Our December report and conference aim to begin
a concerted global effort to fulfil the promises we made
2 years ago. ■ The Lancet
www.thelancet.com Vol 370 August 25, 2007
Comment
The Kampala Forum on working together to overcome
the workforce crisis—a call for papers
www.thelancet.com Vol 370 August 25, 2007
Commonly shared problems in poor and rich countries
are shortages, skill-mix imbalances, geographic maldistribution, negative work environment, and insufficient knowledge base.
The first ever Global Forum on Human Resources for
Health will be held in Kampala, Uganda, on March 4–7,
2008. Announced by the GHWA, the Forum is expected
attract more than 500 participants and health and civil
society leaders, such as WHO Director-General Margaret
Chan, African Union Commissioner Bience Gawanas,
and Realizing Rights’ Mary Robinson. The Forum will
promote analysis of lessons learned on what works,
what has not, and why, and how to accelerate progress.
By promotion of partnerships, the Forum will chart a
roadmap for collaborative action in the coming decade.
GHWA has advocated workforce-friendly policies
among rich and poor countries alike. In countries
of the Organization for Economic Cooperation and
Development, alliance advocates have pushed for
more workers and expanded funding of the workforce,
not the “borrowing from Peter to pay Paul” tactics
used by some narrow projects to sequester workers
for their single-purpose effort. Some initiatives such
as the GAVI Alliance and GFATM have responded by
incorporating broader health-system strengthening
in their funding packages. To reinforce country
The printed journal
includes an image merely
for illustration
Panos Pictures
At the turn of the new millennium, global leaders at
the UN and G8 set ambitious health targets in the
Millennium Development Goals (MDGs), convened
an unprecedented UN General Assembly addressing
a single disease (HIV/AIDS), and launched new global
initiatives such as the Global Alliance for Vaccines
and Immunization (now the GAVI Alliance) and the
Global Fund to fight AIDS, Tuberculosis and Malaria
(GFATM). Much political energy focused on mobilising
the requisite financing. Yet underappreciated was the
weak backbone—ie, the workforce—of health systems
that were supposed to tackle these daunting challenges
in the poorest countries. Two decades of structural
adjustment had capped workforce development, and
low investments in universities and higher education
had dried up the graduate pipeline. HIV/AIDS emerged as
a “triple threat”—increasing work burdens, decimating
the workforce, and demoralising the underpaid and
undersupported workers who remained. Doctors and
nurses were migrating at an alarming rate from poor
to rich countries as part of an expanding global labour
market.
That the workforce was the engine of health work—
obvious to practitioners and leaders on the ground
in affected countries—was simply invisible to global
leaders. African health ministers therefore pushed
workforce issues in a series of World Health Assemblies,
and, in 2004, the Joint Learning Initiative, a consortium
of 100 global health leaders, published its evidencebased report Overcoming the Crisis,1 which documented
the severe shortages of workers globally and especially
in sub-Saharan Africa. If the continent were to have
any hope of achieving the MDGs, Africa would need to
at least double its skilled workforce. The World Health
Report 2006, Working Together for Health,2 issued a
wakeup call urging a decade of action, and launched
the Global Health Workforce Alliance (GHWA)—an
independent stakeholder partnership operating from
a secretariat base at WHO headquarters and consisting
of non-governmental organisations (NGOs), agencies,
professionals, educators, and government leaders. The
two reports presented overwhelming evidence that
the workforce drives health outputs and outcomes.
Nsambye mobile health team visiting patient with AIDS at home, Kampala, Uganda
631
Comment
action, GHWA has launched four global task forces:
fast-tracking training and education, ameliorating
harmful aspects of migration, harmonising HIV
efforts with health systems, and charting feasible
financing options. In partnership with WHO, GHWA
has supported development of tools and guidelines for
workforce planning and management. All of these new
developments will be presented in Kampala.
The Kampala Forum comes at a critical moment
when problem recognition has grown enormously but
effective solutions are only beginning to be tested and
validated. Participants with real-world experiences and
early research results will be welcomed at this Forum.
The Lancet will publish a special issue on human resources
for health the week before the Forum, and now invites
for high-quality research papers to be submitted on this
subject by Nov 1, 2007. The Forum meshes well with
current movements to revitalise primary health care on
the 60th anniversary of WHO, 30 years after Alma Ata.
Kampala promises to be an exciting and critical next
step for achieving global goals and re-energising the
movement for better health in the 21st century.
Lincoln Chen, *Francis Omaswa, Sigrun Mogedal,
Anders Nordstrom, Suwit Wibulpolprasert, Richard Horton
China Medical Board, New York, NY, USA (LC); Global Health
Workforce Alliance, WHO, CH-1211 Geneva 27, Switzerland (FO);
Ministry of Foreign Affairs, Oslo, Norway (SM); WHO, Geneva,
Switzerland (AN); Ministry of Public Health, Nonthaburi ,Thailand
(SW); and The Lancet, London, UK (RH)
[email protected]
We declare that we have no conflict of interest.
1
2
Chen L, Evans T, Anand S, et al. Human resources for health: overcoming
the crisis. Lancet 2004; 364: 1984–90.
WHO. The world health report 2006—working together for health. 2006.
http://www.who.int/whr/2006/whr06_en.pdf (accessed Aug 20, 2007).
Calcium and vitamin D for osteoporotic fracture risk
See Articles page 657
632
Osteoporosis is a chronic progressive bone disease, in
which bone resorption exceeds bone formation, leading
to a reduction in bone-mineral density and disruption
of bone microarchitecture. Calcium and vitamin D deficiency are important factors for bone health, because
reduced calcium absorption increases parathyroid hormone concentration and accelerates the rate of bone
loss, which raises the number and activity of osteoclasts
that release calcium from bone. Although many studies
have investigated the effect of supplementation with
calcium or calcium with vitamin D on fracture risk in
postmenopausal osteoporosis, there are conflicting
outcomes.1–3 However, calcium and vitamin D are still
widely regarded as essential components in osteoporosis
management.4,5
In today’s Lancet,6 Benjamin Tang and colleagues
provide a well-designed systematic review of 29 randomised trials that assessed all fractures and changes
in bone-mineral density after individuals aged 50 years
or older were supplemented with calcium or calcium
plus vitamin D. Unlike previous meta-analyses,7,8 Tang
provides clear answers to several questions which could
be of immediate practical importance for the daily
management of osteoporosis. Their conclusion is that
calcium supplementation, or calcium with vitamin D,
is effective in the prevention of osteoporotic fractures.
This isolated statement, even though supported by
a strong method, would probably not be a major
contribution to an evidence-based management of
patients with osteoporosis. However, the strength and
interest of the study come from the large sample size
(more than almost 64 000 participants), which allows
for meaningful and properly powered subgroup analysis
for fracture effect.
Some recent studies have concluded that calcium
supplementation with or without vitamin D should
not be systematically recommended for the primary or
secondary prevention of fracture in elderly patients.3,9
However, most of these studies had poor long-term
adherence to study drug, which is a common feature for
many anti-osteoporosis drugs that are given daily. In the
main analysis of these studies, 50–60% of the original
population were no longer taking their medication
properly, which might generate concerns about the
overall negative conclusions reported. However, Tang
and colleagues showed that, in a subpopulation with
a compliance rate of at least 80%, the level of risk
reduction was doubled compared with those with poor
compliance, which provides robust scientific grounds for
a worldwide interaction between drug manufacturers
(eg, development of user-friendly formulations), physicians (eg, delivery of positive messages or reinforcement
www.thelancet.com Vol 370 August 25, 2007
through densitometric and biochemical feedbacks), and
patients, to provide optimum adherence to calcium or
calcium with vitamin D supplementation.
30–100 nmol/L is the estimated concentration of
circulating 25-hydroxyvitamin D needed to maintain
normal concentrations of parathyroid hormone and
prevent bone fragility.10,11 The acceptable threshold for
dietary calcium intake, or for calcium pharmacological
supplementation, is unclear—recommendations range
from 400 to 1500 mg per day. These uncertainties have
contributed to a wide range of clinical practices for
calcium with or without vitamin D supplementation.
The recommended minimum daily doses of 1200 mg
of calcium or 800 IU of vitamin D proposed by Tang and
co-workers are consistent with previous reports that
400 IU per day of vitamin D is not sufficient for fracture
prevention, and that doses of 700–800 IU per day are
needed to reduce the risk of hip and non-vertebral
fracture in ambulatory or institutionalised women.12
Such guidance seems conservative and appropriate
to ensure a positive risk-benefit ratio for people at
increased risk of, or who have developed, osteoporosis.
However, two major issues remain. First, should
calcium be added to vitamin D supplementation or
vitamin D to calcium supplementation? Tang and colleagues conclude that addition of vitamin D to calcium
does not significantly affect treatment efficacy. Nevertheless, they report an increased treatment effect of
calcium or calcium with vitamin D in individuals with
low serum 25-hydroxyvitamin D (<25 nmol/L), as well
as in elderly and institutionalised patients. This result
is surprising since these patients have consistently
been reported to have a high prevalence of vitamin D
deficiency. Because another meta-analysis13 recently concluded that oral vitamin D reduces the risk of hip fracture only when calcium supplement is added, it might
be cautious to concentrate on combined calcium and
vitamin D preparations, at least in people at increased risk
of osteoporotic fractures, until further studies are done.
Second, the cost-conscious use of calcium or calcium
with vitamin D supplementation is not fully addressed
by Tang and colleagues. Various treatment options have
been advocated, including systematic supplementation
at the onset of menopause and restriction of calcium
with or without vitamin D to patients receiving other
anti-osteoporotic drugs. Tang suggests that the number
needed to treat (63 patients over 3·5 years) with calcium
www.thelancet.com Vol 370 August 25, 2007
Science Photo Library
Comment
False-colour scanning electron micrograph of trabeculae in spongy bone affected by osteoporosis
or calcium and vitamin D to prevent one osteoporotic
fracture compares favourably with similar calculations
from the cardiovascular field. They suggest that the
cost-effectiveness of calcium supplementation might
even be better in high-risk participants or in those with
optimum or improved compliance. This assumption,
however, should be interpreted carefully.
Number needed to treat is highly dependent on the
baseline absolute risk of the clinical endpoint under
investigation. Tang and co-workers’ analysis emphasises
the need for extensive and sophisticated health-economic
analyses to assess the incremental cost-effectiveness ratio
of calcium or calcium with vitamin D supplementation
in various doses, regimens, populations, and settings.
In conclusion, Tang and colleagues’ contribution is
important because it paves the way for future research
aiming at the best clinical, pharmacological, and economic
use of calcium and vitamin D in patients at increased risk
of osteoporotic fractures.
Jean-Yves Reginster
Bone and Cartilage Metabolism Unit, CHU Centre Ville,
4020 Liege, Belgium
[email protected]
J-YR has received consulting fees from Servier, Novartis, Negma, Lilly, Wyeth,
Amgen, GlaxoSmithKline, Roche, Merckle, Nycomed, NPS, and Theramex;
lecture fees from Merck Sharp and Dohme, Lilly, Rottapharm, IBSA, Genevrier,
Novartis, Servier, Roche, GlaxoSmithKline, Teijin, Teva, Ebewee Pharma, Zodiac,
Analis, Theramex, Nycomed, and Novo-Nordisk; and grant support from Bristol
Myers Squibb, Merck Sharp & Dohme, Rottapharm, Teva, Lilly, Novartis, Roche,
GlaxoSmithKline, Amgen, and Servier.
633
Comment
1
2
3
4
5
6
7
Dawson-Hughes B, Harris SS, Krall EA, Dallal GE. Effect of calcium and
vitamin D supplementation on bone density in men and women 65 years
of age or older. N Engl J Med 1997; 337: 670–76.
Chapuy MC, Arlot ME, Delmas PD, Meunier PJ. Effect of calcium and
cholecalciferol treatment for three years on hip fractures in elderly women.
BMJ 1994; 308: 1081–82.
Jackson RD, LaCroix AZ, Grass M, et al; Women’s Health Initiative
Investigators. Calcium plus vitamin D supplementation and the risk of
fractures. N Engl J Med 2006; 354: 669–83.
Boonen S, Body JJ, Boutsen Y, et al. Evidence-based guidelines for the
treatment of postmenopausal osteoporosis: a consensus document of the
Belgian Bone Club. Osteoporos Int 2005; 16: 239–54.
Delmas PD. Treatment of postmenopausal osteoporosis. Lancet 2002;
359: 2018–26.
Tang BMP, Eslick GD, Nowson C, Smith C, Bensoussan A. Use of calcium or
calcium in combination with vitamin D supplementation to prevent
fractures and bone loss in people aged 50 years and older: a meta-analysis.
Lancet 2007; 370: 657–66.
Avenell A, Gillespie WJ, Gillespie LD, O’Connell DL. Vitamin D and vitamin D
analogues for preventing fractures associated with involutional and
postmenopausal osteoporosis. Cochrane Database Syst Rev 2005; 3: CD000227.
8
9
10
11
12
13
Shea B, Wells G, Cranney A, et al. Calcium supplementation on bone loss in
postmenopausal women. Cochrane Database Syst Rev 2004; 1: CD004526.
Grant AM, Avenell A, Campbell MK, for the RECORD Trial Group. Oral
vitamin D3 and calcium for secondary prevention of low-trauma fractures
in elderly people (Randomised Evaluation of Calcium Or vitamin D,
RECORD): a randomised placebo-controlled trial. Lancet 2005; 365:
1621–28.
Dawson-Hughues B, Heaney RP, Holick MF, Lips P, Meunier PJ, Vieth R.
Estimates of optimal vitamin D status. Osteoporos Int 2005; 16: 713–16.
Garnero P, Munoz F, Sornay-Rendu E, Delmas PD. Associations of
vitamin D status with bone mineral density, bone turnover, bone loss
and fracture risk in healthy postmenopausal women: the OFELY study.
Bone 2007; 40: 716–22.
Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E, Dietrich T,
Dawson-Hughes B. Fracture prevention with vitamin D supplementation:
a meta-analysis of randomized controlled trials. JAMA 2005; 293: 2257–64.
Boonen S, Lips P, Bischoff-Ferrari HA, Vanderschueren D, Haentjens P. Need
for an additional calcium to reduce the risk of hip fracture with vitamin D
supplementation: evidence from a comparative meta-analysis of
randomized controlled trials. J Clin Endocrinol Metab 2007; 92: 1415–23.
Acute myocardial infarction and diabetes
See Articles page 667
634
The idea of acute myocardial infarction as an acute
stress reaction goes back many years, but the association with metabolic changes was probably first
described in The Lancet by Kurien and Oliver in 1966
(figure). In a patient with acute myocardial infarction,
plasma free-fatty-acids and glucose increased rapidly at
the onset.1 Historically, the next step was to show that
the high free-fatty-acid was associated with suppressed
insulin secretion, hyperglycaemia, and glucose
intolerance.2 And in 1986, how high plasma glucose
had risen on admission to hospital was associated
with the enzymatically measured infarct size and
6-month mortality.3 Thus early studies established acute
myocardial infarction as a stress reaction that mediated
glucose intolerance and insulin resistance. Currently,
hyperglycaemia in ST-elevation myocardial infarction is
an established danger sign and worsens the prognosis,
whether or not there is known diabetes mellitus.4
Tighter links have now developed between hyperglycaemia and future coronary heart disease. Although
type 2 diabetes mellitus is a well-known risk factor for
cardiovascular disease, more recently the metabolic
syndrome with subdiabetic blood glucose values
(5·6–7·0 mmol/L) has come to the fore as a risk factor
for myocardial infarction. Thus a prospective 20-year
follow-up of British men with no initial symptoms of
cardiovascular disease or diabetes showed a tripling of
cardiovascular risk in those with three abnormalities of
the metabolic syndrome, and a five-fold increase in the
risk of diabetes.5 Yet, what about the reverse risk from
myocardial infarction to diabetes? If acute myocardial
infarction is a stress reaction, and if stress mediates
hyperglycaemia, would it not be logical to expect that
patients with acute infarction, even in those free of
prevalent diabetes at the onset of infarction, would
be at greater risk of subsequent new-onset diabetes?
In today’s Lancet, Dariush Mozaffarian and colleagues
identify acute myocardial infarction as a prediabetes
risk-equivalent.6 During a mean follow-up of 3·2 years
in this retrospective analysis of the GISSI-PREVENZIONE
trial,7 12% of 8291 patients developed diabetes and up
to 33% developed either diabetes or impaired fasting
glucose. These findings further tie the knot between
myocardial infarction and hyperglycaemia—each causes
the other.
A reasonable hypothesis would be that the
previous acute myocardial infarction was associated
with an undetected modest, but definite tendency
towards prediabetes at the time of the attack.6 Even
glucose values below 5·6 mmol/L, far below the
cutoff of 7·0 mmol/L for the diagnosis of diabetes,
have prognostic value for a diagnosis of diabetes.8
Furthermore, impaired glucose tolerance is common
in acute infarction, suggesting that postchallenge
hyperglycaemia would more accurately have identified
those with occult diabetes at the time of the infarct.9
In the postinfarct period, any further impairment of
glucose metabolism could change occult diabetes to
www.thelancet.com Vol 370 August 25, 2007
Comment
ECG negative
AST
24
1000
colleagues have shown that the Mediterranean diet protects against new diabetes, the data were collected as part
of the large postinfarct GISSI-PREVENZIONE trial.7 The
results were remarkable, in that dietary supplementation
with n-3 polyunsaturated fatty acids, as found in fish oils,
decreased all-cause and cardiovascular deaths. The benefits
of the Mediterranean diet in the study by Mozaffarian and
colleagues were not mediated by n-3 fatty acids. As the
beneficial mechanisms of the Mediterranean diet and
the n-3 fatty acids seem to be different, it is interesting
that combining the Mediterranean diet with intake of
n-3 polyunsaturated fatty acids might give increased
postinfarct dietary protection both from cardiovascular
deaths and new diabetes, potentially a double benefit.
ECG positive
AST
16
AST
160
AST
212
800
600
400
200
0
120
80
40
Elsevier
0
8
am
8
pm
8
am
8
pm
8
am
8
pm
8
am
8
pm
8
am
8
pm
Figure: Metabolic changes at onset of acute myocardial infarction1
Patient admitted with severe chest pain. ECG becomes positive, and asparate
aminotransferase (AST, Reitman-Frankel units, 1 unit=0·482 IU/L) increases. Plasma
free-fatty-acids almost double, and glucose transiently enters diabetic range.
Lionel H Opie
Hatter Cardiovascular Research Institute, Department of Medicine,
Faculty of Health Sciences, University of Cape Town, Observatory,
Cape Town 7925, South Africa
[email protected]
I declare that I have no conflict of interest.
overt diabetes. Such additional factors, identified by
Mozaffarian and colleagues, include the use of β blockers
and diuretics, already known to precipitate diabetes in
hypertension trials.10 Diuretics and β blockers could be
surrogate markers for the development of heart failure,
another diabetogenic disorder.11 Among the lifestyle
factors that were risk factors for diabetes or impaired
fasting glucose were a low score for Mediterranean
diet, high body-mass index (BMI) or BMI gain, smoking,
and heavy consumption of wine.6 However, the study
provides no data on the clinical status of the patients.
Why were they given β blockers and diuretics?
Of the life-style factors, major interest lies in the
potential protective effect of the Mediterranean diet. A
Mediterranean diet score was used, including protective
high concentrations of vegetable intake, fruits, fish,
and olive oil. Barzi and colleagues showed that the
Mediterranean diet was associated with a low odds
ratio for all-cause mortality.12 In another study, there
was an association between the Mediterranean diet and
better survival in a Greek population.13 Furthermore, the
Mediterranean diet was associated with low oxidised
LDL-cholesterol concentrations.14 When the diet was
supplemented with olive oil or nuts, the Mediterranean
diet reduced plasma glucose, systolic blood pressure,
and C-reactive protein concentrations compared with
a standard low-fat diet.15 Although Mozaffarian and
www.thelancet.com Vol 370 August 25, 2007
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Kurien VA, Oliver MF. Serum-free-fatty-acids after acute myocardial infarction
and cerebral vascular occlusion. Lancet 1966; 2: 122–27.
Allison SP, Chamberlain MJ, Hinton P. Intravenous glucose tolerance, insulin,
glucose, and free fatty acid levels after myocardial infarction. BMJ 1969; 4:
776–78.
Tansey MJ, Opie LH. Plasma glucose on admission to hospital as a metabolic
index of the severity of acute myocardial infarction. Can J Cardiol 1986; 2:
326–31.
Zarich SW, Nesto RW. Implications and treatment of acute hyperglycemia in
the setting of acute myocardial infarction. Circulation 2007; 115: 436–39.
Wannamethee SG, Shaper AG, Lennon L, Morris RW. Metabolic syndrome vs
Framingham Risk Score for prediction of coronary heart disease, stroke, and
type 2 diabetes mellitus. Arch Intern Med 2005; 165: 2644–50.
Mozaffarian D, Marfisi RM, Levantesi G, et al. Incidence of new-onset diabetes
and impaired fasting glucose in patients with recent myocardial infarction
and the effect of clinical and lifestyle risk factors. Lancet 2007; 370: 667–75.
Hopper L, Ness A, Higgins JB, Moore T, Ebrahim S. Dietary supplementation
with n-3 polyunsaturated fatty acids and vitamin E after myocardial
infarction: results of the GISSI-Prevenzione trial. Lancet 1999; 354: 447–55.
Tirosh A, Shai I, Tekes-Manova D, et al. Normal fasting plasma glucose levels
and type 2 diabetes in young men. N Engl J Med 2005; 353: 1454–62.
Norhammar A, Tenerz A, Nilsson G, et al. Glucose metabolism in patients with
acute myocardial infarction and no previous diagnosis of diabetes mellitus: a
prospective study. Lancet 2002; 359: 2140–44.
Lam SK, Owen A. Incident diabetes in clinical trials of antihypertensive drugs.
Lancet 2007; 369: 1513–14.
Ashrafian H, Frenneaux MP, Opie LH. Metabolic mechanisms in heart failure.
Circulation 2007; 116: 434–38.
Barzi F, Woodward M, Marfisi RM, Tavazzi L, Valagussa F, Marchioli R.
Mediterranean diet and all-causes mortality after myocardial infarction:
results from the GISSI-Prevenzione trial. Eur J Clin Nutr 2003; 57: 604–11.
Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a
Mediterranean diet and survival in a Greek population. N Engl J Med 2003;
348: 2599–608.
Pitsavos C, Panagiotakos DB, Tzima N, et al. Adherence to the Mediterranean
diet is associated with total antioxidant capacity in healthy adults: the ATTICA
study. Am J Clin Nutr 2005; 82: 694–99.
Estruch R, Martinez-Gonzalez MA, Corella D, et al. Effects of a
Mediterranean-style diet on cardiovascular risk factors: a randomized trial.
Ann Intern Med 2006; 145: 1–11.
635
Comment
Catecholamine treatment for shock—equally good or bad?
Science Photo Library
See Articles page 676
636
The effectiveness and safety of many interventions in
critically ill patients remain poorly validated. Despite
heavy promotion by academic advocates, these different
strategies often show little to no benefit when tested in
randomised multicentre studies. This unfulfilled promise
applies to procedures as diverse as albumin versus saline,
“renal” dopamine, pulmonary-artery catheterisation,
and, more recently, intensive insulin treatment, selective
gut decontamination, and corticosteroids for septic
shock. Why the efficacy of such treatments achieved
in enthusiastic experts’ hands often fails to translate
into general effectiveness merits further study. Possible
explanations include the choice of patient, target or
treatment endpoints, protocol compliance, and potential
antagonism or synergism with one or more concurrent
treatments or procedures unique to some but not other
intensive-care units.
To this ever-expanding collection of busted flushes can
be added the choice of catecholamine in the treatment
of septic shock. Despite studies that highlight the
negative effects of epinephrine on splanchnic blood
flow, metabolism, and acid-base balance, and an absence
of recommendation for its use in adult patients,1 this
catecholamine remains a popular treatment option for
septic and cardiogenic shock. In today’s Lancet, however,
Djillali Annane and colleagues2 report no difference in
clinical outcomes or safety in a prospective comparison
of norepinephrine with or without dobutamine against
epinephrine in patients with septic shock. Arguably,
their sample-size computations were based on an
over-generous anticipation of outcome benefit, yet the
absence of any clear signal after 330 patients is adequate
to convince me that the choice of catecholamine is equally
good or, perhaps more accurately, equally bad.
Why the concern? Longstanding familiarity with catecholamines equates harm with well-recognised and
clinically obvious complications, such as digital ischaemia
and tachyarrhythmias. However, we are not generally
aware of other detrimental effects that, through their
covert nature, are unlikely to be detected in routine
practice but might well be clinically pertinent. Such
negative consequences include stimulation of bacterial
growth,3,4 an effect mediated by removal of iron from
lactoferrin and transferrin by the catechol moiety and
its subsequent acquisition by bacteria.5 This effect has
been shown with epinephrine and norepinephrine, and
synthetic agents such as dobutamine. Catecholamines also
increase factors related to bacterial virulence and biofilm
formation.3 Furthermore, host resistance to bacteria
might be compromised because both catecholamines and
dopaminergic agents, such as dopamine, dobutamine,
and dopexamine, affect activity and survival of most, if not
all, immune-cell populations.6 For example, epinephrine
and norepinephrine decrease the proinflammatory
effect of endotoxin, but enhance production of the
anti-inflammatory cytokine, interleukin 10.7,8 This increase
in interleukin 10 contributes to an immunosuppressive
effect on monocytes and macrophages. Norepinephrine
also has a direct inhibitory effect on the energy metabolism
of monocytes and macrophages.9
Plasma catecholamine concentrations rise up to
20-fold in critical illness.10 Concentrations returned
to normal over 5 days in eventual survivors, but rose
still further in non-survivors, many of whom received
exogenous catecholamines.10 Excess adrenergic stimulation induces metabolic derangements, including
insulin resistance with consequent hyperglycaemia,
and muscle catabolism. Despite increasing whole-body
and myocardial energy expenditure, catecholamines
also reduce metabolic efficiency by suppressing glucose
metabolism and enhancing oxidation of fatty acids.
The ATP yield per oxygen atom is 2·83 with free-fattyacid as substrate, compared with 3·17 with glucose.
This decreased efficiency, combined with the increase
www.thelancet.com Vol 370 August 25, 2007
Comment
in cardiac work induced by adrenergic stimulation and
peripheral vasoconstriction, will place an excess strain on
the failing heart. This effect might be relevant not only
for ischaemic myocardial injury11,12 but also for sepsis,
in which myocardial depression is well recognised.13
Indeed, randomised studies of β-adrenergic agonists
or phosphodiesterase inhibitors in decompensated
heart failure show worse outcomes than do placebo or
inotropic agents that do not increase cAMP, whereas
β-adrenergic blockade has been shown to be beneficial
in burn injury, heart failure, major surgery, and
experimental sepsis.14
We are, therefore, stuck between the Scylla of compromised tissue perfusion in septic shock and the
Charybdis of the complications of currently recommended
first-line treatment.1 Better alternatives to catecholamines
are needed, which might include agents as diverse as
vasopressin, levosimendan, or specific inducible inhibitors
of nitric oxide synthase. We also need to better define the
lowest acceptable blood pressure in individual patients to
minimise the harmful effects of excessive catecholamine
dosing. Additionally, we should limit the use of concurrent
medications that contribute to hypotension (eg, excess
sedative dosing) and vascular hyporeactivity (eg,
etomidate).
I have served on advisory boards for Abbott/Orion (levosimendan) and Ferring
(vasopressin analogues).
Mervyn Singer
13
Bloomsbury Institute of Intensive Care Medicine, Wolfson Institute
for Biomedical Research and Department of Medicine, University
College London, London WC1E 6BT, UK
[email protected]
1
2
3
4
5
6
7
8
9
10
11
12
14
Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines
for management of severe sepsis and septic shock. Crit Care Med 2004; 32:
858–73.
Annane D, Vignon P, Renault A, for the CATS Study Group. Norepinephrine
plus dobutamine versus epinephrine alone for management of septic shock:
a randomised trial. Lancet 2007; 360: 767−84.
Lyte M, Freestone PP, Neal CP, et al. Stimulation of Staphylococcus epidermidis
growth and biofilm formation by catecholamine inotropes. Lancet 2003;
361: 130−35.
Freestone PP, Haigh RD, Lyte M. Specificity of catecholamine-induced growth
in Escherichia coli O157:H7, Salmonella enterica and Yersinia enterocolitica.
FEMS Microbiol Lett 2007; 269: 221−28.
Freestone PP, Lyte M, Neal CP, Maggs AF, Haigh RD, Williams PH.
The mammalian neuroendocrine hormone norepinephrine supplies iron for
bacterial growth in the presence of transferrin or lactoferrin. J Bacteriol 2000;
182: 6091−98.
Oberbeck R. Catecholamines: physiological immunomodulators during
health and illness. Curr Med Chem 2006; 13: 1979−89.
van der Poll T, Jansen J, Endert E, Sauerwein HP, van Deventer SJH.
Noradrenaline inhibits lipopolysaccharide-induced tumor necrosis factor and
interleukin-6 production in human whole blood. Infect Immun 1994; 62:
2046−50.
van der Poll T, Coyle SM, Barbosa K, Braxton CC, Lowry SF. Epinephrine
inhibits tumor necrosis factor-alpha and potentiates interleukin-10
production during human endotoxemia. J Clin Invest 1996; 97: 713−19.
Lunemann JD, Buttgereit F, Tripmacher R, Baerwald CG, Burmester GR,
Krause A. Norepinephrine inhibits energy metabolism of human peripheral
blood mononuclear cells via adrenergic receptors. Biosci Rep 2001; 21: 627−35.
Boldt J, Menges T, Kuhn D, Diridis C, Hempelmann G. Alterations in circulating
vasoactive substances in the critically ill—a comparison between survivors
and non-survivors. Intensive Care Med 1995; 21: 218−25.
Mjos OD, Kjekshus JK, Lekven J. Importance of free fatty acids as a
determinant of myocardial oxygen consumption and myocardial ischemic
injury during norepinephrine infusion in dogs. J Clin Invest 1974; 53: 1290–99.
Beanlands RS, Nahmias C, Gordon E, et al. The effects of β1-blockade on
oxidative metabolism and the metabolic cost of ventricular work in patients
with left ventricular dysfunction: a double-blind, placebo-controlled,
positron-emission tomography study. Circulation 2000; 102: 2070–75.
Rudiger A, Singer M. Sepsis-induced cardiac depression. Crit Care Med 2007;
35: 1599–608.
Suzuki T, Morisaki H, Serita R, et al. Infusion of the β-adrenergic blocker
esmolol attenuates myocardial dysfunction in septic rats. Crit Care Med 2005;
33: 2294–301.
Speaking out about human rights and health in West Papua
The recent Human Rights Watch (HRW) report, Out
of Sight, alerted the international community to the
hidden human-rights abuses in West Papua, Indonesia’s
most easterly province.1 The effect of the crisis on the
health and wellbeing of the indigenous population of
West Papua is an issue that has attracted little attention
in contemporary medical publications.
West Papua occupies half of the island of New Guinea.
Most of its 2 million indigenous inhabitants live in
remote villages scattered across the mountainous and
forested territory. In 1969, after a referendum brokered
by the UN, Indonesia annexed West Papua following a
decision-making process that was widely regarded as
www.thelancet.com Vol 370 August 25, 2007
flawed.2 Since then, independence groups have waged a
low-level guerrilla war against Indonesian rule.
Both restrictions on data gathering by foreigners
and the inaccessible terrain create major obstacles
to undertaking research in West Papua. The HRW
report therefore is invaluable because it provides
documentation of systematic abuses, including torture,
rape, and extrajudicial killings directed against militants
and the civilian population. Police and military personnel
who are accused of violations seem to be immune from
prosecution.1 Refugees fleeing persecution have sought
asylum in Papua New Guinea and in developed countries,
such as the UK and Australia. A participant in our
See Series page 698
637
Getty Images
Comment
Demonstrations and violent clashes followed government’s decision to split province in August, 2003
mental health project of Australian-based refugees, John
(an alias), recounted a story that is consistent not only
with the major human-rights and legal reports from West
Papua,1,2 but also with stories from other participants in
the project. As a child, John witnessed the burning of his
village and the brutal public rape, torture, and murder of
his family. The military apprehended his uncle as he fled
to the border, tearing his finger and toenails off before
forcing him to dig his own grave and shooting him
in public. John suffers from multiple musculoskeletal
complaints and nightmares arising from his torture.
Furthermore, he lives in constant fear for the safety of his
remaining family left in West Papua.
Indonesian rule has brought about major changes
to the demography, ecology, and traditional way of
life in West Papua.3,4 Mining operations that are poorly
regulated are polluting major rivers, while extensive
illegal logging is destroying natural habitats that are
crucial to a traditional land-based culture.3,4 Indonesia’s
transmigration policy has relocated more than
three-quarters of a million ethnically distinct settlers to
West Papua, which is an immense social transformation
that threatens to marginalise the indigenous people,
whose numbers are further threatened by a falling
fertility rate.3 Indigenous Papuans have been displaced
to areas where traditional crops are difficult to grow
and the prevalence of communicable diseases is high.
Questions have been raised about whether these
fundamental disruptions to the traditional way of life
constitute an insidious form of cultural genocide.1
638
Public-health indicators, although incomplete, suggest
that the general health of Papuans is poor.5,6 Malaria,
upper respiratory tract infections, and dysentery are
major causes of childhood morbidity, with infant
mortality ranging from 70 to 200 per 1000.5 More than
50% of children younger than 5 years are undernourished,
and immunisation rates are low.5,6 Maternal mortality is
three times the rate of women in other parts of Indonesia.5
HIV/AIDS rates are 40 times the national average,7
and the epidemic is being fuelled by a burgeoning sex
trade, low levels of literacy, and inadequate services for
prevention and treatment of this disease.7,3 In 2000,
Indonesia acknowledged the parlous state of health in
West Papua, committing US$2·25 billion to enhance
services.6 However, critics continue to comment about
the gross inadequacy of the medical system in relation to
human resources, access, and quality.2,3,7
In response to international criticisms, Indonesia
has offered West Papuans a special autonomy plan
to increase participation of indigenous people in
governance.1 The HRW report suggests, however, that
the political changes have not led to an improvement
in human rights. Vested interests, the remoteness of
the territory, and marginalisation of indigenous people
are obstacles to genuine political change. Nevertheless,
international pressures have prompted improvements
in human rights in other conflict-affected areas of
Indonesia, specifically in East Timor and Aceh. The
international medical profession can play a part in
bringing about change—eg, by engaging with and
supporting progressive Papuan health professionals
in their efforts to improve services, establish training
programmes, and improve standards of care in the
region. Furthermore, gathering more comprehensive
data that focuses on the public-health results of conflict
and socioeconomic neglect is essential. By maintaining
a close scrutiny of health outcomes in West Papua,
medical professionals can have a key role in breaking
the prevailing silence about one of the world’s least
publicised human-rights crises.
*Susan Rees, Derrick Silove
School of Public Health, Tropical Medicine and Rehabilitation
Sciences, James Cook University, Cairns 4870, Australia (SR); and
Centre for Population Mental Health Research, Sydney South West
Area Health Service and School of Psychiatry, University of New
South Wales, Sydney, NSW, Australia (DS)
[email protected]
www.thelancet.com Vol 370 August 25, 2007
Comment
We declare that we have no conflict of interest.
1
2
3
Human Rights Watch. Out of sight: endemic abuse and impunity in Papua’s
central highlands. Human Rights Watch 2007; 19: 1–81.
Brundige E, King W, Vahali P, et al. Indonesian human rights abuses in West
Papua: application of the law of genocide to the history of Indonesian
control. Allard K. Lowenstein International Human Rights Clinic. New
Haven: Yale Law School, 2004.
Wing J, King P. Genocide in West Papua? The role of the Indonesian state
apparatus and a current assessment of the Papuan people. Sydney and
Jayapura: West Papua Project at the Centre for Peace and Conflict Studies,
2005.
4
5
6
7
Environmental Investigation Agency and Telapak. The last frontier: illegal
logging in Papua and China’s massive timber theft. London and Jakarta:
Environmental Investigation Agency and Telapak, 2005: 1–27.
Blair D, Phillips D. Indonesia Commission: peace and progress in Papua.
New York: Council of Foreign Relations, 2003: 76.
Diani H. Health: a specter for Irian Jaya. Jakarta Post Aug 21, 2000: 5.
Butt L, Numbery G, Morin J. The smokescreeen of culture: AIDS and the
Indigenous in Papua, Indonesia. In: Jones R, Finau SA, eds. Pacific health
dialogue: Guam and health transition in the Pacific. Waimauku: Resource
Books, 2002; 9: 283–89.
Where have all the bees gone?
www.thelancet.com Vol 370 August 25, 2007
Bees are the major pollinators of a wide range of
crops, including almonds, cherries, pears, melons, and
cucumbers.8 This role means that they have a massive
commercial value. As well as beekeepers suffering
financially with replacement costs and loss of revenue,
there may well be wider consequences for us all, because
the supply of crops might dwindle. Einstein is often,
though controversially, quoted to have said that without
the bee, mankind would have but 4 years to live.6
Science Photo Library
Beekeepers have been returning to their hives only to
find them deserted. Colony collapse disorder (CCD) is
the name given to this strange disappearance of bees.
Stranger still, no dead bees are found, but honey and
pollen remain. The effects of CCD can be massive, with
some beekeepers reporting losses of 50–90% of bees.1
The problem has been reported across North America,
and, more recently, in European countries, including
France, Sweden and Germany.2
Several theories have been proposed to explain
CCD. Potential suspect pathogens include the Nosema
parasite and Aspergillus fungal infections. Bees could
be more susceptible to disease when their immune
systems are weakened by antibiotics or stresses caused
by apiary overcrowding, poor nutrition, or migratory
stress, because commercial bees are often transported
over long distances for pollination.3 Pesticides have
also been suggested as a cause of CCD, particularly
nicotinoids such as imidacloprid.4 These compounds
have been shown to be safe for honeybees under
field conditions,5 although the sublethal effects are
disputed.
Much hype ensued when a study at Landau University
implicated cell-phone use with CCD. However, coauthor of that study, Jochen Kuhn, is quoted as saying
that there is no proof that electromagnetic radiation
is the cause.6 Global warming has also been blamed,
because bees maintain hive temperature within 3°C, and
varying temperatures affect the bees’ ability to perform
their communicative waggle dance.7 However, CCD is
not a new phenomenon, with records of mysteriously
abandoned hives dating back to 1869.2
Jessica Hamzelou
University College London, London WC1E 6BT, UK
[email protected]
I declare that I have no conflict of interest.
1
2
3
4
5
6
7
8
Frazier M, van Engeldorp D, Caron D. Colony collapse disorder. Apiary News
in Illinois May, 2007. http://www.agr.state.il.us/programs/bees/CCD.pdf
(accessed Aug 14, 2007).
Underwood RM, van Engelsdorp D. Colony collapse disorder: have we seen
this before? Bee Culture. http://www.beeculture.com/content/
ColonyCollapseDisorderPDFs/7%20Colony%20Collapse%20Disorder
%20Have%20We%20Seen%20This%20Before%20-%20Robyn%20M.%20
Underwood%20and%20Dennis%20vanEngelsdorp.pdf (accessed Aug 15,
2007).
Pettis J, van Engelsdorp D, Cox-Foster D. Colony Collapse Disorder Working
Group: pathogen sub-group progress report. May 12, 2007. http://maarec.
cas.psu.edu/CCDPpt/PathogenSub-GroupProjectMay142007.pdf (accessed
Aug 15, 2007).
Cummins J. Requiem for the honeybee. April 24, 2007. http://www.i-sis.
org.uk/requiemForTheHoneybee.php (accessed Aug 15, 2007).
Maus C, Curé G, Schmuck R. Safety of imidacloprid seed dressings to honey
bees: a comprehensive overview and compilation of the current state of
knowledge. Bull Insectol 2003; 56: 51-57. http://www.bulletinof
insectology.org/pdfarticles/vol56-2003-051-057maus.pdf (accessed
Aug 15, 2007).
Sylvers E. Case of the disappearing bees creates a buzz. International Herald
Tribune April 22, 2007. http://www.iht.com/articles/2007/04/22/news/
wireless23.php (accessed Aug 15, 2007).
Tautz J, Maier S, Groh C, Rössler W, Brockmann A. Behavioural performance
in adult honey bees is influenced by the temperature experienced during
their pupal development. Proc Natl Acad Sci USA 2003; 100: 7343–47.
Maheshwari, JK. Endangered pollinators. EnviroNews January, 2003.
http://isebindia.com/01_04/03-01-3.html (accessed Aug 15, 2007).
639
Comment
Clinical update: obstetric anaesthesia
In the developed world, around two-thirds of women
use some form of anaesthesia during childbirth.1,2
Changes in maternal physiology during pregnancy,
and the care of both mother and fetus, present
unique challenges to obstetric anaesthetists. Also,
diverse maternal expectations of the birth experience,
demands for neuraxial block, advancing maternal age,
obesity, coexisting medical morbidities, and caesarean
section rates have all escalated.1 Serious adverse events
associated with obstetric anaesthesia are rare, but are
of particular concern because pregnancy and childbirth
generally has an expectation of little or no risk.
Severe obstetric morbidity occurs in around 1% of
deliveries. Risk factors include maternal age above
34 years, hypertension, previous postpartum haemorrhage, induction of labour, and previous caesarean
section.3 Anaesthetists need to minimise adverse
consequences, especially those relating to obstetric
haemorrhage and severe pre-eclampsia.4 The American
Society of Anesthesiologists (ASA) Task Force on
Obstetric Anesthesia recently updated their guidelines
(panel).5 Many of the recommendations are based
on consensus opinion, indicating a paucity of good
evidence. Cochrane systematic reviews6 also have limited
implications for guiding obstetric anaesthesia practice,
reflecting the small number of women studied and the
very low incidence of anaesthesia-related complications
in this population. Maternal mortality reports provide
valuable insight into these rare events and can influence
obstetric anaesthesia practice.7
Improved safety in obstetric anaesthesia is shown
by the substantial reduction in the incidence of
pulmonary aspiration of gastric contents and its
consequences, and in other risks associated with the
use of general anaesthesia for caesarean section.
This reduction is attributed to the more active role
played by senior experienced staff, the increased use of
regional techniques, fasting regimens, and the use of
non-particulate antacids and H2-receptor antagonists.8
Left-lateral tilt to avoid aortocaval compression and care
in drug administrations to avoid fetal adverse effects
remain standard.
Pregnant women want information about analgesia,
anaesthesia, and the risks associated with central
neuraxial block.9,10 The consent process should include
640
a description of the procedure, alternatives, and risks
and benefits. When possible, such information should
be given antenatally, because the opportunity for
discussion is necessarily shortened during labour and
might be less than optimum.11 Women can be reassured
that long-term postpartum back pain and caesarean
section rates are not increased after the placement of
intrapartum neuraxial block.12
Many techniques are available for analgesia during
labour and can be used alone or in combination.
Many women choose non-drug pain relief in labour.
Interventions such as continuous intrapartum support
and hypnosis or acupuncture could reduce the need
for pharmacological analgesia and other interventions
during labour and childbirth.12 Self-administered nitrous
oxide is simple, safe, inexpensive, and effective in
some women.12 Parenteral analgesia, as intramuscular
pethidine or morphine, is used by almost a quarter
of women in labour.1 Intravenous fentanyl can be
administered by patient-controlled analgesia during
labour when regional anaesthesia is contraindicated,
unavailable, or impossible. Intravenous fentanyl can
also provide short-term analgesia during neuraxial block
placement in a distressed mother or in the second stage
of labour when regional block is unlikely to have time to
be effective before delivery. Concern about the risk of
neonatal respiratory depression at birth limits the dose,
frequency, and timing of opioid administrations during
labour.
Epidural analgesia provides more effective pain
relief during labour than non-epidural methods.12
However, traditional epidurals (bupivacaine 0·25%
or greater) can lengthen the second stage of labour,
and increase the incidence of labour augmentation,
maternal hypotension, motor blockade, and need for
instrumental vaginal birth. Epidural analgesia with low
concentrations of local anaesthetic (eg, ropivacaine 0·2%
or bupivacaine 0·1%) can increase the incidence of
spontaneous vaginal birth compared with a traditional
technique.12 Single-shot spinal analgesia can be useful
before placement of an epidural catheter in a distressed
woman or as the sole analgesic when spontaneous
vaginal birth is imminent.
Combined spinal epidural analgesia is a popular
alternative to epidural alone in some centres. It involves
www.thelancet.com Vol 370 August 25, 2007
Comment
the administration of intrathecal opioid (eg, fentanyl)
or local anaesthetic, or a combination of the two, before
or immediately after epidural catheter placement. This
technique has the advantages of spinal analgesia (eg, fast
onset) with the ability to provide additional analgesia
or anaesthesia as required via the epidural catheter.
Although increased mobility and other advantages
are claimed, comparisons of combined spinal epidural
analgesia with low-dose epidural techniques show that
both provide effective analgesia in labour and a high
maternal satisfaction, with only a slightly faster onset
with the combined technique and less pruritus with
epidurals.13
Epidural analgesia during labour is extended to
provide surgical anaesthesia should caesarean section
be necessary. Spinal anaesthesia has the advantage of
simplicity, rapid onset, low failure rate, minimum drug
dose, and excellent muscle relaxation, which makes it
the technique of choice for both elective and emergency
caesarean section when a functioning epidural catheter
is not in place. Regional anaesthesia has well-recognised
advantages compared with general anaesthesia, including intact protective airway reflexes, the mother
being awake at the baby’s birth, and reduced need for
systemic opioids postoperatively. However, general
anaesthesia is preferred for operative delivery when a
regional technique is contraindicated, such as in major
haemorrhage or coagulopathy. Compared with epidural
anaesthesia, spinal anaesthesia for caesarean section is
associated with a shorter time to starting the operation,
but an increased incidence of sympathetically mediated
maternal hypotension that requires treatment.14
Techniques that reduce the incidence of hypotension
associated with spinal anaesthesia at the time of
caesarean section include administration of intravenous
fluids and vasopressors (eg, phenylephrine or ephedrine),
and lower-limb compression with stockings or inflatable
boots.15
Pre-eclampsia can lead to an exacerbated hypertensive
response at laryngoscopy and airway difficulties,
should general anaesthesia be needed. A single-shot
spinal anaesthetic is the preferred method of providing
anaesthesia for women presenting for caesarean section
without an epidural catheter in situ, provided there is
no hypovolaemia or clinical evidence of coagulopathy.
Although the lower limit varies with the clinical situation,
a platelet count above 80 000×10⁹/L is usually adequate
www.thelancet.com Vol 370 August 25, 2007
Panel: Selected recommendations adapted from ASA guidelines for obstetric
anaesthesia5
• Spinals or epidurals are preferred over general anaesthesia for most caesarean sections
• Early insertion of spinal or epidural catheter for obstetric (eg, twin gestation or
pre-eclampsia) or anaesthetic indications (eg, anticipated difficult airway or obesity)
should be considered to reduce need for general anaesthesia if an emergent procedure
becomes necessary
• If functioning epidural catheter is in place and patient is haemodynamically stable,
epidural anaesthesia is preferred technique for removal of retained placenta
• In major maternal haemorrhage, general anaesthesia is preferred over neuraxial
anaesthesia. Suggested resources for obstetric haemorrhagic emergencies include:
large-bore intravenous catheters, fluid warmer, forced-air body-warmer, blood bank,
uterotonic agents, surgery. Consider off-label use of recombinant activated factor VII
in massive postpartum haemorrhage refractory to conventional treatments
• Sublingual (ie, metered dose spray) or intravenous nitroglycerin can be used as an
alternative to general anaesthesia with halogenated agents for uterine relaxation
• Intravenous ephedrine and phenylephrine are both suitable for treating hypotension
during neuraxial anaesthesia. In absence of maternal bradycardia, phenylephrine
might be preferable because of improved fetal acid-base status in uncomplicated
pregnancies
• For postoperative analgesia after neuraxial anaesthesia for caesarean birth, neuraxial
opioids are preferred over intermittent injections of parenteral opioids
• Anaesthetists should have preformulated strategy for intubation of difficult airway. If
tracheal intubation fails, ventilate with mask and cricoid pressure, or with a laryngeal
mask airway or supraglottic airway device. Suggested resources for difficult intubation
include: rigid laryngoscope blades of alternative design and size from those routinely
used; laryngeal mask airway; endotracheal tubes of assorted size; endotracheal tube
guides (eg, semirigid stylets), light wands, and forceps designed to manipulate distal
portion of endotracheal tube; retrograde intubation equipment; at least one device
suitable for emergency non-surgical airway ventilation (eg, intubating laryngeal mask
airway); fibreoptic intubation equipment; equipment suitable for emergency
cricothyrotomy
• If cardiac arrest occurs during labour and birth, standard resuscitative measures should
be started. Uterine displacement (usually left displacement) should also be
maintained. If maternal circulation is not restored within 4 min of cardiac arrest,
caesarean section should be done by obstetric team
for administration of neuraxial block, as long as there
are no other risk factors16 and the count is not falling.
Regional block should be avoided for at least 12 h after
standard doses of prophylactic low-molecular-weight
heparin. Epidural catheters should not be removed for at
least 12 h after the last dose of such treatment, and any
subsequent dose should not be given until at least 2 h
after catheter removal.12
Although neuraxial techniques are largely safe and
effective, potential complications, such as postdural
puncture headache, can be problematic and fairly
common (around 1%).17 Preventive measures involve
the use of small diameter 25G or 27G pencil-point
spinal needles.5 Rare complications, such as meningitis,
spinal abscess, or epidural haematoma, can be severe.
641
Comment
Other settings suggest that full aseptic precautions,
including mask, gown, and gloves, should be used,18
with meticulous attention to detail when doing regional
techniques in obstetrics.
Anticipation and preparation for managing the
difficult airway is essential at all times in obstetric
anaesthesia and analgesia, but is especially relevant
before induction of general anaesthesia. Oxygenation
without aspiration is the main initial goal. A poor view
of the glottis at laryngoscopy can often be rectified by
simple manoeuvres, such as adjustment of the assistant’s
cricoid-pressure hand while maintaining direct
laryngoscopy. The laryngeal mask airway and bougie are
life-saving in the event of difficult intubation.5
Obstetric units are advised to practice emergency
drills to improve team responses to critical situations,
such as major obstetric haemorrhage.4 Recognised
risk factors for haemorrhage include the presence of
placenta praevia, low-lying anterior placenta (especially
when overlying a previous caesarean section scar), grand
multiparity, and history of postpartum haemorrhage.
Obese women have an increased risk of pregnancy
complications, such as gestational diabetes and hypertensive conditions (including pre-eclampsia) and
labour and birth complications (including obstetric
interventions and caesarean section). Despite technical
difficulties in administering epidural analgesia for this
group of women, early provision of regional techniques
might decrease perinatal and anaesthetic-related complications should emergent delivery be needed.19
Maternal deaths related to anaesthetic complications
have reached a low nadir.8 The CEMACH Report4 and
the ASA Task Force5 both emphasise the importance
of communication and teamwork in providing safe
obstetric care. The multidisciplinary team is often challenged by the urgency of operative birth, especially
when the mother might have an underlying medical
condition or complication of pregnancy.20 Maintaining
hard-won improvements in the safety of obstetric
anaesthesia requires a constant guard against apathy
and complacency.
642
*Allan M Cyna, Jodie Dodd
Department of Women’s Anaesthesia, Women’s and Children’s
Hospital, Adelaide, SA 5006, Australia (AC); and Discipline of
Obstetrics and Gynaecology, University of Adelaide, Adelaide, SA,
Australia (JD)
[email protected]
JD is supported by the Australian National Health and Medical Research Council,
Neil Hamilton Fairley Fellowship. We declare that we have no conflict of interest.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Chan A, Scott J, Nguyen A-M, Sage L. Pregnancy outcome in South
Australia 2005. December, 2006. http://www.dh.sa.gov.au/pehs/PDF-files/
preg-outcome-report-dec06.pdf (accessed Aug 11, 2007).
Clergue F, Auroy Y, Pequignot F, Jougla E, Lienhart A, Laxenaire MC. French
survey of anesthesia in 1996. Anesthesiology 1999; 91: 1509–20.
Waterstone M, Bewley S, Wolfe C. Incidence and predictors of severe
obstetric morbidity: case control study. BMJ 2001; 322: 1089–94.
Lewis G, Drife J. Why mothers die 2000–2002: the sixth report of the
confidential enquiries into maternal deaths in the United Kingdom.
London: Royal College of Obstetricians and Gynaecologists, 2004.
ASA. Practice guidelines for obstetric anesthesia: an updated report by the
American Society of Anesthesiologists Task Force on Obstetric Anesthesia.
Anesthesiology 2007; 106: 843–63.
Cochrane Collaboration. Database of systematic reviews. Cochrane Library
2007. http://www3.interscience.wiley.com/cgi-bin/mrwhome/106568753/
HOME (accessed Aug 11, 2007).
Bolton TJ, Randall K, Yentis SM. Effect of the confidential enquiries into
maternal deaths on the use of syntocinon at caesarean section in the UK.
Anaesthesia 2003; 58: 277–79.
Clyburn PA. Early thoughts on ‘Why Mothers Die 2000–2002’. Anaesthesia
2004; 59: 1157–59.
Kelly G, Blunt C, Moore P, Lewis M. Consent for regional anaesthesia in the
United Kingdom: what is material risk? Int J Obstet Anesth 2004; 13: 71–74.
Jackson A, Henry R, Avery N, VanDenKerkhof E, Milne B. Informed consent
for labour epidurals: what labouring women want to know. Can J Anaesth
2000; 47: 1068–73.
Paech M. “Just put it in!” Consent for epidural analgesia in labour.
Anaesth Intensive Care 2006; 34: 147–49.
Australian and New Zealand College of Anaesthetists and Faculty of Pain
Medicine. Acute pain management: scientific evidence, 2nd edn. 2005.
http://www.anzca.edu.au/publications/acutepain.pdf (accessed Aug 9,
2007).
Simmons S, Cyna A, Dennis A, Hughes D. Combined spinal-epidural versus
epidural analgesia in labour. Cochrane Database Syst Rev 2007; 3:
CD003401.
Ng K, Parsons J, Cyna AM, Middleton P. Spinal versus epidural anaesthesia
for caesarean section. Cochrane Database Syst Rev 2004; 2: CD003765.
Cyna AM, Andrew M, Emmett RS, Middleton P, Simmons SW. Techniques
for preventing hypotension during spinal anaesthesia for caesarean
section. Cochrane Database Syst Rev 2006; 4: CD002251.
Douglas JM. Neuraxial anesthesia in parturients with thrombocytopenia.
In: Halpern SH, Douglas JM, eds. Evidence-based obstetric anesthesia, 1st
edn. London: BMJ Books, Blackwell Publishing, 2005: 165–77.
Turnbull DK, Shepherd DB. Post-dural puncture headache: pathogenesis,
prevention and treatment. Br J Anaesth 2003; 91: 718–29.
Hepner DL. Gloved and masked—will gowns be next? The role of asepsis
during neuraxial instrumentation. Anesthesiology 2006; 105: 241–43.
Saravanakumar K, Rao SG, Cooper GM. Obesity and obstetric anaesthesia.
Anaesthesia 2006; 61: 36–48.
Murray D, Enarson C. Communication and teamwork: essential to learn but
difficult to measure. Anesthesiology 2007; 106: 896–98.
www.thelancet.com Vol 370 August 25, 2007
World Report
Experts disagree over NICE’s approach for assessing drugs
This month, the UK’s High Court ruled in favour of the National Institute for Health and Clinical
Excellence decision to limit access to drugs for Alzheimer’s disease. But the legal win is unlikely
to stem the recent spate of controversies surrounding the institute. Richard Hoey reports.
www.thelancet.com Vol 370 August 25, 2007
For some, NICE is fundamentally
discriminatory, through its use of the
quality-adjusted life-year (QALY) to
assess cost-effectiveness. The QALY
is a statistical measure that assesses
treatments by their effect on both
length and quality of life, and cost per
QALY is intended as a common currency
for comparing value for money across
age-groups or populations.
But John Harris, professor of
bioethics at the University of
Manchester and editor-in-chief of
the Journal of Medical Ethics, believes
use of the QALY discriminates against
the elderly, who have fewer years
of life left to preserve. “I am rather
disheartened by NICE’s approach that
what matters is life-years, not lives.
It is perfectly reasonable to take this
position, but it is not incontrovertibly
true. I believe a tragedy is big or small
in proportion to the lives lost, not the
life-years lost”, he says.
Use of the QALY is far from
universal among equivalent bodies
internationally. The Pharmaceutical
Management Agency of New
Zealand (PHARMAC), established in
1993—6 years before NICE—assesses
cost-effectiveness as only part of its
assessment, and unlike NICE sets no
QALY threshold. Canada’s Common
Drug Review (CDR) does not use
the QALY at all. The CDR prefers to
concentrate on assessing clinical
effectiveness using a more flexible
“commonsense” approach to take into
account cost.
James Wright, managing director of
the University of British Colombia’s
Therapeutics Initiative, which assesses
evidence for the CDR, questions the
evidence for use of QALYs. “The CDR
does not get into QALYs and I think
it is smart not to. I am not convinced
that trying to calculate QALYs is useful,
they are based on assumptions that are
hard for most people to understand
and the science is often questionable.
It is much better to know the benefits
and harms in the [specific] population
a drug has been tested in.”
But Sir Michael Rawlins, chairman
of NICE, believes that the High
Court ruling was a vindication of the
institute’s approach. He concedes
that there are some procedural
lessons to be learnt, to ensure that all
groups are catered for, but vigorously
For more on the NICE case see
Comment Lancet 2007; 370:
547–48
Getty Images
It was a legal ruling greeted with relief
at the offices of the National Institute
for Health and Clinical Excellence (NICE)
but bitter disappointment among
campaigners and the pharmaceutical
industry. After 10 months of legal
wrangles, a London court upheld the
institute’s decision to restrict access
to drugs for Alzheimer’s disease.
But although the ruling will fortify
those who believe medicines should
be rigorously assessed for costeffectiveness, no one is pretending it
will quell the lasting controversy over
how such assessments should be done.
NICE is seen as a world leader in
allocating health-care resources by
evidence-based assessment, and the
legal proceedings were followed by
interested groups far beyond England
and Wales—the area of its remit. But
the institute has also proved a magnet
for controversy, with the Alzheimer’s
furore only the most intense of a series
in recent months, also involving drugs
for macular degeneration, rheumatoid
arthritis, and bowel cancer.
David Wilkinson, a consultant in
old-age psychiatry at the Hampshire
Partnership Trust, is among those who
remain angry at the NICE decision
to limit access to the Alzheimer’s
drugs donepezil, rivastigmine, and
galantamine, believing the institute
to be insensitive to the benefits of
drugs for degenerative illness. ”NICE
has taken the view a response is only
measurable if there is an improvement
over baseline, but with a degenerative
disease that is clearly difficult. It is
facile and simplistic”, he says.
The row vividly illustrates the
difficulty of assessment of treatments
in a way everyone agrees is fair, and that
reconciles the quality of trial evidence
with patients’ and doctors’ personal
beliefs about a treatment’s value.
Alzheimer’s patients’ groups speak to the media following the High Court ruling
643
NICE
World Report
Sir Michael Rawlins believes the High Court ruling vindicates NICE’s approach
rejects the charge that use of the
QALY is discriminatory. “It is a woeful
misunderstanding of the QALY. Almost
invariably it works to the advantage of
the elderly, because often treatments
are more cost effective in them.”
“We have to work out if treatments
are good value for money compared
with what we already have, and do this
across conditions and populations. The
QALY seems the most appropriate, if
not the only appropriate, measure we
have got”, he said.
Although some countries may be
reluctant to adopt the QALY, many
—including Germany, France, and
Sweden—plan to replicate the role
of NICE more generally. The USA’s
Institute of Medicine also wants to
introduce a resource centre for healthcare providers.
Just how successful NICE has been
is open to question. Harris is among
those who criticise it as a crude
rationing body. Others, conversely,
point to its failure to control escalating
drug costs. Between 2000 and 2005,
the UK’s pharmaceutical expenditure
as a proportion of gross domestic
product (GDP) rose by an average of
7·3% a year. This figure compares with
an annual rise of 7·7% in the USA and
10·8% in Canada between 1999 and
2004, but with rises of 6·0% and 6·7%,
respectively, in Germany and France.
Rawlins accepts that NICE has to
take into account the limited pot of
money available for health care, but
644
insists it does not exist to save money.
“I do not think anyone makes any
bones that resources are limited and
are allocated in the most appropriate
fashion. But I have always objected
to rationing as a term. I experienced
rationing as a child. You got your
sweeties for the week and that was
it. We do not behave that way in the
health service.” He estimates NICE has
actually added £1·2 billion each year to
the UK drugs bill. “I never expected to
save money. If we are there to improve
health services, that will cost more.”
But New Zealand’s PHARMAC has
been successful at controlling drug
expenditure, estimating it has kept
annual increases down to 2% or 3%.
Matthew Brougham, PHARMAC’s
acting chief executive, explains:
“Where New Zealand and the UK
diverge is that PHARMAC has a budget
management role, so value for money
and budgetary impact are inextricably
linked. NICE makes recommendations
independently of the impact on
NHS [National Health Service] trust
budgets. We have seen comment that
some NICE recommendations have
caused tension at trusts.”
Such tensions will be one of the
issues that the House of Commons
Health Select Committee addresses
when it reports on its current inquiry
into NICE. Alan Maynard—a specialist
adviser to the committee and
professor of health economics at the
University of York—says trusts are
left to take some difficult rationing
decisions because of differences in
the status of NICE guidance, with a
legal obligation to fund technology
appraisals of new drugs, but not
clinical or public-health guidelines.
”London’s primary care trusts [PCTs]
are trying to get an agreement on
rationing criteria and my view is we
should do that across the country.
For obesity surgery, some PCTs use a
BMI [body mass index] threshold of
30, others 40. We should be looking
at the evidence and coming up with
a decision for everyone. PCTs are too
small to do that”, he says.
The Association of the British
Pharmaceutical Industry (ABPI) is
lobbying the inquiry for changes to the
way NICE assesses new treatments,
particularly drugs for terminal
illnesses such as cancer. It wants NICE
to shift its cost per QALY threshold
for cancer drugs upwards from the
current limit of £30 000. “QALYs are
relative and if your comparator drug is
a cheap generic, it is difficult to hit the
cost-effectiveness threshold. You are
saying to companies do not bother to
research drugs in these areas”, an ABPI
spokesperson told The Lancet.
But Peter Smith, a colleague of
Maynard’s at York, has come to the
opposite conclusion. He recently
did a study showing that PCTs were
using far stricter QALY thresholds for
their rationing decisions than was
NICE. He suggests NICE has been
“if anything too liberal in accepting
some high-cost treatments”, and
advocates reducing the threshold
to £20 000 per QALY. Smith admits
a tougher threshold for new drugs
would be controversial, but argues
public-health interventions, or more
efficient ways of delivering existing
treatments, would be better value
for money for the NHS than would be
many new drugs.
Wright agrees. He warns that “NICE
has lost some of its credibility because
in a number of cases it has come under
political pressure and its decisions
have been reversed. Any group that
uses evidence is going to be making
the judgment that most of the drugs
coming onto the market are not very
valuable”.
But NICE seems unlikely to allow
itself to be pushed towards a stricter
QALY threshold, with Rawlins keenly
aware that such a move would only
intensify the controversy swirling
round the organisation. What he does
want is some help from politicians,
calling for them to ”come clean”
about the difficult decisions that need
to be taken over health-care funding.
Richard Hoey
www.thelancet.com Vol 370 August 25, 2007
World Report
Therapy for autistic children causes outcry in France
A treatment for autistic children with psychiatric problems that has never been tested is
routinely used in France. Some psychiatrists say the technique has produced therapeutic results.
But critics say that it is cruel, unproven, and potentially dangerous. Laura Spinney investigates.
www.thelancet.com Vol 370 August 25, 2007
century, and packing was routinely
applied at Chestnut Lodge—an asylum
in Rockville, Maryland—in the 1950s.
A decade later, American psychiatrist
Michael Woodbury brought it to
France, where it was embraced by the
influential psychoanalytic movement,
whose founder was Sigmund Freud.
Psychoanalysts found that Woodbury’s
philosophy dovetailed with ideas they
had about children’s development.
One psychoanalytic theory holds,
for example, that packing can help
children to dismantle the defensive
behaviours they developed at an early
age, to protect themselves from a
dysfunctional relationship with their
mother.
In 1996, the French National
Consultative Ethics Committee for
Health and Life Sciences published
a report stating that there was no
evidence to substantiate psychoanalytic models of autism, nor that
therapies based on this model were
effective. The authors were also
concerned that, in France, childhood
autism was classified as an infantile
psychosis, whereas the term psychosis
had been dropped from international
and US descriptions. In both WHO’s
International Classification of Diseases
and the US Diagnostic and Statistical
Manual of Mental Disorders, autism
is now described as a pervasive
developmental disorder.
David Cohen heads the child
and psychiatry service at the PitiéSalpêtrière Hospital in Paris, where
packing is used alongside specialised
education and medication for some
severely autistic and schizophrenic
children. He regards it as a valuable
adjunct therapy, and he says there
is no need to evoke psychoanalytic
concepts to account for the “dramatic
improvements” he sees in children
who have received it. Rather, he
says, packing should be viewed as
a form of “bodily mediation”, like
massage, which relaxes the child
while he receives psychotherapy.
He admits, however, that using it
simultaneously with other treatments
makes it impossible to judge whether
any observed improvements can be
credited to the packing.
According to Delion, current
interpretations of the mechanism
of action of packing are academic.
The printed journal
includes an image merely
for illustration
The Wellcome Library
In France, autistic children who
have psychiatric problems routinely
undergo a treatment that has never
been tested in a clinical trial and
that many parents regard as cruel.
Psychiatrists who use the technique
claim that it produces positive
results, but critics argue that it shows
just how far France has fallen out of
step with the international medical
community in its understanding of
the condition.
The therapy, called packing, involves
wrapping a child tightly in wet
sheets that have been placed in the
refrigerator for up to an hour. When
children are encased in this damp
cocoon—with only their head left
free—psychiatrically trained staff talk
to them about their feelings. Typically,
the treatment is repeated several
times a week, and depending on the
results and the severity of the child’s
condition, it can continue for months
or even years.
The man who pioneered packing for
children, Pierre Delion, is head of the
child and adolescent psychiatry unit
at Lille Regional University Hospital
in northern France. He says that it
reinforces childrens’ consciousness
of their bodily limits, which in some
psychiatric
conditions
becomes
fragmented. He recommends that the
technique be used for three types of
patient: severely autistic children who
self-harm; psychotic children; and,
more rarely, children with anorexia.
Referring to the first category, he has
written, “In our experience of packing,
self-harming behaviour very often
disappears”.
Forms of wrapping or envelopment—
for example, in mud or clay—have been
used therapeutically for centuries. The
idea of using it to calm violent patients
was conceived in Germany in the 19th
Some experts believe that state money for autism should be spent on established therapies
645
World Report
Packing is used as an adjunct therapy at the Pitié-Salpêtrière Hospital in Paris
“For me, it combines the body and
the image of the body—the two are
entirely complementary”, he says.
“However, I do not think that debate
is useful, which is why we would now
like to move it onto a more scientific
plane.”
In June, a small clinical trial got
underway in Lille—the first to try to
hone in on the therapeutic effects
of packing. In 120 children, the
investigators will compare one
group wrapped in dry sheets with
one wrapped in wet, cold ones, on a
battery of clinical, electrophysiological,
and other measures. Later, Delion
hopes to identify a neurophysiological
mechanism underlying the technique,
which he speculates may involve
temperature sensors in the brain.
Meanwhile, the technique continues to be practised in French
clinics, often under the supervision
of psychomotriciens—a type of
occupational therapist. Delion, who
has been training professionals for
25 years, thinks that there must be
several hundred teams using it across
the country. A day-case hospital in
Bordeaux that used packing was
the focus of a short documentary
aired on French TV in April, which
provoked outrage from organisations
representing parents of autistic
children. The biggest of these—Autism
France—officially complained to the
TV channel and to the French health
minister.
646
“If you hit someone over the head
with a hammer, obviously when you
stop they will be very happy”, says
Autism France’s president, Mireille
Lemahieu. “But it is not because they
are made happy that hammering them
over the head is a good thing.” She
points to potential negative effects of
the treatment, such as seizures and
heart attacks, which could result from
the thermal shock of being swaddled
in icy sheets.
She also points out that children
who cannot speak—which accounts
for most of the recipients of packing—
cannot withhold their consent for a
procedure that they do not enjoy. For
Delion, this is not an issue, “If a child
is in a road accident”, he says, “you do
not wait to ask his consent to give him
a blood transfusion”. He believes it a
case of doing what is best for the child
in a situation in which the child cannot
help himself, and he emphasises that
parents’ consent is always sought.
Usually, once the parents have agreed,
a psychiatrist explains to the child
what is going to happen, sometimes
with the help of a doll. If the idea
seems not to upset the child too
much, the packing goes ahead. A child
who declines to enter the pack is never
forced.
Olivier Bousquet worked as a
psychiatric nurse for 9 years until he
gave up to care for his autistic son.
As a baby, he says, his son could not
bear to wear clothes or be caressed.
However, like the autistic author and
scientist Temple Grandin, he enjoyed
the firm application of pressure to
his skin. Grandin invented the “hug
machine” to give her the kind of
pressure she liked, and both Cohen
and Delion cite her as evidence
that packing is pleasant for autistic
patients. The difference, says Bousqet,
is that Grandin could choose.
“It is very difficult to imagine my
son wrapped in sheets, without being
able to make a movement”, he says.
“He would be distressed but he would
not be able to express his distress.”
Children often adopt the point of view
that adults expect of them, he says.
Above all, says Bousquet, state money
(and health professionals’ time) spent
on packing would be better spent
on established methods for teaching
autistic children to communicate.
“It is a kind of aggression against
somebody who cannot respond or
defend themselves”, commented
one shocked French psychiatrist,
who preferred to remain nameless,
on learning that the “old-fashioned”
technique was still practised in France.
She said that parents could be placed
in a difficult position, because demand
for places for autistic children at day
hospitals or therapy centres often
outstrips supply. “Whether or not
it is true, they may be afraid that by
saying no, and opposing the doctor’s
recommendations, they will jeopardise
their child’s place”, she says.
Patricia Howlin, professor of clinical
child psychology at the Institute of
Psychiatry, King’s College London, who
specialises in autism, had never heard
of packing, which does not seem to be
practised in the UK. However, Delion
recently gave a course on the technique
at the Tavistock Clinic in London, which
is part of the UK’s National Health
Service. Maria Rhode, a psychotherapist
at the clinic, points out that there
are currently no effective treatments
for autism, and that caring for such
children presents a major, long-term
challenge to health services.
“It is the case for an awful lot of
interventions for children that the
evidence base leaves a lot to be
desired”, says Rhode. “What you are
coming up against here is the very
difficult interface between clinical
judgment and evidence base.” Lack of
proof of efficacy is not the same as lack
of efficacy, she adds, and although she
has concerns about the risk of adverse
effects with packing, she believes
that the benefits that Delion reports
warrant investigation. Along with
many others, she will be watching the
clinical trial with interest.
Laura Spinney
www.thelancet.com Vol 370 August 25, 2007
Perspectives
Book
Much to be done to improve the mental health of young people
Over this past year there has been a
flurry, if not a storm, of media activity
about a supposed crisis in the mental
health of young people. A UNICEF
report presented findings (with a
mixture of the objective and the more
subjective) arguing that the situation
in the UK with respect to the wellbeing
of young people was among the worst
in the industrialised world. The UK’s
Children’s Society launched a 2-year
enquiry into “Good Childhood”, chaired
by Professor Judy Dunn, to examine
whether the pressures of growing up
in modern Britain are as damaging as
some have claimed. Most recently Ian
Duncan Smith and his Conservative
Party colleagues produced a report
arguing that marital breakdown was
the prime cause of children’s problems,
and that the solution lay in providing
substantial rewards for those who
choose to marry.
As if on cue, this four-volume set
of review essays from the USA was
published with the ambitious claim
that not only could the problem be
identified, but that the solutions were
known and should be made available.
The starting point was that the
rising rate of mental and emotional
problems among US children and
teenagers represented a crisis. More
than one in ten had a mental illness
severe enough to cause impairment,
but only one in five of these ever got
treatment. Clearly, the concerns in the
USA closely parallel those in the UK. It
is not just on the invasion of Iraq, the
fostering of terrorism, and the use
of torture that the UK follows the US
lead. But does the US evidence enable
us to assess the issues more clearly or
to determine the solutions that are
needed? Unfortunately not.
The Crisis in Mental Health: Critical
Issues and Effective Programs texts do
provide a mass of useful findings that
certainly carry important messages for
the other countries. Curiously, however,
www.thelancet.com Vol 370 August 25, 2007
the reviews do not provide a considered
critical assessment of the extent to
which things have truly got worse.
The Rutter and Smith review for The
Academia Europaea in 1995 concluded
that there has been a true rise over
the past 50 years in the rates of
substance use/misuse, crime, suicide,
and probably depression among young
people. The more extensive evidence
“Why, then, does the massive
increase in knowledge about
risk and protective factors not
tell us what is going wrong,
and, therefore, what we should
do to improve the mental
health of young people?”
available during the past dozen years
has mainly confirmed the validity of
that conclusion. But, equally, it has
indicated that, despite impressions to
the contrary, there does not seem to
have been a rise in attention deficit
disorders with hyperactivity, eating
disorders, and child abuse. The US
evidence also suggests that there has
been no increase in runaway youth.
Also, there are many problems that
remain a major concern, despite the
lack of evidence of a risk. This would
apply, for example, to the manifold
difficulties experienced by so-called
“looked-after”
children—meaning
those looked after by the state because
of a breakdown in parenting—and the
problems associated with bullying. It
would also apply to the very high rate
of teenage pregnancies and sexually
transmitted diseases in the USA and
the UK. Since the 1995 review, there
has been the new concern over the
massive rise in the rate of diagnoses of
autism spectrum disorders. Is this just
a reflection of better ascertainment
and a broadening of the diagnostic
concept, or has there been the
operation of some new (or increased)
environmental risk factor? The
evidence is inconclusive.
The US evidence is useful in
underlining several points that are
better appreciated now than they were
12 years ago. Most crucially, it is clear
that there is immense heterogeneity
in children’s responses to stress
and adversity. Even with the worst
experiences, some children survive
relatively unscathed. The evidence is
growing that part of the explanation
is to be found in genetic influences
on susceptibility to environmental
hazards and part in the social context
of stress and adversity. These findings
have highlighted the reality and the
importance of resilience.
Why, then, does the massive increase
in knowledge about risk and protective
factors not tell us what is going wrong,
and, therefore, what we should do to
improve the mental health of young
people? The two main reasons are that
there has been too little attempt to
identify the timing and pattern of the
time trends in mental health, and even
fewer attempts to undertake research
to test competing hypotheses on the
possible explanations for the rise in
mental health problems. For example,
the volumes point out that the rate
of juvenile suicide tripled in the USA
between 1955 and 1994 but then
plateaued or declined. The UK pattern
is somewhat similar. Why, in the UK, did
the rate of suicide in young people go up
when the rate in older adults was going
down? Why did the rate of juvenile
crime rise between the 1950s and early
1990s but then fall? Why, however,
was the overall reduction in acquisitive
crime accompanied by a rise in
violence? The rate of childhood poverty
in the UK (as in the USA) remains
deplorably high and the growing gap
between the rich and poor is a matter
of concern. But this does not constitute
a plausible explanation for the rise
over time in mental health problems,
The Crisis in Mental Health:
Critical Issues and Effective
Programs
Volume 1: Childhood Disorders
Hiram Fitzgerald, Barry M Lester,
Barry Zuckerman, eds.
Volume 2: Disorders in
Adolescence
Francisco A Villaruel,
Tom Luster, eds.
Volume 3: Issues for Families,
Schools and Communities
Kristine Freeark,
William S Davidson II, eds.
Volume 4: Early Intervention
Programs and Policies
Norma F Watt, Catherine Ayoub,
Robert H Bradley, Jini E Puma,
Whitney A LeBoeuf, eds.
Four-volume set. Praeger
Publishers, 2006. Pp 1488.
US$300·00. £170·00.
ISBN: 0-275-98480-X.
647
Perspectives
because the main rise occurred during
the time when income levels were
rising and social inequalities narrowing
in the UK. That does not mean that we
should stop being concerned about
the inequalities in modern day Britain,
but it does mean that the causes of the
rise in mental health problems must be
found elsewhere.
In many ways, the best chapter in
these four volumes is the concluding
chapter by Ed Zigler, the father figure
of the Head Start movement in the
USA. He points out that: even the
best interventions cannot eliminate
individual differences; that many of the
evaluations of preventive programmes
have been undertaken by researchers
uncomfortably close to the projects;
that there are major differences
between efficacy (ie, the benefits
when the intervention is delivered
by experts in optimal circumstances)
and effectiveness (ie, the benefits
when the methods are applied under
the conditions of community-wide
services); and that there is a need for
interventions that extend beyond
the preschool years. Most crucially, he
notes that the intervention field suffers
from the “tyranny of the mean”—that
is, it ignores the marked heterogeneity
within groups.
These four volumes are persuasive
on the reality of at least some of the
concerns, they summarise many
useful findings, and they provide some
useful intervention leads. Most of all,
however, they underline the extent to
which we know neither the meaning
of the secular trends nor the possible
solutions. Much remains to be done.
Michael Rutter
[email protected]
In brief
Book A cultural history of
impotence
648
The printed
journal
includes an
image merely
for illustration
The Wellcome Library
Impotence: A Cultural History
Angus McLaren. University of
Chicago Press, 2007. Pp 352.
US$30·00. ISBN 0-226-50076-4.
The farmer sitting next to me on
the aeroplane as I made notes for
this review peered over the jacket.
“Impotence”, he said, “we deal with
impotence in our bulls every day;
the quality of their semen, which
we sell, their vigour and longevity,
the new electric ejaculators, even
their depression when they become
depleted”. But it is not the sexual
performance of bulls that fascinates
Angus McLaren, rather man’s
cultural and scientific responses to
impotence throughout history.
McLaren, a veteran historian of
sex, traces impotence’s biocultural
history from the Greeks to the
present. The language of impotence,
its main personages, theorists,
doctors, and patients, are his métier,
and he shows how it was configured
along all possible physical, mental,
and emotional fault lines. Every
cause was given, from masturbation
to madness, material defect to
mystical intervention. But McLaren’s
real subject is impotence’s public
face: how history construed it
publicly, especially as the humiliated
patient swiftly shut the door
before his urologist could open his
mouth. McLaren also emphasises
remedies for impotence: the
countless nostrums used from time
immemorial, culminating in the
current craze for sildenafil (Viagra).
Viagra’s publicists tout their magic
bullet as proclaiming the end of
impotence as we have known it; they
promote their product as dealing
“compassionately” with partial, semi,
and full erectile dysfunction. But their
justification will not fly smoothly for a
medically informed public that often
knows better during late capitalism.
The biomedical model of impotence
eliminates the psychological causes,
especially issues of attachment and
love.
McLaren chronicles how sexual
performance has also depended on
notions of gender—male, female,
bisexual, and homosexual—and
crises of masculinity. His treatment
is thorough, especially the continuity
of man’s obsession with penetrative
performance, yet conceived largely
from the male’s perspective. Female
fantasies that her male will be forever
erect and hard as steel until she climaxes, receive less attention. If men
in history have had their lives ruined
by charges of impotence, women
have only recently come into their
own with female sexual dysfunction
becoming better understood by the
medical profession (see Lancet 2007;
369: 409–24). In a world for so long
defined by men, it is not surprising
that female impotence—infertility
and sexual dysfunction—has played
second fiddle. But ultimately it is
wrong to think impotence is only
about the male penis. The demonstration of this less than obvious truth
is one of McLaren’s triumphs.
George Rousseau
[email protected]
www.thelancet.com Vol 370 August 25, 2007
Perspectives
Ten Most Wanted
May, 2007
Lunch with The Lancet
Dennis Carlson
1 Rosiglitazone and diabetes (Articles, Sept 23, 2006)
The DREAM (Diabetes REduction Assessment with
ramipril and rosiglitazone Medication) Trial Investigators.
Effect of rosiglitazone on the frequency of diabetes in
patients with impaired glucose tolerance or impaired
fasting glucose. DOI:10.1016/S0140-6736(06)69420-8.
Lancet 2006; 368: 1096–105.
2 Systemic lupus erythematosus (Seminar, Feb 17)
D’Cruz DP, Khamashta MA, Hughes GR. Systemic lupus
erythematosus. DOI:10.1016/S0140-6736(07)60279-7.
Lancet 2007; 369: 587–96.
3 PROactive Study (Articles, Oct 8, 2005)
Dormandy JA, et al on behalf of the PROactive
investigators. Secondary prevention of macrovascular
events in patients with type 2 diabetes in the PROactive
Study (PROspective pioglitAzone Clinical Trial In
macroVascular Events). DOI:10.1016/S01406736(05)67528-9. Lancet 2005; 366: 1279-89.
4 Renin-angiotensin system (Review, April 7)
Schmieder RE, et al. Renin-angiotensin system and
cardiovascular risk. DOI:10.1016/S0140-6736(07)60242-6.
Lancet 2007; 369: 1208–19.
5 Polycystic kidney disease (Seminar, April 14)
Torres VE, Harris PC, Pirson Y. Autosomal dominant
polycystic kidney disease. DOI:10.1016/S01406736(07)60601-1. Lancet 2007; 369: 287–301.
6 Alzheimer’s disease (Seminar, July 29, 2006)
Blennow K, et al. Alzheimer’s disease. DOI:10.1016/
S0140-6736(06)6911. Lancet 2006; 368: 387–403.
7 The incretin system (New Drug Class, Nov 11, 2006)
Drucker DJ, Nauck MA. The incretin system: glucagon-like
peptide-1 receptor agonists and dipeptidyl peptidase-4
inhibitors in type 2 diabetes. DOI:10.1016/S01406736(06)6970. Lancet 2006; 368: 1696–705.
8 Health inequalities (Public Health, March 19, 2005)
Marmot M. Social determinants of health inequalities.
DOI:10.1016/S0140-6736(05)71146-6. Lancet 2005;
365: 1099–104.
9 Drug treatments for obesity (New Drug Class, Jan 6)
Padwal RS, Majumdar SR. Drug treatments for obesity:
orlistat, sibutramine, and rimonabant. DOI:10.1016/
S0140-6736(07)6003. Lancet 2007; 369: 71–77.
10 Postextubation laryngeal oedema (Articles, March 31)
François B, et al for the Association des Réanimateurs du
Centre-Ouest (ARCO). 12-h pretreatment with
methylprednisolone versus placebo for prevention of
postextubation laryngeal oedema. DOI:10.1016/S01406736(07)60526-1. Lancet 2007; 369: 1083–89.
In 1965, His Imperial Majesty Haile Selassie I, Conquering
Lion of Judah, King of Kings and Elect of God, invited Queen
Elizabeth II to Ethiopia. No effort was spared: trees were
planted, potholes repaired, and an army plane dropped
flower petals on the procession. A state dinner was held near
Kossoye, 20 miles north of Gondar. Dennis Carlson, alas, was
not invited. Carlson, who graduated from the University of
Washington’s medical school, in 1955, was then Dean of the
Public Health College in Gondar. Over a glass of wine in Addis
Ababa, where we were both travelling with the Carter Center,
he described his career, which, save for the disappointment
of the royal banquet, has been full and rewarding.
Carlson is not only an MD; he has long been “dying by
degrees”. He studied behavioural sciences at Berkeley, and
was a fellow in the history of medicine at Johns Hopkins
University. He originally went to Ethiopia with a group from
his religious tradition, the Swedish-American Baptists. He did
primary care and general surgery and trained health assistants
at a mission hospital. “My interests soon shifted away from
clinical medicine to community health”, Carlson said. “It was
very clear to me that training and education were the best
long-term investments.” He pursued his interests in history
and anthropology, learning Amharic because “language has
to go with culture, and culture has to go with history”. After
his Gondar deanship and teaching at Johns Hopkins, he joined
Save the Children, with his wife, Beulah, to do nutritional
relief and develop health services during the famine in the
1980s. Later he worked at Addis Ababa University in publichealth training. This background made him a natural to
head up the new Ethiopia Public Health Training Initiative,
a multi-institutional collaboration supported by the Carter
Center, which promotes teaching staff development in
seven university health science faculties and creates local
educational materials for training health students.
Carlson has lived in Ethiopia for about 20 years, although
his home base is Washington State. His primary concern
is “personal and community development—integrating
material and spiritual aspects”. He and Beulah are putting
that passion to work in the Ethiopian village of Kossoye,
where illiteracy was once the norm. Now there is an
elementary school of 1100 students, and a modernisation
project that includes household vegetable gardening.
He and his historian son, Andrew, have just completed a
history of Kossoye. I am lost for a moment at the richness of
Carlson’s life. Say, I ask, was Selassie really descended from
King Solomon? “Oh, details!” he says, grinning broadly and
dismissing my question with a genial wave of his hand.
The ten most wanted Lancet articles downloaded from
ScienceDirect (see Lancet 2003; 361: 1265. DOI:10.1016/
S0140-6736(03)1298) in May, 2007.
www.thelancet.com Vol 370 August 25, 2007
Faith McLellan
[email protected]
649
Obituary
John R Hogness
First president of the US Institute of Medicine,
medical leader at University of Washington,
Seattle, WA, USA. Born on June 27, 1922, in
Oakland, CA, USA, he died of heart and kidney
failure on July 2, 2007, in Seattle, aged 85 years.
John Hogness had been dean of the University of
Washington’s medical school in Seattle for 4 years when
he disappeared in 1968. Literally. None of the faculty
knew where he was until later, when they found out that
he had arranged to cover the work of a family physician in
Omak, a small town in eastern Washington. “A primary care
physician needed a vacation”, recalled Paul Ramsey, who has
been dean of the school since 1997, “John took his practice
for 2 weeks. He took his call, even performed an emergency
appendectomy.” The sojourn gave Hogness “first-hand
experience in what the issues are in providing care in the
small communities in our region”, Ramsey said. “It helped
guide his efforts as dean, and helped drive his educational
agenda to partner with communities and to begin training
students outside of the teaching hospitals in the cities.”
Those experiences also helped in his and others’ efforts to
create a regional medical education system for Washington,
Wyoming, Alaska, Montana, and Idaho, called WWAMI for
the states’ initials. Of the five states, only Washington has a
medical school, where students from all five states study.
Hogness is perhaps best known for the role he took on in
1971 as the first president of the Institute of Medicine (IOM).
“I think he really was a very good stabilising influence”,
said Queta Bond, former executive officer of the IOM. “He
650
was a visionary. Those early years were really critical to
the institution. He helped to put it on the right course”,
establishing innovative policies and procedures. Hogness
“was this huge man”, Bond said. “But he was very gentle,
very articulate, and a good spokesperson.”
Rheba de Tornyay met Hogness when she was elected to
the IOM in the early 1970s. After Hogness returned to the
University of Washington in 1974 as president, he recruited
de Tornyay, then at the University of California, Los Angeles,
as dean of the school of nursing. “I had no intention of
changing jobs, and I came to UW specifically because of
him”, said de Tornyay. She remembers how “He worked
toward consensus repeatedly. He could run a committee
meeting like no one. I’ve worked for a lot of people and I’ve
never worked for anyone who was less of a male chauvinist
than John Hogness. He was colour blind and gender blind,
and he really paid attention to what was being said rather
than who was saying it.”
“He was very non-confrontational”, said Tom Grayston,
who was a medical student with him in the 1940s and
then a colleague at the University of Washington in later
decades. “He was a very friendly person and very easy to
like and I think that he used those traits to be a successful
administrator. He was not a driven administrator by any
means, he tended to be quiet and gave people a chance to
talk and express their opinions, and got people to come
together.”
Hogness earned his bachelor’s and medical degrees from
the University of Chicago, Chicago, IL, and completed an
internship at Columbia Presbyterian Medical Center in
New York. After a year in the Army, he finished his training
in medicine and endocrinology in Seattle at King County
Hospital. He was medical director of the University of
Washington’s hospital before he became dean in 1964.
Hogness “was something of an institution builder”,
Bond said, moving from the presidency of the University of
Washington in 1979 to serve as president of the Association
of Academic Health Centers. “He had a vision that it would be
broader in purview than the Association of American Medical
Colleges”, Bond said, “that it would incorporate health care
profession issues, not just medicine”. Hogness returned to
the University of Washington again after his time at the
Association of Academic Health Centers, as a faculty member
at the School of Public Health. For about 10 years, he focused
his work on health-care policy. He was passionate about a
national health insurance plan, de Tornyay said. Hogness also
had a tremendous sense of humour, said Grayston. “He was
Norwegian, and he could tell a Norwegian joke better than
Garrison Keillor”, he said. Hogness is survived by his second
wife, Margaret; five children; four stepchildren; and a brother.
His first wife, Katherine, predeceased him.
Ivan Oransky
[email protected]
www.thelancet.com Vol 370 August 25, 2007
Correspondence
Folic acid for stroke
prevention
The meta-analysis by Xiaobin
Wang and colleagues (June 2,
p 1876)1 concludes that folic acid
supplementation reduces by 18%
the relative risk of stroke in primary
prevention. There are two major
limitations to this meta-analysis.
First, it is based on eight studies,
three of which concerned patients
with end-stage renal disease (ESRD).2–4
In a further study,5 homocysteine
concentrations were not reported.
ESRD is itself a well known determinant
of hyperhomocysteinaemia, as is
clearly shown in Wang and colleagues’
table 1 where such patients have
on average at least two-fold higher
homocysteine concentrations than
patients with coronary heart disease
(CHD). Although the reduction in
homocysteine concentrations after
folic acid supplementation is similar
between patients with CHD and ESRD,
absolute levels are still largely higher
in those with ESRD.
The second limitation is intrinsic
to ESRD, for which there is a higher
incidence of cardiovascular morbidity
and a proportional increase during
follow-up compared with almost
all other chronic diseases. Vascular
access infection and thrombosis in
ESRD also increase the risk of stroke.
These variables were not considered
by Wang and colleagues as criteria for
trial exclusion.
Data from ESRD trials should be
assessed separately because too
many confounding variables can bias
the interpretation of the benefits
associated
with
homocysteinelowering treatment. In our opinion,
the relative risk of stroke reduction
in CHD is probably higher than 18%
because it is not significant in ESRD.
We declare that we have no conflict of interest.
*Vincenzo Sepe, Gabriella Adamo,
Maria Grazia Giuliano,
Carmelo Libetta, Antonio Dal Canton
[email protected]
www.thelancet.com Vol 370 August 25, 2007
Unit of Nephrology, Dialysis and Transplantation,
Fondazione IRCCS Policlinico San Matteo, P le Golgi
2, 27100 Pavia, Italy
1
2
3
4
5
Wang X, Qin X, Demirtas H, et al. Efficacy of folic
acid supplementation in stroke prevention: a
meta-analysis. Lancet 2007; 369: 1876–82.
Wrone EM, Hornberger JM, Zehnder JL, et al.
Randomized trial of folic acid for prevention of
cardiovascular events in end stage renal
disease. J Am Soc Nephrol 2004; 15: 420–26.
Righetti M, Serbelloni P, Milani S, et al.
Homocysteine-lowering vitamin B treatment
decreases cardiovascular events in hemodialysis
patients. Blood Purif 2006; 24: 379–86.
Zoungas S, McGrath BP, Branley P, et al.
Cardiovascular morbidity and mortality in the
Atherosclerosis and Folic Acid Supplementation
Trial (ASFAST) in chronic renal failure: a
multicenter, randomized, controlled trial.
J Am Coll Cardiol 2006; 47: 1108–16.
Mark SD, Wang W, Fraumeni JF Jr, et al. Lowered
risks of hypertension and cerebrovascular
disease after vitamin/mineral supplementation:
the Linxian Nutrition Intervention. Am J
Epidemiol 1996; 143: 658–64.
Authors’ reply
We agree with Vincenzo Sepe and
colleagues that, in addition to the
factors examined in our meta-analysis
(treatment duration, homocysteine
reduction, grain fortification, and
history of stroke), other pre-existing
conditions that could affect the
findings should also be considered. Of
the eight trials included in our metaanalysis, three were in populations
with end-stage renal disease (ESRD),
three were in coronary heart disease
(CHD), one was in stroke, and one
was in oesophageal dysplasia. We
found that folic acid supplementation
significantly reduced the risk of stroke
by 18% in the overall analysis, and
by 25% in primary prevention in a
stratified analysis.
We have now done an additional
stratified analysis by pre-existing
conditions (table), with exclusion
of the trial with pre-existing stroke
(secondary prevention). Compared
with the six trials with ESRD or CHD,
the trial without ESRD or CHD had a
lower relative risk. Despite the limited
number of trials and sample size,
the data suggest that the beneficial
effect of folic acid supplementation
on stroke prevention could be even
greater if the trials were done in
populations without ESRD or CHD.
We expect that populations with high
concentrations of homocysteine,
low concentrations of folic acid, and
primary hypertension but without
pre-existing conditions such as ESRD
or CHD could benefit most from folic
acid supplementation in the primary
prevention of stroke. Future studies in
populations with those characteristics
are highly recommended.
We declare that we have no conflict of interest.
*Xiaobin Wang, Xianhui Qin,
Hakan Demirtas, Xiping Xu
[email protected]
Mary Ann and J Milburn Smith Child Health
Research Program, Children’s Memorial Hospital
and Children’s Memorial Research Center,
Northwestern University Feinberg School of
Medicine, Chicago, IL 60614, USA (XW); Institute
for Biomedicine, Anhui Medical University, Hefei,
China (XQ); and Division of Epidemiology and
Biostatistics, University of Illinois at Chicago School
of Public Health, Chicago, IL, USA (HD, XX)
Stroke events/total patients Relative risk
(95% CI)
Intervention
group
Trials with ESRD
(Zoungas et al; Wrone et al; Righetti et al)
p
Control
group
31/535
36/378
0·68 (0·37–1·25)
0·21
168/4930
186/4000
0·78 (0·63–0·96)
0·0187
Trials with ESRD or CHD
199/5465
(Zoungas et al; Wrone et al; Righetti et al;
Liem et al; Lonn et al; Bonaa et al)
222/4378
0·76 (0·63–0·93)
0·0062
35/1661
0·63 (0·37–1·07)
0·09
Trials with CHD
(Liem et al; Lonn et al; Bonaa et al)
Trial with oesophageal dysplasia
(Mark et al)
22/1657
ESRD=end-stage renal disease. CHD=coronary heart disease.
Table: Risk of stroke in stratified analysis by pre-existing conditions
Submissions should be
made via our electronic
submission system at
http://ees.elsevier.com/
thelancet/
651
Correspondence
Recurrent anaphylaxis
to synthetic folic acid
Science Photo Library
Food fortification with synthetic folic
acid (pteroylmonoglutamic acid)
remains a source of debate in terms
of benefit versus issues of safety and
consumer choice.1 Several countries
have adopted mandatory folic acid
fortification. A decision to proceed in
Australasia was made in June, 2007.
We report the case of a woman who
had three episodes of type I hypersensitivity, including anaphylaxis, after
synthetic folic acid exposure. Her first
episode occurred within minutes of
taking a 5 mg folic acid tablet. She
developed an itchy throat, nausea,
generalised rash, diarrhoea, and lightheadedness; she was treated with
antihistamines. The second episode
followed consumption of 800 mL
lime-flavoured water fortified with
20 μg/100 mL folic acid. Within minutes of finishing the drink she developed
an itchy throat, generalised pruritus, and
nausea. She was treated with adrenaline
and antihistamines. A further episode
occurred within minutes of drinking
150 mL of a beverage containing feijoa
(a fruit of the Myrtaceae family) and
supplements including 53·5 μg/100 mL
folic acid. She developed generalised
rash, vomiting, and lightheadedness.
Adrenaline was given en route to
hospital, with good response.
Intradermal testing with folic acid
0·05 μg/mL solution containing folic
acid, bicarbonate, and water was positive (9 mm wheal, 35 mm flare). A control patient was negative. Skin-prick
tests to other food and beverage products were negative. A graded, blinded
challenge to the folic acid solution led to
widespread urticaria at a dose of 160 μg.
Before her first episode she had
taken a multivitamin B supplement
and recalled recurrent episodes of
urticaria, and presumably sensitisation
to folic acid occurred at this time.
She seems to tolerate dietary folates
(pteroylpolyglutamates).
Hypersensitivity to synthetic folic
acid has been rarely described.2-4 One
652
report documents sensitivity to synthetic folic acid in medication, as a
food supplement, and possibly to
dietary folate.2 In a further case of
anaphylaxis after folic acid exposure
in multivitamin preparations, development of IgE antibody to folic acid was
shown by in-vivo and in-vitro testing.3
In IgE-mediated reactions, folic acid,
with a molecular weight of only
441 D, probably acts as a hapten by
conjugation with self-proteins.3
Folic acid fortification must be
accompanied by clear food labelling
to enable those who develop allergy to
avoid life-threatening reactions. Folic
acid allergy should be considered in
the differential diagnosis of idiopathic
anaphylaxis and suspected cereal
allergy where skin-prick or RAST testing
to standard grains is inconclusive.
We declare that we have no conflict of interest.
*Julie Smith, Marianne Empson,
Clare Wall
[email protected]
Immunology Department, Auckland Hospital, Park
Road, Private Bag 92-024, Auckland, New Zealand
(JS, ME); and Faculty of Medical and Health Sciences,
Discipline of Nutrition, University of Auckland,
Auckland, New Zealand (CW)
1
2
3
4
Eichholzer M, Tönz O, Zimmermann R. Folic
acid: a public health challenge. Lancet 2006;
367: 1352–61.
Sanders GM, Fritz SB. Allergy to natural and
supplemental folic acid as a cause of chronic,
intermittent urticaria and angioedema.
Ann Allergy Asthma Immunol 2004;
93: S51–52.
Dykewicz MS, Orfan NA, Sun W. In vitro
demonstration of IgE antibody to folatealbumin in anaphylaxis from folic acid.
J Allergy Clin Immunol 2000; 106: 386–89.
Sesin GP, Kirschenbaum. Folic acid
hypersensitivity and fever: a case report.
Am J Hosp Pharm 1979; 36: 1565–67.
DFID’s health strategy
In your June 16 Editorial (p 1973),1
you praise the new health strategy of
the UK’s Department for International
Development (DFID),2 and its “rather
unusual, but much needed, donor
practice of budget support”. However,
this strategy fails to address one
crucial limitation: the International
Monetary Fund (IMF) policies which
dictate that a significant proportion of
budget support remains unabsorbed
or unspent.
A report of the Independent
Evaluation Office of the IMF reveals that,
since 1999, around 37% of additional
foreign assistance to countries in subSaharan Africa under IMF programmes
was explicitly designed to increase
international reserves.3 Of the 63%
designed to be absorbed, only 28% was
supposed to be spent; the other 72%
was programmed to increase public
savings. In all, the IMF programmed
28% of 63% of additional foreign
assistance—a ridiculous 17·64%—to
be both absorbed and spent. This “IMF
tax”, as one observer has described
it,4 reduces the real amount budget
support that can actually be spent by
82·36%.
At a seminar in London on April 2,
2007, the lead author of the independent evaluation recommended that
the IMF should be transparent about
this policy of limiting countries’ ability
to spend foreign assistance. However,
the Board and management of the IMF
rejected the recommendation, on the
grounds that secrecy gave them more
flexibility.5
DFID’s new health strategy paper
cites the integration of grants from the
Global Fund to fight AIDS, Tuberculosis
and Malaria to Mozambique into the
health Sector-Wide Approach (SWAp)
as an example of the “potential of
country-led aid instruments to achieve
aid harmonization and alignment”.2
However, it could also mean that the
Global Fund grants are now subjected
to the IMF tax, severely limiting the
amount that can actually go towards
improving health care. Since the IMF
prefers to keep this “tax” secret, we
might never find out.
DFID, the IMF, and the Global Fund
should state clearly how they will
guarantee that budget support to the
health sector will be entirely absorbed
and spent, rather than “taxed” by
the IMF. Without a clear position on
this issue, the amount of foreign
assistance provided in the form of
www.thelancet.com Vol 370 August 25, 2007
Correspondence
budget support must be examined
with extreme caution.
We declare that we have no conflict of interest.
*Gorik Ooms, Nathan Ford
[email protected]
Médecins Sans Frontières Belgium, 94 Rue Dupré,
1090 Brussels, Belgium (GO); and Médecins Sans
Frontières South Africa, Johannesburg, South Africa
(NF)
1
2
3
4
5
The Lancet. DFID’s health strategy. Lancet
2007; 369: 1973.
DFID. Working together for better health:
evidence for action. London: DFID, 2007.
http://www.dfid.gov.uk/pubs/files/healthstrategy07-evidence.pdf (accessed June 29,
2007).
Independent Evaluation Office of the IMF. The
IMF and aid to sub-Saharan Africa.
Washington, DC: International Monetary
Fund, 2007. http://www.imf.org/external/np/
ieo/2007/ssa/eng/pdf/report.pdf (accessed
June 29,2007).
Hanlon J. IMF makes major concessions on
wages & aid. http://www.open.ac.uk/
technology/mozambique/pics/d55391.doc
(accessed June 29, 2007).
Hanlon J. IMF does cap aid, says independent
evaluation. http://www.open.ac.uk/
technology/mozambique/pics/d75975.doc
(accessed June 29,2007).
We join you in congratulating the
UK government‘s Department for
International Development (DFID) for
its new health strategy1 and recommend that readers peruse DFID’s
second progress report, Reducing
maternal deaths: evidence and action.2
The report’s many examples show how
to approach health-system strengthening by promoting reproductive health.
For example, DFID has committed
£252 million to reducing geographic
and social disparities in access to reproductive and child health services in
India by building new health facilities
and expanding core medical training,
not just by funding vertical family
planning and child health services. This
comprehensive approach has increased
the proportion of institutional deliveries to 19% in one Indian state over
a 7-year period, and laid a strong foundation for progress towards reducing
136 000 maternal deaths annually.2
The benefits for women, families, and
communities will be immeasurable.
We, like you, urge the UK government to use such evidence to hold
www.thelancet.com Vol 370 August 25, 2007
other large development agencies, in
particular the World Bank and WHO,
accountable for their commitments
to reproductive health.
I declare that I have no conflict of interest.
Adrienne Germain
[email protected]
President, International Women’s Health Coalition,
New York, NY 10001, USA
1
2
The Lancet. DFID’s health strategy. Lancet
2007; 369: 1973.
DFID’s maternal health strategy. Reducing
maternal deaths: evidence and action. Second
progress report. London: DFID, 2007. http://
www.dfid.gov.uk/pubs/files/maternal-healthprogress-report.pdf (accessed Aug 2, 2007).
Since 135 of the 137 patients were
observed for 13 months, it is easy
to calculate a CI by the exact binomial
method, which will be roughly correct,
just using the data in the published
paper. If we observe zero events in
135 patients, the exact 95% CI is
0 to 0·027. The data are thus consistent with a probability of a bleed of as
large as 0·027 (2·7%).
I declare that I have no conflict of interest.
Martin Bland
[email protected]
University of York, York YO10 5DD, UK
1
Misleading confidence
intervals
In their randomised trial of a
combination of a cyclo-oxygenase-2
inhibitor and a proton-pump inhibitor
for prevention of recurrent ulcer
bleeding in patients at very high
risk, Francis Chan and colleagues
(May 12, p 1621)1 use the Kaplan-Meier
method to estimate the probability
of recurrent bleeding at 13 months. In
the intervention group, no events were
observed. Chan and colleagues give
the estimated probability of a bleed
as 0 (95% CI 0 to 0). This CI implies
that there is no error in the estimate
and that, in the entire population of
patients that this sample represents, no
bleed can ever occur within 13 months.
We cannot draw this conclusion and
the CI must be wrong. What Chan and
colleagues seem to have done is to use a
large sample method for calculation of
the CI for a sample which is far too small
in that no events have been observed.
Chan FKL, Wong VWS, Suen BY, et al.
Combination of a cyclo-oxygenase-2 inhibitor
and a proton-pump inhibitor for prevention of
recurrent ulcer bleeding in patients at very
high risk: a double-blind, randomised trial.
Lancet 2007; 369: 1621–26.
Authors’ reply
We thank Martin Bland for providing a
more accurate method of estimating
the 95% CI when no events have been
observed. We have now recalculated
the probability of recurrent bleeding
using the exact binomial method, and
the corrected 95% CIs in the combinedtreatment group are shown in the
table. The between-group differences
remain significant and the conclusions
of the paper are unchanged.
FKLC has received an independent research grant
and a consulting fee from Pfizer and paid lecture
fees by TAP Pharmaceuticals. BYS declares no
conflict of interest.
*Francis Ka Leung Chan,
Bing Yee Suen, for the research team
[email protected]
Department of Medicine and Therapeutics,
Chinese University of Hong Kong, Prince of Wales
Hospital, 30–32 Ngan Shing Street, Shatin,
Hong Kong SAR, China
Probability of recurrent bleeding (95% CI) p
Combined–treatment
group
All patients
Control group
0% (0·0–2·6)
8·9% (4·1–13·7)
Patients who did not take concomitant aspirin 0% (0·0–3·1)
7·1% (2·4–11·8)
0·0004
0·004
Patients who took concomitant aspirin
0% (0·1–14·8)
19·0% (2·2–35·8)
0·03
Per–protocol analysis
0% (0·0–3·4)
6·0% (1·4–10·6)
0·01
Table: Exact binomial estimates of the likelihood of recurrent ulcer bleeding at 13 months
653
Correspondence
Minimisation of
immunosuppression in
transplantation
Science Photo Library
We were surprised at important
omissions in the Clinical Update on
immunosuppression
minimisation
by Mohamed Sayegh and Giuseppe
Remuzzi (May 19, p 1676).1
In the early days of immunosuppression with azathioprine and
steroids, increasing the dose of
steroids resulted in stunting of growth
and serious cushingoid deformities
in children and a high incidence of
femoral neck fractures in adults.
In 1977, McGeown and colleagues2
started a successful protocol of steroid
reduction, and steroid avoidance was
attempted in living donor recipients
by a Paris group in 1973. A randomised
trial of low-dose versus high-dose
steroids in Oxford showed similar
patient and graft survival rates in renal
transplant recipients, with those on a
low steroid dose avoiding the severe
side-effects.
Sayegh and Remuzzi suggest that
the benefit of ciclosporin compared
with azathioprine regimens lasts for
only 1 year, but the first multicentre
randomised controlled trial3 showed
that ciclosporin had a 10-year
advantage over azathioprine.
When Sayegh and Remuzzi refer
to more recent trials, they omit the
substantial reduction of maintenance
immunosuppression reported after
alemtuzumab induction,4 and an even
greater restriction of maintenance
immunosuppression reported by the
Pittsburg group.5
The level of maintenance immunosuppression in relation to closeness
of tissue type is very important
and does not feature in the Clinical
Update, which on the whole gives a
false impression that advances in dose
reduction in immunosuppression have
not been impressive.
Transplantation has been an
enormous therapeutic advance,
and the graft survival half-life of
654
all organs continues to lengthen.
A minimum dose of maintenance
immunosuppression is a goal already
achieved, to the benefit of many
patients, without incurring the penalty
of increased rejection.
not been promising.3 Even in Calne’s
own series, a subsequent publication
showed late acute rejection and
chronic rejection.4
We declare that we have no conflict of interest.
We declare that we have no conflict of interest.
Mohamed H Sayegh,
*Giuseppe Remuzzi
*R Y Calne, K O Lee
[email protected]
[email protected]
Department of Surgery, Douglas House Annex,
18 Trumpington Road, Cambridge CB2 2AH, UK
(RYC); and Department of Medicine, National
University of Singapore, Singapore (KOL).
Transplantation Research Center, Brigham and
Women’s Hospital and Children’s Hospital, Harvard
Medical School, Boston, MA, USA (MHS); and Mario
Negri Institute for Pharmacological Research, Negri
Bergamo Laboratories, 24125 Bergamo, Italy (GR)
1
1
2
3
4
5
Sayegh MH, Remuzzi G. Clinical update:
immunosuppression minimisation. Lancet
2007; 369: 1676–78.
McGeown MG, Kennedy JA, Loughbridge WGG,
et al. One hundred kidney transplants in the
Belfast City Hospital. Lancet 1977; 2: 648–51.
Beveridge T, Calne RY, for the European
multicentre trial group. Cyclosporine
(Sandimmun) in cadaveric renal
transplantation. Transplantation 1995; 59:
1568–70.
Calne RY, Friend PJ, Moffatt S, et al. Prope
tolerance, perioperative campath IH, and lowdose cyclosporin monotherapy in renal allograft
recipients. Lancet 1998; 351: 1701–02.
Shapiro R, Basu A, Tan HP, et al. Campath
preconditioning and tacrolimus monotherapy
with subsequent weaning in renal transplant
recipients. XX International Congress of the
Transplantation Society; Sept 5–10, 2004;
Vienna, Austria (abstr 0144).
Authors’ reply
We certainly agree with R Calne
and K Lee that we omitted some
references. This was mandated by the
journal because of space limitations.
We were specifically asked to focus on
recent publications.
There have been several trials of
steroid withdrawal since 1977, as well
as in the newer immunosuppression
era, with mixed results. We believe
that our conclusions are correct.
We note that ciclosporin benefit is
restricted to the short term because
the rate of graft loss as measured by
half-life has not changed. It is true
that overall graft survival is better with
ciclosporin, but that is due to a higher
success rate early.
Again, the studies with alemtuzumab
are mixed. Some have been more
successful than others. A high rate of
vascular rejection has been noted,1,2
and our personal experience has also
2
3
4
Knechtle SJ, Pirsch JD, H Fechner J, et al.
Campath-1H induction plus rapamycin
monotherapy for renal transplantation: results
of a pilot study. Am J Transplant 2003; 3:
722–30.
Ciancio G, Burke GW, Gaynor JJ, et al. The use of
Campath-1H as induction therapy in renal
transplantation: preliminary results.
Transplantation 2004; 78: 426–33.
Hill P, Gagliardini E, Ruggenenti P, Remuzzi G.
Severe early acute humoral rejection resulting
in allograft loss in a renal transplant recipient
with Campath-1H induction therapy.
Nephrol Dial Transplant 2005; 20: 1741–44.
Watson CJ, Bradley JA, Friend PJ, et al.
Alemtuzumab (CAMPATH 1H) induction
therapy in cadaveric kidney transplantation—
efficacy and safety at five years. Am J Transplant
2005; 5: 1347–53.
Human resources in
developing countries
In his Comment on human resources
in developing countries (April 14,
p 1238),1 Claudio Lanata states that
0–18% of nurses and midwives trained
in sub-Saharan Africa are working in
developed countries—mainly Canada,
Denmark, Finland, Ireland, Portugal,
the UK, and the USA.
In view of the strict language and
other licensing requirements for health
professionals in Nordic countries, it is
surprising to see Finland and Denmark
in Lanata’s list. Indeed, a review of
his references gives little support to
the claim. The WHO Report 20062
mentions Finland and Denmark as
recipients of nurses from sub-Saharan
Africa, but no numbers are given and
the data source is undefined. The
other reference does not mention
Nordic countries at all, but describes
the pivotal role of the UK in the health
www.thelancet.com Vol 370 August 25, 2007
Correspondence
personnel exodus.3 On the other hand,
a search from a comprehensive health
professional register at the National
Authority for Medicolegal Affairs in
Finland indicated that, at the end
of 2006, none of the 86 162 nurses
registered in Finland and only seven
of 21 899 physicians had been trained
in sub-Saharan Africa (Elisa Alanen,
personal communication). Thus,
Lanata’s information is incorrect, at
least concerning Finland.
Earlier research has shown that health
professionals migrate particularly
to English-speaking countries with
insufficient domestic training of nurses
or physicians.3 Finland has a unique
language and it currently trains about
1% of every age-group to be doctors
and 4% to be nurses. The situation is
similar in other Nordic countries. Thus,
significant African health professional
immigration into this region is highly
unlikely.
I declare that I have no conflict of interest.
Per Ashorn
per.ashorn@uta.fi
University of Tampere Medical School and Tampere
University Hospital, Finn-Medi 3, FIN-33014
Tampere, Finland
1
2
3
Lanata C. Human resources in developing
countries. Lancet 2007; 369: 1238–39.
World Health Organization. The world health
report 2006—working together for health.
2006. http://www.who.int/whr/2006/en/
(accessed April 20, 2007).
Eastwood JB, Conroy RE, Naicker S, West PA,
Tutt RC, Plange-Rhule J. The loss of health
professionals from sub-Saharan Africa:
the pivotal role of the UK. Lancet 2005;
365: 1893–900.
Author’s reply
Per Ashorn highlights an important
point in understanding the complexities of migration of trained
health professionals from developing
to developed countries. His claim
that language has an important
role is correct. Health professionals
who decide to migrate, seeking
better wages, working conditions,
and professional development,
generally select countries with
strong commercial trade, similar
languages, and with cultural and
www.thelancet.com Vol 370 August 25, 2007
former colonial ties.1 Developed
countries
with
shortages
of
skilled health professionals apply
several mechanisms that facilitate
migration—eg, easy access to
information, passing of legislation to
facilitate migration, and aggressive
recruitment.1,2
Although it is true that skilled
health professionals from subSaharan Africa migrate less often to
Finland and Denmark than to the
USA or UK, it does occur. These two
countries provide tax incentives
for highly skilled professionals of
foreign origin paid above specific
salary levels.3 More than 5% of the
highly skilled professionals and 8% of
doctors in Finland are foreign-born,
mainly in eastern Europe; less than 50
doctors are from sub-Saharan Africa.1
WHO monitors these figures from
data obtained from official sources
such as the National Authority for
Medicolegal Affairs in Finland and
the Danish National Board of Health
(both sources of the 2006 World
Health Report).
Developed countries benefit by
hiring highly skilled migrant health
professionals to fill labour deficits,
with important savings on the cost
of training. For developing countries,
it creates a major loss, and what has
been called a subsidy from the poor
to the rich. It has been estimated
that, in Kenya, for example, for
every migrating doctor who costs
about US$66 000 to educate from
primary school to university, the
country loses about $518 000 worth
of returns from investments.4 These
losses might be partly compensated
by migrant remittances, with
amounts that are hard to estimate.
In addition to these economic
losses, many developing countries
are left with severe problems in
providing appropriate care to their
population. And the trends for
migration are increasing at a striking
rate, particularly for nurses.2,5 There
is an urgent need for better tools to
monitor this important public-health
problem, and for policies to mitigate
them in both the developed countries
that hire highly trained foreign health
professionals and the developing
countries that cannot retain them.
I declare that I have no conflict of interest.
Claudio F Lanata
[email protected]
Instituto de Investigacion Nutricional, Av. La
Molina 1885, Lima-12, Peru
1
2
3
4
5
Connell J, Zurn P, Stilwell B, Awases M,
Braichet JM. Sub-Saharan Africa: beyond the
health worker migration crisis? Soc Sci Med
2007; published online Feb 20. DOI: 10.1016/
j.socscimed.2006.12.013.
Buchan J, Sochalski J. The migration of nurses:
trends and policies. Bull World Health Organ
2004; 82: 587–94.
Dumont JC, Lemaitre G. Counting immigrants
and expatriates in OECD countries: a new
perspective. OECD Social, Employment and
Migration Working Papers (in press).
Kirigia JM, Gbary AR, Muthuri LK, Nyoni J,
Seddoh A. The cost of health professionals’
brain drain in Kenya. BMC Health Service Res
2006, 6: 89.
Pond B, McPake B. The health migration
crisis: the role of four Organisation for
Economic Cooperation and Development
countries. Lancet 2006; 367: 1448–55.
MMC, Britain’s aid
deficit, and overseas
health-care shortages
Several
topics
have
become
increasingly prominent in current
medical discourse. There are concerns
over a well documented migration
of doctors from lower and middle
income countries,1 and, for example,
the UK National Health Service’s
(NHS’s) complicit role in this
movement. This problem is balanced
against an individual’s human right
to migrate and the understandable
desire for many doctors to move
to well funded health-care systems
within stable democracies.
Another concern in the UK is the
large projected number of junior
doctors (estimated to be more than
10 0002) who will fail to secure a
training post under the government’s
Modernising Medical Careers scheme.
Finally, the British government has
655
Correspondence
Science Photo Library
persistently failed to reach the UN’s
mandated goal of 0·7% of gross
national income (GNI) for overseas
development aid, managing only
0·47% in 2005.3
Although seemingly unrelated,
all might have a simple remedy. We
argue that the UK Department of
Health should allow doctors a period
of 1–3 years of service in lower and
middle income countries, remunerated
at 33–50% of their NHS salary and
adjusted for the cost of living in
the target country. The benefits
are manifold. First, it would allow a
substantial proportion of the NHS
workforce to continue their practice in
a testing environment, continuing the
precedent set by the memorandum
of understanding between the UK
and South Africa for migration of
health-care professionals.4 Second, it
would indirectly compensate lower
and middle income countries for
the financial loss accrued by training
health professionals now in service
in the NHS. Third, it would increase
the doctor–population ratio in the
most poorly resourced countries
to the requisite WHO minimum of
100 physicians per 100 000 people
(in 2004 the UK had a ratio of 166·5
per 100 000 compared with the
lowest ratio of 1·1 per 100 000 in
Malawi5). Finally, it would increase the
amount Britain spends on overseas
development aid, although even with
a projected spend of £200 million
(10 000 doctors earning £20 000
each) in addition to the £5916 million
2005–06 overseas development aid
budget, this would only increase
spending to 0·486% of GNI.
This proposal, relatively smallscale in both costs and planning
compared with some development
initiatives in lower and middle income
countries, would nonetheless provide
a significant service to the lives of
many people in the developing
world. In so doing, it would also
provide a challenging and stimulating
environment for young British doctors
to improve their clinical acumen in
656
resource-limited environments. The UK
has promised to increase the amount
spent on overseas development aid to
the UN target of 0·7% by 2013. This
initiative would provide one practical
step towards that goal.
We declare that we have no conflict of interest.
*Colin Brown, Anjum Khan
[email protected]
Royal Free Hospital, London NW3 2QG, UK
1
2
3
4
5
Stilwell B, Diallo K, Zurn P, Vuljicic M, Adams O,
Dal Poz M. Managing brain drain and brain
waste of health workers in Nigeria.
Bull World Health Organ 2004; 82: 595–600.
Brown M, Barnes P, Hodgson H, et al. Junior
doctors. The Times (London), May 28, 2007.
Department for International Development.
Statistics on international development 2006.
http://www.dfid.gov.uk/pubs/files/sid2006/
section2.asp (accessed June 16, 2007).
Sidley P. South Africa and Britain reach
agreement to curb poaching of healthcare
staff. BMJ 2004; 329: 532.
World Health Organization. Global health atlas:
physicians, density per 100 000 population. By
Country, latest available as of 2004. http://
globalatlas.who.int/ (accessed June 16, 2007).
Quality of life in children
with cerebral palsy
Scope welcomes the findings of the
study by Heather Dickinson and
colleagues (June 30, p 2171),1 which
provides convincing evidence that
children with cerebral palsy largely
enjoy a similar quality of life to that of
their non-disabled peers.
Assumptions are often made that
disabled children have a lesser quality
of life, and indeed there is little solid
research on this issue. We commend
the methods used for this study—ie,
assessment of the views and opinions
of disabled children. Both this aspect
and the emphasis on social and
environmental factors that affect
disability mark a huge cultural shift for
a study by medical professionals on
cerebral palsy.
Scope fully supports the study’s
central
recommendation
for
further action to support social and
educational policies that recognise the
similarity between disabled and nondisabled children. However, this action
must include tackling the specific
barriers faced by disabled children,
particularly not being able to attend
a local school near their family or to
access vital equipment in order to be
able to communicate.
We recognise that social barriers are
one element affecting quality of life
for disabled children. Children with
additional support needs might require
a higher level of practical support,
including specialist equipment and
aids, as well as increased access to
health services.
However, we would like to see more
studies that include the views and
experiences of disabled children and
which recognise the immense social
barriers they and their families face.
It would also be useful for there to be
comparative studies of adults and older
people with cerebral palsy, particularly
those going through transitional life
stages—such as teenagers—to explore
their experiences of age, cerebral palsy,
and social barriers.
Scope is a national disability organisation which
focuses on people with cerebral palsy (based in
the UK).
*Jon Sparkes, David Hall
[email protected]
Chief executive (JS) and patron (DH), Scope,
6 Market Road, London N7 9PW, UK
1
Dickinson HO, Parkinson KN,
Ravens-Sieberer U, et al. Self-reported quality
of life of 8–12-year-old children with cerebral
palsy: a cross-sectional European study. Lancet
2007; 369: 2171–78.
Department of Error
Jain P. ADHD: from childhood to adulthood.
Lancet 2007; 369: 1788—The authors of this
Correspondence letter (May 26) should have
been Pratima Jain, Pir Shah, and Sam Ramaiah.
www.thelancet.com Vol 370 August 25, 2007
Articles
Use of calcium or calcium in combination with vitamin D
supplementation to prevent fractures and bone loss in
people aged 50 years and older: a meta-analysis
Benjamin M P Tang, Guy D Eslick, Caryl Nowson, Caroline Smith, Alan Bensoussan
Summary
Background Whether calcium supplementation can reduce osteoporotic fractures is uncertain. We did a meta-analysis
to include all the randomised trials in which calcium, or calcium in combination with vitamin D, was used to prevent
fracture and osteoporotic bone loss.
Methods We identified 29 randomised trials (n=63 897) using electronic databases, supplemented by a hand-search of
reference lists, review articles, and conference abstracts. All randomised trials that recruited people aged 50 years or
older were eligible. The main outcomes were fractures of all types and percentage change of bone-mineral density
from baseline. Data were pooled by use of a random-effect model.
Findings In trials that reported fracture as an outcome (17 trials, n=52 625), treatment was associated with a 12% risk
reduction in fractures of all types (risk ratio 0·88, 95% CI 0·83–0·95; p=0·0004). In trials that reported bone-mineral
density as an outcome (23 trials, n=41 419), the treatment was associated with a reduced rate of bone loss of 0·54%
(0·35–0·73; p<0·0001) at the hip and 1·19% (0·76–1·61%; p<0·0001) in the spine. The fracture risk reduction was
significantly greater (24%) in trials in which the compliance rate was high (p<0·0001). The treatment effect was better
with calcium doses of 1200 mg or more than with doses less than 1200 mg (0·80 vs 0·94; p=0·006), and with vitamin
D doses of 800 IU or more than with doses less than 800 IU (0·84 vs 0·87; p=0·03).
Interpretation Evidence supports the use of calcium, or calcium in combination with vitamin D supplementation, in
the preventive treatment of osteoporosis in people aged 50 years or older. For best therapeutic effect, we recommend
minimum doses of 1200 mg of calcium, and 800 IU of vitamin D (for combined calcium plus vitamin D
supplementation).
Introduction
The social and economic burden of osteoporotic fractures
is increasing worldwide, as the population ages. In
the USA, osteoporosis affects more than 10 million
individuals,1 and the yearly expenditure on osteoporotic
fractures has exceeded that on breast cancer.2 The
prevention of fractures has therefore become a major
public-health priority. However, preventive drugs can be
as expensive as the treatment of fractures themselves,
albeit far less painful to the patients.3 Furthermore, as the
demographic trend of ageing shifts to Asia, Africa, and
Latin America, much of the rising cost of prevention of
fractures will be disproportionately borne by some of the
poorest health-care systems in the world. As a result, the
development of an affordable preventive therapy will have
a great effect on health, and its economic management,
worldwide.
Calcium alone, or in combination with vitamin D, has
been suggested as an inexpensive treatment to prevent
osteoporotic bone loss and fractures, costing as little as
€0·41 per day in one European study.4 However, there has
been substantial uncertainty about its efficacy in lowering
the fracture rate. Data from earlier clinical trials showed
that it reduced the fracture rate,5,6 but this finding was not
confirmed in subsequent multicentre trials.7–9 Moreover,
results from meta-analyses have been inconsistent.10–13 All
www.thelancet.com Vol 370 August 25, 2007
Lancet 2007; 370: 657–66
See Comment page 632
Centre for Complementary
Medicine Research, University
of Western Sydney, New South
Wales, Australia
(B M P Tang MD, C Smith PhD,
Prof A Bensoussan PhD);
University of Sydney, Nepean
Hospital, Penrith, New South
Wales, Australia (B M P Tang,
G D Eslick PhD); and School of
Exercise and Nutrition
Sciences, Deakin University,
Victoria, Australia
(Prof C Nowson PhD)
Correspondence to:
Dr Benjamin Tang, Western
Clinical School, University of
Sydney, Nepean Hospital, PO Box
63, Penrith, NSW 2751, Australia
[email protected]
meta-analyses have included a different subset of the
available trials, but none has offered a comprehensive
review of all the relevant evidence. Consequently, the role
of calcium supplementation in the preventive treatment
of osteoporotic fractures has remained uncertain.
7867 potentially relevant
references screened
7593 excluded because they were irrelevant
(eg, animal studies, diagnostic studies,
physiological or pharmacological studies)
274 abstracts for assessment
215 studies excluded
(Editorials, non-randomised trials, younger age
groups, vitamin D only studies, observational
or epidemiological studies, narrative reviews)
59 studies for full text review
29 studies included in analysis
12 had fracture and BMD data
5 had fracture data only
12 had BMD data only
30 studies excluded
(Duplicate studies, used dietary calcium,
calcium as nutritional supplementation,
incomplete outcome data, calcium combined
with exercise or calcium compared with other
antiresorptive therapies)
Figure 1: Study selection
BMD=bone-mineral density.
657
Articles
For the National Institute of
Health repository see
http://www.clinicaltrials.gov
For the National Research
Register repository see
http://www.nrr.nhs.uk
For the Current Controlled
Trials repository see
http://www.controlled-trials.
com
For the Trials Central
repository see http://www.
trialscentral.org
For the International
Osteoporosis Foundation see
http://www.osteofound.org
For the National Guideline
Clearinghouse see
http://www.guideline.gov
For the American College
of Physicians see
http://www.acpjc.org
Year
We aimed to do a systemic review to quantitatively
assess all the published randomised controlled trials that
assessed the effect of calcium, or calcium in combination
with vitamin D supplementation, on osteoporotic
fractures and bone-mineral density, in adults aged
50 years and older.
Methods
Search strategy and selection criteria
The study was done with a prospectively developed
protocol, which prespecified the research objective,
search strategy, study eligibility criteria, and the methods
of data extraction and statistical analysis. All subgroup
variables were defined before analysis. The reporting of
the study’s findings was in accordance with the Quality
of Reporting of Meta-analyses (QUOROM) conference
statement.14
We searched, without language restrictions, for all
publications on calcium and vitamin D between January,
1966, and January, 2007, using electronic databases,
Country
Total
(n)
Mean age
(years)*
including Medline, Embase, Current Content, Cumulative
Index to Nursing and Allied Health Literature (CINAHL),
Database of Abstracts of Reviews of Effects (DARE),
Cochrane Central Register of Controlled Trials (CENTRAL),
and the Cochrane Database of Systematic Reviews. We
also searched for unpublished trials and those in progress
using clinical trials repositories, including that of the
National Institute of Health, the National Research
Register, Current Controlled Trials, and Trials Central. The
search was supplemented by use of resource websites,
including those of the International Osteoporosis
Foundation, National Guideline Clearinghouse, American
College of Physicians, and Computer Retrieval of
Information on Scientific Projects.
The search strategy used the following MeSH search
terms: (1) “osteoporosis”; (2) “bone density”, “bone
strength”, or “bone loss”; (3) “fracture” or “bone fracture”;
(4) “calcium carbonate”, “calcium”, or “dietary calcium”;
(5) “vitamin D”; and (6) “cholecalciferol” or
“colecalciferol”.
Participants
Treatment
Dose(Ca/Vit D)
Outcomes
Trial quality†
Chapuy-15
1992 France
2790
84 (6)
Mobile elderly women in nursing homes
Ca+vit D
1200 mg/800 IU
Fracture & BMD
B,R
Reid-127
1993 New Zealand
122
58 (5)
Healthy, postmenopausal women
Ca
1000 mg
Fracture & BMD
B,C
Chevalley28
1994 Switzerland
156
72 (7)
Healthy, elderly men and women
Ca
800 mg
Fracture & BMD
B,C
Recker29
1996 USA
197
74 (7)
Independent postmenopausal women
Ca
1200 mg
Dawson-Hughes-16
1997 USA
389
71
Healthy, ambulatory men and women
Ca+vit D
Riggs30
1998 USA
236
66 (3)
Healthy, postmenopausal women
Ca
Peacock31
2000 USA
261
75 (8)
Independent, mobile elderly men and women
Ca
Chapuy-225
2002 USA
583
85
Ambulatory, institutionalised women
Ca+vit D
Larsen24
2004 Denmark
Harwood32
2004 U.K.
Fujita26
2004 Japan
RECORD-17
2005 UK
2638
RECORD-27
2005 UK
2643
Porthouse33
2005 UK
Jackson8
2006 USA
Reid-234
9605
150
19
Fracture
B,C
Fracture & BMD
B,C,R
1600 mg
Fracture & BMD
B,C
750 mg
Fracture & BMD
R
1200 mg/800 IU
Fracture & BMD
None reported
500 mg/700 IU
74 (66–103) Elderly men and women
Ca+vit D
1000 mg/400 IU
Fracture
None reported
81 (67–92)
Elderly women with previous fractures
Ca+vit D
1000 mg/800 IU
Fracture & BMD
R
81
Elderly, institutionalised women
Ca
900 mg
Fracture & BMD
None reported
78 (6)
Elderly men and women with previous fractures Ca
1000 mg
Fracture
B,C,R
77 (6)
Elderly men and women with previous fractures Ca+vit D
1000mg / 800IU
Fracture
B,C,R
3314
77 (5)
Women with risk factors for hip fracture
Ca+vit D
1000mg / 800IU
Fracture
R
36 282
62 (7)
Healthy, postmenopausal women
Ca+vit D
1000 mg/400 IU
Fracture & BMD
B,R
2006 New Zealand
1471
74 (4)
Healthy, postmenopausal women
Ca
1000 mg
Fracture & BMD
B,C,R
Prince-19
2006 Australia
1460
75 (3)
Healthy, elderly women
Ca
1200 mg
Fracture & BMD
C,R
None reported
Prince-246
1995 Australia
84
62 (5)
Healthy, postmeopausal women
Ca
1000 mg
BMD
Lamke35
1978 Sweden
40
60 (3)
Elderly women with previous fractures
Ca
1000 mg
BMD
None reported
Orwell36
1990 USA
77
58 (12·5)
Healthy men
Ca+vit D
1000 mg
BMD
None reported
Dawson-Hughes-237 1990 USA
301
58 (5)
Healthy, postmenopausal women
Ca
500 mg
BMD
None reported
Elders38
295
50 (46–55)
Healthy, postmenopausal women
Ca
1500 mg
BMD
None reported
60
76
Elderly women from nursing home
Ca
800 mg
BMD
R
118
52
Healthy, postmenopausal women
Ca+vit D
600 mg/400 IU
BMD
B,C,R
None reported
1991 Netherlands
Lau39
1992 Hong Kong
Aloia40
1994 USA
Strause41
1994 USA
59
66 (7)
Healthy, postmenopausal women
Ca
1000 mg
BMD
Storm42
1998 USA
60
72 (6)
Healthy, postmenopausal women
Ca
1000 mg
BMD
None reported
Baeksgaard43
1998 Denmark
240
62 (6)
Healthy, postmenopausal women
Ca+vit D
1000 mg/560 IU
BMD
None reported
Grados44
2003 France
192
Meier45
2004 Germany
55
75 (68–79)
Female outpatients
Ca+vit D
500 mg/400 IU
BMD
None reported
56 (11)
Healthy men and women
Ca+vit D
200 mg/200 IU
BMD
None reported
R=randomisation. C=allocation concealment. B=blinding of outcome assessment. BMD=bone-mineral density. Ca=calcium. Vit=vitamin. *The SD or the range of the age is shown in parantheses for studies that
have this information available. †Reported in the study or confirmed by the author.
Table 1: Study and participant summary characteristics
658
www.thelancet.com Vol 370 August 25, 2007
Articles
We hand-searched the reference lists of every primary
study for additional publications. Further searches were
done by reviewing abstract booklets and review articles.
We included all randomised trials that used calcium, or
calcium with vitamin D supplementation, versus placebo,
reported fracture or bone-mineral density as an outcome,
and included patients aged 50 years or older.
Studies were excluded if they were duplicated studies,
did not use a placebo or control group, used dietary
calcium as an intervention, used calcium as part of a
complex nutritional supplementation regimen, used
calcium in combination with other treatment (eg,
fluoride, hormones, or antiresorptive therapy), enrolled
patients with secondary osteoporosis (eg, long-term
glucocorticoid use), or used vitamin D without calcium.
The primary outcome was fracture of any site, including
hip, vertebra, and wrist. The secondary outcome was
bone-mineral density, expressed as percentage change
from baseline. For the purpose of analysis, more than
one fracture suffered by the same patient was counted as
one event, with the first fracture regarded as the primary
outcome.
Data extraction
Data were extracted independently by two reviewers;
disagreements were resolved by consensus. To quantify
the level of agreement between reviewers, a κ statistic
was calculated. The κ statistic is a chance-corrected
proportional index, with values ranging from +1 (perfect
agreement) to –1 (complete disagreement). Information
extracted included year of publication, country of origin,
clinical setting, trial duration, participant demographics,
sample size, calcium and vitamin D doses, baseline
calcium
intake,
serum
25-(OH)-vitamin
D3
concentration, and drug formulation. Authors were
contacted if further study details were needed,
discrepancies or inaccuracies were detected, or duplicate
publication was suspected.
Quality assessment
We assessed the method of every study using a four-item
checklist—namely, reporting of randomisation method,
allocation concealment, blinding of outcome assessment,
and completeness of follow-up. The criteria were drawn
from the Cochrane Collaboration guidelines.15 To assess
the effect of trial quality on the effect size, sensitivity
analysis was done by comparison of studies that fulfilled
quality criteria with those that did not.
Statistical analysis
In every study, we calculated the risk ratio (RR) for the
primary outcome (the fracture rate) and the mean
percentage difference between groups for the secondary
outcome (bone-mineral density), along with the 95% CIs.
The outcome measures were pooled by use of the
random-effects model. On the basis of pooled RR and the
baseline risk, the number needed to treat was calculated,
www.thelancet.com Vol 370 August 25, 2007
along with its 95% CI.16 Heterogeneity was assessed with
Cochran’s Q statistic and quantified using the I2 statistic,
which indicated the proportion of variability across
studies that was due to heterogeneity rather than
sampling error. Metaregression was used to assess the
effect of age, baseline fracture risk, bodyweight, trial
duration, and compliance on treatment efficacy. The
baseline fracture risk—ie, that in the control group—was
calculated. The effect of individual studies on the pooled
effect size was assessed with influence analysis, in which
the analysis was repeated omitting one study at a time, to
establish the contribution of each study to the effect
size.
We anticipated the presence of clinical heterogeneity,
based on the findings that the effects of calcium, or
calcium with vitamin D, seemed to vary depending on
dose and various treatment factors (eg, treatment
duration, previous fractures), and since participant
demographics and clinical settings differed greatly
between studies. Since the test for heterogeneity had low
statistical power, we assumed the presence of
heterogeneity a priori, and used the random effects
model in all analyses.
To assess whether the treatment effect (reduction in
fracture risk) of calcium, or calcium with vitamin D, was
modified by clinical and demographic variables, we
prespecified a list of 12 variables for subgroup analysis.
Variables were chosen on the basis of either biological
plausibility (eg, age, baseline vitamin D concentration, or
drug dose) or known risk factors (eg, institutional settings
or history of previous fractures). We choose 1200 mg
calcium and 800 IU vitamin D as threshold levels based
on the following reasons. For calcium, the recommended
dietary intake is 1200 mg for both men and women older
than 50 years, according to the US Department of
RR (95% CI)
Chapuy-15
Reid-127
Chevalley28
Recker29
Dawson-Hughes-16
Riggs30
Peacock31
Chapuy-225
Larsen24
Harwood32
Fujita26
RECORD-17
Porthouse33
RECORD-27
Jackson8
Reid-234
Prince-19
Overall
For the Computer Retrieval of
Information on Scientific
Projects see http://www.crisp.
cit.nih.gov/
RR (95% CI)
Relative weight (%)
0·75 (0·64–0·87)
0·40 (0·08–1·98)
0·96 (0·35–2·66)
0·85 (0·56–1·30)
0·46 (0·23–0·90)
0·89 (0·51–1·57)
0·81 (0·46–1·43)
0·85 (0·64–1·13)
0·84 (0·72–0·98)
0·49 (0·03–7·67)
0·31 (0·07–1·39)
0·94 (0·77–1·15)
0·96 (0·70–1·33)
0·94 (0·77–1·15)
0·97 (0·92–1·03)
0·92 (0·75–1·14)
0·87 (0·69–1·10)
0·88 (0·83–0·95)
12·73
0·18
0·44
2·40
0·97
1·37
1·38
4·92
12·24
0·06
0·20
8·72
3·91
8·74
27·14
7·90
6·69
0·1
Test for overall effect: Z=–3·55, p=0·0004
Test for heterogeneity: p=0·20, I2=20%
0·2
0·5
Favours treatment
1
2
5
10
Favours control
Figure 2: Effect of calcium and calcium in combination with vitamin D on fracture risk
RR=risk ratio. Size of data markers are proportional to the weight of every study in the forest plot. Horizontal
bars=95% CI.
659
Articles
A
Chapuy-15
Dawson-Hughes-237
Chapuy-225
Jackson8
Reid-234
Prince-246
Riggs30
Chevalley28
Reid-127
Peacock31
Harwood32
Meier45
Grados44
Baeksgaard43
Aloia
Dawson-Hughes-16
Lamke35
Lau39
Prince-19
Overall
Difference in means (95% CI)
Difference in means (95% CI)
2·79
2·77
0·37
5·82
3·89
7·92
6·93
0·79
11·76
3·41
12·63
13·03
0·96
2·84
10·95
0·39
0·19
0·64
11·92
1·10 (0·10 to 2·10)
1·20 (0·19 to 2·21)
2·65 (0·39 to 5·69)
0·84 (0·25 to 1·42)
1·60 (0·80 to 2·40)
0·43 (0·00 to 0·86)
0·70 (0·20 to 1·20)
2·10 (0·04 to 4·16)
0·20 (0·00 to 0·40)
1·20 (0·32 to 2·08)
0·25 (0·11 to 0·39)
0·10 (0·00 to 0·20)
3·19 (1·34 to 5·04)
1·00 (0·01 to 1·99)
0·28 (0·03 to 0·53)
2·98 (0·02 to 5·94)
4·40 (0·14 to 8·66)
2·42 (0·13 to 4·71)
0·19 (0·00 to 0·38)
0·54 (0·35 to 0·73)
–8·00
Test for overall effect: Z=7·07, p<0·0001
Test for heterogeneity: p=<0·0001, I2=73%
B
Dawson-Hughes-237
Reid-233
Riggs20
Reid-127
Peacock31
Harwood32
Meier45
Strause41
Storm42
Baeksgaard43
Aloia40
Dawson-Hughes-16
Elders38
Orwell36
Lau39
Fujita26
Overall
Relative weight (%)
–4·00
0·00
4·00
Favours control
Difference in means (95% CI)
8·00
Favours treatment
Difference in means (95% CI)
Relative weight (%)
0·90 (0·04 to 1·76)
1·80 (0·90 to 2·70)
1·00 (0·24 to 1·76)
0·61 (0·03 to 1·19)
2·66 (0·45 to 4·87)
1·10 (–0·97 to 3·17)
0·80 (0·08 to 1·52)
2·28 (0·10 to 4·46)
3·50 (0·97 to 6·03)
1·60 (0·40 to 2·80)
0·80 (0·01 to 1·59)
2·47 (0·02 to 4·92)
2·50 (–0·12 to 5·12)
–0·40 (–1·35 to 0·55)
2·40 (0·13 to 4·67)
6·55 (0·47 to 12·63)
1·19 (0·76 to 1·61)
9·77
9·35
10·68
12·41
3·04
3·39
11·03
3·11
2·44
7·11
10·40
2·57
2·30
8·99
2·93
0·48
–8·00
Test for overall effect: Z=5.41, p<0·0001
Test for heterogeneity: p=0·02, I2=49%
–4·00
0·00
4·00
Favours control
8·00
Favours treatment
Figure 3: Effect of calcium and calcium in combination with vitamin D on hip (A) and vertebral (B)
bone-mineral density
RR (95% CI)
Chapuy-15
Reid-127
Chevalley28
Dawson-Hughes-16
Riggs30
Peacock31
Chapuy-225
Fujita26
Overall
Recker29
RECORD-17
Porthouse33
RECORD-27
Jackson8
Reid-234
Prince-19
Overall
RR (95% CI)
0·75 (0·64–0·87)
0·40 (0·08–1·98)
0·96 (0·35–2·66)
0·46 (0·23–0·90)
0·89 (0·51–1·57)
0·81 (0·46–1·43)
0·85 (0·64–1·13)
0·31 (0·07–1·39)
0·76 (0·67–0·86)
0·85 (0·56–1·30)
0·94 (0·77–1·15)
0·96 (0·70–1·33)
0·94 (0·77–1·15)
0·97 (0·92–1·03)
0·92 (0·75–1·14)
0·87 (0·69–1·10)
0·96 (0·91–1·01)
80% compliance rate
Lower compliance rate
0·1
0·2
0·5
Favours treatment
Figure 4: Effect of compliance on fracture risk reduction
660
1
2
5
10
Favours control
Agriculture’s dietary reference intakes guidelines.17
Furthermore, most of the trials included in our analysis
had a threshold level between 1000 mg and 1200 mg. For
vitamin D, the lowest dose thought to have an effect is
800 IU per day. A recent meta-analysis showed that for a
lower dose of 400 IU per day, there was no significant
treatment benefit.18 However, a higher dose (eg, 800 IU per
day) reduced fractures in both ambulatory and institutionalised elderly people.19 A test of interaction was done
on all subgroups, to establish if the difference in effect size
between subgroups was statistically significant.
We did cumulative meta-analysis by undertaking
sequential random-effects pooling, starting with the
earliest studies.20 Each successive meta-analysis then
summarised all the trials in the preceding years. Results
were presented as a series of mini meta-analyses, which
were ordered chronologically in a forest plot to show the
consequence of adding studies on the effect size.
We assessed publication bias using the Egger’s regression
model.21 If publication bias was detected, the effect of such
bias was assessed with the fail-safe number method.22,23
The fail-safe number was the number of unpublished
studies that would be needed to nullify the observed result
to statistical non-significance at the α=0·05 level.
Publication bias is generally regarded as a concern if the
fail-safe number is less than 5n+10, where n is the number
of studies included in the meta-analysis.
Results were regarded as statistically significant if
p<0·05. All analyses were done with Comprehensive
Meta-analysis (version 2.0) and GraphPad Prism.
Role of the funding source
The sponsor of the study had no role in study design, data
collection, data analysis, data interpretation, or writing of
the report. The corresponding author had full access to all
the data in the study and had final responsibility for the
decision to submit for publication.
Results
Of the 7867 references screened, 29 studies were included
in the final analysis (figure 1).5–9,24–46 17 trials reported
fractures as an outcome, 5–9,24–34 and 24 reported bone-mineral
density.5,6,8,9,25–28,30–32,34–46 Some trials reported both outcomes
(table 1). One study7 had a four-group trial, with one placebo
and three experimental groups. Its calcium group, and
calcium in combination with vitamin D group, were
analysed separately and therefore treated as two studies; as
a result, the placebo group was counted twice.
In total, 63 897 individuals were analysed, most of whom
were women (n=58 785 [92%]) with a mean age of 67·8 years
(SD 9·7). The median baseline risk for fracture was 16%
(10–22). In 13 trials, participants received calcium and
vitamin D combination supplementation, whereas in all
other trials they received calcium-only supplementation
(table 1).
Trial quality was better in studies reporting fractures as
outcome than it was in those reporting bone-mineral
www.thelancet.com Vol 370 August 25, 2007
Articles
Subtotal (n)* RR (95% CI)
Reid-127
Supplementation
Calcium
6517
0·90 (0·80–1·00)
46 108
0·87 (0·77–0·97)
No
46 919
0·86 (0·78–0·95)
Yes
5706
0·93 (0·82–1·06)
49 233
0·94 (0·90–0·99)
3392
0·76 (0·66–0·88)
Calcium and vitamin D
0·63
Previous fractures
0·85
Clinical setting
Community
Institutionalised
0·003
Serum 25(OH) vitamin D3 concentration†
Low
10 144
0·86 (0·78–0·93)
Normal
39 167
0·94 (0·90–0·99)
Hip
51 935
0·87 (0·75 – 0·99)
Vertebral
45 184
0·87 (0·75 – 1·01)
0·06
Fracture sites
0·40 (0·08–1·98)
0·96 (0·35–2·66)
0·85 (0·56–1·30)
0·89 (0·51–1·57)
0·81 (0·46–1·43)
0·31 (0·07–1·39)
0·94 (0·77–1·15)
0·92 (0·75–1·14)
0·87 (0·69–1·10)
0·90 (0·80–1·00)
0·75 (0·64–0·87)
0·46 (0·23–0·90)
0·85 (0·64–1·13)
0·84 (0·72–0·98)
0·49 (0·03–7·67)
0·94 (0·77–1·15)
0·96 (0·70–1·33)
0·97 (0·92–1·03)
0·87 (0·77–0·97)
Calcium only
Calcium and vitamin D
0·1
0·2
0·5
1
Favours treatment
Low
Normal
7272
0·80 (0·71–0·89)
45 241
0·95 (0·91–1·00)
0·008
Calcium dose
<1200 mg
47 359
0·94 (0·89–0·99)
≥1200 mg
5266
0·80 (0·72–0·89)
<800 IU
36 671
0·87 (0·71–1·05)
≥800 IU
9437
0·84 (0·75–0·94)
0·006
Vitamin D dose
0·03
Sex
46 586
0·88 (0·80–0·97)
6039
0·88 (0·80–0·96)
<1%
38 212
0·96 (0·91–1·02)
≥1%
5621
0·80 (0·70–0·91)
50–69
36 640
0·97 (0·92–1·02)
70–79
12 481
0·89 (0·82–0·96)
3504
0·76 (0·67–0·87)
Men and women studies
0·33
Percentage change in BMD
0·007
Age (years)
≥80
Chevalley28
Recker29
Riggs30
Peacock31
Fujita26
RECORD-27
Reid-234
Prince-19
Overall
Chapuy-15
Dawson-Hughes-16
Chapuy-225
Larsen24
Harwood32
RECORD-17
Porthouse33
Jackson8
Overall
RR (95% CI)
0·72
Dietary calcium intake‡
Women-only studies
RR (95% CI)
p value
0·003
Compliance§
≥80%
4508
0·76 (0·67–0·86)
60–69%
3511
0·92 (0·71–1·19)
50–59%
44 494
0·96 (0·91–1·01)
0·002
BMD=bone-mineral density. *Number may not add up to 100% of total because
of missing data in some variables. †A serum 25(OH) vitamin D3 concentration
<25 nmol/L was regarded as below normal. Results were similar on higher cut-off
points. ‡Dietary intake was regarded as low if <700 mg per day. Result was similar
on higher cut-off point (1000–1200 mg/day). §Compliance means the use of 80%
or more of the study drug. There was no study that had less than 50% compliance,
and there was no study in the 70–80% subgroup.
Table 2: Subgroup analysis for fracture
density only. In fracture trials, 14 of the 17 trials (82%)
reported data for methodological quality, whereas in
bone-mineral density only trials, only two of the 11 trials
(18%) reported such data (table 1).
Of the 17 trials reporting fracture as an outcome
(n=52 625), calcium and calcium in combination with
vitamin D were associated with a 12% risk reduction in
www.thelancet.com Vol 370 August 25, 2007
2
5
10
Favours control
Figure 5: Effect of calcium and calcium in combination with vitamin D on fracture risk reduction
fractures of all types (RR 0·88, 95% CI 0·83–0·95;
p=0·0004; figure 2). The direction of effect was consistent
in all studies, with all favouring treatment in its effect on
fractures. Of the 24 trials reporting bone-mineral density,
calcium and calcium in combination with vitamin D were
associated with a reduced bone loss of 0·54% (0·35–0·73;
p<0·0001) at the hip and 1·19% (0·76–1·61; p<0·0001) in
the spine (figure 3). A positive treatment effect on
bone-mineral density was evident in most studies.
Trials with higher compliance showed a significantly
greater risk reduction than did those with lower
compliance rates (figure 4). Of the eight trials (n=4508)
that reported a compliance rate of 80% or more, the
treatment was associated with a 24% risk reduction in
fractures of all types (p<0·0001). We found no relation
between compliance and an increased dose of
calcium (p=0·57).
We found that the treatment effect was similar across
fracture sites and sex, all showing a risk reduction
of 12–13% (table 2). Similarly, a history of previous
fractures did not change the treatment effect (p=0·85).
The addition of vitamin D to calcium did not change
treatment effect significantly (figure 5). The difference in
RR between calcium-only supplementation and calcium
with vitamin D combination was very small (0·87 vs 0·90)
and was not significant (p=0·63).
People with low vitamin D serum concentration
(25-(OH)-vitamin D3<25 nmol/L), had a greater risk
reduction compared with those whose serum
25-(OH)-vitamin D3 was normal (RR 0·86 vs 0·94);
however, the result was not significant (p=0·06). Results
were the same at 35 nmol/L. At an even higher cut-off
point (50 nmol/L) there was no longer a difference in
risk reduction (RR 0·82 vs 0·89; p=0·46). The treatment
effect was greater in people who were institutionalised
than in those living in the community (RR 0·76 vs 0·94;
p=0·003).
661
Articles
A
B
0
–0·20
–0·40
Log risk ratio
–0·60
–0·80
–1·00
–1·20
–1·40
–1·60
–1·80
–2·00
0·95
1·61
2·27
2·93
3·59
4·25
4·91
5·57
6·23
6·89 7·55
49·95 55·41 60·87 66·33 71·79 77·25 82·71 88·17 93·63 99·09 100
Compliance (%)
Trial duration (years)
Figure 6: Meta-regression analysis of trial duration (A) and compliance (B)
Size of the circles corresponds to the weight of each study.
The treatment effect was also greater in participants
whose daily calcium intake was low (defined as less than
700 mg per day) (0·80 vs 0·95; p=0·008). There was a
non-significant association between being institutionalised and having a low daily calcium intake (p=0·08).
The treatment effect was significantly better in trials
with high compliance rate than in those with a rate lower
than 80% (table 2). The treatment effect was also best
with calcium doses of 1200 mg or more, or vitamin D
doses of 800 IU or more (table 2).
Age was an important determinant of treatment
efficacy. Risk reduction was less in people aged
50–70 years than in those who were older than 70 years
100
Spearman r=–0·8
Compliance rate (%)
p=0·0003
75
50
25
0
2
4
Trial duration (years)
6
8
Figure 7: Relation between compliance rate and trial duration
Slope (95% CI)
Trial duration
0·045 (0·02 to 0·07)
0·002
Compliance
0·007 (0·003 to 0·01)
0·001
–0·01 (–0·013 to –0·003)
0·002
Age
Bodyweight
0·01 (0·004 to 0·017)
0·002
Baseline risk
–0·01 (–0·02 to –0·002)
0·02
Table 3: Summary of metaregression analysis
662
p value
(table 2). For people older than 70 years, the extent of risk
reduction was statistically significant. As we expected, a
lesser reduction in bone-mineral density was associated
with a greater treatment effect (table 2).
We used metaregression analysis to examine the
variation in treatment effect (reduction in fracture risk)
attributable to prespecified variables. We noted that a
smaller treatment effect was associated with an increased
trial duration and lower participant compliance (figure 6).
There was a strong correlation between trial duration and
participant compliance, with reduced compliance
recorded in trials of long duration (figure 7), suggesting
that the reduced treatment effect in these trials may well
have been related to poor participant compliance. We
also found a greater treatment effect with older patients,
a lower bodyweight, and a higher baseline risk (table 3).
Egger’s regression analysis showed that publication
bias was present (p=0·01; figure 8). We therefore used
the fail-safe methods to estimate the number of potential
missing studies needed to significantly change the
conclusion of our findings. This analysis showed that, to
nullify our estimated effect size, 100 studies with
non-significant findings or 22 studies showing harmful
treatment effect would be needed. In view of the fact that
there have been no more than 30 studies published over
the past 15 years, it is highly improbable that such a large
number of similar studies would have gone unpublished
or have been missed by our extensive search strategy.
Furthermore, the missing studies are likely to be small,
the effect of which is probably very negligible.
The influence analysis showed that no particular trial
affected the pooled effect size (figure 9). Cumulative
meta-analysis showed that the evidence was consistent
over time (figure 10). The point estimates and their CIs
stabilised over the past year (2006) and remained
unchanged, even when three large studies8,9,31 were added.
This result suggested that the addition of any future
www.thelancet.com Vol 370 August 25, 2007
Articles
40
Precision (1/SE)
30
20
10
0
–2·0
–1·5
–1·0
–0·5
0
0·5
1·0
1·5
2·0
Log risk ratio
Figure 8: Funnel plot to assess publication bias
Circles indicate individual studies. Diamond indicates summary estimate. SE=standard error.
study, even if it included many thousands of participants,
would add little to the cumulative body of evidence.
In a sensitivity analysis, we analysed separately the
studies that did not meet methodological criteria and
compared them with studies of better quality. For each of
the four criteria, studies of lesser quality showed a slightly
more optimistic estimate of the RR than did those of
high quality (table 4). However, the results did not differ
significantly. The κ statistic was 0·882, suggesting good
agreement between reviewers in data extraction.
Discussion
Our meta-analysis has shown that calcium
supplementation, alone or in combination with vitamin
D, is effective in the preventive treatment of osteoporotic
fracture. Over an average treatment duration of 3·5 years,
the risk of fracture was reduced and was accompanied by
a reduction of bone loss at the hip and spine. The fracture
risk reduction was greater in individuals who were
elderly, lived in institutions, had a low bodyweight, had a
low calcium intake, or were at a high baseline risk than it
was in others. The treatment effect was consistent
irrespective of sex, fracture sites, or history of previous
fracture. Moreover, the treatment was similarly effective
whether the person used calcium or calcium in
combination with vitamin D supplementation. However,
the treatment was less effective if compliance was poor.
For calcium-only supplementation, a minimum dose of
1200 mg is needed for best therapeutic effect. For calcium
in combination of vitamin D supplementation, a
minimum dose of 800 IU of vitamin D is recommended.
Although addition of vitamin D supplementation was
not shown to offer additional risk reduction over and
above the use of calcium alone, a significant difference
was observed between the effects of different vitamin D
doses. This discrepancy could be due to statistical artifact.
However, we would like to point out that our analysis was
www.thelancet.com Vol 370 August 25, 2007
Study removed
RR (95% CI)
Chapuy-15
Reid-127
Chevalley28
Recker29
Dawson-Hughes-16
Riggs30
Peacock31
Chapuy-225
Larsen24
Harwood32
Fujita26
RECORD-17
Porthouse33
RECORD-27
Jackson8
Reid-234
Prince-19
Overall
0·94 (0·90–0·98)
0·89 (0·83–0·95)
0·88 (0·82–0·95)
0·88 (0·82–0·95)
0·90 (0·85–0·95)
0·88 (0·82–0·95)
0·88 (0·82–0·95)
0·88 (0·82–0·95)
0·89 (0·83–0·96)
0·88 (0·82–0·95)
0·89 (0·83–0·95)
0·88 (0·81–0·94)
0·88 (0·82–0·94)
0·88 (0·81–0·94)
0·85 (0·80–0·91)
0·88 (0·81–0·95)
0·88 (0·82–0·95)
0·88 (0·83–0·95)
RR (95% CI)
0·5
Favours treatment
1
2
Favours placebo
Figure 9: Influence analysis
Study added
RR (95% CI)
Chapuy-15
Reid-127
Chevalley28
Recker29
Dawson-Hughes-16
Riggs30
Peacock31
Chapuy-225
Larsen24
Harwood32
Fujita26
RECORD-17
Porthouse33
RECORD-27
Jackson8
Reid-234
Prince-19
Overall
0·75 (0·64–0·87)
0·74 (0·64–0·86)
0·75 (0·64–0·87)
0·76 (0·66–0·87)
0·74 (0·65–0·85)
0·75 (0·66–0·86)
0·75 (0·66–0·86)
0·77 (0·68–0·87)
0·79 (0·72–0·87)
0·79 (0·72–0·87)
0·79 (0·72–0·87)
0·82 (0·75–0·89)
0·83 (0·76–0·90)
0·84 (0·78–0·91)
0·88 (0·81–0·95)
0·88 (0·82–0·95)
0·88 (0·83–0·95)
0·88 (0·83–0·95)
Cumulative RR (95% CI)
Year
1992
1993
1994
1996
1997
1998
2000
2002
2004
2004
2004
2005
2005
2005
2006
2006
2006
0·5
Favours treatment
1
2
Favours placebo
Figure 10: Cumulative meta-analysis
663
Articles
Studies reported criteria
RR (95% CI)
Studies not reported criteria
RR (95% CI)
p value
Randomisation
0·89 (0·82–0·98)
0·84 (0·74–0·95)
0·12
Allocation concealment
0·91 (0·82–1·01)
0·87 (0·78–0·96)
0·79
Blind assessment of outcome
0·89 (0·80–0·98)
0·86 (0·77–0·95)
0·15
Loss to follow-up*
0·88 (0·81–0·95)
0·89 (0·73–1·08)
0·69
*Loss to follow-up was regarded as significan t if 30% or more.
Table 4: Sensitivity analysis by methodological criteria
limited by the scarcity of data for vitamin D doses higher
than 800 IU. It is possible that vitamin D does have a
beneficial effect when the dose is large enough (ie,
>800 IU). In the absence of such data, we recommend
that if vitamin D is to be used as an adjunct
supplementation to calcium, its dose should be at least
800 IU or more.
Our study on calcium is a large and exhaustive
meta-analysis. An earlier meta-analysis by Shea and
colleagues10 included studies on calcium-only
supplementation. However, it was restricted to
postmenopausal women and had a substantially smaller
sample size (n=1806) than our study. Two other
meta-analyses12,13 included studies on calcium in
combination with vitamin D supplementation. The first,
by Avenell and co-workers,12 had only 10 376 participants.12
The more recent trial, by Boonen and colleagues,13 was
restricted because it reported only data for hip fracture
and had omitted a large trial24 of 9605 people. By contrast,
our study reported on fracture rate and bone-mineral
density, and contained all relevant trials, including those
omitted by previous meta-analyses. It therefore includes
all the available evidence for the efficacy of calcium
supplementation in the treatment of osteoporotic
fracture.
Poor compliance is a major obstacle to obtaining the
full benefit of calcium supplementation. When we
analysed trials with a compliance rate of at least
80% separately, the risk reduction was doubled. We also
noted that there was a substantially greater number of
people enrolled in trials that reported a low compliance
rate than in those with a high compliance rate. Therefore,
the pooled risk reduction we reported is probably an
underestimation of the treatment efficacy; the risk
reduction associated with a high compliance rate may
show more accurately the therapeutic efficacy of
calcium.
We noted a greater treatment effect in individuals with
low dietary calcium intake than in those whose dietary
intake was high. This result is important since inadequate
dietary calcium is prevalent throughout the world,
especially in elderly people and women.47,48 It is also
consistent with what is understood about the
pathophysiology of osteoporosis. After midlife there is an
age-related yearly loss of bone in both sexes of about 1%,49
which is accelerated to 2% for up to 14 years in women
664
around the age of menopause.50 This bone loss is
characterised by the loss of calcium from bone. To keep
bone loss to a minimum, increased dietary calcium is
needed to offset the continuing loss. It is possible that
the role of supplementation might diminish with
increased intake. It can be difficult, however, for many
elderly people to maintain a calcium intake above
1000 mg per day, especially when energy intake falls with
increasing age.
The treatment effect was previously reported to be
higher for patients in institutionalised care,12 and this
result was confirmed by our study. Some investigators
have suggested that the greater treatment effect could be
attributed to an increased prevalence of vitamin D
deficiency in elderly people who are institutionalised.13,51
Although a lower concentration of serum 25-(OH)-vitamin
D3 was noted in institutionalised people in our study, it is
unlikely to be the only cause. On the basis of the strong
relation between treatment effect and compliance, as
shown by our data, the supervised care in institutions
could feasibly have contributed to increased compliance,
and hence a greater treatment effect. In fact, of the three
studies using institutionalised patients, two reported the
presence of nursing staff to ensure compliance5,25 and
one was undertaken in a hospital26 in which drugs were
routinely given by nurses.
Our study has implications for both clinicians and
policymakers. The estimated number needed to treat
shows that 63 patients will need to be treated over
3·5 years to prevent one fracture. This result is comparable
to other preventive treatments such as statins, in which
40 people would need to be treated for 5 years to prevent
one major cardiovascular event,52 and it is substantially
better than aspirin treatment, in which more than
270 participants would need to be treated for 6 years to
prevent one cardiovascular event.53 Furthermore, the
number needed to treat decreased to 30 or fewer in
individuals who were elderly, had low dietary calcium
intake, or were compliant with calcium supplementation.
On the basis of our recommended minimum dose of
1200 mg of calcium or 800 IU of vitamin D, many
formulations of calcium or combined calcium with
vitamin D tablets that are available contain insufficient
quantities of the active ingredients. In view of the large
number of calcium supplementation tablets sold
worldwide, adequate dosage is an important issue to
address if best possible public-health benefits are to be
realised. Our study also has implications for future
studies of cost-effectiveness. Although our findings
confirmed that therapeutic effect generally increased
with age, it also suggested that the effect becomes much
greater and clinically significant after the age of 70 years.
The cost-effectiveness of selecting a specific age group,
such as people aged 70 years or older, will therefore need
to be addressed in future studies.
Our study has several strengths. First, the large number
of patients provided us with adequate statistical power to
www.thelancet.com Vol 370 August 25, 2007
Articles
detect a treatment effect, whereas most individual trials
were unable to do so because of their small sample size.
Second, we undertook extensive analysis in exploring
important variables that could affect clinical management,
hence providing clinicians with an evidence base for their
decisionmaking. Third, our results are robust and
consistent, as shown by our extensive search for potential
bias by use of influence analysis, publication bias analysis,
cumulative meta-analysis, and sensitivity analysis.
Our study also has limitations. We have excluded trials
that studied calcium as part of a dietary intake or nutritional
supplementation regimen. However, these studies have
been reviewed elsewhere.54 We did not include any
observational studies, in which evidence suggests an
association between low dietary calcium intake and
increased fracture risk, since these studies were beyond
the scope of our review. Similarly, there were trials
investigating the use of calcium in secondary osteoporosis,
or patients with major comorbidities, which we omitted
since our entry criteria excluded them. Furthermore, there
were no men-only trials. Hence, our findings on the effect
of sex on treatment efficacy were inferred indirectly by
comparison of women-only to mixed-sex trials. Additionally,
we did multiple comparisons in our subgroup analysis.
Caution is therefore needed in the interpretation of our
findings, in view of the increased likelihood of type 1 error.
Lastly, we were unable to assess the interaction of physical
exercise on the treatment effect, because trials reporting
physical exercise used widely differing units, so we could
not calculate a summary estimate.
Contributors
BMPT had full access to all the data in the study and takes responsibility
for the integrity of the data and the accuracy of the data analysis. AB and
CN organised the study concept and design. BMPT acquired, analysed,
and interpreted the data. BMPT, GDE, CN, CS, and AB drafted the
manuscript. BMPT did statistical analysis.
Conflict of interest statement
We declare that we have no conflict of interest.
Acknowledgments
This project was supported by a grant from the Australian Government.
The opinions expressed in this publication do not necessarily reflect those
of the Commonwealth of Australia, which does not accept any liability for
any loss, damage, or injury incurred by the use of or reliance on the
information contained herein.
References
1
US Department of Health and Human Services: bone health and
osteoporosis: a report of the Surgeon-General. Rockville: US
Department of Health and Human Services, 2004.
2
Hoerger T, Downs K, Lakshmanan M, et al. Healthcare use among
US women aged 45 and older: total cost and costs for selected
postmenopausal health risks. J Womens Health Gend Based Med 1999;
8: 1077–89.
3
Cummings S, Melton L. Epidemiology and outcomes of osteoporotic
fractures. Lancet 2002; 359: 1761–67.
4
Lilliu H, Pamphile R, Chapuy M, Schulten J, Arlot M, Meunier P.
Calcium-vitamin D3 supplementation is cost-effective in hip fractures
prevention. Maturitas 2003; 44: 299–305.
5
Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to
prevent hip fractures in the elderly women. N Engl J Med 1992; 327:
1637–42.
6
Dawson-Hughes B, Harris SS, Krall EA, Dallal GE. Effect of
calcium and vitamin D supplementation on bone density in men
and women 65 years of age or older. N Engl J Med 1997; 337: 670–76.
www.thelancet.com Vol 370 August 25, 2007
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
The RECORD Trial Group. Oral vitamin D3 and calcium for
secondary prevention of low-trauma fractures in elderly people
(Randomised Evaluation of Calcium Or vitamin D, RECORD): a
randomised placebo-controlled trial. Lancet 2005; 365: 1621–28.
Jackson RD, LaCroix AZ, Gass M, et al. Calcium plus vitamin D
supplementation and the risk of fractures. N Engl J Med 2006; 354:
669–83.
Prince RL, Devine A, Dhaliwal SS, Dick IM. Effects of calcium
supplementation on clinical fracture and bone structure: results of
a 5-year, double-blind, placebo-controlled trial in elderly women.
Arch Intern Med 2006; 166: 869–75.
Shea B, Well G, Cranney A, et al. Meta-analysis of calcium
supplementation for the prevention of postmenopausal
osteoporosis. Endocr Rev 2002; 23: 552–59.
Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E,
Dietrich T, Dawson-Hughes B. Fracture prevention with vitamin D
supplementation: a meta-analysis of randomized controlled trials.
JAMA 2005; 293: 2257–64.
Avenell A, Gillespie W, Gillespie L, O’Connell D. Vitamin D and
vitamin D analogues for preventing fractures associated with
involutional and post-menopausal osteoporosis.
Cochrane Database Syst Rev 2005; 3: CD000227.
Boonen S, Lips P, Bouillon R, Bischoff-Ferrari HA,
Vanderschueren D, Haentjens P. Need for additional calcium to
reduce the risk of hip fracture with vitamin D supplementation:
evidence from a comparative meta-analysis of randomized
controlled trials. J Clin Ednocrinol Metab 2007; 92: 1415–23.
Moher D, Cook DJ, Eastwood S, et al. Improving the quality of
reports of meta-analyses of randomised controlled trials: the
QUOROM statement. Lancet 1999; 354: 1896–900.
Collaboration TC. Cochrane handbook for systematic reviews of
interventions 4.2.6: The Cochrane Collaboration, 2006.
Altman D, Bland M. Interaction revisited: the difference between
two estimates. BMJ 2003; 326: 219.
United States Department of Agriculture: dietary reference intakes:
elements. National agricultural library, 2007.
Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E,
Dietrich T, Dawson-Hughes B. Fracture prevention with vitamin D
supplementation: a meta-analysis of randomized controlled trials.
JAMA 2005; 293: 2257–64.
Bischoff-Ferrari HA. How to select the doses of vitamin D in the
management of osteoporosis. Osteoporos Int 2007;18: 401–07.
Lau J, Schmid C, Chalmers T. Cumulative meta-analysis of clinical
trials builds evidence for examplary medical care. J Clin Epidemiol
1995; 48: 45–57.
Egger M, Davey S, Schneider M, Minder C. Bias in meta-analysis
detected by a simple, graphical test. BMJ 1997; 315: 629–34.
Rosenthal R. The ‘file drawer problem’ and tolerance for null result.
Psychol Bull 1979; 86: 638–41.
Orwin R. A fail-safe N for effect size in meta-analysis. J Educ Stat
1983; 8: 157–59.
Larsen E, Mosekilde L, Foldspang A. Vitamin D and calcium
supplementation prevents osteoporotic fractures in elderly
community dwelling residents: a pragmatic population-based 3-year
intervention study. J Bone Miner Res 2004; 19: 370–78.
Chapuy M, Pamphile R, Paris E, et al. Combined calcium and
vitamin D3 supplementation in elderly women: confirmation of
reversal of secondary hyperparathyroidism and hip fracture risk:
the Decalyos II study. Osteoporos Int 2002; 13: 257–64.
Fujita T, Ohue M, Fujii Y, Miyauchi A, Takagi Y. Reappraisal of
Katsuragi calcium study, a prospective, double-blind,
placebo-controlled study of the effect of active absorbable algal
calcium (AAACa) on vertebral deformity and fracture.
J Bone Miner Res 2004; 22: 32–38.
Reid I, Ames R, Evans M, Gamble G, Sharpe S. Effect of calcium
supplementation on bone loss in post-menopausal women.
N Engl J Med 1993; 328: 460–64.
Chevalley T, Rizzoli R, Nydegger V, et al. Effects of calcium
supplements on femoral bone mineral density, vertebral fracture
rate in vitamin-D-replete elderly patients. Osteoporos Int 1994; 4:
245–52.
Recker R, Hinders S, Davies M, et al. Correcting calcium nutritional
deficiency prevents spine fractures in elderly women.
J Bone Miner Res 1996; 11: 1961–66.
665
Articles
30
31
32
33
34
35
36
37
38
39
40
41
42
666
Riggs B, O’Fallon W, Muhs J, O’Connor M, Kumar R, Melton L.
Long-term effects of calcium supplementation on serum
parathyroid hormone level, bone turnover, and bone loss in elderly
women. J Bone Miner Res 1998; 13: 168–74.
Peacock M, Liu G, Carey M, et al. Effect of calcium or 25OH
vitmain D3 dietary supplementation on bone loss at the hip in men
and women over the age of 60. J Clin Ednocrinol Metab 2000; 85:
3011–19.
Harwood R, Sahota O, Gaynor K, Masud T, Hosking D.
A randomised, controlled comparison of different calcium and
vitamin D supplementaton regimens in elderly women after hip
fracture: the Nottingham Neck of Femur (NoNOF) Study.
Age Ageing 2004; 33: 45–51.
Porthouse J, Cockayne S, King C, et al. Randomised controlled trial
of supplementation with calcium and cholecalciferol (vitamin D3)
for prevention of fractures in primary care. BMJ 2005; 330: 1003–06.
Reid I, Mason B, Horne A, et al. Randomized controlled trial of
calcium in healthy older women. Am J Med 2006; 119: 777–85.
Lamke B, Sjoberg H, Sylven M. Bone mineral content in women
with Colles’ fracture: effect of calcium supplementation.
Acta Orthop Scand 1978; 49: 143–46.
Orwoll E, Oviatt S, McClung M, Deftos L, Sexton G. The rate of
bone mineral loss in normal men and the effects of calcium and
cholecalciferol supplementation. Ann Intern Med 1990; 112: 29–34.
Dawson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N,
Tannenbaum S. A controlled trial of the effect of calcium
supplementation on bone density in postmenopausal women.
N Engl J Med 1990; 323: 878–83.
Elders P, Netelenbos J, Lips P, et al. Calcium supplementation
reduces vertebral bone loss in perimenopausal women: a controlled
trial in 248 women between 46 and 55 years of age.
J Clin Endocrinol Metab 1991; 73: 533–40.
Lau E, Woo J, Leung P, Swaminathan R, Leung D. The effects of
calcium supplementation and exercise on bone density in elderly
Chinese women. Osteoporos Int 1992; 2: 168–73.
Aloia J, Vaswani A, Yeh J, Ross P, Flaster E, Dilmanian F. Calcium
supplemntation with and without hormone replacement therapy to
prevent postmenopausal bone loss. Ann Intern Med 1994; 120: 98–103.
Strause L, Saltman P, Smith K, Bracker M, Andon M. Spinal bone
loss in postmenopausal women supplemented with calcium and
trace minerals. J Nutr 1994; 124: 1060–64.
Storm D, Eslin R, Porter E, et al. Calcium supplementation prevents
seasonal bone loss and changes in biochemical markers of bone
turnover in elderly New England women: a randomized
placebo-controlled trial. J Clin Endocrinol Metab 1998; 83: 3817–25.
43
44
45
46
47
48
49
50
51
52
53
54
Baeksgaard L, Andersen K, Hyldstrup L. Calcium and vitamin D
supplementation increases spinal BMD in healthy, postmenopausal
women. Osteoporos Int 1998; 8: 255–60.
Grados F, Brazier M, Kamel S, et al. Effects on bone mineral
density of calcium and vitamin D supplementation in elderly
women with vitamin D deficiency. Joint Bone Spine 2003; 70:
203–08.
Meier C, Woitge H, Witte K, Lemmer B, Seibel M. Supplementation
with oral vitamin D3 and calcium during winter prevents seasonal
bone loss: a randomized controlled open-label prospective trial.
J Bone Miner Res 2004; 19: 1221–30.
Prince R, Devine A, Dick I, et al. The effects of calcium
supplementation (milk powder or tablets) and exercise on bone
density in postmenopausal women. J Bone Miner Res 1995; 10:
1068–75.
Department of Health. Nutrition and bone health. London: The
Stationery Office, 1998.
Ervin R, Wang C, Wright J, Kennedy-Stephenson J. Dietary intake
of selected minerals for the United States population: 1999–2000.
Advance Data 2004; 341: 1–8.
Jones G, Nguyen T, Sambrook P, Kelly P, Eisman J. Progressive loss
of bone in the femoral neck in elderly people: longitudinal findings
from the Dubbo osteoporosis epidemiology study. BMJ 1994; 309:
691–95.
Ahlborg H, Johnell O, Turner C, Rannevik G, Karlsson M. Bone
loss and bone size after menopause. N Engl J Med 2003; 349:
327–34.
Francis R. Calcium, vitamin D and involutional osteoporosis.
Curr Opin Clin Nutr Metab Care 2006; 9: 13–17.
Cholesterol Treatment Trialis’ (CTT) Collaborators. Efficacy and
safety of cholesterol-lowering treatment: prospective meta-analysis
of data from 90 056 participants in 14 randomised trials of statins.
Lancet 2005; 366: 1267–78.
Berger J, Roncaglioni M, Avanzini F, Pangrazzi I, Tognoni G,
Brown D. Aspirin for the primary prevention of cardiovascular
events in women and men. JAMA 2006; 295: 306–13.
Cumming R, Nevitt M. Calcium for prevention of osteoporotic
fractures in postmenopausal women. J Bone Miner Res 1997; 12:
1321–29.
www.thelancet.com Vol 370 August 25, 2007
Articles
Incidence of new-onset diabetes and impaired fasting
glucose in patients with recent myocardial infarction and
the effect of clinical and lifestyle risk factors
Dariush Mozaffarian, RosaMaria Marfisi, Giacomo Levantesi, Maria G Silletta, Luigi Tavazzi, Gianni Tognoni, Franco Valagussa*,
Roberto Marchioli
Summary
Background Individuals with diabetes are at higher risk of myocardial infarction than non-diabetics. However, much
less is known about the incidence of, and risk factors for, development of diabetes and impaired fasting glucose in
patients who have had a myocardial infarction. We set out to estimate this incidence and investigate whether lifestyle
factors such as dietary habits might alter this risk.
Methods We used prospectively obtained data for 8291 Italian patients with a myocardial infarction within the previous
3 months, who were free of diabetes (determined by medication use, a physician-reported diagnosis, or fasting glucose
≥7 mmol/L) at baseline. Incidence of new-onset diabetes (new diabetes medication or fasting glucose ≥7 mmol/L)
and impaired fasting glucose (fasting glucose ≥6·1 mmol/L and <7 mmol/L) were assessed at follow-up at 0·5, 1·0,
1·5, 2·5, and 3·5 years. Baseline data for body-mass index (BMI), other risk factors, dietary habits, and medications
were updated during follow-up. A Mediterranean diet score was assigned according to consumption of cooked and
raw vegetables, fruit, fish, and olive oil. Associations of demographic, clinical, and lifestyle risk-factors with incidence
of diabetes and impaired fasting glucose were assessed with multivariable Cox proportional hazards.
Findings During 26 795 person-years (mean follow-up 3·2 years [SD 0·9]), 998 individuals (12%) developed new-onset
diabetes (incidence 37 cases per 1000 person-years). Of the 7533 without impaired fasting glucose at baseline,
2514 (33%) developed new-onset impaired fasting glucose or diabetes (incidence 123 cases per 1000 person-years),
rising to 3859 (62%) of 6229 with the lower cutoff for impaired fasting glucose of 5·6 mmol/L (incidence 321 cases
per 1000 person-years). Independent risk factors for new-onset diabetes or impaired fasting glucose included older
age, hypertension, use of beta-blockers, lipid-lowering medications (protective), and diuretic use. Independent lifestyle
risk-factors included higher BMI, greater BMI gain during follow-up, current smoking, a lower Mediterranean dietary
score, and wine consumption of more than 1 L/day. Data for physical activity were unavailable, but inability to perform
exercise testing was associated with higher incidence of diabetes and impaired fasting glucose.
Interpretation Compared with population-based cohorts, patients with a recent myocardial infarction had a higher annual
incidence rate of impaired fasting glucose (1·8 vs 27·5% in our study) and diabetes (0·8–1·6% compared with 3·7%) in
this study. Thus, our results indicate that myocardial infarction could be a prediabetes risk equivalent. Smoking cessation,
prevention of weight gain, and consumption of typical Mediterranean foods might lower this risk, which emphasises the
need for guidance on diet and other lifestyle factors for patients who have had a myocardial infarction.
Introduction
Individuals with diabetes mellitus and no known
coronary heart disease have rates of acute myocardial
infarction and mortality similar to those for non-diabetic
individuals with established coronary heart disease.1–5
However, individuals with both coronary heart disease
and diabetes are at far higher risk of myocardial
infarction and death than those with either condition.1–5
Although the incidence of, and risk factors for, the
development of myocardial infarction in diabetic
individuals have been given great attention,6 much less
is known about the converse relationship: development
of diabetes in patients with myocardial infarction.
It is plausible that a confluence of predisposing risk
factors would increase the incidence of diabetes or
impaired fasting glucose (IFG), a prediabetic condition,
in patients with myocardial infarction. However, although
www.thelancet.com Vol 370 August 25, 2007
Lancet 2007; 370: 667–75
See Comment page 634
Departments of Medicine,
Epidemiology, and Nutrition,
Harvard Medical School and
Harvard School of Public
Health, Boston, MA, USA
(D Mozaffarian MD); Consorzio
Mario Negri Sud, Santa Maria
Imbaro, Chieti, Italy
(R Marfisi MS, G Levantesi MD,
M G Silletta MS, G Tognino MD,
R Marcholi MD); IRCCS
Policlinico San Matteo, Pavia,
Italy (L Tavazzi MD); and the
Ospedale San Gerardo, Monza,
Italy (F Valagussa MD)
*F Valagussa (the
GISSI-Prevenzione Chairman)
died in December 2006
Correspondence to
Dr Dariush Mozaffarian,
Departments of Medicine,
Epidemiology, and Nutrition,
Harvard Medical School and
Harvard School of Public Health,
665 Huntington Ave, Boston,
MA 02115
dmozaff[email protected]
the cross-sectional prevalence of impaired glucose
tolerance is known to be high in patients who recently
had a myocardial infarction,7,8 the incidence of new-onset
impaired glucose homoeostasis in post-MI patients has
not been established.
We aimed to investigate the incidence of diabetes and
IFG in a large cohort of patients who had had a
myocardial infarction within the previous 3 months9,10
and to assess the independent demographic, clinical,
and lifestyle risk factors related to an increased risk. We
were particularly interested in modifiable lifestyle
behaviours that might affect the incidence of diabetes
and IFG after a myocardial infarction, including
components of a traditional Mediterranean dietary
pattern that in short-term randomised experimental
trials improves insulin sensitivity and other
vascular-metabolic risk factors.11–25
667
Articles
Methods
Patients
Patients
We used information from the GISSI-Prevenzione
study,9,10 a randomised trial of fish oil and vitamin E in
patients who had had a myocardial infarction. The study
design and primary results have been described.9,10
11 323 men and women with recent (≤3 months)
myocardial infarction were enrolled from 172 centers in
Italy between October, 1993 and September, 1995. Ethics
approval was obtained, and all patients provided written
informed consent. For this study, we excluded
3032 individuals: 2139 with prevalent diabetes at
enrolment (defined by physician diagnosis, use of
diabetic medication, or fasting glucose ≥7 mmol/L) and
893 with missing information about diet or weight at
baseline, or about glucose at baseline or during
follow-up, leaving 8291 participants for analysis.
Procedures
At baseline, information was obtained for demographics,
cardiovascular risk factors, medications, dietary habits,
and results of cardiac echocardiography, exercise
testing, and coronary angiography. We measured
height, weight, heart rate, and blood pressure. Fasting
blood samples were drawn for assessment of glucose,
lipids, fibrinogen, and complete blood count. The
baseline assessment was done on average 3·5 weeks
(mean 25 days [SD 21]; <5 days in 14 people) after
myocardial infarction, which would reduce to a
minimum the effects of the peri-myocardial infarction
period on glucose levels.
At enrolment, patients were advised to consume
low-fat dairy products, poultry, fish, lean meat, fruits,
vegetables, legumes, whole grains, and olive oil instead
of butter or other oils. Because this was not a dietary
intervention trial, the advice was not systematically
reinforced; thus, the dietary habits of participants
during the study were probably their usual habits and
were unlikely to have been greatly altered because of
this general advice. Participants were followed-up at
regular clinic visits at 0·5, 1·0, 1·5, 2·5, and 3·5 years.
At each visit, they were weighed, had fasting blood
drawn for laboratory evaluation; up-to-date information
was obtained on smoking habits, medications, and
interim events. Dietary habits were assessed at baseline,
0·5 years, and 1·5 years.
In view of the benefits of a traditional Mediterranean
dietary pattern on markers of insulin sensitivity and
other vascular-metabolic risk factors,11–25 we assessed
the combined effect of consumption of several typical
components of a traditional Mediterranean diet. The
GISSI-Prevenzione study was designed as a pragmatic
trial to assess patients in the real-world setting who
were enrolled and followed-up by their own cardiologists.
Thus, full-scale diet questionnaires were not
administered at numerous follow-up visits by
cardiologists in the busy clinic setting.
668
Age (years)
59 (11)
Sex (female)
1075 (13%)
Body mass index (kg/m²)
26·3 (3·4)
Obese (body mass index ≥30 kg/m²)
1076 (13%)
Physician-diagnosed hypertension
3702 (45%)
Previous acute myocardial infarction
Intermittent claudication
907 (11%)
287 (3%)
Current smoker
3699 (45%)
Former smoker
2922 (35%)
Systolic blood pressure (mm Hg)
123 (15)
Diastolic blood pressure (mm Hg)
77 (9)
Heart rate (beats per minute)
68 (10)
Ejection fraction
53 (10)
Heart Failure
None
2924 (35%)
NYHA Class 1
4647 (56)
NYHA Class 2 to 3
708 (9%)
Angina
None
5006 (60%)
CCS 1
2656 (32%)
CCS 2 to 4
482 (6%)
Exercise stress test
Negative
4137 (50%)
Positive
1632 (20%)
Not done
2522 (30%)
Exercise cycloergometer test (Watts)
Exercise treadmill test (mins)
101 (33)
8·1 (3·2)
Laboratory
Fasting glucose (mg/dl)
92 (12) [5·1 (0·7) mmol/L]
LDL cholesterol (mg/dl)
139 (38)
HDL cholesterol (mg/dl)
42 (12)
Triglycerides (mg/dl)
159 (80)
Fibrinogen (mg/dl)
389 (137)
Leukocyte count (×10⁹ L)
7·7 (2·2)
Medications
Fish oil assignment
4166 (50%)
Antiplatelet medication
7652 (92%)
ACE inhibitors
3698 (45%)
Beta-blockers
3897 (47%)
Diuretics
694 (8%)
Cholesterol-lowering medication*
399 (5%)
Data are mean (SD) or number (%). ACE=angiotensin-converting-enzyme.
CCS=Canadian Cardiovascular Society. NYHA=New York Heart Association. *Use
of lipid-lowering medications increased during follow-up (coincident with a
statin subrandomisation29 and publication of several statin trials) to 31% at 6
months, 42% at 1·5 years, and 45% at 3.5 years. Time-varying covariates were
used to account for all changes in medication use.
Table 1: Baseline characteristics of non-diabetic patients with recent
myocardial infarction
Therefore, we administered a brief questionnaire about
specific food items that would give a good indication of
the dietary variation in Italian adults.26 Widely consumed
foods such as pasta, pasta sauces, or bread did not feature
www.thelancet.com Vol 370 August 25, 2007
Articles
Statistical analysis
We used Kaplan-Meier analysis to evaluate time to
development of new-onset diabetes or IFG, and the Cox
proportional hazards model to evaluate the independent
associations of demographic, clinical, and lifestyle risk
factors, with time-at-risk until first event, death, or last
follow-up visit. The proportional hazards assumption
was confirmed with the use of Schoenfeld residuals.
Multivariable models were adjusted for age, sex, smoking,
and other potential risk factors on the basis of biological
plausibility and associations with exposures or outcomes
in this population.
We also assessed factors that might mediate higher risk
of diabetes, including a rise in body-mass index (BMI)
during follow-up, serum lipid levels, systemic
inflammation assessed by fibrinogen and leukocyte
count, and dietary intake of butter or other oils.
Time-varying analyses were used to update covariate
information on diet, BMI, smoking, medications, lipids,
and other laboratory measures. Indicator variables were
used for missing data on baseline covariates or on
smoking; values were otherwise carried forward for
missing time-varying covariates. For parsimony in model
construction, we excluded from the final models several
covariates (eg, blood pressure, intermittent claudication,
use of antiplatelet medication) that were neither
independently associated with diabetes risk nor materially
www.thelancet.com Vol 370 August 25, 2007
1·0
0·9
Diabetes (medications or ≥7 mmol/L)
Disease-free survival
on the questionnaire; instead, we included questions on
the usual consumption of cooked and raw vegetables,
fruit, fish, olive oil and other oils, butter, cheese, wine,
and coffee. The questionnaire did not assess other
components of a Mediterranean-type diet, such as grains,
nuts, or legumes. Each item was scored on a scale from 0
to 3 on the basis of frequency of consumption,26 and
scores were summed to obtain a Mediterranean diet
score (range 0–15) that was evaluated according to
prespecified categories. We also evaluated the benefits of
each food component separately. Dietary habits were
updated over time (baseline, 0·5 years, and 1·5 years)
using cumulative updating.27
At each clinic visit, cases of type II diabetes mellitus
were recorded by either new use of diabetes medications
(insulin or oral hypoglycemic agents) or a fasting glucose
of 7 mmol/L or higher.28 IFG was defined as fasting
glucose 6·1 mmol/L or higher but lower than 7 mmol/L,
in the absence of diabetic medication; we also evaluated
IFG using the lower cutpoint (5·6 mmol/L) (in November,
2003 the American Diabetes Association expert
committee on the diagnosis and classification of diabetes
mellitus suggested a revision of the diagnostic criteria for
IFG, lowering the diagnostic threshold from 6·1 to
5·6 mmol/L).28 Risk of IFG (≥6·1 mmol/L) was assessed
in the 7533 individuals with fasting glucose <6·1 mmol/L
at baseline, and risk of IFG (≥5·6 mmol/L) was assessed
in the 6229 individuals with fasting glucose <5·6 mmol/L
at baseline.
0·8
0·7
Diabetes or IFG110 (≥6·1 mmol/L)
0·6
0·5
0·4
Diabetes or IFG100 (≥5·6 mmol/L)
0·3
Number at risk:
Diabetes
Diabetes/IFG110
Diabetes/IFG100
0
1
8291
7533
6229
7691
5934
3481
Years
2
3
7383
5364
2747
7080
5007
2442
Figure 1: Time to development of new-onset diabetes or impaired fasting glucose
Multivariable-adjusted* p
Hazard ratio (95% CI)
Risk Factors
Age (per 5-year increase)
1·07 (1·03–1·11)
Female sex (compared with male)
0·82 (0·66–1·02)
<0·001
Body mass index (each increasing unit, kg/m²)
1·09 (1·07–1·11)
<0·001
Physician–diagnosed hypertension
1·22 (1·07–1·39)
0·003
Previous acute myocardial infarction
1·04 (0·85–1·27)
Current smoker (compared to never)
1·60 (1·34–1·90)
Former smoker (compared to never)
1·06 (0·90–1·24)
Days from acutemyocardial infarction to enrollment (each day)
1·00 (0·99–1·01)
NYHA Class 1 (compared with 0)
1·11 (0·96–1·29)
NYHA Class 2–3 (compared with 0)
1·10 (0·85–1·43)
Angina CCS 1 (compared with 0)
0·95 (0·81–1·10)
Angina CCS 2–4 (compared with 0)
0·96 (0·72–1·28)
Positive exercise stress test
1·07 (0·91–1·26)
Exercise capacity (comparing high with low quintiles)
0·84 (0·64–1·10)
Inability to undergo exercise testing
1·52 (0·48–4·75)
<0·001
Medications
Fish oil
0·98 (0·86–1·11)
Vitamin E
1·04 (0·92–1·18)
ACE inhibitor
1·11 (0·97–1·26)
Beta–blocker
1·27 (1·12–1·45)
Diuretic
1·08 (0·87–1·34)
<0·001
Lipid-lowering medication
0·78 (0·67–0·90)
0·001
Mediterranean diet score (11–15 vs 0–5)
0·65 (0·49–0·85)
0·002
Cheese consumption (regularly vs never)
1·05 (0·73–1·52)
Wine consumption (≥1 L/day vs never or rarely)
1·08 (0·64–1·83)
Coffee consumption (≥5 cups/day vs never or rarely)
1·19 (0·88–1·61)
Dietary habits
*Includes each of the variables listed in this table.
Table 2: Risk factors for incident diabetes patients with recent myocardial infarction
altered the relation between the Mediterranean diet score
and incidence of diabetes. We calculated the tests for
trend by determining the median value of each category
669
Articles
and evaluating this as a continuous variable. Analyses
were done with Stata (version 8.2).
2·0
Relative risk of death
1·8
Role of the funding source
1·6
This report was supported by the National Heart, Lung,
and Blood Institute (K08-HL-075628), which had no role
in the study design, the collection, analysis, or
interpretation of data, writing of the report, or the
decision to submit the report for publication. The
corresponding author had full access to all the data in the
study and had final responsibility for the decision to
submit for publication.
1·4
1·2
1·0
0·8
p trend=0·04
0·6
<5.6
5·6–6·1
6·1–7
>7 or medication
Fasting glucose (mmol/L)
Figure 2: Multivariable-adjusted relative risk of death associated with
development of diabetes or IFG
Baseline dietary score
4·1 (0–5)*
(n=1104)
6·6 (6–7)*
(n=2030)
8
(n=1439)
9·4 (9–10)* 11·7 (11–15)* p for
(n=2504)
(n=1214)
trend†
Risk factors
Age (years)
56 (11)
58 (11)
59 (11)
60 (10)
59 (10)
<0·001
Sex (female)
115 (10%)
221 (11%)
199 (14%)
346 (14%)
194 (16%)
<0·001
0·002
Body-mass index (kg/m²)
26·0(3·4)
26·3(3·4)
26·5(3·4)
26·3(3·4)
26·4 (3·4)
Obese (body mass index
≥30 kg/m²)
133 (12%)
256 (13%)
191 (13%)
329 (13%)
167 (14%)
Physician-diagnosed
hypertension
458 (41%)
873 (43%)
639 (44%)
1156(46%)
576 (47%)
98 (9%)
208 (10%)
150 (10%)
297 (12%)
154 (13%)
49 (4%)
66 (3%)
Previous acute myocardial
infarction
Intermittent claudication
0·01
47 (3%)
78 (3%)
47 (4%)
Current smoker
663 (60%) 1045 (51%)
612 (43%)
983 (39%)
396 (33%)
<0·001
Former smoker
297 (27%)
666 (33%)
503 (35%)
963(38%)
493 (41%)
<0·001
Systolic blood pressure
(mm Hg)
123 (15)
123 (15)
123 (15)
123 (15)
123 (15)
Diastolic blood pressure
(mm Hg)
77 (9)
76 (9)
77 (9)
77 (8)
77 (8)
Heart rate (beats per minute)
68 (10)
68 (10)
68 (11)
68 (10)
68 (10)
Ejection fraction percentage
53 (10)
53 (10)
53 (10)
53 (10)
53 (11)
None
432 (39%)
700 (35%)
481 (33%)
826 (33%)
485 (40%)
NYHA Class 1
595 (54%) 1149 (57%)
809 (56%) 1449(58%)
645 (53%)
Heart failure
NYHA Class 2–3
75 (7%)
179 (9%)
147 (10%)
226 (9%)
81 (7%)
Angina
None
689 (62%) 1199 (59%)
831 (58%) 1487 (59%)
800 (66%)
CCS 1
329 (30%)
690 (34%)
511 (36%)
811 (32%)
315 (26%)
72 (7%)
110 (5%)
77 (5%)
159 (6%)
64 (5%)
CCS 2–4
Exercise stress test
Negative
561 (51%) 1009 (50%)
691 (48%)
1252 (50%)
624 (51%)
Positive
229 (21%)
390 (19%)
262 (18%)
489 (20%)
262 (22%)
Not completed
314 (28%)
631 (31%)
486 (34%)
763 (30%)
328 (27%)
103 (33)
100 (31)
101 (35)
100 (34)
101 (31)
Exercise cycloergometer
test (Watts)
Exercise treadmill test (mins)
8·3(3·6)
8·2(3·2)
8·0(3·0)
8·2(3·2)
8·0 (2·9)
(Continues on next page)
670
Results
Table 1 shows baseline characteristics of participants.
The average age was 59 years (range 20–90), and 13% of
participants were women. Patients were, on average,
overweight (mean BMI 26·3 kg/m²), and 1078 (13%)
were obese. Substantial proportions of patients had
hypertension, were current smokers, or had heart failure
or angina symptoms. Nearly all patients were taking
antiplatelet medications, and about half were taking
ACE-inhibitors, beta-blockers, or (by the end of
follow-up) lipid-lowering medications.
Of the 8291 patients without diabetes at baseline
followed-up for 26 795 person-years (mean 3·2 years
[SD 0·9]), 998 (12%) developed diabetes—an incidence
of 37 cases per 1000 person-years (figure 1). Four (0·05%)
of the 8291 participants were lost to follow-up. Among
the 7533 patients with fasting glucose lower than
6·1 mmol/L at baseline, 2514 (33%) either developed
diabetes (n=769) or IFG (n=1745) during follow-up—an
incidence of IFG alone of 85 cases per 1000 person-years
and of diabetes and IFG of 123 cases per 1000 person-years.
With the lower cutpoint for IFG (5·6 mmol/L) and
evaluating the 6229 patients with fasting glucose lower
than 5·6 mmol/L at baseline, 3856 (62%) either
developed diabetes (n=548) or developed IFG (n=3308)
during follow-up—an incidence of IFG100 alone of
275 cases per 1000 person-years and of diabetes and IFG
of 321 cases per 1000 person-years.
Independent clinical risk factors for diabetes included
older age, higher BMI, hypertension, and current
smoking (table 2). Each unit of higher BMI was
associated with 9% greater risk (95% CI 7–11).
Hypertension was associated with a 22% higher
risk (7–39), and current smoking was associated with a
60% higher risk (34–90). Female sex and greater exercise
capacity were associated with trends toward lower risk,
although these were not significant. β-blocker use was
associated with 27% (12–45%) higher risk and
lipid-lowering medications were associated with a 22%
(10–33%) lower risk of incident diabetes. A higher
Mediterranean diet score was associated with
35% (15–51%) lower risk of diabetes, consumption of
cheese, wine, and coffee were not associated with
diabetes incidence.
www.thelancet.com Vol 370 August 25, 2007
Articles
Findings for each of these risk factors were generally
similar for incidence of diabetes and IFG (data not shown),
except significantly higher risk was seen for inability to
perform exercise testing (HR 2·43 [95% CI 1·21–4·91]),
use of diuretics (1·15 [1·00–1·33]), and wine consumption
of more than 1 litre a day (1·45 [1·09–1·91]). Treatment
assignment (fish oil or vitamin E) did not reduce incidence
of diabetes or IFG (data not shown).
Adjusting for other risk factors (as in table 1), significant
independent potential mediators of diabetes incidence
included greater BMI gain (for each unit increase, HR
1·17 [95% CI 1·11–1·22], p<0·001), higher triglycerides
(interquintile HR 1·61 [95% CI 1·30–2·00], p<0·001),
lower high-density lipoprotein (HDL) cholesterol (1·46
[1·17–1·81], p=0·001), higher leucocyte count (1·23
[1·00–1·53], p=0·05), and higher consumption of butter
or other oils (1·26 [1·02–1·56], p=0·03). Fibrinogen levels
were not associated with risk (0·98 [0·80–1·21], p=0·87).
We assessed whether the incidence of new-onset
diabetes or IFG was associated with risk of adverse clinical
outcomes. During 28 885 person-years of follow-up,
475 people died. Compared with individuals with normal
fasting glucose (<5·6 mmol/L), there was a trend toward
10% higher risk of death after development of IFG between
5·6–6·05 mmol/L, a trend toward 15% higher risk of
death after development of IFG between 6·1–7 mmol/L,
and 44% higher risk of death after development of diabetes
(p for trend <0·05) (figure 2). Findings were similar for
risk of death or recurrent myocardial infarction (data not
shown).
To determine the potential effects of a Mediterranean-type
diet, we investigated this association in more detail. A
higher score, indicating greater intake of traditional
Mediterranean dietary components, was associated with
older age, being female, slightly greater BMI, hypertension,
previous acute myocardial infarction, and former rather
than current smoking (table 3). There were no major
differences between the different scores in terms of blood
pressure, heart rate, ejection fraction, heart failure severity,
anginal symptoms, or medications. Higher Mediterranean
diet scores were associated with slightly higher HDL, and
lower triglycerides and leucocyte count. Moderate intake
of coffee (<2 cups a day) and wine (≤ 0·5 L/day) was more
common, in people with higher Mediterranean diet scores.
Consumption of butter and oils other than olive oil was
less common with higher Mediterranean diet scores.
After adjustment for age, sex, and smoking, greater
intake of traditional Mediterranean dietary components
was associated with lower incidence of diabetes (p for
trend=0·001). Those with the highest score had a
37% lower risk of developing diabetes than did those with
the lowest scores (table 4). Findings were similar after
additional adjustment for other risk factors (table 4).
Evaluating individual components of the score, several
were associated with trends toward lower risk, but only
the association for cooked vegetables was significant:
comparing high to low quartiles of intake, the
www.thelancet.com Vol 370 August 25, 2007
(Continued from previous page)
Laboratory tests
Fasting glucose (mmol/L)
5·1 (0·7)
5·1 (0·7)
5·1 (0·7)
5·1 (0·7)
LDL cholesterol (mmol/L)
3·6 (1·0)
3·6 (1·0)
3·6 (1·0)
3·6 (1·0)
5·1 (0·7)
3·6 (1·0)
HDL cholesterol (mmol/L)
1·06(0·3)
1·07(0·3)
1·07(0·3)
1·09(0·3)
1·10 (0·3)
0·03
Triglycerides (mmol/L )
1·9 (1·0)
1·8 (1·0)
1·8 (1·0)
1·7 (1·0)
1·7 (1·0)
0·003
Fibrinogen (mg/L)
3·9 (1·4)
3·9 (1·4)
3·8 (1·3)
3·9 (1·4)
3·9 (1·4)
Leukocyte count (×10⁹ L)
7·9 (2·2)
7·8 (2·3)
7·7 (2·3)
7·6 (2·2)
7·5 (2·1)
1016 (50%)
734 (51%)
1241 (50%)
612 (50%)
<0·001
Medications
Fish oil assignment
Antiplatelet medication
563 (51%)
1035 (94%)
1874 (92%) 1328 (92%)
2303(92%) 1112 (92%)
ACE inhibitors
471 (43%)
913 (45%)
641 (45%)
1155(46%)
518 (43%)
Beta-blockers
545 (49%)
961 (47%)
679 (47%)
1155 (46%)
564 (46%)
Diuretics
60 (5%)
164 (8%)
125 (9%)
239 (10%)
106 (9%)
Cholesterol-lowering
medication‡
73 (7%)
92 (5%)
55 (4%)
114 (5%)
65 (5%)
304 (28%)
675 (33%)
554 (39%)
1031 (41%)
601 (50%)
Sometimes
412 (37%)
965 (48%)
679 (47%)
1144(46%)
486 (40%)
Often
188 (17%)
223 (11%)
139 (9%)
218 (9%)
80 (7%)
Regularly
200 (18%)
167 (8%)
67 (5%)
111 (4%)
47 (4%)
Almost never
377 (34%)
855 (42%)
632(44%)
1213(48%)
684 (56%)
Sometimes
416 (38%)
899 (44%)
666(46%) 1069 (43%)
448 (37%)
Often
187 (17%)
192 (10%)
104 (7%)
157 (6%)
61 (5%)
Regularly
124 (11%)
84 (4%)
37 (3%)
65 (3%)
21 (2%)
57 (5%)
101 (5%)
48 (3%)
109 (4%)
60 (5%)
0·03
Other dietary habits
Other oil consumption
Almost never
<0·001
Butter consumption
<0·001
Cheese consumption
Almost never
<0·001
Sometimes
337 (31%)
591 (29%)
467 (33%)
809 (32%)
422 (35%)
Often
384 (35%)
855 (42%)
628 (44%)
993(40%)
428 (35%)
Regularly
326 (30%)
483 (24%)
296 (21%)
593 (24%)
304 (25%)
Almost never
378 (34%)
647 (32%)
468 (33%)
874 (35%)
489 (40%)
≤0·5 L/day
459 (42%) 1004 (50%)
714 (50%)
1301 (52%)
616 (51%)
0·5–1·0 L/day
206 (19%)
302 (15%)
223 (16%)
279 (11%)
95 (8%)
61 (6%)
77 (4%)
34 (2%)
50 (2%)
14 (1%)
121 (11%)
242 (12%)
210 (15%)
392 (16%)
252 (21%)
Wine consumption
>1 lL/day
<0·001
Coffee consumption
Almost never
<0·001
<2 cups per day
261 (24%)
637 (31%)
493 (34%)
905 (36%)
439 (36%)
2-4 cups per day
470 (43%)
768 (38%)
539 (38%)
882 (35%)
393 (32%)
>4 cups per day
252 (23%)
383 (19%)
197 (14%)
325 (13%)
130 (11%)
Values are mean (SD) or n (%). ACE=angiotensin-converting-enzyme. CCS=Canadian Cardiovascular Society.
NYHA=New York Heart Association.*Mean (range). †Age-adjusted for differences across categories of the
Mediterranean dietary score. ‡Use of lipid-lowering medications increased during follow-up (coincident with a statin
subrandomisation30 and publication of several statin trials) to 31% at 6 months, 42% at 1·5 years, and 45% at 3·5 years.
Time-varying covariates were used to account for all changes in medication use.
Table 3: Patients’ characteristics categorised by baseline consumption of a Mediterranean-type diet in
patients with recent myocardial infarction
multivariable-adjusted HR for cooked vegetables was 0·65
(95% CI 0·43–0·99); for raw vegetables 1·03 (0·77–1·38);
for fruit 0·82 (0·60–1·11); for fish 0·81 (0·63–1·05); for
olive oil 0·78 (0·51–1·21).
671
Articles
Dietary score
p
0–5
6–7
8
9–10
11–15
Diabetes
Person-years
1423
4530
4289
10 264
6289
Number of cases
83
197
161
378
179
Incidence rate per 1000 person-years
58
43
38
37
28
HR (95% CI)
Adjusted for age, sex, and smoking
1·0
0·81 (0·63–1·05)
0·74 (0·57–0·97)
0·76 (0·60–0·97)
0·63 (0·48–0·82)
0·001
Multivariable*
1·0
0·81 (0·63–1·05)
0·74 (0·57–0·97)
0·77 (0·60–0·98)
0·65 (0·49–0·85)
0·004
Diabetes and IFG
Person-years
1150
3552
3288
7640
4852
Number of cases
222
490
438
905
459
Incidence rate per 1000 person-years
193
138
133
118
95
HR (95% CI)
Adjusted for age, sex, and smoking
1·0
0·77 (0·66–0·91)
0·79 (0·67–0·93)
0·74 (0·64–0·86)
0·64 (0·54–0·75)
<0·001
Multivariable*
1·0
0·78 (0·67–0·92)
0·80 (0·68–0·94)
0·75 (0·65–0·88)
0·66 (0·56–0·78)
<0·001
*Adjusted for age, sex, smoking (current, former, never), time from myocardial infarction to enrolment, treatment assignment (four categories), BMI (kg/m²), maximum
exercise tolerance during stress testing (quintiles), ischaemia during stress testing (present, absent, not done), New York Heart Association heart failure symptoms (none,
class 1, class 2–3), Canadian Cardiovascular Society angina symptoms (none, class 1, class 2–4), history of hypertension (yes/no), prior myocardial infarction previous to index
myocardial infarction (yes/no), angiotensin-converting-enzyme inhibitor use (yes/no), β-blocker use (yes/no), diuretic use (yes/no), lipid-lowering medication use (yes/no),
and consumption of cheese, wine, and coffee (each in four categories).
Table 4: Risk of diabetes or diabetes and IFG according to consumption of a Mediterranean-type diet in patients with recent myocardial infarction
Comparing multivariable models with and without
adjustment for potential mediators (BMI gain, serum
lipid levels, fibrinogen and leucocyte count, intake of
butter or oils other than olive oil), the risk associated with
the Mediterranean-type diet was about one-fifth lower
including these factors (28% risk reduction) than not
including them (35% risk reduction; table 4), suggesting
that these factors might mediate about one-fifth of the
observed potential effect. Based on the changes in the
risk estimates, the most important potential mediators
seemed to be better HDL and triglyceride levels and
lower consumption of butter or oils other than olive oil.
Consumption of traditional Mediterranean dietary
components was also associated with lower incidence of
diabetes and IFG (p for trend <0·001) (table 4). The
magnitude (34%) of the lower incidence was similar to
the lower risk (35%) seen for incidence of diabetes alone.
Investigating potential mediators, differences in weight
gain, serum lipid levels, inflammatory markers, and
intake of butter or oils other than olive oil seemed to
mediate about 25% of the association. Based on changes
in the risk estimates, the most important factor in
preventing the development of diabetes and IFG seemed
to be lower consumption of butter or oils other than olive
oil. Findings were similar using the lower IFG cutpoint
(5·6 mmol/L). A Mediterranean-type diet was associated
with lower incidence of diabetes and IFG100 (p<0·001),
with 28% lower risk comparing the highest to the lowest
Mediterranean dietary scores (multivariable HR 0·72
[95% CI 0·63–0·82], table 4). Comparing absolute rates,
individuals with the highest Mediterranean-type diet
scores (11–15) developed far fewer cases of diabetes or
672
IFG (incidence 247 cases per 1000 person-years)than did
individuals with the lowest scores (0–5) (incidence
458 cases per 1000 person-years).
Discussion
These results show that the incidence of prediabetes
(IFG) and diabetes is high in patients with recent acute
myocardial infarction free of glucose abnormalities in
the peri-myocardial infarction period. One-third of
patients developed new diabetes or IFG (≥6·1 mmol/L)
during 3·5 years’ follow-up, a proportion that rose to
nearly two-thirds when the lower IFG cutpoint
(5·6 mmol/L) was used. This finding was not due to
undiagnosed prevalent disease at the time of myocardial
infarction, since we measured baseline fasting glucose to
exclude prevalent cases. These findings indicate that, just
as diabetes can be considered a coronary heart disease
risk-equivalent,31 acute myocardial infarction should
potentially be considered a prediabetes risk-equivalent.
Two previous studies investigated the cross-sectional
prevalence of impaired glucose tolerance in patients with
acute coronary syndromes or recent myocardial
infarction,7,8 but these represented the burden of
undiagnosed (predisposing) disease at the time of
myocardial infarction, not the development of new cases
in ensuing years. Of 2499 patients with stable angina or
remote (>6 months previously) myocardial infarction,32
22% developed new-onset diabetes or IFG (≥6·1 mmol/L)
during 6 years’ follow-up, an incidence rate of about
4·1% per year.
In comparison, in our study, 33% developed diabetes
or IFG during 3·5 years’ follow-up, an incidence rate of
www.thelancet.com Vol 370 August 25, 2007
Articles
12·3% per year. This higher incidence suggests that,
compared with patients with stable angina or remote
myocardial infarction, patients with recent myocardial
infarction have a stronger propensity towards prediabetes
or diabetes. Compared to the more general population,
the contrast is even greater. In population-based cohorts
of middle-aged white adults followed-up in roughly the
same years as our study, annual incidence rates of
diabetes ranged from 0·8–1·6%33–35 (vs 3·7% in our study)
and the annual incidence rate of IFG alone (5·6–7 mmol/L)
was 1·8%34 (vs 27·5% in our study). The high risk of
incident IFG and diabetes in patients who had a
myocardial infarction has clinical implications, indicating
the importance of both surveillance and prevention of
prediabetes and diabetes in these individuals. Regular
screening might be appropriate for all such patients,
using fasting glucose measurements, possibly along with
other screening tests such as oral glucose tolerance
testing.
These results also highlight the need for investigation
of mechanisms and pathways that might account for this
risk. It seems less likely that acute myocardial infarction
itself would increase the risk of subsequent diabetes. We
postulate that high rates of diabetes and IFG in the years
after myocardial infarction are at least partly due to
shared pathways between risk factors for acute plaque
rupture and a propensity toward metabolic dysfunction.
In our analysis, independent risk factors for diabetes
included BMI, hypertension, current smoking, and in
analyses of potential mediators, greater BMI gain, higher
triglycerides, lower HDL cholesterol, and higher leucocyte
count. Each of these could be markers of or risk factors
for both plaque instability and metabolic dysfunction.
Medications commonly used to treat myocardial
infarction patients (such as β blockers, lipid-lowering
drugs, and diuretics) were also independently associated
with diabetes incidence. These findings were consistent
with earlier reports that specific drugs might affect
insulin sensitivity and diabetes risk.36–39 The inverse
association with lipid-lowering agents could also reflect
selections of patients prescribed such medications,
generally those with high LDL (and lower risk of IFG or
diabetes), compared with those with low HDL or high
triglycerides (at higher risk of IFG or diabetes).
ACE-inhibitor use was not associated with diabetes risk
in this analysis, which could be due to confounding by
indication—ie, prescription of ACE-inhibitors to
higher-risk patients. Additional attention to the possible
effects of these medications on risk of IFG, as well as
diabetes, is needed.
Patients with both coronary heart disease and diabetes
have significantly worse outcomes than people with only
one of these conditions.1–5 In this study, patients who
have had a myocardial infarction and then developed
diabetes had a higher risk of adverse events than those
who maintained normal fasting glucose, even though the
follow-up period was relatively short and measured only
www.thelancet.com Vol 370 August 25, 2007
the initial impact of new onset—and largely subclinical—
IFG and diabetes. Given these adverse outcomes,
identification of modifiable risk factors to prevent
development of IFG and diabetes in patients who have
had a myocardial infarction is imperative.
Addressing lifestyle behaviours can be particularly
important in preventing disease. The high risks we
noted from both high baseline BMI and BMI gain are
warnings that the growing obesity pandemic will
continue to offset gains in treatment of cardiac
patients.39 The 60% higher risk associated with current
smoking also provides another powerful motivation for
providers to emphasise smoking cessation and for
patients to heed their advice. Although we did not have
information on physical activity, inability to perform
exercise testing was associated with more than two-fold
higher incidence of diabetes and IFG, and it is clear
that greater physical activity, combined with modest
weight loss, lowers incidence of diabetes.39 Coffee intake
was not linked with risk of diabetes or IFG, possibly
due to misclassification on the limited diet
questionnaire. In additional exploratory analyses,
compared with never or rare consumption, any coffee
intake was associated with a trend towards lower
incidence of diabetes and IFG100 (multivariable
HR=0·92, 95% CI 0·83–1·02).
The lower risk associated with a Mediterranean-type
diet suggests that diet could help reduce incidence of
prediabetes and diabetes after a myocardial infarction.
Many,11–24 though not all,25 trials have indicated that a
Mediterranean-type diet lowers risk factors linked to
insulin resistance and diabetes, including serum
triglycerides, HDL cholesterol, systemic inflammation,
endothelial function, and insulin sensitivity. These
physiological effects in short-term randomised trials
provide biological plausibility for the inverse association
between consumption of a Mediterranean-type diet and
incidence of IFG and diabetes in this study.
Also consistent with these trials, the lower incidence of
diabetes and IFG seemed partly related to differences in
serum triglycerides, HDL cholesterol, and inflammatory
marker levels. Typical components of a Mediterranean-type
diet—eg, olive oil, fruits, vegetables, whole grains,
legumes, nuts, and fish, have each been linked to
reductions in risk factors, but it is unclear whether one
specific food, or a combination of foods, is most
important. The results of one study suggest that different
combinations of these foods may be similarly beneficial.24
Our findings indicate that a combination of such foods
could more strongly affect risk or better identify
individuals consuming a Mediterranean-type diet than
any single food. Assignment to fish oil treatment did not
reduce IFG or diabetes risk, suggesting that findings
were not related to n-3 fatty acids in fish (but other
factors, such as protein and selenium, are present in fish
but not fish oil). The beneficial effects of a
Mediterranean-type diet could also be related to what is
673
Articles
not consumed: fewer processed meats, other processed
foods, refined grains, and trans fats. Consistent with this,
the lower incidence of diabetes and IFG seemed partly
related to lower consumption of butter or oils other than
olive oil. Notably, however, modern diets in Mediterranean
regions are increasingly inconsistent with traditional
Mediterranean-type diets.40
Our analysis had several strengths. Participants were
enrolled and followed-up by general cardiology
practitioners, representing a real-world cross-section of
patients who have had a myocardial infarction. Fasting
glucose levels were measured at baseline and at multiple
times during follow-up, enabling accurate diagnosis of
prevalent diabetes or IFG at baseline and development of
diabetes or IFG during follow-up. Data for exposures, risk
factors, and events were obtained prospectively, which
kept selection or recall bias to a minimum. Also, few
patients were lost to follow-up, thus minimising bias from
differential censoring. Data on risk factors, dietary habits,
and potential mediators were obtained prospectively at
multiple time points during follow-up, allowing
time-varying multivariable adjustment. The large numbers
of new cases of diabetes and IFG provided substantial
statistical power.
There were also several potential limitations. The use of
oral glucose tolerance tests would probably have captured
additional new cases of diabetes or IFG; thus, the true
incidence of abnormal insulin-glucose homoeostasis was
probably underestimated. Detailed information on
medications, such as type of diuretics, was not available.
Misclassification of both clinical and lifestyle exposures
could have occurred due to measurement error or
biological variation; because data were collected
prospectively, such errors would probably be random with
respect to the outcomes and cause underestimation of the
observed associations.
Lifestyle habits could have been partially driven by
changes in response to chronic disease development and
physician guidance (eg, individuals consuming a
Mediterranean-type diet were more likely to have a history
of hypertension and higher BMI at baseline); because
such individuals might be more likely to develop diabetes
or IFG, this would also drive the observed beneficial
associations toward the null. We did not have information
on physical activity, although we adjusted for exercise
tolerance during stress testing, which is a well-validated
measure of underlying physical fitness. The serum lipid
and inflammatory measures might not have perfectly
captured the activity of these physiological pathways,
underestimating the degree to which such pathways
mediated risk. Although we adjusted for a variety of risk
factors, residual confounding by unmeasured or
incompletely measured factors cannot be excluded. The
results were seen in Italian patients with recent myocardial
infarction and might not be generalisable to other
populations. For example, the range of dietary differences
in an Italian population may be less than in other
674
populations, which would cause an underestimation of
the magnitude of benefit of a Mediterranean-type diet
compared with other Western diets.
Our findings indicate that incidence of IFG and diabetes
is high in the years after myocardial infarction, suggesting
that acute myocardial infarction could be a prediabetes
risk-equivalent. The results highlight the need to screen
such patients and investigate potential pathways (eg,
shared risk factors, medication use, lifestyle behaviours)
that might mediate this relation. Our findings also suggest
that smoking cessation, prevention of weight gain, and
consumption of typical Mediterranean foods could
substantially lower this risk, which has important
implications for counselling patients soon after they have
a myocardial infarction—an opportune time to institute
lifestyle changes in patients motivated by a life-changing
event.
Contributors
DM participated in the conception and design, statistical analysis and
data interpretation, manuscript drafting, critical revision of the
manuscript for important intellectual content, and approval of the final
manuscript for submission. R Marfisi and R Marchioli participated in
the conception and design, data collection, statistical analysis and data
interpretation, critical revision of the manuscript for important
intellectual content, and approval of the final manuscript for
submission. GL, MS, LT, GT, and FV participated in the data collection,
critical revision of the manuscript for important intellectual content, and
(except for FV) approval of the final manuscript for submission.
Conflict of interest statement
We declare that we have no conflict of interest.
Acknowledgments
We thank the GISSI-Prevenzione Investigators and the patients who
participated in the trial. This analysis was supported by the National
Heart, Lung, and Blood Institute, National Institutes of Health
(K08-HL-075628). The authors warmly remember the many
contributions and kind friendship of Dr Valagussa (the
GISSI-Prevenzione Chairman), who passed away in December 2006.
References
1
Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Mortality
from coronary heart disease in subjects with type 2 diabetes and in
nondiabetic subjects with and without prior myocardial infarction.
N Engl J Med 1998; 339: 229–34.
2
Mukamal KJ, Nesto RW, Cohen MC, et al. Impact of diabetes on
long-term survival after acute myocardial infarction: comparability
of risk with prior myocardial infarction. Diabetes Care 2001; 24:
1422–27.
3
Hu FB, Stampfer MJ, Solomon CG, et al. The impact of diabetes
mellitus on mortality from all causes and coronary heart disease in
women: 20 years of follow-up. Arch Intern Med 2001; 161: 1717–23.
4
Lotufo PA, Gaziano JM, Chae CU, et al. Diabetes and all-cause and
coronary heart disease mortality among US male physicians.
Arch Intern Med 2001; 161: 242–47.
5
Cho E, Rimm EB, Stampfer MJ, Willett WC, Hu FB. The impact of
diabetes mellitus and prior myocardial infarction on mortality from
all causes and from coronary heart disease in men.
J Am Coll Cardiol 2002; 40: 954–60.
6
American Diabetes Association. The link between diabetes and
cardiovascular disease. Available at: http://www.diabetes.org/
uedocuments/Factsheetthelink2.pdf (accessed Feb 8, 2006).
7
Norhammar A, Tenerz A, Nilsson G, et al. Glucose metabolism in
patients with acute myocardial infarction and no previous diagnosis
of diabetes mellitus: a prospective study. Lancet 2002; 359: 2140–44.
8
Hashimoto K, Ikewaki K, Yagi H, et al. Glucose intolerance is
common in Japanese patients with acute coronary syndrome who
were not previously diagnosed with diabetes. Diabetes Care 2005; 28:
1182–86.
www.thelancet.com Vol 370 August 25, 2007
Articles
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto
miocardico (GISSI). Dietary supplementation with n-3
polyunsaturated fatty acids and vitamin E after myocardial
infarction: results of the GISSI-Prevenzione trial. Lancet 1999; 354:
447–55.
Marchioli R, Barzi F, Bomba E, et al. Early protection against
sudden death by n-3 polyunsaturated fatty acids after myocardial
infarction: time-course analysis of the results of the Gruppo Italiano
per lo Studio della Sopravvivenza nell’Infarto Miocardico
(GISSI)-Prevenzione. Circulation 2002; 105: 1897–903.
de Lorgeril M, Renaud S, Mamelle N, et al. Mediterranean
alpha-linolenic acid-rich diet in secondary prevention of coronary
heart disease. Lancet 1994; 343: 1454–59.
Ryan M, McInerney D, Owens D, Collins P, Johnson A,
Tomkin GH. Diabetes and the Mediterranean diet: a beneficial
effect of oleic acid on insulin sensitivity, adipocyte glucose transport
and endothelium-dependent vasoreactivity. Q JM 2000; 93: 85–91.
Fuentes F, Lopez-Miranda J, Sanchez E, et al. Mediterranean and
low-fat diets improve endothelial function in hypercholesterolemic
men. Ann Intern Med 2001; 134: 1115–19.
Perez-Jimenez F, Lopez-Miranda J, Pinillos MD, et al. A
Mediterranean and a high-carbohydrate diet improve glucose
metabolism in healthy young persons. Diabetologia 2001; 44:
2038–43.
Jula A, Marniemi J, Huupponen R, Virtanen A, Rastas M,
Ronnemaa T. Effects of diet and simvastatin on serum lipids,
insulin, and antioxidants in hypercholesterolemic men: a
randomized controlled trial. JAMA 2002; 287: 598–605.
Bemelmans WJ, Broer J, Feskens EJ, et al. Effect of an increased
intake of alpha-linolenic acid and group nutritional education on
cardiovascular risk factors: the Mediterranean Alpha-linolenic
Enriched Groningen Dietary Intervention (MARGARIN) study.
Am J Clin Nutr 2002; 75: 221–27.
Sondergaard E, Moller JE, Egstrup K. Effect of dietary intervention
and lipid-lowering treatment on brachial vasoreactivity in patients
with ischemic heart disease and hypercholesterolemia. Am Heart J
2003; 145: E19.
Ambring A, Friberg P, Axelsen M, et al. Effects of a
Mediterranean-inspired diet on blood lipids, vascular function and
oxidative stress in healthy subjects. Clin Sci (Lond) 2004; 106:
519–25.
Esposito K, Marfella R, Ciotola M, et al. Effect of a
mediterranean-style diet on endothelial dysfunction and markers of
vascular inflammation in the metabolic syndrome: a randomized
trial. JAMA 2004; 292: 1440–46.
Rodriguez-Villar C, Perez-Heras A, Mercade I, Casals E, Ros E.
Comparison of a high-carbohydrate and a high-monounsaturated
fat, olive oil-rich diet on the susceptibility of LDL to oxidative
modification in subjects with Type 2 diabetes mellitus. Diabet Med
2004; 21: 142–49.
Fito M, Cladellas M, de la Torre R, et al. Antioxidant effect of virgin
olive oil in patients with stable coronary heart disease: a
randomized, crossover, controlled, clinical trial. Atherosclerosis 2005;
181: 149–58.
Vincent-Baudry S, Defoort C, Gerber M, et al. The Medi-RIVAGE
study: reduction of cardiovascular disease risk factors after a 3-mo
intervention with a Mediterranean-type diet or a low-fat diet.
Am J Clin Nutr 2005; 82: 964–71.
Lapointe A, Couillard C, Lemieux S. Effects of dietary factors on
oxidation of low-density lipoprotein particles. J Nutr Biochem 2006;
17: 645–58.
www.thelancet.com Vol 370 August 25, 2007
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Estruch R, Martinez-Gonzalez MA, Corella D, et al. Effects of a
Mediterranean-style diet on cardiovascular risk factors: a
randomized trial. Ann Intern Med 2006; 145: 1–11.
Michalsen A, Lehmann N, Pithan C, et al. Mediterranean diet has
no effect on markers of inflammation and metabolic risk factors in
patients with coronary artery disease. Eur J Clin Nutr 2006; 60:
478–85.
Barzi F, Woodward M, Marfisi RM, Tavazzi L, Valagussa F,
Marchioli R. Mediterranean diet and all-causes mortality after
myocardial infarction: results from the GISSI-Prevenzione trial.
Eur J Clin Nutr 2003; 57: 604–11.
Mozaffarian D, Ascherio A, Hu FB, et al. Interplay between
different polyunsaturated fatty acids and risk of coronary heart
disease in men. Circulation 2005; 111: 157–64.
Genuth S, Alberti KG, Bennett P, et al. Follow-up report on the
diagnosis of diabetes mellitus. Diabetes Care 2003; 26: 3160–67.
GISSI Prevenzione Investigators. Results of the low-dose (20 mg)
pravastatin GISSI Prevenzione trial in 4271 patients with recent
myocardial infarction: do stopped trials contribute to overall
knowledge? GISSI Prevenzione Investigators (Gruppo Italiano per
lo Studio della Sopravvivenza nell’Infarto Miocardico). Ital Heart J
2000; 1: 810–20.
Sofi F, Vecchio S, Giuliani G, et al. Dietary habits, lifestyle and
cardiovascular risk factors in a clinically healthy Italian population:
the ‘Florence’ diet is not Mediterranean. Eur J Clin Nutr 2005; 59:
584–91.
Executive summary of The third report of the National Cholesterol
Education Program (NCEP) expert panel on detection, evaluation,
and treatment of high blood cholesterol in adults (Adult Treatment
Panel III). JAMA 2001; 285: 2486–97.
Arcavi L, Behar S, Caspi A, Reshef N, Boyko V, Knobler H. High
fasting glucose levels as a predictor of worse clinical outcome in
patients with coronary artery disease: results from the Bezafibrate
Infarction Prevention (BIP) study. Am Heart J 2004; 147: 239–45.
Brancati FL, Kao WH, Folsom AR, Watson RL, Szklo M. Incident
type 2 diabetes mellitus in African American and white adults: the
Atherosclerosis Risk in Communities Study. JAMA 2000; 283:
2253–59.
Bonora E, Kiechl S, Willeit J, et al. Population-based incidence rates
and risk factors for type 2 diabetes in white individuals: the Bruneck
study. Diabetes 2004; 53: 1782–89.
Cugati S, Wang JJ, Rochtchina E, Mitchell P. Ten-year incidence of
diabetes in older Australians: the Blue Mountains Eye Study.
Med J Aust 2007; 186: 131–35.
Stas S, Appesh L, Sowers J. Metabolic safety of antihypertensive
drugs: myth versus reality. Curr Hypertens Rep 2006; 8: 403–08.
Sarafidis PA, McFarlane SI, Bakris GL. Antihypertensive agents,
insulin sensitivity, and new-onset diabetes. Curr Diab Rep 2007; 7:
191–99.
Padwal R, Majumdar SR, Johnson JA, Varney J, McAlister FA.
A systematic review of drug therapy to delay or prevent type 2
diabetes. Diabetes Care 2005; 28: 736–44.
Ford ES, Ajani UA, Croft JB, et al. Explaining the decrease in U.S.
deaths from coronary disease, 1980–2000. N Engl J Med 2007; 356:
2388–98.
Ratner R, Goldberg R, Haffner S, et al. Impact of intensive lifestyle
and metformin therapy on cardiovascular disease risk factors in the
diabetes prevention program. Diabetes Care 2005; 28: 888–94.
675
Articles
Norepinephrine plus dobutamine versus epinephrine alone
for management of septic shock: a randomised trial
Djillali Annane, Philippe Vignon, Alain Renault, Pierre-Edouard Bollaert, Claire Charpentier, Claude Martin, Gilles Troché, Jean-Damien Ricard,
Gérard Nitenberg, Laurent Papazian, Elie Azoulay, Eric Bellissant, for the CATS Study Group*
Summary
Lancet 2007; 370: 676–84
See Comment page 636
*Listed at end of report
Raymond Poincaré Hospital
(AP-HP), University of
Versailles Saint Quentin, PRES
UniverSud, Paris, France
(Prof D Annanne MD); Service
de Réanimation Polyvalente,
Hôpital Dupuytren, Limoges,
France (Prof P Vignon MD);
Centre d’Investigation Clinique
INSERM 0203, Unité de
Biométrie, Université de
Rennes 1, Rennes, France
(A Renault PhD); Service de
Réanimation Médicale
(Prof P-E Bollaert MD) and
Service de Réanimation
Chirurgicale (C Charpentier MD),
Hôpital Central, Nancy,
France; Département
d’Anesthésie-Réanimation,
CHU Nord (AP-HM), Marseille,
France (Prof C Martin MD);
Service de Réanimation, CH,
Versailles, France
(G Troché MD); Service de
Réanimation Médicale, Hôpital
Louis Mourier, Colombes,
France (Prof J-D Ricard MD);
Service de Réanimation,
Institut Gustave Roussy,
Villejuif, France
(G Nitenberg MD); Service de
Réanimation Médicale, CHU
Sainte-Marguerite (AP-HM),
Marseille, France
(Prof L Papazian MD); Service de
Réanimation Médicale, Hôpital
Saint-Louis (AP-HP), Paris,
France (Prof E Azoulay MD); and
Centre d’Investigation Clinique
INSERM 0203, Unité de
Pharmacologie Clinique,
Hôpital de Pontchaillou,
Université de Rennes 1, Rennes,
France (Prof E Bellissant MD)
Correspondence to:
Prof Djillali Annane, Raymond
Poincaré Hospital (AP-HP),
University of Versailles Saint
Quentin, PRES UniverSud, Paris,
France
[email protected]
676
Background International guidelines for management of septic shock recommend that dopamine or norepinephrine
are preferable to epinephrine. However, no large comparative trial has yet been done. We aimed to compare the
efficacy and safety of norepinephrine plus dobutamine (whenever needed) with those of epinephrine alone in septic
shock.
Methods This prospective, multicentre, randomised, double-blind study was done in 330 patients with septic shock
admitted to one of 19 participating intensive care units in France. Participants were assigned to receive epinephrine
(n=161) or norepinephrine plus dobutamine (n=169), which were titrated to maintain mean blood pressure at
70 mm Hg or more. The primary outcome was 28-day all-cause mortality. Analyses were by intention to treat. This
trial is registered with ClinicalTrials.gov, number NCT00148278.
Findings There were no patients lost to follow-up; one patient withdrew consent after 3 days. At day 28, there were
64 (40%) deaths in the epinephrine group and 58 (34%) deaths in the norepinephrine plus dobutamine group (p=0·31;
relative risk 0·86, 95% CI 0·65–1·14). There was no significant difference between the two groups in mortality rates
at discharge from intensive care (75 [47%] deaths vs 75 [44%] deaths, p=0·69), at hospital discharge (84 [52%] vs
82 [49%], p=0·51), and by day 90 (84 [52%] vs 85 [50%], p=0·73), time to haemodynamic success (log-rank p=0·67),
time to vasopressor withdrawal (log-rank p=0·09), and time course of SOFA score. Rates of serious adverse events
were also similar.
Interpretation There is no evidence for a difference in efficacy and safety between epinephrine alone and norepinephrine
plus dobutamine for the management of septic shock.
Introduction
Sepsis places a huge burden on health-care systems. In the
USA, the annualised increase in the incidence of sepsis is
estimated to be about 9% and its associated mortality is
about 18%.1 Septic shock, the most severe form of sepsis,
accounts for about 9% of admissions to intensive care
units, and its short-term mortality ranges between 40%
and 60%.2 Septic shock is commonly defined by the need
for vasopressors to reverse sepsis-induced hypotension.3 At
the time this study was designed, guidelines from the
French Society of Intensive Care Medicine recommended
the use of dopamine as first-line treatment for septic shock
and norepinephrine plus dobutamine (in patients with low
cardiac output despite adequate fluid resuscitation) or
epinephrine alone in dopamine-resistant shock.4 Both
strategies enable induction of vascular and cardiac effects
but the combination of norepinephrine and dobutamine
has the theoretical advantage over epinephrine in allowing
a precise modulation of these two types of effect. More
recent international guidelines recommend dopamine or
norepinephrine as first-line drugs for the management of
septic shock and epinephrine in patients who respond
poorly to dopamine or norepinephrine.5,6 Indeed, when
compared with norepinephrine in small randomised trials,
epinephrine has shown deleterious effects on splanchnic
blood flow7–10 and on acid–base balance.7,11–13 However, these
adverse effects were transient,7 and a recent systematic
review on vasopressor therapy for management of septic
shock concluded that there was no evidence for any
difference on short-term mortality between epinephrine
and norepinephrine.14 However, there were few patients
included in that review.
The question of an advantage of norepinephrine plus
dobutamine (whenever needed) over epinephrine alone
thus remains unanswered. To address this question, we
did a large multicentre randomised controlled trial to
assess and compare the efficacy and safety of
norepinephrine plus dobutamine with those of epinephrine
alone in the treatment of septic shock.
Methods
Patients
Patients over the age of 18 years admitted to participating
intensive care units between Oct 12, 1999, and Dec 31,
2004, were eligible for assessment. The inclusion criteria
were the presence, for less than 7 days, of: evidence of
infection; at least two of the four criteria for systemic
inflammatory response syndrome (temperature above 38°C
or below 36°C, heart rate above 90 bpm, respiratory rate
above 20 cycles per min and arterial CO2 tension below
32 mm Hg or need for mechanical ventilation, polymorphonuclear neutrophil count above 12×10⁹ cells per L
or below 4×10⁹ cells per L); and at least two signs of tissue
hypoperfusion or organ dysfunction. These signs were
www.thelancet.com Vol 370 August 25, 2007
Articles
Pulmonary capillary wedge pressure
≤10 mm Hg or hypovolaemia
at echocardiography
Fluid challenge 15–20 mL/kg of
colloids or crystalloids
<70 mm Hg
Cardiac index >2·5 L/min per m²
↑ Norepinephrine or epinephrine
by 0·2 μg/kg per min
Mean
blood
pressure
Pulmonary capillary wedge pressure
>10 mm Hg or no sign of hypovolaemia
at echocardiography
SWITCH TO
EPINEPHRINE/NOREPINEPHRINE
IF DOBUTAMINE/PLACEBO FAILED
OR VICE VERSA
Norepinephrine 0·2 μg/kg per min*
±dobutamine 5 μg/kg per min*
RANDOMISATION
Mean
blood
pressure
Epinephrine 0·2 μg/kg per min*
±placebo
Cardiac index ≤2·5 L/min per m²
↑ Dobutamine or placebo
by 5 μg/kg per min
↓ Norepinephrine or epinephrine
by 0·2 μg/kg per min
↓ Dobutamine or placebo by
5 μg/kg per min
≥70 mm Hg
Figure 1: Treatment algorithm
*Starting dose.
1591 assessed for eligibility
defined as a ratio of arterial oxygen tension over inspired
fraction of oxygen of less than 280 mm Hg (if patient was
mechanically ventilated), urinary output below 0·5 mL
per kg of bodyweight per h or below 30 mL/h (for at least
1 h), or arterial lactate concentration above 2 mmol/L,
platelet count below 100×10⁹ cells per L. Additionally,
patients had to meet the three following criteria for less
than 24 h: systolic blood pressure below 90 mm Hg or
mean blood pressure below 70 mm Hg; administration of
fluid bolus of at least 1000 mL or capillary wedge pressure
between 12 and 18 mm Hg; and need for more than 15 µg
per kg of bodyweight per min of dopamine or any dose of
epinephrine or norepinephrine. Reasons for exclusion
were pregnancy; evidence of obstructive cardiomyopathy,
acute myocardial ischaemia, or pulmonary embolism;
advanced stage cancer, malignant haemopathy, or AIDS
with a decision to withhold or withdraw aggressive
therapies; persistent (longer than a week) polymorphonuclear neutrophil count of less than 0·5×10⁹ cells per L;
and inclusion in another clinical trial.
The protocol was approved by the ethics committee of
the French Society of Intensive Care and by the Consultative
Committee for the Protection of People in Biomedical
Research of Saint-Germain en Laye, France. Written
informed consent was obtained from the patients
themselves or their closest relatives.
Procedures
In this prospective, multicentre, randomised,
double-blind study, eligible patients were randomly
assigned, in a 1:1 ratio, to receive either epinephrine
www.thelancet.com Vol 370 August 25, 2007
330 randomised
1261 excluded
674 did not meet inclusion criteria
178 met exclusion criteria
409 other reasons
283 consent refusal
98 physician refusal or missing
161 assigned epinephrine
156 received epinephrine
5 did not receive epinephrine
0 early death
5 other reasons:
3 consent withdrawal
2 problem in drug supply
169 assigned norepinephrine
166 received norepinephrine
3 did not receive norepinephrine
1 early death
2 other reasons:
1 consent withdrawal
1 problem in drug supply
0 discontinued intervention
0 lost to follow-up
1 discontinued intervention
1 withdrew consent at day 3
0 lost to follow-up
161 analysed
161 main endpoint (28 day)
161 90 day follow-up
169 analysed
169 main endpoint (28 day)
169 90 day follow-up
Figure 2: Trial profile
alone or norepinephrine plus dobutamine (whenever
needed) according to a computer-generated random list.
Randomisation was done centrally by an independent
statistician to ensure appropriate concealment, was
stratified by centre, and equilibrated by blocks of six. To
ensure masking of treatment allocation, patients
677
Articles
Overall (n=330)
Age (years)
Sex (M/F)
63 (50–73)
202 (61%)/128 (39%)
Epinephrine (n=161)
Norepinephrine plus dobutamine
(n=169)
65 (53–75)
60 (47–72)
103 (64%)/58 (36%)
99 (59%)/70 (41%)
McCabe classification
0: no fatal underlying disease
196 (59%)
98 (61%)
98 (58%)
1: life expectancy ≤5 years
101 (31%)
45 (28%)
56 (33%)
2: life expectancy <1 year
33 (10%)
18 (11%)
15 (9%)
Patient’s location before ICU admission
Home
93 (28%)
43 (27%)
50 (30%)
Ward
190 (58%)
95 (59%)
95 (56%)
38 (12%)
17 (11%)
21 (12%)
9 (3%)
6 (4%)
3 (2%)
2 (2–4)
2 (2–3)
2 (2–4)
53 (40–65)
54 (42–67)
52 (38–64)
Other ICU
Long-term care facility
ICU admission to randomisation delay (days)
SAPS II at admission
SIRS criteria
Temperature (°C)
38·4 (36·8–39·1) (n=329)
38·4 (37·0– 39·1) (n=160)
38·4 (36·8–39·2) (n=169)
Heart rate (bpm)
114 (99–130)
112 (98–124)
118 (100–133)
Mechanically ventilated (yes)
312 (95%)
153 (95%)
159 (94%)
Leucocyte count (×10⁹/L)
12·9 (7·6–20·7)
12·7 (7·3–21·7)
12·9 (7·7–20·2)
Tissue hypoperfusion/organ dysfunction
PaO2/FiO2 (mm Hg)
151 (102–220) (n=316)
156 (96–225) (n=153)
150 (103–210) (n=163)
Urinary output (mL/24 h)
600 (200–1195) (n=328)
520 (200–1150) (n=159)
670 (210–1250) (n=169)
Lactate (mmol/L)
Platelet counts (×10⁹/L)
3·2 (1·9–5·1) (n=319)
176 (93–267)
2·9 (1·7–5·0) (n=155)
193 (92–275)
3·3 (2·1–5·1) (n=164)
167 (99–236)
Haematocrit (%)
31·5 (6·2) (n=326)
32·4 (6·1) (n=158)
Glasgow Coma Score
14 (7–15)
13 (7–15)
30·8 (6·1) (n=168)
14 (8–15)
SOFA score
11 (9–14)
11 (9–13)
11 (9–14)
100 (82–124)
100 (82–124)
100 (84–120)
Shock criteria
Systolic blood pressure (mm Hg)
Mean arterial blood pressure (mm Hg)
Fluid loading (mL)
69 (19)
1750 (1000–3500) (n=328)
70 (19)
68 (19)
1500 (1000–3500) (n=160)
2000 (1000–3625) (n=168)
Catecholamine requirements
Dopamine >15 µg/kg per min
63 (19%)
38 (24%)
25 (15%)
Epinephrine
137 (42%)
61 (38%)
76 (45%)
Norepinephrine
102 (31%)
48 (30%)
54 (32%)
Dopamine and epinephrine
11 (3%)
6 (4%)
5 (3%)
Dopamine and norepinephrine
11 (3%)
6 (4%)
5 (3%)
6 (2%)
2 (1%)
4 (2%)
Adequate initial antibiotics
250 (76%)
119 (74%)
131 (78%)
Renal replacement therapy
31 (9%)
15 (9%)
16 (10%)
39 (23%)
Epinephrine and norepinephrine
Specific concomitant therapies
Corticosteroids
None
67 (20%)
28 (17%)
Hydrocortisone alone
148 (45%)
73 (45%)
75 (44%)
Hydrocortisone and fludrocortisone
115 (35%)
60 (37%)
55 (33%)
119
57
62
11 (19%)
14 (23%)
Eligible for activated protein C
Treated with activated protein C
25 (21%)
Data are mean (SD) or median (IQR) for continuous variables (samples sizes are also reported in case of missing data) and n (%) for categorical variables. ICU=intensive care
unit. PaO2/FiO2=ratio of arterial oxygen tension to inspired fraction of oxygen.
Table 1: General characteristics at randomisation
randomly assigned to the epinephrine group were given
epinephrine plus a placebo in place of dobutamine. Study
678
treatments were provided by the pharmacist at each site
as identical syringes for norepinephrine and epinephrine
www.thelancet.com Vol 370 August 25, 2007
Articles
(labelled in blue) and for dobutamine and its placebo
(labelled in white). Treatments were titrated according to
an algorithm designed to maintain mean blood pressure
of 70 mm Hg or more (figure 1). Adherence to this
algorithm was checked at each meeting of investigators.
Haemodynamic assessment required systematic
continuous invasive monitoring of mean arterial
blood pressure and central venous pressure, together
with systematic assessment of hypovolaemia and
cardiac index. Assessment of cardiac index was done in
accordance with the standard of care of each
participating intensive care unit—ie, by right heart
catheterisation, echocardiography doppler, pulse
contour cardiac output, or oesophageal doppler.
Allocation of study treatments was concealed from
patients, primary investigators, and all co-investigators
until the release of the final statistical analysis (June 29,
2005). Concomitant treatments were left to the
discretion of patients’ primary physicians.
The following data were recorded at baseline: general
characteristics; severity of underlying co-morbidities by
the McCabe and Jackson15 classification; severity of
illness on the SAPS II16 and SOFA17 scores; systemic
and pulmonary haemodynamics (when the patient had
a pulmonary artery catheter); arterial lactate
concentrations and blood gases; and results of standard
laboratory tests, blood cultures, and cultures of
specimen sampled in each presumed site of infection.
Haemodynamics, arterial lactate concentrations, and
blood gases were recorded twice daily and the SOFA
score was calculated once a day from randomisation to
day 28 (or to discharge from the intensive care unit, or
death, depending on which occurred first).
For the safety assessment, careful neurological and
cardiac examinations and a 12-lead electrocardiograph
were done every day. Cardiac enzymes, echocardiography,
coronary angiography, brain CT, or MRI were done
whenever indicated by physical examination. Survival
status was recorded during the 90 days after
randomisation. For patients who left the hospital before
90 days, survival status was systematically confirmed by
visit at an outpatient clinic or by telephone call.
The primary endpoint was day 28 all-cause mortality.
Secondary endpoints were survival distribution from
randomisation to day 90; mortality rates at day 7, day 14,
at discharge from intensive care and from hospital, and
at day 90; systemic haemodynamics; arterial pH and
lactate; SOFA score; time to haemodynamic success (ie,
a mean blood pressure above 70 mm Hg for at least 12 h
consecutively); and time to vasopressor withdrawal (ie,
the first interruption of study drugs for at least 24 h).
Safety was assessed daily from randomisation to
day 28 (or to discharge from intensive care unit or
death, depending on which occurred first) and mainly
focused on the occurrence of serious adverse events
such as supraventricular arrhythmias with ventricular
rate above 150 bpm, ventricular arrhythmias, acute
www.thelancet.com Vol 370 August 25, 2007
Overall
(n=330)
Epinephrine
(n=161)
Norepinephrine
plus dobutamine
(n=169)
Type of infection
Community acquired
185 (56%)
88 (55%)
97 (57%)
Hospital acquired, postoperative
57 (17%)
31 (19%)
26 (15%)
Hospital acquired, others
88 (27%)
42 (26%)
46 (27%)
Primary source of infection
Lung
155 (47%)
74 (46%)
81 (48%)
84 (25 %)
45 (28%)
39 (23%)
Primary septicaemia
67 (20%)
28 (17%)
39 (23%)
Urinary tract
40 (12%)
19 (12%)
21 (12%)
Bones/joints/soft tissues
34 (10%)
12 (8%)
22 (13%)
Mediastinum/endocarditis
Abdomen
10 (3%)
6 (4%)
4 (2%)
Central nervous system
8 (2%)
4 (3%)
4 (2%)
Catheter related
6 (2%)
4 (3%)
2 (1%)
Head and neck
2 (0·6%)
1 (0·6%)
1 (0·6%)
Others
5 (2%)
3 (2%)
2 (1%)
118 (36%)
64 (40%)
54 (32%)
Positive blood cultures
Causal microorganism
None
63 (19%)
30 (19%)
33 (20%)
One
174 (53%)
88 (55%)
86 (51%)
More than one
93 (28%)
43 (27%)
50 (30%)
Gram-positive bacteria
154 (47%)
69 (43%)
85 (50%)
Gram-negative bacteria
158 (48%)
83 (52%)
75 (44%)
28 (9%)
11 (7%)
17 (10%)
3 (1%)
2 (1%)
28 (9%)
12 (8%)
Anaerobes
Mycobacterium
Fungi
1 (0·6%)
16 (10%)
Parasite
1 (0·3%)
1 (0·6%)
0 (0%)
Virus
3 (1%)
3 (2%)
0 (0%)
Data are number of patients (%).
Table 2: Characteristics of infections
Epinephrine (n=161)
Norepinephrine plus
dobutamine (n=169)
p
At day 7
40 (25%)
34 (20%)
0·30
At day 14
56 (35%)
44 (26%)
0·08
At day 28
64 (40%)
58 (34%)
0·31
At discharge from intensive care
75 (47%)
75 (44%)
0·69
At discharge from hospital
84 (52%)
82 (49%)
0·51
At day 90
84 (52%)
85 (50%)
0·73
Data are number of deaths (%).
Table 3: All-cause mortality rates
OR (logistic regression)
HR (Cox regression)
All covariates (n=308)
0·90 (0·54–1·49); p=0·67
0·87 (0·59–1·28); p=0·47
All covariates except appropriateness of antibiotic
treatment (n=319)
0·82 (0·51–1·34); p=0·44
0·84 (0·58–1·22); p=0·36
All covariates except blood lactate concentration and 0·82 (0·51–1·31); p=0·40
appropriateness of antibiotic treatment (n=330)
0·87 (0·61–1·24); p=0·43
Data are risk estimate (95% CI); p value.
Table 4: Adjusted treatment effects on mortality rates at day 28
679
Articles
coronary events, limb ischaemia or skin necrosis, and
acute cerebrovascular events (whether haemorrhagic or
ischaemic). These events were deemed to be related to
catecholamine infusion when they occurred while patients
were receiving study drugs.
Pharmacoeconomic analysis was done on the basis of a
model that computes the medical cost of patients in
intensive care.18 This model takes into account the rate of
invasive or surgical procedures and estimates the mean
medical cost per patient in intensive care.
The data and safety monitoring board met three times
during the study to analyse the conduct of the study, the
results of interim analyses, and to review serious adverse
events. Interim analyses were done on Nov 18, 2002, and
Oct 1, 2003, after the assessment of 185 and 232 patients,
respectively. After each analysis, the independent data and
safety monitoring board advised the study chairmen to
continue the study. A diagnosis validation committee also
met three times during the study to grade, without
knowledge of treatment allocation, the patients as having
definite septic shock, probable septic shock, and probable
non-septic shock,19 and to assess the appropriateness of
antibiotic treatments.
Statistical analysis
We expected an all-cause mortality rate at day 28 of 60% in
the epinephrine group, on the basis of data from another
trial we were doing in patients with septic shock when we
planned this protocol.20 We calculated that we would need
a sample size of 340 patients to be able to detect, in a
two-sided test done with a 0·05 type I error, an absolute
reduction of 20% in the mortality rate at day 28 with
95% probability.
The two interim analyses were planned with an
O’Brien and Fleming stopping boundary.21 With this
1·0
Epinephrine
Norepinephrine plus dobutamine
0·9
Survival probability
0·8
0·7
0·6
0·5
0·4
0·3
Log-rank
χ²=0·39
p=0·53
0·2
Mortality at 3 months
Epinephrine: 84/161 (52%)
Norepinephrine plus dobutamine: 85/169 (50%)
χ²=0·12
p=0·73
0·1
0
0
10
20
30
40
50
60
70
80
90
81
91
79
85
79
84
74
84
Role of the funding source
Time (days)
Number at risk
Epinephrine
161
Norepinephrine 169
plus dobutamine
117
131
102
117
96
108
Figure 3: Survival from randomisation to day 90
680
88
98
84
92
procedure, the differences between the two groups were
considered significant if the critical Z values were higher
than 3·471, 2·454, and 2·004 at the first, second, and final
analyses, respectively (corresponding to nominal two-sided
p values of <0·0005, <0·0141, and <0·0451).
The statistical analysis, prospectively defined, was done
by intention to treat (ie, in all analyses, patients were
grouped according to their original randomised treatment)
with SAS statistical software (version 8.2; Cary, NC, USA).
For continuous variables, the means and SD or the median
(IQR), in case of significant non-normality of the
distribution, are reported. The number of patients in each
category and the corresponding percentages are given for
categorical variables.
The effects of treatments on the frequency of fatal events
(mortality rates at day 7, day 14, day 28, at discharge from
intensive care or from hospital, and day 90) were compared
between groups by χ² tests. Corresponding relative risks
(RR), together with their 95% CI, were estimated.
Cumulative event curves (censored endpoints) were
estimated with the Kaplan-Meier procedure. The effects of
treatments on these endpoints were compared between
groups with the log-rank test. For the primary endpoint,
we did additional analyses with logistic and Cox regression
models, adjusting for the main baseline factors that
predict outcomes22 (ie, McCabe and Jackson classification,
SAPS II, SOFA, arterial lactate concentrations, and
appropriateness of antibiotic treatments).23 For these
analyses, continuous variables were broken into two
classes on the basis of their median values. Odds ratios
(OR) and hazard ratios (HR), together with 95% CI, were
estimated with these models. For day 90 survival, patients
who were still alive at 90 days were treated as censored.
For time to vasopressor therapy withdrawal, among
patients who had more than one outcome event during
the 28 days from randomisation, time to first event was
used in the analyses. For this endpoint, the patients who
died before vasopressor therapy could be withdrawn and
those for whom vasopressor therapy could not be
withdrawn during the 28 days from randomisation were
treated as censored. The frequency of serious adverse
events was compared between groups with the χ² test or
Fisher’s exact test as appropriate. In the pharmacoeconomic
analysis, the rates of invasive or surgical procedures were
compared between groups by the χ² test and the mean
medical costs per patient in intensive care were compared
between groups by the Wilcoxon test. All reported p values
are two-sided.
This trial is registered with ClinicalTrials.gov, number
NCT00148278.
The funding source had no role in the conduct of the
study, the collection and interpretation of the data, or in
the drafting of the report. All authors had full access to all
the data of the study, and all agreed to submit the final
manuscript for publication.
www.thelancet.com Vol 370 August 25, 2007
Articles
Results
www.thelancet.com Vol 370 August 25, 2007
A
120
Epinephrine
Norepinephrine plus dobutamine
Mean blood pressure (mm Hg)
100
80
60
40
20
0
B
7·6
7·5
7·4
Arterial pH
7·3
7·2
7·1
7·0
6·9
6·8
C
11
9
Arterial lactate (mmol/L)
1591 patients were assessed for eligibility, of whom
330 were randomly assigned to treatment (figure 2).
The two treatment groups were well balanced at baseline
except that the median age was slightly higher in the
epinephrine group than in the norepinephrine plus
dobutamine group (table 1). More than half of the
patients had community-acquired infections; the lung
was the commonest site of infection (table 2). The
causal pathogen was identified in 267 (80%) cases
(table 2). Antibiotic treatment was deemed to be
appropriate in 250 (76%) cases. Use of corticosteroids
and recombinant human activated protein C was much
the same in the two groups (table 1).
At day 28, there was no significant difference in
mortality rates between the two groups (table 3; RR
0·86, 95% CI 0·65–1·14; p=0·31). Adjusted analyses
yielded similar results (table 4). There was no significant
difference between the two groups for mortality rates at
day 7 (0·81, 0·54–1·21; p=0·30), day 14 (0·75, 0·54–1·04;
p=0·08), discharge from intensive care (0·95, 0·75–1·21;
p=0·69) or hospital (0·93, 0·75–1·15; p=0·51), and at
day 90 (0·96, 0·78–1·19; p=0·73). Survival until day 90
did not differ significantly between the two groups
(p=0·53; figure 3).
Mean blood pressure increased to much the same
extent in both groups after randomisation (figure 3).
Compared with the norepinephrine plus dobutamine
group, arterial pH was significantly lower on
day 1 (p<0·0001), day 2 (p=0·0008), day 3 (p=0·0019),
and day 4 (p=0·0007) in the epinephrine group
(figure 3). Arterial lactate concentrations were also
significantly increased at day 1 (p=0·0003) in patients
given epinephrine only, compared with those given
norepinephrine plus dobutamine (figure 4). SOFA score
improved over time in both groups to much the same
extent (data not shown). There were no significant
differences between the two groups for the length of
stay (p=0·71), the number of days not in intensive care
until day 28 (p=0·14), the number of days not in
intensive care until day 90 (p=0·31), the number of
pressor-free days until day 28 (p=0·05), or the number
of pressor-free days until day 90 (p=0·18; table 5).
Likewise, there were no significant differences between
the two groups in terms of the time to haemodynamic
success (log-rank p=0·67; figure 5) and the time to
vasopressor withdrawal (log-rank p=0·09, figure 4). The
doses of vasopressors needed to achieve these effects
were not different between the two treatment groups
(figure 6).
There was no significant difference between the two
groups in the rates of invasive or surgical procedures:
69 (43%) patients in the epinephrine group and 71 (42%)
in the norepinephrine plus dobutamine group
underwent such procedures (p=0·88). There was no
significant difference between the two groups in mean
costs per patient: £5439 (SD 5715) in the epinephrine
7
5
3
1
-1
0
2
3
4
5
6
7
8
9
10
Time (days)
Figure 4: Effects of treatment
(A) Mean blood pressure. (B) Arterial pH. (C) Arterial lactate concentration. Error bars are SD.
group and £5672 (5258) in the norepinephrine plus
dobutamine group (p=0·73).
There were no significant differences between the
two groups in the rate of severe arrhythmias,
cerebrovascular or myocardial events, limb ischaemia,
or any other side-effect possibly related to catecholamine
administration (table 6).
681
Articles
Length of stay, days
Epinephrine
Norepinephrine plus p*
dobutamine
15 (7–31)
16 (6–32)
0·71
Days not in intensive care until day 28
0 (0–13)
0 (0–14)
0·14
Days not in intensive care until day 90
9 (0–68)
31 (0–75)
0·31
Pressor-free days until day 28
20 (0–24)
22 (6–25)
0·05
Pressor-free days until day 90
53 (0–86)
66 (6–86)
0·18
Data are median (IQR). *Wilcoxon’s test.
Table 5: Length of stay in intensive care, days not in intensive care, and pressor-free days until day 28
and day 90
A
1·0
Epinephrine
Norepinephrine plus dobutamine
0·9
0·8
Probability
0·7
0·6
0·5
0·4
0·3
Log-rank
χ²=0·18
p=0·67
0·2
01
0
B
0
2
1
3
4
5
6
7
8
9
10
1·0
0·9
0·8
Probability
0·7
0·6
0·5
0·4
0·3
0·2
Log-rank
χ²=2·93
p=0·09
0·1
0
0
2
4
6
8
10
12
14
16
Time (days)
18
20
22
24
26
28
Figure 5: Kaplan-Meier plots of time to (A) haemodynamic success and (B) vasopressor withdrawal
Discussion
We found no evidence for a difference in all-cause
mortality, in either the short term or the long term,
between patients with septic shock treated with
epinephrine and those treated with norepinephrine
plus dobutamine (whenever needed). Furthermore, we
found no evidence for a difference between the two
therapeutic options in terms of delay in haemodynamic
682
stabilisation, resolution of organ dysfunction, or adverse
events.
Our findings accord with those of a systematic review
on vasopressor therapy for management of septic
shock.14 That review identified two small (52 patients in
total), single-centre randomised trials that compared
mortality from septic shock between patients treated
with norepinephrine and dobutamine and those treated
with epinephrine. There were 13 (50%) deaths in the
norepinephrine plus dobutamine group and 13 (50%)
in the epinephrine group, with a relative risk of death
of 0·98 (95% CI 0·57–1·67).
Our study population was similar to that reported in
most recent clinical investigations in terms of
demographic data, severity of illness, source of
infection, type of pathogens, and crude mortality rate.19
At baseline, treatment groups were well balanced,
except that patients allocated epinephrine were slightly
older on average, which could have favoured the
norepinephrine plus dobutamine group.2,24 Logistic and
Cox’s proportional hazard regression analyses did not
show any evidence for a difference between groups in
the main endpoint with adjustment for the main
baseline factors that predict outcome; age was not
incorporated into this model because it was included in
SAPS II. The numbers of patients treated with
corticosteroids or activated protein C, which could
favourably affect septic shock mortality,20,25 were
comparable between the two groups.
This study was designed with the expectation of a
mortality rate of 60% in patients treated with
epinephrine on the basis of previous work and studies
we were doing when this study was planned.2,7,20 This
assumption was supported by the results of a European
cohort study that showed a 30-day mortality rate close
to 60% in patients with shock who received epinephrine
alone or in combination with other catecholamines.26
We expected a 20% absolute reduction in 28-day
mortality with norepinephrine combined with
dobutamine. Indeed, we thought that potentially
deleterious regional haemodynamic and metabolic
effects of epinephrine could have been associated with
a substantial increase in mortality. Additionally, given
that norepinephrine and dobutamine are more
expensive than epinephrine, and because administration
and titration of two catecholamines is more cumbersome
and requires monitoring of cardiac output, we hoped
that such treatment would be associated with substantial
improvement in clinical outcomes. In fact, the observed
crude mortality rate in the epinephrine group was
40% at day 28. This difference from our expectation
could be related to the generalised use of corticosteroids
and, to a lesser extent, of activated protein C. However,
the lower than expected mortality rate with epinephrine
slightly increased the study power, allowing, for
instance, the detection of a 15% difference between
groups with 85% probability. In fact, to test the
www.thelancet.com Vol 370 August 25, 2007
Articles
Contributors
DA, PEB, and EB designed the study protocol. DA and EB obtained
funds from the French Ministry of Health to conduct the study, and
wrote the report. AR and EB were responsible for data management and
statistical analyses. All authors contributed to the report.
CATS Study Group
Study chairs—Djillali Annane, Eric Bellissant Statisticians—Véronique
Sébille, Alain Renault. Steering committee—Djillali Annane, Eric
Bellissant, Pierre Edouard Bollaert, Claude Martin, Gérard Nitenberg,
Gilles Troché, Philippe Vignon. Independent data and safety monitoring
board—Christian Brun-Buisson, Claude Perret, Christian Perronne.
Independent diagnosis validation committee—Daniel Debacker, JeanFrançois Timsit, Jean-Paul Mira. Monitoring—Caroline Fisch, Chanfi
Maoulida, Nathalie Zinsou. Data management—Delphine Serra. Quality
assurance—Catherine Mouchel. Pharmacists—Annick Tibi, Blandine
Lehmann. Other investigators—D Caen (CH-Sud Francilien, Corbeil
Essonne, France); C Guérot (Hôpital Européen Georges Pompidou,
AP-HP, Paris, France); J Carlet (Hôpital Saint Joseph, Paris, France);
H Outin (Hôpital de Poissy Saint Germain, Poissy, France); A Parrot
(Hôpital Tenon AP-HP, Paris, France); L Brinquin (Hôpital du Val de
Grâce, Paris, France); J S Lemaire (CH de Roubaix, Roubaix, France).
www.thelancet.com Vol 370 August 25, 2007
A
Epinephrine or norepinephrine dose (µg/kg per min)
3·5
Epinephrine
Norepinephrine plus dobutamine
3·0
2·5
2·0
1·5
1·0
0·5
0
–0·5
1
2
3
4
5
6
7
8
9
10
Time (days)
Epinephrine
155/161
Norepinephrine 163/169
plus dobutamine
B
146/151 106/132
153/161 110/148
85/129
78/139
60/120
62/130
45/116
46/123
37/114
32/119
31/104
23/116
19/100
17/109
13/94
11/101
4
5
6
7
8
9
10
13/114
9/119
10/104
6/116
9/100
3/109
5/94
2/101
12
Placebo or dobutamine dose (µg/kg per min)
hypothesis that the absolute difference observed
between groups in the day 28 all-cause mortality rate
was real, 5000 patients would be needed to show
statistical significance with a 95% probability. Such a
large clinical trial is very unlikely to be done in patients
with septic shock, since they account for only 9% of
admissions to intensive care;2 even if half of all such
patients were eligible for a study, recruitment of an
adequate sample size would require 100 intensive care
units to recruit for 2–3 years or 20 units to recruit for
10–15 years.
Epinephrine was associated with some delays in the
normalisation of arterial pH and lactate concentrations
compared with norepinephrine plus dobutamine. This
finding is in keeping with those of previous studies that
have compared the short-term effects of epinephrine
with those of norepinephrine plus dobutamine,7,12
dopamine,11 or dopexamine.13 These metabolic effects
are probably the result of exaggerated aerobic glycolysis
through Na+K+ ATPase stimulation within the muscles,
rather than persistent tissue dysoxia.27 The metabolic
effects recovered within 4 days and had no effect on the
time to haemodynamic stabilisation, recovery of organ
dysfunction, or on survival.
There is some evidence that epinephrine might
induce serious myocardial side-effects, thus favouring
the use of norepinephrine.6 However, we found no
evidence for a difference in the occurrence of
arrhythmias, ischaemic damage to the brain or the
heart, or any other serious adverse event related to the
two therapeutic strategies.
In practice, physicians could use either epinephrine
alone, or norepinephrine alone or in combination with
dobutamine in patients with low cardiac index. Future
trials should compare the efficacy and safety of
epinephrine, or norepinephrine, with those of
dopamine, and more importantly should clarify the
optimum haemodynamic goals of a vasopressor therapy
in septic shock.
10
8
6
4
2
0
–2
1
2
3
Time (days)
Epinephrine
122/161 101/151
Norepinephrine 129/169 100/161
plua dobutamine
52/132
54/148
35/129
32/139
26/120
24/130
18/116
17/123
Figure 6: Drug doses over time
(A) Epinephrine or norepinephrine. (B) Placebo or dobutamine. Error bars are SD.
Overall
(n=330)
Epinephrine
(n=161)
Norepinephrine plus
dobutamine (n=169)
Supraventricular tachycardia >150 bpm
41 (12%)
19 (12%)
22 (13%)
Ventricular arrhythmias
20 (6%)
12 (7%)
8 (5%)
Acute coronary event
8 (2%)
5 (3%)
3 (2%)
Limb ischaemia
8 (2%)
2 (1%)
6 (4%)
Stroke
4 (1%)
2 (1%)
2 (1%)
Central nervous system bleeding
3 (0·9%)
3 (2%)
0 (0%)
During catecholamine infusion
After catecholamine infusion
Arrhythmias
Stroke
13 (4%)
6 (4%)
7 (4%)
4 (1%)
2 (1%)
2 (1%)
Other neurological sequelae
2 (0·6%)
1 (0·6%)
1 (0·6%)
Others
6 (2%)
3 (2%)
3 (2%)
Data are n (%).
Table 6: Serious adverse events
683
Articles
Conflict of interest statement
We declare that we have no conflict of interest.
Acknowledgments
The French Ministry of Health provided financial support (1997
Clinical Research Hospital Programme PHRC 1997, AOM 97123). The
study does not necessarily reflect the views of the Ministry and in no
way anticipates the Ministry’s future policy in this area.
References
1
Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of
sepsis in the United States from 1979 through 2000. N Engl J Med
2003; 348: 1546–54.
2
Annane D, Aegerter P, Jars-Guincestre MC, Guidet B, for the
CUB-Réa Network. Current epidemiology of septic shock: the
CUBRea Network. Am J Respir Crit Care Med 2003; 168: 165–72.
3
American College of Chest Physicians/Society of Critical Care
Medicine Consensus Conference. Definitions for sepsis and organ
failure and guidelines for the use of innovative therapies in
sepsis. Crit Care Med 1992; 20: 864–74.
4
Raphael JC, Antony I, Bellissant E, et al. Utilisation des
cathécolamines au cours du choc septique (adultes-enfants).
Réan Urg 1996; 5: 441–50.
5
Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis
Campaign guidelines for management of severe sepsis and septic
shock. Crit Care Med 2004; 32: 858–73.
6
Hollenberg SM, Ahrens TS, Annane D, et al. Practice parameters
for hemodynamic support of sepsis in adult patients: 2004 update.
Crit Care Med 2004; 32: 1928–48.
7
Levy B, Bollaert PE, Charpentier C, et al. Comparison of
norepinephrine and dobutamine to epinephrine for
hemodynamics, lactate metabolism, and gastric tonometric
variables in septic shock: a prospective, randomized study.
Intensive Care Med 1997; 23: 282–87.
8
Meier-Hellmann A, Reinhart K, Bredle DL, Specht M, Spies CD,
Hannemann L. Epinephrine impairs splanchnic perfusion in
septic shock. Crit Care Med 1997; 25: 399–404.
9
Duranteau J, Sitbon P, Teboul JL, et al. Effects of epinephrine,
norepinephrine, or the combination of norepinephrine and
dobutamine on gastric mucosa in septic shock. Crit Care Med
1999; 27: 893–900.
10 De Backer D, Creteur J, Silva E, Vincent JL. Effects of dopamine,
norepinephrine, and epinephrine on the splanchnic circulation in
septic shock: which is best? Crit Care Med 2003; 31: 1659–67.
11 Day NP, Phu NH, Bethell DP, et al. The effects of dopamine and
adrenaline infusions on acid-base balance and systemic
hemodynamics in severe infection. Lancet 1996; 348: 219–23.
12 Seguin P, Bellissant E, Le Tulzo Y, et al. Effects of epinephrine
compared with the combination of dobutamine and
norepinephrine on gastric perfusion in septic shock.
Clin Pharmacol Ther 2002; 71: 381–88.
684
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Seguin P, Laviolle B, Guinet P, Morel I, Mallédant Y, Bellissant E.
Dopexamine and norepinephrine vs epinephrine on gastric
perfusion in septic shock patients: a randomized study. Crit Care
2006; 10: R32.
Mullner M, Urbanek B, Havel C, Losert H, Waechter F,
Gamper G. Vasopressors for shock. Cochrane Database Syst Rev
2004; 3: CD003709.
McCabe WA, Jackson GG. Gram negative bacteremia. I. Etiology
and ecology. Arch Intern Med 1962; 110: 847–55.
Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute
Physiology Score (SAPS II) based on a European/North American
multicenter study. JAMA 1993; 270: 2957–63.
Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related
Organ Failure Assessment) score to describe organ
dysfunction/failure. Intensive Care Med 1996; 22: 707–10.
Chaix C, Durand-Zaleski I, Alberti C, Brun-Buisson C. A model to
compute the medical cost of intensive care patients.
Pharmacoeconomics 1999; 15: 573–82.
Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005;
365: 63–78.
Annane D, Sebille V, Charpentier C, et al. Effect of treatment with
low doses of hydrocortisone and fludrocortisone on mortality in
patients with septic shock. JAMA 2002; 288: 862–71.
O’Brien PC, Fleming TR. A multiple testing procedure for clinical
trials. Biometrics 1979; 35: 549–56.
Pocock SJ, Assmann SE, Enos LE, Kasten LE. Subgroup analysis,
covariate adjustment and baseline comparisons in clinical trial
reporting: current practice and problems. Stat Med 2002; 21: 2917–30.
Annane D, Sebille V, Troche G, Raphael JC, Gajdos P,
Bellissant E. A 3-level prognostic classification in septic shock
based on cortisol levels and cortisol response to corticotropin.
JAMA 2000; 283: 1038–45.
Martin GS, Mannino DM, Moss M. The effect of age on the
development and outcome of adult sepsis. Crit Care Med 2006; 34:
15–21.
Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of
recombinant human activated protein C for severe sepsis.
N Engl J Med 2001; 344: 699–09.
Sakr Y, Reinhardt K, Vincent JL, et al. Does dopamine
administration in shock influence outcome? Results of the Sepsis
Occurrence in Acutely Ill Patients (SOAP) Study. Crit Care Med
2006; 34: 589–97.
Levy B, Gibot S, Franck P, Cravoisy A, Bollaert PE. Relation
between muscle Na+K+ ATPase activity and raised lactate
concentrations in septic shock: a prospective study. Lancet 2005;
365: 871–75.
www.thelancet.com Vol 370 August 25, 2007
Seminar
Polycystic ovary syndrome
Robert J Norman, Didier Dewailly, Richard S Legro, Theresa E Hickey
Polycystic ovary syndrome is a heterogeneous endocrine disorder that affects about one in 15 women worldwide. The
major endocrine disruption is excessive androgen secretion or activity, and a large proportion of women also have
abnormal insulin activity. Many body systems are affected in polycystic ovary syndrome, resulting in several health
complications, including menstrual dysfunction, infertility, hirsutism, acne, obesity, and metabolic syndrome.
Women with this disorder have an established increased risk of developing type 2 diabetes and a still debated increased
risk of cardiovascular disease. The diagnostic traits of polycystic ovary syndrome are hyperandrogenism, chronic
anovulation, and polycystic ovaries, after exclusion of other conditions that cause these same features. A conclusive
definition of the disorder and the importance of the three diagnostic criteria relative to each other remain controversial.
The cause of polycystic ovary syndrome is unknown, but studies suggest a strong genetic component that is affected
by gestational environment, lifestyle factors, or both.
Polycystic ovary syndrome is one of the most common
endocrine disorders in women of reproductive age, and
the most frequent cause of hyperandrogenism and
oligo-anovulation,1,2 both of which have substantial
psychological, social, and economic consequences.3 An
increased awareness of this disorder in the general
population and medical communities has taken place in
recent years with the knowledge that women with polycystic
ovary syndrome are susceptible to metabolic syndrome4,5
and its associated comorbidities. Because of the heterogeneity in its presentation, the definition of polycystic
ovary syndrome has been controversial in disciplines as
diverse as internal medicine, gynaecology, and psychiatry.
Therefore, polycystic ovary syndrome is a persisting
challenge for clinical and basic research scientists who try
to elucidate its origins and distinguish primary pathological
changes from secondary environmental disruptions.
Classification and epidemiology
Three key diagnostic features of polycystic ovary syndrome
have been proposed with various degrees of emphasis,
dependent on the perspective of the investigator. These
features are hyperandrogenism, chronic anovulation, and
polycystic ovaries on ultrasonography.6,7 Importantly,
other conditions are known to cause or to mimic the
features of polycystic ovary syndrome, and must be
excluded prior to diagnosis. These include congenital
adrenal hyperplasia, Cushing’s syndrome, and androgensecreting tumours for hyperandrogenism and raised
prolactin or insufficient luteinising hormone for
anovulation. Although obesity, insulin resistance, and
metabolic syndrome are frequently present in women
with polycystic ovary syndrome, they are not regarded as
intrinsic disturbances of the disorder.
At present, there are two main definitions for polycystic
ovary syndrome, which are the topic of intense debate.8,9
The 1990 National Institutes of Health (NIH) criteria
require the presence of chronic anovulation plus clinical or
biochemical signs of hyperandrogenism, whereas the
2003 Rotterdam criteria require the presence of two or
more of: chronic anovulation, clinical or biochemical signs
of hyperandrogenism, and polycystic ovaries. By inclusion
www.thelancet.com Vol 370 August 25, 2007
of polycystic ovaries as a diagnostic criterion, the Rotterdam
definition recognises four phenotypes of polycystic ovary
syndrome (table 1). This definition has raised controversial
and unsolved issues that have implications for clinical
diagnosis and study design. The Androgen Excess Society
recently reported the results of an evidence-based review of
phenotypes for polycystic ovary syndrome.11 The results
suggested that polycystic ovary syndrome should primarily
be regarded as a disorder of excessive androgen
biosynthesis, use, or metabolism. In terms of phenotypes
of polycystic ovary syndrome, ovulatory women with
hyperandrogenism and polycystic ovaries have a mild form
of the disorder,12 but women with polycystic ovaries in the
absence of anovulation or hyperandrogenism do not.
Disagreement continues as to whether chronic anovulation
and polycystic ovaries without overt hyperandrogenism is
a form of polycystic ovary syndrome. Although preliminary
data suggest that women of this phenotypic group have
subtle endocrine and metabolic abnormalities consistent
with a mild form of the disorder,12 the metabolic features of
these women are regarded by some to be too mild or
distinct to be associated with the increased risk of
developing metabolic disease that is characteristic of
women with polycystic ovary syndrome.13,14
The prevalence of polycystic ovary syndrome, as defined
by the 1990 National Institutes of Health (NIH) criteria,
in unselected populations of women of reproductive age
is between 6∙5 and 8%. Mexican American women might
have a higher prevalence of polycystic ovary syndrome
than white or black American women.15 Immigrant
Lancet 2007; 370: 685–97
Research Centre for
Reproductive Health, School of
Paediatrics and Reproductive
Health, University of Adelaide,
Adelaide, South Australia,
Australia (R J Norman MD,
T E Hickey PhD); Department of
Endocrine Gynaecology and
Reproductive Medicine,
Hôpital Jeanne de Flandre,
Centre Hospitalier de Lille, Lille,
France (D Dewailly MD); and
Department of Obstetrics and
Gynecology, Penn State College
of Medicine, Milton S Hershey
Medical Center, Hershey, PA,
USA (R S Legro MD)
Correspondence to:
Dr Theresa E Hickey, Medical
School North, Frome Road,
Adelaide, South Australia 5005,
Australia
[email protected]
Search strategy and selection criteria
We searched the Cochrane library (up to 2005), Medline via
PubMed (up to November, 2006) and EMBASE (up to July,
2006). We used the terms “PCOS”; “PCOD”; “PCO”;
“Stein-Leventhal syndrome”; “hirsutism” not “PCOS”. We
selected articles in the past 5 years, but also used highly
regarded older articles and several relevant review articles.
The reference list was modified by each author and in
response to comments from reviewers.
685
Seminar
Severe PCOS
Hyperandrogenism Ovulatory
and chronic
PCOS
anovulation
Mild PCOS
Periods
Irregular
Irregular
Normal
Irregular
Ovaries on ultrasonography
Polycystic
Normal
Polycystic
Polycystic
Androgen concentrations
High
High
High
Mildly raised
Insulin concentrations
Increased
Increased
Increased
Normal
Risks
Potential long-term Potential long-term
Unknown
Unknown
16%
16%
Prevalence in affected women10 61%
7%
PCOS=polycystic ovary syndrome.
Table 1: Phenotypes for polycystic ovary syndrome based on 2003 Rotterdam criteria
populations from the Indian subcontinent to the UK,
and Australian women of Aboriginal heritage also have a
high prevalence of polycystic ovary syndrome.16,17 Adoption
of the 2003 Rotterdam criteria for the diagnosis of this
disorder will presumably increase the prevalence of
polycystic ovary syndrome because the scope for inclusion
is broader than it is with the 1990 NIH criteria.8 Indeed,
in a study of women with normogonadotropic anovulatory
(WHO type 2) infertility, the prevalence of polycystic
ovary syndrome was 1∙5-fold higher when defined by
the 2003 Rotterdam criteria than when defined by the
1990 NIH criteria.13
Clinical features and diagnosis
Polycystic ovary syndrome has many signs and features;
three main characteristics must be assessed to establish
whether a woman conforms to one of the recognised
phenotypes of the syndrome that are summarised in
table 1.
Hyperandrogenism
Hyperandrogenism is the most constant and prominent
diagnostic component of polycystic ovary syndrome, but
reliable detection of this feature is not straightforward,
and indices vary substantially dependent on ethnic origin,
bodyweight, and age. Hyperandrogenism is assessed by
clinical features, biochemical indices, or both. Clinically,
hyperandrogenism is diagnosed by the mostly subjective
assessment of cutaneous manifestations of excessive
androgen activity, such as hirsutism, acne (especially in
young women), and female-pattern alopecia (more
apparent in old women). Hirsutism is the most common
of these symptoms, being present in about 60% of
women with polycystic ovary syndrome,18–20 although it is
rarely present in Asian women.21 Degrees of hirsutism
vary greatly in different ethnic populations, and the
threshold of abnormality should be measured on a
population basis.22 Debate exists as to whether the
frequency of acne and alopecia in women with polycystic
ovary syndrome is higher than in the general population,
and present recommendations regard them as unreliable
clinical signs of hyperandrogenism, particularly if they
are the only diagnostic feature.23
686
Biochemically, hyperandrogenaemia is most commonly
assessed by measurement of serum total testosterone (T)
and sex hormone binding protein (SHBG), followed by
calculation of the free or bioavailable (free and weakly
bound to albumin) fraction by the free androgen index
(T/SHBG×100) or the mass action equation, respectively.24–26
The mass action equation is regarded as the method of
choice to calculate free serum testosterone, if reliable
assays are used and normative data specific to each assay
are developed.25,26 Radioimmunoassays that claim to
measure free testosterone directly are available and
widespread, but are highly unreliable and should not be
used.25,26 The concentrations of other serum androgens
such as androstenedione or the adrenal androgen
prasterone sulfate (known as DHEAS) are often high in
women with polycystic ovary syndrome, but their
measurement is of little value in the average clinical
setting. However, some workers have suggested that
ethnic groups, even distinct populations of caucasian
ethnic origin, might differ greatly with respect to the
concentrations of specific androgens in the serum of
women with polycystic ovary syndrome.27
Unfortunately, serum analysis fails to measure the
biochemical hyperandrogenism of polycystic ovary
syndrome in about 20–40% of patients,20 and even
semiquantitative measurements such as the modified
Ferriman-Gallwey score for hirsutism28 might underestimate clinical hyperandrogenism.22 Most commercial
assays for total serum testosterone are not designed or
validated for detection within the range for women,26
raising concern about their real diagnostic value.
Moreover, the range that is regarded as healthy for
women by commercial laboratories is very broad, and
has been shown to include many hyperandrogenic
women, even those with severe hirsutism.29 Until more
accurate methods of measurement are developed, many
investigators think that failure to detect biochemical or
clinical hyperandrogenism should not exclude diagnosis
of polycystic ovary syndrome in the presence of other
clinical signs.
Chronic anovulation
Diagnosis of chronic anovulation is easier than diagnosis
of hyperandrogenism, because the major clinical
signs—namely, oligomenorrhoea or amenorrhoea—vary
in duration but are generally unambiguous.
Oligomenorrhoea is defined as less than eight periods
per year, or cycles that are longer than 35 days, and
amenorrhoea is absence of menstruation for more than
3 months without pregnancy. However, a high rate of
false negatives is possible if menstrual history alone is
investigated. Regular cycles do not exclude chronic
anovulation without evidence of a progesterone
concentration in serum during the luteal phase of the
menstrual cycle that is consistent with a recent ovulation.
When chronic anovulation is present, serum prolactin
and luteinising hormone (LH) assays should be done to
www.thelancet.com Vol 370 August 25, 2007
Seminar
exclude hypothalamic and pituitary diseases, which
would cause hyperprolactinaemia (prolactin >20–30 μg/L),
gonadotropin deficiency (LH <2 IU/L), or both.
Additionally, chronic anovulation due to polycystic ovary
syndrome should not be confounded with some forms of
functional hypothalamic amenorrhoea that are caused by
extreme caloric restriction, exercise, or both, in which
amenorrhoea is associated with low plasma oestrogen, is
not responsive to progestagen withdrawal to induce
menstruation, and is characterised by normal or low
gonadotropin.30
excessive secretion persists in cultured cells over many
passages,39 suggesting a genetic association, but up to now
none of the genes implicated in steroid biosynthesis has
been linked to polycystic ovary syndrome through relevant
polymorphisms or mutations.40 However, the expression
and activity of many steroidogenic enzymes is increased
in thecal cells from patients with the disorder, and this
hyperactivity might be caused by a disturbance of
intracellular signalling pathways that have not previously
been implicated in the pathogenesis of this disorder.41
Abnormalities of folliculogenesis
Polycystic ovaries on ultrasonography
The definition of the diagnostic features for polycystic
ovaries by ultrasonography has been controversial
because technical advancements, including highfrequency vaginal probes and image-enhancing software,
have improved resolution and measurement capabilities.
Previous definitions, which were based on transabdominal
ultrasonography,31 have now been revised on the basis of
transvaginal techniques,32 and state that in the follicular
phase ovary (lacking follicles larger than 10mm in
diameter), the presence of 12 or more follicles measuring
2–9 mm in diameter, or increased ovarian volume
(>10 mL) suffice to make a diagnosis of polycystic ovaries
(figure). Although there are other characteristic features,
priority has been given to follicle number and ovarian
volume because both have the advantage of being
measured in real time and are regarded as key and
consistent features of polycystic ovaries. The assessment
of polycystic ovaries in adolescent girls should be done by
transabdominal ultrasonography with measurement of
ovarian volume only, because the criterion based on
follicles is much less reliable by the abdominal route,
especially in obese individuals.32 The adult upper healthy
threshold volume of 10 mL seems to be also appropriate
for post-menarchal adolescents.33 Measurement of serum
anti-Mullerian hormone (AMH), which is secreted by
granulosa cells of developing follicles, is emerging as a
potential surrogate for ultrasonography, because values
correlate closely with antral follicle count in pilot
investigations.34 This assay might facilitate the diagnosis
of polycystic ovary syndrome in circumstances in which
ultrasonography is inappropriate or unavailable, although
the assay is not valid for women older than 35 years.
Polycystic ovaries have two to six times more primary,
secondary, and small antral follicles than do healthy
ovaries.42–44 These early developing follicles seem to be
functionally normal in most respects. The mechanism
that determines excess numbers of follicles is unknown,
but several lines of evidence implicate abnormal
androgen signalling. As defined by strict consensus
criteria, 90–100% of women with polycystic ovary
syndrome have polycystic ovaries,12,14 and several studies
have reported a positive correlation between follicle
number and serum testosterone and androstenedione
concentrations in these women.14,45,46 Administration of
dihydrotestosterone to female rhesus monkeys induces a
polycystic-ovary-like morphology, including increased
ovarian volume and increased follicle numbers,
suggesting a direct action of androgens on ovarian cells.47
Although a similar result has been reported in
female-to-male trans-sexuals after long-term testosterone
treatment,48 such women seemed to have a high frequency
of polycystic ovaries before hormone administration.49,50
However, the notion of androgen-induced polycystic
ovaries is supported by the findings of one study in which
monotherapy with the non-steroidal anti-androgen
flutamide reduced ovarian volume and improved the
abnormal follicle profile in adolescent girls with polycystic
ovary syndrome.51 Additionally, a polymorphism in the
androgen receptor that affects the potency of its activity
has been implicated in the disorder.52,53 Although the
increase in follicle numbers in polycystic ovaries might
Normal ovary
Polycystic ovary
Pathogenesis
Androgen abnormalities
60–80% of women with polycystic ovary syndrome have
high concentrations of circulating testosterone,19,20,35 and
about 25% have high concentrations of prasterone sulfate
(DHEAS),36 leading some investigators to surmise that
uncontrolled steroidogenesis might be the primary
abnormality in this disorder.37 Polycystic ovaries have a
thickened thecal layer, and thecal cells derived from these
ovaries secrete excessive androgens in vitro under basal
conditions or in response to LH stimulation.38 The
www.thelancet.com Vol 370 August 25, 2007
Figure: Normal and polycystic ovary shown by transvaginal ultrasonography during the follicular phase of a
menstrual cycle
The fluid-filled antrum of a developing follicle appears as a dark circle. When compared with a normal ovary, the
polycystic ovary is typically enlarged and contains an abnormally increased number of developing follicles.
687
Seminar
be due to a trophic effect of androgens on primate
follicular cells,47,54 it might also be that the follicles of a
polycystic ovary grow very slowly, creating a stockpiling
effect. This slow growth might be mediated by deficient
growth signals from the oocyte44 or by the inhibitory
effect of excess AMH.55
In anovulatory women with polycystic ovary syndrome,
antral follicle growth stops when the follicle is less than
10 mm in diameter, which is the stage just before
emergence of a dominant follicle. Follicular arrest is
associated with excessive stimulation of follicular cells by
insulin, LH, or both, in addition to a hyperandrogenic
environment.56 Insulin enhances the responsiveness of
granulosa cells to LH, and in the ovaries of hyperinsulinaemic women with polycystic ovary syndrome,
arrested follicles show signs of premature luteinisation.57
Granulosa cells from women with polycystic ovary
syndrome also seem to be selectively insulin resistant,
whereby insulin-stimulated glucose metabolism is
impaired but insulin-stimulated steroidogenesis is
normal,58,59 suggesting that deficient energy mobilisation
within the follicle contributes to anovulation. The association between hyperinsulinaemia, insulin resistance,
and anovulation in women with the syndrome inspired
the use of insulin sensitisers such as metformin as a
therapeutic approach to induce ovulation.
Gonadotropin abnormalities
Women with polycystic ovary syndrome display abnormal
patterns of gonadotropin pulsatility, which is characterised
by excessive secretion of LH but normal secretion of
follicle-stimulating hormone (FSH).60 This pattern of
secretion gives rise to an abnormal circulating LH to
FSH ratio in some patients, mostly those of normal
weight.61 The pattern is exacerbated by gonadotropinreleasing hormone challenge tests, which further
increase circulating LH and 17-hydroxyprogesterone
concentrations in women with the disorder.62 Overall,
these data suggest the presence of a defect of the
hypothalamic–pituitary axis in polycystic ovary syndrome,
which is further supported by evidence of increased
pituitary sensitivity to stimulation with corticotropinreleasing factor, resulting in an excessive adrenocorticotropic hormone and cortisol response in women
with this disorder.63 However, high concentrations of
androgens desensitise the hypothalamus to negative
feedback by progesterone,64 suggesting that the
abnormalities of gonadotropin release in polycystic ovary
syndrome are secondary to abnormal steroid release by
the ovaries or adrenal glands.
Insulin action abnormalities
Women with polycystic ovary syndrome seem to have a
level of peripheral insulin resistance that is much like
that of women with type 2 diabetes, which is characterised
by a 35–40% decrease in insulin-mediated glucose
uptake.65 Normoglycaemic women with the syndrome
688
display both fasting and glucose-challenged hyperinsulinaemia,66 and a β-cell compensation that is
inadequate for their degree of peripheral insulin
resistance,67,68 suggesting that they are at high risk of
type 2 diabetes.69 Indeed, about 40% of obese women
with polycystic ovary syndrome have impaired glucose
tolerance after a standard challenge with 75 g oral
glucose,70–72 although the frequency is lower in thinner
women with the syndrome.73
Insulin resistance might contribute to hyperandrogenism and gonadotropin abnormalities through
several mechanisms. High concentrations of insulin
reduce circulating SHBG values, thereby increasing the
bioavailability of testosterone,74,75 and might also serve as a
co-factor to stimulate adrenal and ovarian androgen
biosynthesis, thereby contributing to abnormal
gonadotropin concentrations.76–79 Insulin might also act
directly in the hypothalamus, pituitary gland, or both, to
regulate gonadotropin release,80 but the contribution of
insulin resistance to manifestation of gonadotropin
abnormalities in polycystic ovary syndrome remains
uncertain.81 Insulin resistance in the disorder is
characterised by selective-tissue insulin sensitivity, in
which some tissues seem highly resistant (ie, skeletal
muscle) and others sensitive (ie, adrenal and ovary). In
affected tissues, metabolic pathways are generally
resistant to stimulation by insulin, but mitogenic or
steroidogenic pathways are not.82
The reconfiguration of polycystic ovary syndrome as a
metabolic syndrome with reproductive implications has
led to detailed studies of women affected by this disorder
for signs of insulin resistance. Women with polycystic
ovary syndrome have also proved to have dyslipidaemia,83–85
high levels of inflammatory markers,86,87 endothelial
dysfunction,88,89 and an increased frequency of sleep
apnoea.90–92
Causes and risk factors
The cause of polycystic ovary syndrome remains
unknown, although, like most complex heterogeneous
diseases, both environmental and genetic factors are
implicated. With time and technological advances, focus
has shifted from the ovary93 to the hypothalamic–pituitary
axis,60 to some primary defects of insulin activity94,95 as the
primary pathological cause of the syndrome. Compelling
evidence exists to implicate all three sources, and because
they form an interacting system of factors that regulate
ovarian function, it is possible that there are many
different primary disturbances that result in the same
pathological outcome. A priority for the future is to
understand the relation between the phenotypes of
polycystic ovary syndrome and their underlying
pathophysiology.
Familial aggregation of polycystic ovary syndrome has
been recognised for many years.96–98 In a twin study, the
genetic component of the syndrome as a single variable
characterised by self-reported oligomenorrhoea in the
www.thelancet.com Vol 370 August 25, 2007
Seminar
presence of either hirsutism or acne was estimated to
be 79%.99 Polycystic ovary syndrome does not show clear
Mendelian inheritance, is regarded as a complex disorder,
and poses unique challenges to geneticists.100 Genetic
analysis is hampered by low fecundity, absence of a male
phenotype, absence of an animal model, age-related
changes in the reproductive phenotype, and variation in
the diagnostic criteria. Epigenetic variation has also been
implicated as a confounding factor.101 Various genetic
associations of uncertain meaning have been reported,
most of which have not been replicated. Linkage
disequilibrium has consistently been reported between
polycystic ovary syndrome and a microsatellite marker
(D19S884) on chromosome 19p13.2, located in nonconserved intronic sequences between exons 55 and 56 of
the fibrillin 3 (FBN3) gene,100,102,103 which encodes an
extracellular matrix protein implicated in the regulation of
the activity of specific growth factors. The association has
been shown by some investigators,104 but not by others,105
although none of them tested for genetic linkage. The
meaning of D19S884 is still under investigation. Overall,
several genetic associations have been shown in polycystic
ovary syndrome, but at present there is no clinical
indication for genetic testing in women with this disorder.
Obesity has a substantial effect on the manifestation of
polycystic ovary syndrome.106 Excess weight exacerbates
metabolic and reproductive abnormalities in women
with the syndrome, and family studies suggest that
weight gain might promote the phenotype of polycystic
ovary syndrome in a susceptible population.107 Obesity in
women with polycystic ovary syndrome is prevalent in
North America.69,108,109 In an unselected population from
Alabama, 24% of women with the syndrome were
overweight (body mass index [BMI] 25∙0–29∙9) and
42% were obese (BMI >30).110 Women with the syndrome
from other countries tend to be thinner: in a study from
the Netherlands, the mean BMI was 28–29, and
prevalence studies have shown BMIs in the range of
25–28 in the UK, Greece, and Finland.19,111–113 The reasons
for the prevalence of excessive weight in women with
polycystic ovary syndrome in the USA might be due to
insufficient physical exercise or diet,114 and the high
prevalence of obesity in the general population.
Evidence that the syndrome is caused by factors in the
environment which mimic hormones, and are able to
disrupt endocrine activity, is scarce. However, research is
in progress, and compelling evidence exists that exposure
of pregnant non-human primates and sheep to excess
androgens predisposes female offspring to develop a
syndrome similar to polycystic ovary syndrome.115 A
similar effect in human beings is difficult to ascertain,
although female offspring of women with congenital
adrenal virilising disorders have a masculinisation of
neuroendocrine function that shares some similarities
with the abnormalities in women with polycystic ovary
syndrome.116 Theoretically, the gestational environment
and lifestyle factors in early childhood mediate the effect
www.thelancet.com Vol 370 August 25, 2007
of predisposing genetic factors, and thereby programme
individuals for development of reproductive disorders
such as polycystic ovary syndrome later in life.117 The
effects of fetal programming, lifestyle factors, or both, in
the cause of polycystic ovary syndrome might arise
through epigenetic modifications of DNA.101
Health implications
The potential health consequences of polycystic ovary
syndrome are a lifelong issue (table 2). There is little doubt
that the prevalence of impaired glucose tolerance and
diabetes mellitus is increased substantially in women with
polycystic ovary syndrome,70,71 although the magnitude of
the increase depends on the prevalence of obesity in the
population,118 and racial influences are evident. Conversion
rates of glucose tolerance from normal to abnormal are
accelerated in polycystic ovary syndrome,71,119,120 and up to
10% of women with this disorder develop diabetes during
the third or fourth decade.70,71 The evidence for increased
risk of cardiovascular disease in women with polycystic
ovary syndrome is less clear, although cardiovascular risk
factors are substantially increased, including hyperlipidaemia, hyperandrogenaemia, hypertension, markers
of a prothrombotic state, and markers of inflammation.121
Altered vascular and endothelial function in young women
with polycystic ovary syndrome is well documented, and
increased death rates from cardiovascular disease have
been shown in women with menstrual irregularity
(possibly with polycystic ovary syndrome) in the Nurses’
Health Study.122
There have been several definitions of metabolic
syndrome123 and several reports of an increased prevalence
of metabolic syndrome in women with polycystic ovary
syndrome.5,124 However, whether this increase is caused by
a feature specific to polycystic ovary syndrome or is merely
a consequence of adiposity is still unclear. An increase in
central fat, hyperinsulinaemia, glucose intolerance,
increased blood pressure, and other isolated features of
metabolic syndrome are more common in women with
polycystic ovary syndrome than they are in the general
Prenatal or childhood
Adolescence, reproductive years
Postmenopausal
Reproductive
Premature adrenarche
Early menarche
Menstrual irregularity
Hirsutism
Acne
Infertility
Endometrial cancer
Miscarriage
Pregnancy complications
Delayed menopause?
Metabolic
Abnormal fetal growth
Obesity
Impaired glucose tolerance
Insulin resistance
Dyslipidaemia
Type 2 diabetes
Obesity
Impaired glucose tolerance
Insulin resistance
Dyslipidaemia
Type 2 diabetes
Other
..
Sleep apnoea
Fatty liver
Depression
Cardiovascular disease?
Table 2: Lifelong health complications
689
Seminar
population. Although there is insufficient evidence to
suggest that the increased prevalence can be explained by
polycystic ovary syndrome alone, excess androgen in
women has been shown to be a risk factor for metabolic
syndrome independent of obesity and insulin resistance.125
Management
insulin-sensitising agents such as thiazolidinediones108
and metformin138 might be useful in the treatment of
hirsutism and acne because insulin resistance affects
both disorders, but the recommendation of these drugs
for cosmetic purposes is premature. For androgenic
alopecia in women, 2–5% topical minoxidil is regarded
as the most effective treatment.139,140
Lifestyle changes
The association between excessive weight, hyperandrogenaemia, impaired glucose tolerance, menstrual
abnormalities, and infertility emphasises the need to
address lifestyle issues in women with polycystic ovary
syndrome, particularly nutrition and exercise. Realistic
and achievable weight loss can be set to improve an
individual’s reproductive and metabolic fitness because
only a small (2–5%) reduction of bodyweight can greatly
improve these indices.126 A small reduction of bodyweight
was sufficient to restore ovulation and increase insulin
sensitivity by 71% in obese anovulatory women.127 Loss of
abdominal fat, which is associated with insulin resistance,
seems to be crucial to restore ovulation in these women.
Weight loss also increases SHBG concentration, reduces
testosterone concentration and androgenic stimulation
of the skin, improves menstrual function and conception
rates, and reduces miscarriage rates.128–131 Although drugs
to increase insulin sensitivity in diabetics are used to
treat women with polycystic ovary syndrome, weight
reduction is more effective and should be the initial
treatment for obese women with this disorder.132 Little is
known about the best exercise patterns, but evidence-based
dietary approaches exist in short-term studies. Caloric
restriction seems more important than macronutrient
composition, and there is little evidence that a
high-protein diet is better than a high-carbohydrate
diet.133,134 Although acute weight loss can be achieved with
severe caloric restriction, long-term weight maintenance
is rare, and acute weight loss potentially has dangerous
effects for reproduction.135
Cosmetic issues
Skin and hair disorders can be substantial in women
with polycystic ovary syndrome, and are physically and
psychologically very damaging. Although standard
commercial cosmetic approaches are used initially,
ovarian suppression through oral contraceptives is widely
prescribed for hirsutism and acne, especially in the
adolescent population. This treatment option has the
advantage of regulating the menstrual cycle and providing
contraception. Cyproterone in combination with
oestrogen is one of the most effective treatments of
hirsutism, although side-effects such as tiredness,
reduced libido, and changes in liver function are
common. Laser electrolysis alone or in combination with
topical application of eflornithine cream to retard hair
growth is also very effective to reduce hirsutism.136 Acne
often responds well to oral contraceptives with low doses
of cyproterone or drospirenone.137 Evidence exists that
690
Menstrual irregularity
The abnormal hormonal concentrations characteristic of
polycystic ovary syndrome might predispose women with
this disorder to development of endometrial cancer,
although the data that support this finding are not very
convincing.141 The number of menstrual cycles is less
important than the avoidance of endometrial hyperplasia,
and intermittent induction of menstruation by any
means, most commonly by progestagens or
administration of oral contraceptives, either cyclically or
continuously, prevents abnormal uterine proliferation.
Use of combined oral contraceptives is the most
common treatment for symptoms of polycystic ovary
syndrome because they interfere with androgen activity
via several mechanisms, including reduced androgen
production, increased hepatic SHBG synthesis, and
competitive binding to androgen receptors by some
progestagens. However, the potential long-term metabolic
side-effects of combined oral contraceptives in women
with polycystic ovary syndrome is being debated,
especially since women with this disorder have a
propensity for development of obesity and metabolic
abnormalities. Combined oral contraceptives have been
shown to decrease insulin sensitivity, impair glucose
tolerance, and alter lipid profiles in healthy populations
of women, but seemingly not to a degree that affects the
risk of diabetes mellitus or cardiovascular disease.142
Published work on the metabolic consequences of
combined oral contraceptives in women with polycystic
ovary syndrome is equivocal, and studies generally do
not satisfy the criteria for evidence-based medicine.142
Therefore, the assumption that the use of combined oral
contraceptives in women with the syndrome is safe is
premature, especially because women with this disorder
often start taking oral contraceptives early in adolescence,
continue taking them for long periods, and are already
susceptible to metabolic perturbations. Treatments that
couple combined oral contraceptives with insulin
sensitisers, antiandrogens, or both, are emerging with
potential beneficial effects on metabolic abnormalities,
especially in young women with the syndrome.142,143
Infertility
Women with polycystic ovary syndrome form the largest
group of women with WHO class 2 ovulatory dysfunction,
which is characterised by chronic anovulation in the
presence of normal FSH and oestradiol concentrations.
Induction of ovulation is the first-line treatment for this
class of anovulation, and is aimed at the introduction of
www.thelancet.com Vol 370 August 25, 2007
Seminar
an endocrine milieu that promotes growth and ovulation
of a single dominant follicle with consequent singleton
pregnancy.
Clomifene is a selective oestrogen-receptor modulator
that antagonises the negative feedback of endogenous
oestrogen on the hypothalamic–pituitary axis. Treatment
with clomifene should return LH to normal and increase
FSH secretion, and thereby stimulate follicle growth and
ovulation. Clomifene has been the gold standard
treatment for induction of ovulation in women with
polycystic ovary syndrome for many decades,144,145 and a
meta-analysis has shown that the use of clomifene is six
times more likely to result in pregnancy than is placebo
in such women (numbers needed to treat=5∙9, 95% CI
3∙6–16∙7).145 A prospective follow-up of thin women with
ovulatory dysfunction has shown high conception rates
in ovulatory responders treated with clomifene,
approaching 50% after three cycles of treatment, and
75% within nine cycles.146 The examination of follicle
development by ultrasonography and measurement of
serum concentrations of oestradiol might help to avoid
multifollicular development. One of the side-effects of
clomifene is increased risk of multiple pregnancy,147
which is possibly reduced by tailoring the treatment
regimen to account for patients’ characteristics that
predict specific outcomes.146 Despite a high degree of
efficacy, some women with polycystic ovary syndrome
are resistant to clomifene and do not ovulate, or fail to
achieve pregnancy despite ovulation. Failure to achieve
pregnancy might be due to adverse effects of clomifene
on the endometrium.148 Factors that affect resistance to
clomifene or failure to achieve pregnancy are, in order of
importance, hyperandrogenaemia, obesity, ovarian
volume, and menstrual dysfunction. A prediction model
has been developed to assess the chance of a woman
treated with clomifene being able to ovulate and become
pregnant, taking these variables into account.149
Like clomifene, aromatase inhibitors reduce oestrogen
stimulation of the hypothalamic–pituitary axis, but do so
by reducing oestrogen biosynthesis. Patients who are
resistant to clomifene are allegedly more sensitive to
induction of ovulation with aromatase inhibitors such as
letrozole that have less side-effects on endometrial
thickness than has clomifene, and possibly a lower risk
of multiple pregnancy.150 A randomised controlled trial
has shown that there is less ovarian stimulation with
letrozole than with clomifene,151 which might contribute
to a lower risk of multiple pregnancy. However, concern
about the possibility of fetal abnormalities as a result of
aromatase inhibition has led to avoidance of these drugs
in some countries.
Induction of ovulation with gonadotropins has been
shown to be very effective as a low-dose regimen with
gradual increase in dose as needed after close
examination of hormone and ultrasound response,152
and some investigators suggest that this regimen is
preferable to clomifene for first-line treatment in
www.thelancet.com Vol 370 August 25, 2007
women with polycystic ovary syndrome.153 Alternative
methods of gonadotropin induction, such as treatment
onset with a high dose followed by a gradual reduction
(step-down protocol), demand more skills and are not
more effective than the low-dose regimen.154 Overall,
ovulation induction with gonadotropins has a reasonable
success rate both in terms of ovulation and cumulative
pregnancy rates.147,155 As with clomifene, multiple
pregnancies remain a major drawback of gonadotropins,156–158 but this complication can be substantially
reduced with adequate examination and a low threshold
for readiness to cancel the stimulation. In addition,
polycystic ovaries are very sensitive to gonadotropic
stimulation and ovarian hyperstimulation syndrome is
a serious, potentially life-threatening, outcome of
induction of ovulation with gonadotropins in patients
with polycystic ovaries.159
Ovarian drilling with laser or diathermy has also been
shown to be very effective in the induction of ovulation in
women with polycystic ovary syndrome,160 but has risks
related to the operation and development of intrapelvic
adhesions. Benefits might be longlasting, and the risks
of multiple pregnancies are not increased.
Insulin sensitisers, including thiazolidinediones108
and D-chiro-inositol,161 have also been shown to increase
ovulation and reduce hyperandrogenism in women with
polycystic ovary syndrome, but metformin remains the
most commonly used agent. Metformin is not approved
by the US Food and Drug Administration (FDA) to
induce ovulation, and the best possible dose is unknown.
However, metformin does not seem to be associated
with any known fetal toxic effect or teratogenicity. In a
meta-analysis, metformin was shown to have a
substantial benefit in the induction of ovulation in
women with polycystic ovary syndrome.162 Ovulation
was achieved in 46% of women who received metformin,
compared with 24% in the placebo group (numbers
needed to treat=4∙4), which is similar to the benefit
conferred by clomifene.162 However, a 6-month
multicentre trial that directly compared the effects of
clomifene and metformin as single-agents found
clomifene was better than metformin overall for
treatment of infertility.164 This trial showed that the
livebirth rate in women given clomifene (22∙5%, 47 of
209) was higher than in those given metformin (7∙2%,
15 of 208; p<0∙001).164 The first adequately powered
multicentre trial to examine the combined effect of
clomifene and metform showed no benefit for ovulation
or pregnancy rates compared with clomifene alone
(cumulative pregnancy rate 40% vs 46%; risk difference
−6%; 95% CI −20 to 7).163 This finding was recently
supported in a study that also showed no additional
beneficial effect of combination treatment on livebirth
rate compared with clomifene alone.164 Although
multiple pregnancies occurred only in women treated
with clomifene and not with metformin, the overall rate
(about 5%) was low.164 These studies have shown the
691
Seminar
need for increased scrutiny of the role of metformin in
the treatment of infertility in women with polycystic
ovary syndrome.
Induction of ovulation with clomifene or gonadotropins
might be associated with a higher rate of early pregnancy
loss in women with polycystic ovary syndrome than in
women who ovulate and conceive normally, but this
effect is difficult to prove definitively. Similarly, whether
women with the syndrome have a higher rate of early
pregnancy loss because of endocrine disruptions that are
inherent to the disorder or whether early pregnancy loss
is higher in women with polycystic ovary syndrome
because of treatment for induction of ovulation per se is
debatable, although mounting evidence favours the first
hypothesis.165 Hyperinsulinaemia, insulin resistance, or
both, might have a key role in the pathological cause of
early pregnancy loss, prompting studies to examine the
potential benefit of metformin treatment to reduce its
occurrence. However, at present there are no adequately
designed studies to address the role of metformin in the
reduction of the putative increased frequency of early
pregnancy loss in women with polycystic ovary
syndrome,166 although some randomised trials have
shown decreased early pregnancy loss in groups treated
with metformin.167,168 By contrast, a large multicentre trial
showed a non-significant but concerning increased rate
of first-trimester pregnancy loss in the group treated with
metformin (10 of 25 individuals) compared with that of
clomifene-containing groups (14 of 62 individuals in the
group treated only with clomifene, and 20 of 80 individuals
in the group treated with both clomifene and
metformin).164 In this study, metformin treatment was
stopped with confirmation of pregnancy.
In vitro fertilisation is the last option that should be
considered in the treatment of infertility in anovulatory
women with polycystic ovary syndrome, but is often
needed when infertility is related to men or to unrelated
additional female factors. By contrast with protocols to
induce ovulation that aim to produce a single dominant
follicle in anovulatory women, hyperstimulation with
gonadotropins before in-vitro fertilisation (IVF) aims to
inhibit dominant follicle selection and promote
multifollicular growth for subsequent surgical aspiration
of mature oocytes, whether a woman is anovulatory or
not. Similar to induction of ovulation with gonadotropins,
ovarian hyperstimulation syndrome is a common
complication of ovarian hyperstimulation in women with
polycystic ovaries.159 Lower doses of FSH, early
cancellation, and coasting (ie, avoidance of FSH for
several days) might be needed to avoid ovarian
hyperstimulation syndrome.169 Retrieval of immature
oocytes followed by in-vitro maturation without
gonadotropin stimulation is an emerging alternative
option for infertile women with polycystic ovary syndrome
who are prone to ovarian hyperstimulation syndrome.170
A meta-analysis of early pregnancy in women with
polycystic ovary syndrome after IVF showed that
692
pregnancy rates, miscarriage rates, and birthweight were
equivalent to those in controls.171 However, the pregnancies
of women with the syndrome are more likely to be
complicated by gestational diabetes, pre-eclampsia,
pregnancy hypertension, and preterm labour, and
neonates are more likely to be admitted to intensive care
with a higher perinatal mortality rate, unrelated to
multiple pregnancy166 (panel). These data were supported
by a large prospective trial164 of women with polycystic
ovary syndrome treated with metformin, clomifene, or
their combination to induce ovulation, who were followed
up from conception to delivery. The study showed that
the most common pregnancy complications (in
descending order) were: pre-eclampsia, gestational
diabetes, and preterm labour.164 Overall, the rate of
pregnancy complications after fetal heart motion
approached 40%. The status of polycystic ovary syndrome
of women undergoing fertility treatment should therefore
be established before starting treatment protocols.
Issues for peripuberty
Overweight children are more likely to have premature
puberty than normal-weight children, and those with a
low birthweight, premature pubarche, or both, are
particularly prone to an early menarche and development
of polycystic ovary syndrome in adolescence.172 Ibanez
and de Zegher143 prevented insulin resistance and
features of polycystic ovary syndrome in young girls
with premature pubarche by administration of
metformin with very low doses of an androgen
antagonist, flutamide, alone or in combination with an
oral contraceptive containing drospirenone as the
progestagen. These findings need verification through
adequately powered randomised controlled trials.
Whether this effect can be generalised to adult women
and to women of diverse ethnicities remains to be
Panel: Complications of infertility treatment, pregnancy,
and the perinatal period that are significantly increased in
women with polycystic ovary syndrome
Infertility treatment
• Multiple pregnancy after ovulation induction
• Ovarian hyperstimulation syndrome
• IVF cycle cancellation
Pregnancy
• Early pregnancy loss
• Gestational diabetes
• Pregnancy-induced hypertension
• Pre-eclampsia
• Delivery by caesarean section
Perinatal period
• Admission to a neonatal intensive care unit
• Perinatal mortality
• Premature delivery
www.thelancet.com Vol 370 August 25, 2007
Seminar
ascertained. Caution should be applied before
administration of these drugs to children and
adolescents because of potential teratogenicity in an
unplanned pregnancy.
Health issues for family members
Although the genetics of polycystic ovary syndrome
remain unclear, a strong familial component exists, as
shown by family studies and twin records. The discovery
that insulin resistance and hyperandrogenaemia are
more common in the sisters of women with polycystic
ovary syndrome173 than in other women led to additional
studies which show that first-degree relatives of women
with polycystic ovary syndrome have similar metabolic
disorders, possibly predisposing to metabolic and
cardiovascular disease.174–177 In a large family study of
336 women with polycystic ovary syndrome and
307 probands, indicators of hyperinsulinaemia were
more common in the sisters of women with the syndrome
than in control women, and hyperandrogenaemia was
the major predictor of insulin resistance.107 In
162 non-Hispanic white mothers of women with
polycystic ovary syndrome, the total cholesterol and LDL
cholesterol concentrations were higher than in 62 control
women, whereas triglyceride and HDL cholesterol
concentrations did not change.177 Therefore, the diagnosis
of polycystic ovary syndrome should initiate a thorough
review of family members in addition to investigation of
the patient.
Additional clinical investigations
Because of its genetic and metabolic implications, clinical
investigation of polycystic ovary syndrome should include
examination of family history of diabetes mellitus,
cardiovascular disease, and hyperlipidaemia, possibly
with assessment of relevant risk factors in siblings and
older family members. Lifestyle issues, including history
of diet and exercise, should be investigated. Clinical
measurements might include calculation of BMI, relative
waist circumference (waist to hip ratio), serum lipids
(cholesterol, triglycerides, and HDL cholesterol), and
glucose metabolism. Assessment of insulin either as a
fasting hormone or as a surrogate of insulin resistance
(eg, homoeostasis model assessment) is of little clinical
value, although widely used for research studies. Repeated
measurements of glucose and lipid status should take
place more regularly in women with polycystic ovary
syndrome than in women without the syndrome, because
conversion from healthy to pathological status happens
more frequently in the disorder.71,119,120
Health-related quality of life is generally worse in
women with polycystic ovary syndrome than in women
without the disorder, and appropriate counselling might
be needed for some individuals.178 A questionnaire of
health-related quality of life specific for women with
polycystic ovary syndrome has been developed and
validated for this purpose.179,180 Psychological studies have
www.thelancet.com Vol 370 August 25, 2007
shown higher frequency of depression and psychological
and psychosexual morbidity in women with polycystic
ovary syndrome than in women without the disorder,
women with other non-reproductive diseases, or both.181
Obesity and hirsutism have a major effect on
health-related quality of life in women with polycystic
ovary syndrome, and improvement of these physical
symptoms might greatly improve the psychosocial and
psychosexual situation for these women.182
Future prospects
Polycystic ovary syndrome is a diverse and complex
female endocrine disorder, which is presently recognised
as a major economic health burden that is likely to
expand together with obesity.3 Future priorities in relation
to polycystic ovary syndrome include the development of
evidence-based criteria for diagnosis and treatment, and
determination of the natural history, cause, long-term
consequences, and prevention of the disorder.183
Conflict of interest statement
RSL has served as a consultant to Glaxo Smith Kline and Ferring, and
has received lecture fees from Serono, meeting support from Abbott,
and grant support from Pfizer. Other authors declare that they have no
conflict of interest.
References
1
Carmina E, Rosato F, Janni A, Rizzo M, Longo RA. Extensive
clinical experience: relative prevalence of different androgen excess
disorders in 950 women referred because of clinical
hyperandrogenism. J Clin Endocrinol Metab 2006; 91: 2–6.
2
Ehrmann DA. Polycystic ovary syndrome. N Engl J Med 2005; 352:
1223–36.
3
Azziz R, Marin C, Hoq L, Badamgarav E, Song P. Health care-related
economic burden of the polycystic ovary syndrome during the
reproductive life span. J Clin Endocrinol Metab 2005; 90: 4650–58.
4
Ehrmann DA, Liljenquist DR, Kasza K, Azziz R, Legro RS,
Ghazzi MN. Prevalence and predictors of the metabolic syndrome
in women with polycystic ovary syndrome. J Clin Endocrinol Metab
2006; 91: 48–53.
5
Essah PA, Nestler JE. The metabolic syndrome in polycystic ovary
syndrome. J Endocrinol Invest 2006; 29: 270–80.
6
Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop
Group. Revised 2003 consensus on diagnostic criteria and
long-term health risks related to polycystic ovary syndrome.
Fertil Steril 2004; 81: 19–25.
7
Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop
Group. Revised 2003 consensus on diagnostic criteria and
long-term health risks related to polycystic ovary syndrome (PCOS).
Hum Reprod 2004; 19: 41–47.
8
Azziz R. Controversy in clinical endocrinology: diagnosis of
polycystic ovarian syndrome: the Rotterdam criteria are premature.
J Clin Endocrinol Metab 2006; 91: 781–85.
9
Franks S. Controversy in clinical endocrinology: diagnosis of
polycystic ovarian syndrome: in defense of the Rotterdam criteria.
J Clin Endocrinol Metab 2006; 91: 786–89.
10 Dewailly D, Catteau-Jonard S, Reyss AC, Leroy M, Pigny P.
Oligoanovulation with polycystic ovaries but not overt
hyperandrogenism. J Clin Endocrinol Metab 2006; 91: 3922–27.
11 Azziz R, Carmina E, Dewailly D, et al. Positions statement: criteria
for defining polycystic ovary syndrome as a predominantly
hyperandrogenic syndrome: an Androgen Excess Society guideline.
J Clin Endocrinol Metab 2006; 91: 4237–45.
12 Carmina E, Lobo RA. Do hyperandrogenic women with normal
menses have polycystic ovary syndrome? Fertil Steril 1999; 71: 319–22.
13 Broekmans FJ, Knauff EA, Valkenburg O, Laven JS, Eijkemans MJ,
Fauser BC. PCOS according to the Rotterdam consensus criteria:
change in prevalence among WHO-II anovulation and association
with metabolic factors. BJOG 2006; 113: 1210–17.
693
Seminar
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
694
Welt CK, Gudmundsson JA, Arason G, et al. Characterizing discrete
subsets of polycystic ovary syndrome as defined by the Rotterdam
criteria: the impact of weight on phenotype and metabolic features.
J Clin Endocrinol Metab 2006; 91: 4842–48.
Goodarzi MO, Quinones MJ, Azziz R, Rotter JI, Hsueh WA,
Yang H. Polycystic ovary syndrome in Mexican-Americans:
prevalence and association with the severity of insulin resistance.
Fertil Steril 2005; 84: 766–69.
Rodin DA, Bano G, Bland JM, Taylor K, Nussey SS. Polycystic
ovaries and associated metabolic abnormalities in Indian
subcontinent Asian women. Clin Endocrinol 1998; 49: 91–99.
Davis SR, Knight S, White V, Claridge C, Davis BJ, Bell R.
Preliminary indication of a high prevalence of polycystic ovary
syndrome in indigenous Australian women. Gynecol Endocrinol
2002; 16: 443–46.
Conway GS, Honour JW, Jacobs HS. Heterogeneity of the polycystic
ovary syndrome: clinical, endocrine and ultrasound features in
556 patients. Clin Endocrinol 1989; 30: 459–70.
Balen AH, Conway GS, Kaltsas G, et al. Polycystic ovary syndrome:
the spectrum of the disorder in 1741 patients. Hum Reprod 1995; 10:
2107–11.
Chang WY, Knochenhauer ES, Bartolucci AA, Azziz R. Phenotypic
spectrum of polycystic ovary syndrome: clinical and biochemical
characterization of the three major clinical subgroups. Fertil Steril
2005; 83: 1717–23.
Carmina E, Koyama T, Chang L, Stanczyk FZ, Lobo RA. Does
ethnicity influence the prevalence of adrenal hyperandrogenism
and insulin resistance in polycystic ovary syndrome?
Am J Obstet Gynecol 1992; 167: 1807–12.
DeUgarte CM, Woods KS, Bartolucci AA, Azziz R. Degree of facial
and body terminal hair growth in unselected black and white
women: toward a populational definition of hirsutism.
J Clin Endocrinol Metab 2006; 91: 1345–50.
Azziz R, Carmina E, Dewailly D, et al. Criteria for defining
polycystic ovary syndrome as a predominantly hyperandrogenic
syndrome: an androgen excess Society guideline.
J Clin Endocrinol Metab 2006; 91: 4237–45.
Mathur RS, Moody LO, Landgrebe S, Williamson HO. Plasma
androgens and sex hormone-binding globulin in the evaluation of
hirsute females. Fertil Steril 1981; 35: 29–35.
Miller KK, Rosner W, Lee H, et al. Measurement of free testosterone
in normal women and women with androgen deficiency:
comparison of methods. J Clin Endocrinol Metab 2004; 89: 525–33.
Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of
simple methods for the estimation of free testosterone in serum.
J Clin Endocrinol Metab 1999; 84: 3666–72.
Welt CK, Arason G, Gudmundsson JA, et al. Defining constant
versus variable phenotypic features of women with polycystic ovary
syndrome using different ethnic groups and populations.
J Clin Endocrinol Metab 2006; 91: 4361–68.
Ferriman D, Gallwey JD. Clinical assessment of body hair growth in
women. J Clin Endocrinol Metab 1961; 21: 1440–47.
Ayala C, Steinberger E, Smith KD, Rodriguez-Rigau LJ, Petak SM.
Serum testosterone levels and reference ranges in reproductive-age
women. Endocr Pract 1999; 5: 322–29.
Jonard S, Pigny P, Jacquesson L, Demerle-Roux C, Robert Y,
Dewailly D. The ovarian markers of the FSH insufficiency in
functional hypothalamic amenorrhoea. Hum Reprod 2005; 20: 101–07.
Adams J, Franks S, Polson DW, et al. Multifollicular ovaries: clinical
and endocrine features and response to pulsatile gonadotropin
releasing hormone. Lancet 1985; 2: 1375–79.
Balen AH, Laven JS, Tan SL, Dewailly D. Ultrasound assessment of
the polycystic ovary: international consensus definitions.
Hum Reprod Update 2003; 9: 505–14.
Bridges NA, Cooke A, Healy MJ, Hindmarsh PC, Brook CG.
Standards for ovarian volume in childhood and puberty. Fertil Steril
1993; 60: 456–60.
Pigny P, Jonard S, Robert Y, Dewailly D. Serum anti-Mullerian
hormone as a surrogate for antral follicle count for definition of
the polycystic ovary syndrome. J Clin Endocrinol Metab 2006; 91:
941–45.
Hahn S, Tan S, Elsenbruch S, et al. Clinical and biochemical
characterization of women with polycystic ovary syndrome in North
Rhine-Westphalia. Horm Metab Res 2005; 37: 438–44.
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Kumar A, Woods KS, Bartolucci AA, Azziz R. Prevalence of adrenal
androgen excess in patients with the polycystic ovary syndrome
(PCOS). Clin Endocrinol 2005; 62: 644–49.
Zhang LH, Rodriguez H, Ohno S, Miller WL. Serine
phosphorylation of human P450c17 increases 17,20-lyase activity:
implications for adrenarche and the polycystic ovary syndrome.
Proc Natl Acad Sci USA 1995; 92: 10619–23.
Gilling-Smith C, Willis DS, Beard RW, Franks S. Hypersecretion of
androstenedione by isolated thecal cells from polycystic ovaries.
J Clin Endocrinol Metab 1994; 79: 1158–65.
Nelson VL, Legro RS, Strauss JF 3rd, McAllister JM. Augmented
androgen production is a stable steroidogenic phenotype of
propagated theca cells from polycystic ovaries. Mol Endocrinol 1999;
13: 946–57.
Escobar-Morreale HF, Luque-Ramirez M, San Millan JL. The
molecular–genetic basis of functional hyperandrogenism and the
polycystic ovary syndrome. Endocr Rev 2005; 26: 251–82.
Wood JR, Ho CK, Nelson-Degrave VL, McAllister JM,
Strauss JF 3rd. The molecular signature of polycystic ovary
syndrome (PCOS) theca cells defined by gene expression profiling.
J Reprod Immunol 2004; 63: 51–60.
Hughesdon PE. Morphology and morphogenesis of the
Stein-Leventhal ovary and of so-called “hyperthecosis”.
Obstet Gynecol Surv 1982; 37: 59–77.
Webber LJ, Stubbs S, Stark J, et al. Formation and early development
of follicles in the polycystic ovary. Lancet 2003; 362: 1017–21.
Maciel GA, Baracat EC, Benda JA, et al. Stockpiling of transitional
and classic primary follicles in ovaries of women with polycystic
ovary syndrome. J Clin Endocrinol Metab 2004; 89: 5321–27.
Pache TD, de Jong FH, Hop WC, Fauser BC. Association between
ovarian changes assessed by transvaginal sonography and clinical
and endocrine signs of the polycystic ovary syndrome. Fertil Steril
1993; 59: 544–99.
Jonard S, Robert Y, Cortet-Rudelli C, Pigny P, Decanter C,
Dewailly D. Ultrasound examination of polycystic ovaries: is it
worth counting the follicles? Hum Reprod 2003; 18: 598–603.
Vendola KA, Zhou J, Adesanya OO, Weil SJ, Bondy CA. Androgens
stimulate early stages of follicular growth in the primate ovary.
J Clin Invest 1998; 101: 2622–29.
Pache TD, Chadha S, Gooren LJ, et al. Ovarian morphology in
long-term androgen-treated female to male transsexuals. A human
model for the study of polycystic ovarian syndrome? Histopathology
1991; 19: 445–52.
Balen AH, Schachter ME, Montgomery D, Reid RW, Jacobs HS.
Polycystic ovaries are a common finding in untreated female to
male transsexuals. Clin Endocrinol 1993; 38: 325–29.
Bosinski HA, Peter M, Bonatz G, et al. A higher rate of
hyperandrogenic disorders in female-to-male transsexuals.
Psychoneuroendocrinology 1997; 22: 361–80.
De Leo V, Lanzetta D, D’Antona D, la Marca A, Morgante G.
Hormonal effects of flutamide in young women with polycystic
ovary syndrome. J Clin Endocrinol Metab 1998; 83: 99–102.
Mifsud A, Ramirez S, Yong EL. Androgen receptor gene CAG
trinucleotide repeats in anovulatory infertility and polycystic
ovaries. J Clin Endocrinol Metab 2000; 85: 3484–88.
Hickey T, Chandy A, Norman RJ. The androgen receptor CAG
repeat polymorphism and X-chromosome inactivation in Australian
Caucasian women with infertility related to polycystic ovary
syndrome. J Clin Endocrinol Metab 2002; 87: 161–65.
Weil S, Vendola K, Zhou J, Bondy CA. Androgen and
follicle-stimulating hormone interactions in primate ovarian follicle
development. J Clin Endocrinol Metab 1999; 84: 2951–56.
Jonard S, Dewailly D. The follicular excess in polycystic ovaries, due
to intra-ovarian hyperandrogenism, may be the main culprit for the
follicular arrest. Hum Reprod Update 2004; 10: 107–17.
Franks S, Mason H, White D, Willis D. Etiology of anovulation in
polycystic ovary syndrome. Steroids 1998; 63: 306–07.
Willis DS, Watson H, Mason HD, Galea R, Brincat M, Franks S.
Premature response to luteinizing hormone of granulosa cells from
anovulatory women with polycystic ovary syndrome: relevance to
mechanism of anovulation. J Clin Endocrinol Metab 1998; 83: 3984–91.
Wu XK, Zhou SY, Liu JX, et al. Selective ovary resistance to insulin
signaling in women with polycystic ovary syndrome. Fertil Steril
2003; 80: 954–65.
www.thelancet.com Vol 370 August 25, 2007
Seminar
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
Rice S, Christoforidis N, Gadd C, et al. Impaired insulin-dependent
glucose metabolism in granulosa-lutein cells from anovulatory
women with polycystic ovaries. Hum Reprod 2005; 20: 373–81.
Rebar R, Judd HL, Yen SS, Rakoff J, Vandenberg G, Naftolin F.
Characterization of the inappropriate gonadotropin secretion in
polycystic ovary syndrome. J Clin Invest 1976; 57: 1320–29.
DeVane GW, Czekala NM, Judd HL, Yen SS. Circulating
gonadotropins, estrogens, and androgens in polycystic ovarian
disease. Am J Obstet Gynecol 1975; 121: 496–500.
Ehrmann DA, Rosenfield RL, Barnes RB, Brigell DF, Sheikh Z.
Detection of functional ovarian hyperandrogenism in women with
androgen excess. N Engl J Med 1992; 327: 157–62.
Lanzone A, Petraglia F, Fulghesu AM, Ciampelli M, Caruso A,
Mancuso S. Corticotropin-releasing hormone induces an
exaggerated response of adrenocorticotropic hormone and cortisol
in polycystic ovary syndrome. Fertil Steril 1995; 63: 1195–99.
Blank SK, McCartney CR, Marshall JC. The origins and sequelae of
abnormal neuroendocrine function in polycystic ovary syndrome.
Hum Reprod Update 2006; 12: 351–61.
Dunaif A, Segal KR, Shelley DR, Green G, Dobrjansky A, Licholai T.
Evidence for distinctive and intrinsic defects in insulin action in
polycystic ovary syndrome. Diabetes 1992; 41: 1257–66.
Dunaif A, Graf M, Mandeli J, Laumas V, Dobrjansky A.
Characterization of groups of hyperandrogenic women with
acanthosis nigricans, impaired glucose tolerance, and/or
hyperinsulinemia. J Clin Endocrinol Metab 1987; 65: 499–507.
Ehrmann DA, Sturis J, Byrne MM, Karrison T, Rosenfield RL,
Polonsky KS. Insulin secretory defects in polycystic ovary syndrome.
Relationship to insulin sensitivity and family history of
non-insulin-dependent diabetes mellitus. J Clin Invest 1995; 96: 520–27.
Dunaif A, Finegood DT. Beta-cell dysfunction independent of
obesity and glucose intolerance in the polycystic ovary syndrome.
J Clin Endocrinol Metab 1996; 81: 942–47.
Kahn SE, Prigeon RL, McCulloch DK, et al. Quantification of the
relationship between insulin sensitivity and beta-cell function in
human subjects. Evidence for a hyperbolic function. Diabetes 1993;
42: 1663–72.
Legro RS, Kunselman AR, Dodson WC, Dunaif A. Prevalence and
predictors of risk for type 2 diabetes mellitus and impaired glucose
tolerance in polycystic ovary syndrome: a prospective, controlled
study in 254 affected women. J Clin Endocrinol Metab 1999; 84:
165–69.
Ehrmann DA, Barnes RB, Rosenfield RL, Cavaghan MK, Imperial J.
Prevalence of impaired glucose tolerance and diabetes in women
with polycystic ovary syndrome. Diabetes Care 1999; 22: 141–46.
Ehrmann DA, Kasza K, Azziz R, Legro RS, Ghazzi MN. Effects of
race and family history of type 2 diabetes on metabolic status of
women with polycystic ovary syndrome. J Clin Endocrinol Metab
2005; 90: 66–71.
Gambineri A, Pelusi C, Manicardi E, et al. Glucose intolerance in a
large cohort of mediterranean women with polycystic ovary syndrome:
phenotype and associated factors. Diabetes 2004; 53: 2353–58.
Nestler JE, Powers LP, Matt DW, et al. A direct effect of
hyperinsulinemia on serum sex hormone-binding globulin levels in
obese women with the polycystic ovary syndrome.
J Clin Endocrinol Metab 1991; 72: 83–89.
Pugeat M, Crave JC, Elmidani M, et al. Pathophysiology of sex
hormone binding globulin (SHBG): relation to insulin.
J Steroid Biochem Mol Biol 1991; 40: 841–49.
Azziz R, Ehrmann DA, Legro RS, Fereshetian AG, O’Keefe M,
Ghazzi MN. Troglitazone decreases adrenal androgen levels in
women with polycystic ovary syndrome. Fertil Steril 2003; 79: 932–37.
Nestler JE, Jakubowicz DJ. Decreases in ovarian cytochrome
P450c17 alpha activity and serum free testosterone after reduction of
insulin secretion in polycystic ovary syndrome. N Engl J Med 1996;
335: 617–23.
Adashi EY, Rock JA, Guzick D, Wentz AC, Jones GS, Jones HW Jr.
Fertility following bilateral ovarian wedge resection: a critical
analysis of 90 consecutive cases of the polycystic ovary syndrome.
Fertil Steril 1981; 36: 320–25.
Willis D, Mason H, Gilling-Smith C, Franks S. Modulation by
insulin of follicle-stimulating hormone and luteinizing hormone
actions in human granulosa cells of normal and polycystic ovaries.
J Clin Endocrinol Metab 1996; 81: 302–09.
www.thelancet.com Vol 370 August 25, 2007
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
Adashi EY, Hsueh AJ, Yen SS. Insulin enhancement of luteinizing
hormone and follicle-stimulating hormone release by cultured
pituitary cells. Endocrinology 1981; 108: 1441–49.
Dunaif A. Insulin resistance and the polycystic ovary syndrome:
mechanism and implications for pathogenesis. Endocr Rev 1997; 18:
774–800.
Diamanti-Kandarakis E, Papavassiliou AG. Molecular mechanisms
of insulin resistance in polycystic ovary syndrome. Trends Mol Med
2006; 12: 324–32
Wild RA, Applebaum-Bowden D, Demers LM, et al. Lipoprotein
lipids in women with androgen excess: independent associations
with increased insulin and androgen. Clin Chem 1990; 36:
283–89.
Talbott E, Clerici A, Berga SL, et al. Adverse lipid and coronary heart
disease risk profiles in young women with polycystic ovary syndrome:
results of a case-control study. J Clin Epidemiol 1998; 51: 415–22.
Legro RS, Kunselman AR, Dunaif A. Prevalence and predictors of
dyslipidemia in women with polycystic ovary syndrome. Am J Med
2001; 111: 607–13.
Kelly CC, Lyall H, Petrie JR, Gould GW, Connell JM, Sattar N. Low
grade chronic inflammation in women with polycystic ovarian
syndrome. J Clin Endocrinol Metab 2001; 86: 2453–55.
Diamanti-Kandarakis E, Paterakis T, Alexandraki K, et al. Indices
of low-grade chronic inflammation in polycystic ovary syndrome
and the beneficial effect of metformin. Hum Reprod 2006; 21:
1426–31.
Paradisi G, Steinberg HO, Hempfling A, et al. Polycystic ovary
syndrome is associated with endothelial dysfunction. Circulation
2001; 103: 1410–15.
Orio F Jr, Palomba S, Cascella T, et al. Early impairment of endothelial
structure and function in young normal-weight women with polycystic
ovary syndrome. J Clin Endocrinol Metab 2004; 89: 4588–93.
Vgontzas AN, Legro RS, Bixler EO, Grayev A, Kales A,
Chrousos GP. Polycystic ovary syndrome is associated with
obstructive sleep apnea and daytime sleepiness: role of insulin
resistance. J Clin Endocrinol Metab 2001; 86: 517–20.
Fogel RB, Malhotra A, Pillar G, Pittman SD, Dunaif A, White DP.
Increased prevalence of obstructive sleep apnea syndrome in obese
women with polycystic ovary syndrome. J Clin Endocrinol Metab
2001; 86: 1175–80.
Tasali E, Van Cauter E, Ehrmann DA. Relationships between sleep
disordered breathing and glucose metabolism in polycystic ovary
syndrome. J Clin Endocrinol Metab 2006; 91: 36–42.
Stein IF, Leventhal ML. Amenorrhea associated with polycystic
ovaries. Am J Obstet Gynaecol 1935; 29: 181–91.
Chang RJ, Nakamura RM, Judd HL, Kaplan SA. Insulin resistance
in nonobese patients with polycystic ovarian disease.
J Clin Endocrinol Metab 1983; 57: 356–59.
Dunaif A, Segal KR, Futterweit W, Dobrjansky A. Profound
peripheral insulin resistance, independent of obesity, in polycystic
ovary syndrome. Diabetes 1989; 38: 1165–74.
Givens JR. Ovarian hyperthecosis. N Engl J Med 1971; 285: 691.
Ferriman D, Purdie AW. The inheritance of polycystic ovarian
disease and a possible relationship to premature balding.
Clin Endocrinol 1979; 11: 291–300.
Lunde O, Magnus P, Sandvik L, Hoglo S. Familial clustering in
the polycystic ovarian syndrome. Gynecol Obstet Invest 1989;
28: 23–30.
Vink JM, Sadrzadeh S, Lambalk CB, Boomsma DI. Heritability of
polycystic ovary syndrome in a Dutch twin-family study.
J Clin Endocrinol Metab 2006; 91: 2100–04.
Stewart DR, Dombroski B, Urbanek M, et al. Fine mapping of
genetic susceptibility to polycystic ovary syndrome on chromosome
19p13.2 and tests for regulatory activity. J Clin Endocrinol Metab 2006;
91: 4112–17.
Hickey TE, Legro RS, Norman RJ. Epigenetic modification of the
X chromosome influences susceptibility to polycystic ovary
syndrome. J Clin Endocrinol Metab 2006; 91: 2789–91.
Urbanek M, Legro RS, Driscoll DA, et al. Thirty-seven candidate
genes for polycystic ovary syndrome: strongest evidence for linkage
is with follistatin. Proc Natl Acad Sci USA 1999; 96: 8573–78.
Urbanek M, Woodroffe A, Ewens KG, et al. Candidate gene region
for polycystic ovary syndrome on chromosome 19p13.2.
J Clin Endocrinol Metab 2005; 90: 6623–29.
695
Seminar
104 Tucci S, Futterweit W, Concepcion ES, et al. Evidence for association of
polycystic ovary syndrome in caucasian women with a marker at the
insulin receptor gene locus. J Clin Endocrinol Metab 2001; 86: 446–49.
105 Villuendas G, Escobar-Morreale HF, Tosi F, Sancho J, Moghetti P,
San Millan JL. Association between the D19S884 marker at the
insulin receptor gene locus and polycystic ovary syndrome.
Fertil Steril 2003; 79: 219–20.
106 Norman RJ, Masters SC, Hague W, Beng C, Pannall P, Wang JX.
Metabolic approaches to the subclassification of polycystic ovary
syndrome. Fertil Steril 1995; 63: 329–35.
107 Legro RS, Bentley-Lewis R, Driscoll D, Wang SC, Dunaif A. Insulin
resistance in the sisters of women with polycystic ovary syndrome:
association with hyperandrogenemia rather than menstrual
irregularity. J Clin Endocrinol Metab 2002; 87: 2128–33.
108 Azziz R, Ehrmann D, Legro RS, et al. Troglitazone improves
ovulation and hirsutism in the polycystic ovary syndrome: a
multicenter, double blind, placebo-controlled trial.
J Clin Endocrinol Metab 2001; 86: 1626–32.
109 Legro RS, Myers ER, Barnhart HX, et al. Pregnancy in polycystic
ovary syndrome (PPCOS) baseline characteristics: effect of race
and ethnicity on the PCOS phenotype. J Soc Gynecol Investig 2006;
13: 268a–69a.
110 Azziz R, Woods KS, Reyna R, Key TJ, Knochenhauer ES, Yildiz BO.
The prevalence and features of the polycystic ovary syndrome in an
unselected population. J Clin Endocrinol Metab 2004; 89: 2745–49.
111 Asuncion M, Calvo RM, San Millan JL, Sancho J, Avila S,
Escobar-Morreale HF. A prospective study of the prevalence of the
polycystic ovary syndrome in unselected Caucasian women from
Spain. J Clin Endocrinol Metab 2000; 85: 2434–38.
112 Diamanti-Kandarakis E, Kouli CR, Bergiele AT, et al. A survey of the
polycystic ovary syndrome in the Greek island of Lesbos: hormonal
and metabolic profile. J Clin Endocrinol Metab 1999; 84: 4006–11.
113 Taponen S, Martikainen H, Jarvelin MR, et al. Metabolic
cardiovascular disease risk factors in women with self-reported
symptoms of oligomenorrhea and/or hirsutism: Northern Finland
Birth Cohort 1966 Study. J Clin Endocrinol Metab 2004; 89: 2114–18.
114 Carmina E, Legro RS, Stamets K, Lowell J, Lobo RA. Difference in
body weight between American and Italian women with polycystic
ovary syndrome: influence of the diet. Hum Reprod 2003; 18: 2289–93.
115 Abbott DH, Barnett DK, Bruns CM, Dumesic DA. Androgen excess
fetal programming of female reproduction: a developmental aetiology
for polycystic ovary syndrome? Hum Reprod Update 2005; 11: 357–74.
116 Barnes RB, Rosenfield RL, Ehrmann DA, et al. Ovarian
hyperandrogynism as a result of congenital adrenal virilizing
disorders: evidence for perinatal masculinization of neuroendocrine
function in women. J Clin Endocrinol Metab 1994; 79: 1328–33.
117 Davies MJ, Norman RJ. Programming and reproductive
functioning. Trends Endocrinol Metab 2002; 13: 386–92.
118 Boudreaux MY, Talbott EO, Kip KE, Brooks MM, Witchel SF. Risk
of T2DM and impaired fasting glucose among PCOS subjects:
results of an 8-year follow-up. Curr Diab Rep 2006; 6: 77–83.
119 Norman RJ, Masters L, Milner CR, Wang JX, Davies MJ. Relative
risk of conversion from normoglycaemia to impaired glucose
tolerance or non-insulin dependent diabetes mellitus in polycystic
ovarian syndrome. Hum Reprod 2001; 16: 1995–98.
120 Legro RS, Gnatuk CL, Kunselman AR, Dunaif A. Changes in glucose
tolerance over time in women with polycystic ovary syndrome: a
controlled study. J Clin Endocrinol Metab 2005; 90: 3236–42.
121 Orio F, Palomba S, Colao A. Cardiovascular risk in women with
polycystic ovary syndrome. Fertil Steril 2006; 86 (suppl 1): S20–21.
122 Solomon CG, Hu FB, Dunaif A, et al. Long or highly irregular
menstrual cycles as a marker for risk of type 2 diabetes mellitus.
JAMA 2001; 286: 2421–26.
123 Alberti KG, Zimmet P, Shaw J. The metabolic syndrome—a new
worldwide definition. Lancet 2005; 366: 1059–62.
124 Carmina E, Napoli N, Longo RA, Rini GB, Lobo RA. Metabolic
syndrome in polycystic ovary syndrome (PCOS): lower prevalence
in southern Italy than in the USA and the influence of criteria for
the diagnosis of PCOS. Eur J Endocrinol 2006; 154: 141–45.
125 Coviello AD, Legro RS, Dunaif A. Adolescent girls with
polycystic ovary syndrome have an increased risk of the
metabolic syndrome associated with increasing androgen
levels independent of obesity and insulin resistance.
J Clin Endocrinol Metab 2006; 91: 492–97.
696
126 Moran LJ, Brinkworth G, Noakes M, Norman RJ. Effects of lifestyle
modification in polycystic ovarian syndrome. Reprod Biomed Online
2006; 12: 569–78.
127 Huber-Buchholz MM, Carey DG, Norman RJ. Restoration of
reproductive potential by lifestyle modification in obese polycystic
ovary syndrome: role of insulin sensitivity and luteinizing
hormone. J Clin Endocrinol Metab 1999; 84: 1470–74.
128 Kiddy DS, Hamilton-Fairley D, Bush A, et al. Improvement in
endocrine and ovarian function during dietary treatment of obese
women with polycystic ovary syndrome. Clin Endocrinol 1992; 36:
105–11.
129 Pasquali R, Gambineri A, Biscotti D, et al. Effect of long-term
treatment with metformin added to hypocaloric diet on body
composition, fat distribution, and androgen and insulin levels in
abdominally obese women with and without the polycystic ovary
syndrome. J Clin Endocrinol Metab 2000; 85: 2767–74.
130 Crosignani PG, Colombo M, Vegetti W, Somigliana E, Gessati A,
Ragni G. Overweight and obese anovulatory patients with
polycystic ovaries: parallel improvements in anthropometric
indices, ovarian physiology and fertility rate induced by diet.
Hum Reprod 2003; 18: 1928–32.
131 Moran LJ, Noakes M, Clifton PM, Tomlinson L, Galletly C,
Norman RJ. Dietary composition in restoring reproductive and
metabolic physiology in overweight women with polycystic ovary
syndrome. J Clin Endocrinol Metab 2003; 88: 812–19.
132 Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the
incidence of type 2 diabetes with lifestyle intervention or
metformin. N Engl J Med 2002; 346: 393–403.
133 Stamets K, Taylor DS, Kunselman A, Demers LM, Pelkman CL,
Legro RS. A randomized trial of the effects of two types of
short-term hypocaloric diets on weight loss in women with
polycystic ovary syndrome. Fertil Steril 2004; 81: 630–37.
134 Moran LJ, Noakes M, Clifton PM, Wittert GA, Williams G,
Norman RJ. Short-term meal replacements followed by dietary
macronutrient restriction enhance weight loss in polycystic ovary
syndrome. Am J Clin Nutr 2006; 84: 77–87.
135 Tsagareli V, Noakes M, Norman RJ. Effect of a very-low-calorie diet
on in vitro fertilization outcomes. Fertil Steril 2006; 86: 227–29.
136 Shapiro J, Lui H. Treatments for unwanted facial hair. Skin Ther Lett
2005; 10: 1–4.
137 van Vloten WA, van Haselen CW, van Zuuren EJ, Gerlinger C,
Heithecker R. The effect of 2 combined oral contraceptives
containing either drospirenone or cyproterone acetate on acne and
seborrhea. Cutis 2002; 69 (suppl 4): 2–15.
138 Harborne L, Fleming R, Lyall H, Norman J, Sattar N. Descriptive
review of the evidence for the use of metformin in polycystic ovary
syndrome. Lancet 2003; 361: 1894–901.
139 Ellis JA, Sinclair R, Harrap SB. Androgenetic alopecia:
pathogenesis and potential for therapy. Expert Rev Mol Med 2002;
2002: 1–11.
140 Lucky AW, Piacquadio DJ, Ditre CM, et al. A randomized,
placebo-controlled trial of 5% and 2% topical minoxidil solutions in
the treatment of female pattern hair loss. J Am Acad Dermatol 2004;
50: 541–53.
141 Hardiman P, Pillay OC, Atiomo W. Polycystic ovary syndrome and
endometrial carcinoma. Lancet 2003; 361: 1810–12.
142 Vrbikova J, Cibula D. Combined oral contraceptives in the
treatment of polycystic ovary syndrome. Hum Reprod Update 2005;
11: 277–91.
143 Ibanez L, de Zegher F. Low-dose flutamide-metformin therapy for
hyperinsulinemic hyperandrogenism in non-obese adolescents and
women. Hum Reprod Update 2006; 12: 243–52.
144 Hughes E, Collins J, Vandekerckhove P. Gonadotrophin-releasing
hormone analogue as an adjunct to gonadotropin therapy for
clomiphene-resistant polycystic ovarian syndrome.
Cochrane Database Syst Rev 2000; 2: CD000097.
145 Beck JI, Boothroyd C, Proctor M, Farquhar C, Hughes E. Oral
anti-oestrogens and medical adjuncts for subfertility associated with
anovulation. Cochrane Database Syst Rev 2005; 1: CD002249.
146 Imani B, Eijkemans MJ, te Velde ER, Habbema JD, Fauser BC.
Predictors of chances to conceive in ovulatory patients during
clomiphene citrate induction of ovulation in normogonadotropic
oligoamenorrheic infertility. J Clin Endocrinol Metab 1999; 84:
1617–22.
www.thelancet.com Vol 370 August 25, 2007
Seminar
147 van Santbrink EJ, Fauser BC. Ovulation induction in
normogonadotropic anovulation (PCOS).
Best Pract Res Clin Endocrinol Metab 2006; 20: 261–70.
148 Palomba S, Russo T, Orio F Jr, et al. Uterine effects of clomiphene
citrate in women with polycystic ovary syndrome: a prospective
controlled study. Hum Reprod 2006; 21: 2823–29.
149 Imani B, Eijkemans MJ, te Velde ER, Habbema JD, Fauser BC. A
nomogram to predict the probability of live birth after clomiphene
citrate induction of ovulation in normogonadotropic
oligoamenorrheic infertility. Fertil Steril 2002; 77: 91–97.
150 Casper RF, Mitwally MF. Review: aromatase inhibitors for ovulation
induction. J Clin Endocrinol Metab 2006; 91: 760–71.
151 Fatemi HM, Kolibianakis E, Tournaye H, Camus M,
Van Steirteghem AC, Devroey P. Clomiphene citrate versus
letrozole for ovarian stimulation: a pilot study. Reprod Biomed Online
2003; 7: 543–46.
152 Franks S, Gilling-Smith C. Advances in induction of ovulation.
Curr Opin Obstet Gynecol 1994; 6: 136–40.
153 Lopez E, Gunby J, Daya S, Parrilla JJ, Abad L, Balasch J. Ovulation
induction in women with polycystic ovary syndrome: randomized
trial of clomiphene citrate versus low-dose recombinant FSH as
first line therapy. Reprod Biomed Online 2004; 9: 382–90.
154 Christin-Maitre S, Hugues JN. A comparative randomized
multicentric study comparing the step-up versus step-down
protocol in polycystic ovary syndrome. Hum Reprod 2003; 18:
1626–31.
155 Mulders AG, Eijkemans MJ, Imani B, Fauser BC. Prediction of
chances for success or complications in gonadotrophin ovulation
induction in normogonadotrophic anovulatory infertility.
Reprod Biomed Online 2003; 7: 170–78.
156 Schenker JG, Yarkoni S, Granat M. Multiple pregnancies following
induction of ovulation. Fertil Steril 1981; 35: 105–23.
157 Navot D, Goldstein N, Mor-Josef S, Simon A, Relou A, Birkenfeld A.
Multiple pregnancies: risk factors and prognostic variables during
induction of ovulation with human menopausal gonadotrophins.
Hum Reprod 1991; 6: 1152–55.
158 Dickey RP, Olar TT. Hormone treatment for infertility. Restrictions
won’t prevent multiple pregnancies. BMJ 1993; 307: 1281–82.
159 Tummon I, Gavrilova-Jordan L, Allemand MC, Session D. Polycystic
ovaries and ovarian hyperstimulation syndrome: a systematic
review*. Acta Obstet Gynecol Scand 2005; 84: 611–16.
160 Farquhar C, Lilford RJ, Marjoribanks J, Vandekerckhove P.
Laparoscopic “drilling” by diathermy or laser for ovulation
induction in anovulatory polycystic ovary syndrome.
Cochrane Database Syst Rev 2005; 3: CD001122.
161 Nestler JE, Jakubowicz DJ, Reamer P, Gunn RD, Allan G. Ovulatory
and metabolic effects of D-chiro-inositol in the polycystic ovary
syndrome. N Engl J Med 1999; 340: 1314–20.
162 Lord JM, Flight IH, Norman RJ. Metformin in polycystic ovary
syndrome: systematic review and meta-analysis. BMJ 2003; 327:
951–53.
163 Moll E, Bossuyt PM, Korevaar JC, Lambalk CB, van der Veen F.
Effect of clomifene citrate plus metformin and clomifene citrate
plus placebo on induction of ovulation in women with newly
diagnosed polycystic ovary syndrome: randomised double blind
clinical trial. BMJ 2006; 332: 1485.
164 Legro RS, Barnhart HX, Schlaff WD, et al. Clomiphene, metformin,
or both for infertility in the polycystic ovary syndrome. N Engl J Med
2007; 356: 551–66.
165 Homburg R. Pregnancy complications in PCOS.
Best Pract Res Clin Endocrinol Metab 2006; 20: 281–92.
166 Boomsma CM, Eijkemans MJ, Hughes EG, Visser GH, Fauser BC,
Macklon NS. A meta-analysis of pregnancy outcomes in women
with polycystic ovary syndrome. Hum Reprod Update 2006; 12:
673–83.
www.thelancet.com Vol 370 August 25, 2007
167 Palomba S, Orio F, Jr., Nardo LG, et al. Metformin administration
versus laparoscopic ovarian diathermy in clomiphene
citrate-resistant women with polycystic ovary syndrome: a
prospective parallel randomized double-blind placebo-controlled
trial. J Clin Endocrinol Metab 2004; 89: 4801–09.
168 Palomba S, Falbo A, Orio F Jr, et al. A randomized controlled trial
evaluating metformin pre-treatment and co-administration in
non-obese insulin-resistant women with polycystic ovary syndrome
treated with controlled ovarian stimulation plus timed intercourse
or intrauterine insemination. Hum Reprod 2005; 20: 2879–86.
169 Delvigne A, Rozenberg S. Epidemiology and prevention of ovarian
hyperstimulation syndrome (OHSS): a review. Hum Reprod Update
2002; 8: 559–77.
170 Chian RC, Buckett WM, Abdul Jalil AK, et al. Natural-cycle in vitro
fertilization combined with in vitro maturation of immature oocytes
is a potential approach in infertility treatment. Fertil Steril 2004; 82:
1675–78.
171 Heijnen EM, Eijkemans MJ, Hughes EG, Laven JS, Macklon NS,
Fauser BC. A meta-analysis of outcomes of conventional IVF in
women with polycystic ovary syndrome. Hum Reprod Update 2006;
12: 13–21.
172 Ibanez L, Valls C, Potau N, Marcos MV, de Zegher F. Polycystic
ovary syndrome after precocious pubarche: ontogeny of the
low-birthweight effect. Clin Endocrinol 2001; 55: 667–72.
173 Norman RJ, Masters S, Hague W. Hyperinsulinemia is common in
family members of women with polycystic ovary syndrome.
Fertil Steril 1996; 66: 942–47.
174 Yildiz BO, Yarali H, Oguz H, Bayraktar M. Glucose intolerance,
insulin resistance, and hyperandrogenemia in first degree relatives
of women with polycystic ovary syndrome. J Clin Endocrinol Metab
2003; 88: 2031–36.
175 Kaushal R, Parchure N, Bano G, Kaski JC, Nussey SS. Insulin
resistance and endothelial dysfunction in the brothers of Indian
subcontinent Asian women with polycystic ovaries. Clin Endocrinol
2004; 60: 322–28.
176 Leibel NI, Baumann EE, Kocherginsky M, Rosenfield RL.
Relationship of adolescent polycystic ovary syndrome to parental
metabolic syndrome. J Clin Endocrinol Metab 2006; 91: 1275–83.
177 Sam S, Legro RS, Essah PA, Apridonidze T, Dunaif A. Evidence for
metabolic and reproductive phenotypes in mothers of women with
polycystic ovary syndrome. Proc Natl Acad Sci USA 2006; 103:
7030–35.
178 Elsenbruch S, Benson S, Hahn S, et al. Determinants of emotional
distress in women with polycystic ovary syndrome. Hum Reprod
2006; 21: 1092–99.
179 Guyatt G, Weaver B, Cronin L, Dooley JA, Azziz R. Health-related
quality of life in women with polycystic ovary syndrome, a
self-administered questionnaire, was validated. J Clin Epidemiol
2004; 57: 1279–87.
180 Coffey S, Bano G, Mason HD. Health-related quality of life in
women with polycystic ovary syndrome: a comparison with the
general population using the Polycystic Ovary Syndrome
Questionnaire (PCOSQ) and the Short Form-36 (SF-36).
Gynecol Endocrinol 2006; 22: 80–86.
181 Coffey S, Mason H. The effect of polycystic ovary syndrome on
health-related quality of life. Gynecol Endocrinol 2003; 17: 379–86.
182 Hahn S, Benson S, Elsenbruch S, et al. Metformin treatment of
polycystic ovary syndrome improves health-related quality-of-life,
emotional distress and sexuality. Hum Reprod 2006; 21: 1925–34.
183 Legro RS, Azziz R, Giudice L. A twenty-first century research
agenda for polycystic ovary syndrome.
Best Pract Res Clin Endocrinol Metab 2006; 20: 331–36.
697
Series
Health and Human Rights 4
Violations of human rights: health practitioners as witnesses
James Orbinski, Chris Beyrer, Sonal Singh
Lancet 2007; 370: 698–704
This is the fourth in a Series of
four papers about health and
human rights
See Comment page 637
Department of Family and
Community Medicine,
St Michael’s Hospital, Toronto,
Ontario, Canada
(J Orbinski MD); Bloomberg
School of Public Health, Johns
Hopkins University, Baltimore,
MD, USA (C Beyrer MD); and
Wake Forest University School
of Medicine, Winston-Salem,
NC, USA (S Singh MD)
Correspondence to:
James Orbinski, Department of
Family and Community
Medicine, St Michael’s Hospital,
Toronto, Ontario, Canada,
M5B 1W8
[email protected]
For humanitarian health-care practitioners bearing witness to violations of human dignity has become synonymous
with denunciations, human rights advocacy, or lobbying for political change. A strict reliance on legal interpretations
of humanitarianism and human rights is inadequate for fully understanding the problems inherent in political
change. With examples from the HIV/AIDS epidemic in the USA, the Rwandan genocide, and physician-led political
activism in Nepal, we describe three cases in which health practitioners bearing witness to humanitarian and
human-rights issues have had imperfect outcomes. However these acts of bearing witness have been central to the
promotion of humanitarianism and human rights, to the pursuit of justice that they have inevitably and implicitly
endorsed, and thus to the politics that have or might yet address these issues. Despite the imperfections, bearing
witness, having first-hand knowledge of humanitarian and human-rights principles and their limitations, and
systematically collecting evidence of abuse, can be instrumental in tackling the forces that constrain the realisation of
human health and dignity.
Introduction
The humanitarian community’s longstanding principles
of neutrality and impartiality have been challenged by a
host of developments since the 1990s—including an often
open disregard for international law by war criminals,
direct targeting of civilians and aid workers, the use of
foreign aid to fuel conflicts, and the emergence of
pandemics and epidemics with unprecedented rates of
transmission.1 Organisations with interests in humanitarianism and human rights historically sought to reduce
human suffering and save lives, and took a studiously
apolitical stance. Lately, however, non-governmental
organisations have been challenged by both their constituents and their benefactors to support or condemn government interventions and at times, have acquiesced in, and
therefore implicitly accepted, military intervention.
Such a position would have been quite inconceivable a
couple of decades ago. With case examples from the
HIV/AIDS epidemic in the USA, genocide in Rwanda,
and the role of health professionals witnessing human
rights abuses in Nepal, we discuss the shift in
humanitarian and human rights ideology, specifically the
movement away from neutrality towards an activist
approach. We also discuss the responsibility of medical
practitioners to document and bear witness to violations
of human rights, and to intervene to alleviate suffering if
possible. We also assess the roles of individual
organisations, bearing in mind that any organisation is
the sum of its participants, and that its emerging
philosophies represent the beliefs of the individuals
involved with these groups, or the voice of a dominant
few that might or might not have great experience and
knowledge.
Historical institutional perspectives
Many people understand that the inherent monopoly on
the power of the state needs a method of monitoring
governments to prevent them lapsing into tyranny.
698
Effective monitoring mechanisms must, by their very
nature, be independent of the state. Modern
non-governmental organisations tend to follow one of
two main models in their relations with the governments
that they monitor and their role as witnesses to human
rights’ abuses.
First, is what we will call the state-reportage model, in
which groups such as the International Committee of the
Red Cross and Red Crescent Societies (ICRC), use their
influence within the corridors of power and mainly report
directly to government officials. This strategy places the
onus on the government officials to react to the reports
and ensures that ICRC can continue their humanitarian
efforts within the country. However, government officials
can choose to ignore these reports and then the option
for ICRC is to make the reports public. This action also
threatens their perceived neutrality, and yet is possible
because of their independent status, their access to
vulnerable populations and can damage their relations
with the government. Examples of such indifference to
ICRC reports include those about abuses at the US
military prison at Guantanamo Bay, Cuba and the prison
Abu Ghraib in Iraq that were provided repeatedly to US
officials, without any remedial action being taken by the
US government.2
By contrast with the state-reportage model, what we
will call the wide-dissemination model, in which
organisations, such as Amnesty International and
Human Rights Watch, believe that accountability extends
beyond the closed doors of governments and seek to
disseminate information widely about countries’
human-rights behaviours to a worldwide audience. This
strategy is based on the belief that the power of the state
can be limited by providing accurate information to its
citizens and the worldwide community. Authoritarian
regimes, in addition to several democracies, have been
systematically embarrassed in the mass media as a result
of evidence brought to light by these organisations. The
www.thelancet.com Vol 370 August 25, 2007
Series
result of these tactics is often an antagonistic relationship
between these organisations and the states they monitor,
and these groups often need to sacrifice a degree of
mobility for absolute autonomy, and therefore often rely
on a network of confidential informants and unpaid
volunteers.
Médecins Sans Frontières works in a hybrid of both
models; it delivers aid and bears witness. Originally this
organisation was committed equally to direct humanitarian assistance and to bearing witness when human
rights or humanitarian principles are violated. In practice,
Médecins Sans Frontières encounters both the difficulties
experienced by state-reportage groups such as ICRC and
organisations that use the wide-dissemination approach.
They seek to meet humanitarian needs and to live in
such dilemmas rather than always to solve them. Bearing
witness is central to the humanitarian cause in that it
shows the reality of human suffering at the hands of
states violating the rules of war or because of state policies
(eg, intellectual property issues affecting drug access,
and health worker shortages) that contribute to
suffering.
However, for Médecins Sans Frontières, bearing
witness is not a priority, but becomes imperative if no
other alternative is available. In the present crisis zone in
Darfur, Sudan, they chose to break their silence over the
systematic rape of clinic attendees by government forces.
In a report in May, 2005, Médecins Sans Frontières
described the number of rape victims attending their
clinics in Darfur and extrapolated these findings to the
general Darfur area.3 As a consequence of disseminating
this information, Paul Foreman, Médecins Sans
Frontières’ head-of-mission for Darfur was arrested by
Sudanese police and charged with crimes against the
state.4 Médecins Sans Frontières made a difficult decision
to disseminate this information, even though their
delivery of care might be jeapordised, and indeed a
consequence of this decision has been a reduced presence
in the Darfur region. Yet, this report was the only available
systematic collection of information about rape, although
the use of rape as a weapon in the region was the subject
of much speculation. The decision to release this
information, and its consequences, were a direct response
to the challenges presented by the issues of contemporary
crimes against humanity including ethnic cleansing.
Each of these approaches has strengths and limitations. Human Rights Watch and Amnesty International
have been criticised for supporting mainly civil and
political rights without giving pragmatic aid in the form
of health or economic services.5 The approach of ICRC,
is to deliver aid while avoiding publicly criticising the
states in which it operates, so they can continue to
provide pragmatic aid. Intuitively, one might think that
having direct access to decisionmakers would benefit
the goals of the organisation; but, in reality we see time
and time again organisations lauded by their host
countries as long as their objectives do not interfere
www.thelancet.com Vol 370 August 25, 2007
with those of the host. As soon as the conformity lapses,
so too does all the goodwill accumulated over time—as
was seen with Médecins Sans Frontières in Sudan.
Neither of these strategies can force changes in state
policy for the benefit of its people.
Dominant philosophical paradigms often seem so
immovable that they are like the very air we breathe, they
surround us yet for the most part we are unaware of their
existence, or at best simply take them for granted. Only a
massive shift in the perception of an issue or discipline
can compel us to see these matters in a new light, and
recognise our unconscious biases. The emergence of
three such threats to human rights—pandemics such as
that of HIV/AIDS, genocide and crimes against humanity
as seen in Rwanda and Darfur, and gross violations of
human rights and intimidation of caregivers—might yet
force such a necessary shift in thinking by the
humanitarian and medical community. In the following
section, we bear witness to the violations of human rights
and humanitarian law that we have encountered as
medical workers.
Chris Beyrer, New York City, USA, 1986
For people who witnessed and survived the early years of
HIV/AIDS in the USA, the unconscionable reluctance of
the Reagan administration to respond to the unfolding
epidemic remains a source of anger, pain, and grief.6 In
New York City, the HIV prevalence among homosexual
and bisexual men was roughly 50% by the time the first
HIV tests became available in 1985. Half of us.
In 1986, I was a third-year medical student at the State
University of New York Downstate Medical Center in
East Flatbush, Brooklyn. Our clinical training was based
at Kings County Hospital, the public hospital of last
resort for the people of Brooklyn, and the largest in the
city. This was a difficult time to be studying medicine.
Needle sticks were terrifying, the wards at Downstate
were overflowing with gravely ill and dying AIDS patients,
the old tuberculosis ward had reopened and was past
capacity, and all we could offer was treatment for
opportunistic infections—our patients’ AIDS progressed,
essentially unaffected by our interventions. At that time,
East Flatbush was the largest Haitian community outside
Haiti, and was desperately poor. The fiscal crisis of New
York City had brutal effects on the public hospital system,
but these effects were felt most acutely, as always, by poor
and minority communities in the outer boroughs—Kings
County was short of everything from soap and toilet
paper, to medical supplies and support staff. Short of
everything but patients.
The challenges of trying to learn medicine at a centre
of the New York AIDS epidemic were shared by all my
classmates, but a special poignancy was felt by the gay
men among us—since we knew we had a one in two
chance of becoming one of the febrile, coughing young
men we admitted night after night. Our professors and
699
Series
residents were generally respectful, but morning rounds
would be laced with phrases like “risk factor—
homosexual”. We were doctors and they were patients,
and the patients were the homosexuals—not the medical
staff—the presumption was that we were all straight and
had nothing to fear. And then in that autumn of 1986, my
lover starting having night sweats, weight loss, fevers,
and a persistent cough. I surreptitiously palpated his
lymph nodes—yes, greatly enlarged. He was never really
well again. And for the next 5 years, as I finished medical
school, internship, and residency, we lived as a discordant
couple, lived with AIDS, and lost, as so many other gay
men did, friend after friend.
Although the personal difficulties we faced were large
and real, the toughest challenges for us were those related
to access to health care and to benefits, issues that still
haunt same-sex couples in countless places. My partner
was an actor, and had no health insurance. I was a
resident doctor when he was most ill, and the training
programme that insured me and the other physicians
in-training had family benefits packages, but these
packages did not cover same-sex couples. He could not
be covered by my benefits, and I, in turn, could take no
“family” leave time to care for him. Indeed, my managers
insisted that I make up all the time (about 3 weeks) that I
took off when he was dying, and the days of the funeral
and memorial services. Our largest problem was that he
did qualify for Medicaid (after we spent all of our savings
on his first long hospitalisation) but to keep those benefits
he had to stay registered in the state in which he became
sick (New York), and not the state in which we lived
during my training (Maryland). Every illness meant a
long drive across three states, usually in the middle of
night, to get an emergency room in New York for
treatment. Our lives changed substantially for the better
when HIV/AIDS was included in the Americans With
Disabilities Act,7 and people with AIDS became eligible
for disability benefits.
Corbis
The printed journal
includes an image merely
for illustration
Rwandan refugees forced to flee to safety during the genocide
700
As a junior physician treating AIDS patients while
living with my lover through the last tough stages of the
disease, the roles of witness, partner, caregiver, and
physician blurred. Our last challenge was an AIDS-related
malignancy. When we finally got into the AIDS ward at a
wonderful hospital in New York City, my partner’s
physician recognised my dilemma and, for the first time
in 5 years of dealing with health-care providers, addressed
it openly. He pulled me aside once we were clearly out of
clinical options, and said: “Let me be the doctor here. You
just be his lover.”
Although physicians have some training in
compassionate care, and learn quickly that a patient’s
social and family supports are essential for coping with
their disease—to understand, to know how unjust and
dysfunctional the US health-care system can be is difficult
until you experience it. More than a decade after our
struggle, the University Hospital where I trained
implemented domestic-partner benefits for same-sex
couples. With that simple act, the huge struggles that we
went through by not having those benefits were stopped
with a signature. Rights such as domestic-partner
benefits remain rare in of the rest of the world, and are
threatened increasingly in the USA. Access to care
remains such a fundamental right—and its denial is one
of the most agonising insults to dignity one can
experience.
James Orbinski, Kigali, Rwanda, 1994
I was Head of Mission for Médecins Sans Frontières in
Kigali, Rwanda in the last weeks of the genocide in 1994.
This story could have taken place in Kosovo, Bosnia, East
Timor, Chechnya, Sri Lanka, Darfur, Burma, Uganda, or
any number of other places where human dignity is
assaulted daily by criminal acts in war.
Rwanda’s genocide was the first in which the international
community had the political freedom, because of the end
of the cold war, to act to prevent and stop the killing. Yet as
the world watched the genocide on television, the UN
Security Council equivocated on its responsibility, and
states with influence in the region chose to abstain from
assisting victims. The genocide was mostly done before
the launch of the French military’s Operation Turquoise,
which created a safe-zone covering a fifth of the country.
Acts of war and revenge killings followed in Rwanda and
in neighbouring Zaire in 1996 and 1997, and war engulfed
the region, in what some commentators have called
Africa’s World War, until 2003.8
One night in Kigali during the genocide, after many
long hours of surgery, from the hospital balcony we
watched packs of dogs roaming the streets. They were
well fed with the taste of human flesh, and wild. They
were fighting with each other over the remains of a
corpse that lay in the street, and threatening to attack a
man who had ventured outside of the hospital fence in
search of firewood. Later that night, among the thousands
of people we either treated or gave shelter to at the
www.thelancet.com Vol 370 August 25, 2007
Series
hospital, a little girl aged about nine years told me
through a translator how she had escaped murder at the
hands of the Interahamwe killing squads. She told me:
“My mother hid me in the latrine. I saw through the hole.
I watched them hit her with machetes. I watched my
mother’s arm fall into my fathers’ blood on the floor, and
I cried without noise in the toilet.” During that time,
around Rwanda, Tutsis, and moderate Hutus were being
butchered in a systematic and deliberate manner. That is
genocide.
Some people, very powerful people—people who
controlled the army, the economy, and the government,
decided that a group of people were no longer allowed to
exist. These powerful people used the apparatus of the
state to achieve their aims. They were more effective and
far more efficient than Pol Pot, or even the Nazis. People
were killed in their homes, or after being assembled in
churches, schools, and hospitals, or taken by bus or
marched to mass graves where they were not shot, but
had their hands and feet cut off—to bleed to death,
unable to climb out of the graves. People often begged,
even paid, to have their children shot, rather than for
them to suffer this particular cruelty. Between 500 000
and 800 000 individuals were murdered in 14 weeks.
This period was terrible—in the truest sense of the
word, it excited terror—an extreme human fear that I
could literally feel in the trembling hands and quivering
lips of the people I met in the city who were in hiding, or
among some of the 6000 people who lived in the hallways
and stairwells of the hospital that Médecins Sans Frontières
reopened in Kigali at that time. The people of Kigali were
now the living dead. Their trembling revealed not simply
their fear of pain, or physical trauma, but worse. It revealed
that they knew deeply, and well beyond what mere words
can ever describe, that they were truly alone, and they
knew the malevolence and the sheer calculated cruelty of
not only the killers, but also of the policies of nations such
as France, the USA, and the UK.
These states pursued their foreign policies—knowingly
—through genocide. They themselves did not hold the
machetes, did not rape women, and did not sever the
hands and feet of children. But they are culpable. They are
guilty of not only being silent bystanders as were many
other nations, and they are guilty not only of acquiescence,
but they are also guilty of complicity. France armed and
trained the genocidal army, and all three countries
obstructed, and denied the political and material means to
support the UN Assistance Mission for Rwanda. Theirs is
a complicity that is revealed by the remark of François
Mitterrand, then president of France, who said to an aide
in the summer of 1994: ”In such countries, genocide is not
too important…”9
As doctors, we could not stop the genocide, but we
refused to remain silent. We were witnesses, and we
spoke out to the world about the horror of genocide as it
was happening. Unlike the little girl in the toilet, we had
a voice and could not watch in silence. Nor could we turn
www.thelancet.com Vol 370 August 25, 2007
away in acquiescence, or become complicit by reducing
our presence to the simple silent technical medicine of
suturing the wounds of deliberate savagery. In Rwanda,
we demanded a military intervention to stop the genocide.
Why? Because humanitarianism cannot end war, create
peace, or clear consciences of the politically indifferent. It
is a human response to political failure—an immediate,
short-term act that cannot absolve political responsibility
for public security either nationally or internationally.
Respect for both the living and the dead necessitates that
we demand international intervention and local
accountability in cases of mass human rights abuses, and
refuse to accept otherwise.
Sonal Singh, Nepal, 2006
The Himalayan kingdom of Nepal has been undergoing
rapid and tumultuous political change in the past decade.
The political struggle has been accompanied by a violent
insurgency, in which nearly 20 000 lives have been lost,
200 000 people have been internally displaced, and an
estimated 2 million people have migrated to India.10
The present conflict brought widespread health and
human rights abuses, many directly affecting children.11
Health workers including doctors have borne witness to
this historical transformation in Nepal, and have
endeavoured to ensure health as a human right in the
overall macroeconomic and political picture.12 The
activities of the 1990 revolution that ushered in multiparty
democracy in Nepal were sparked by a group of dedicated
health professionals.13 Health providers in Nepal have
shown the power of, and dangers faced by, witnesses of
health and human rights violations.
Although health-care facilities had not been directly
targeted by insurgents, services housed in government
buildings were targeted, such as Village Development
Committee offices.14 Several community-health centres
were destroyed and several health workers lost their lives
in the conflict.15 Health workers also reported intimidation,
harassment, extortion, and threats by insurgents.15 They
were also threatened by the Nepalese Government for
treating victims of the violence. In November, 2001, the
Nepalese Ministry of Health issued a directive that
doctors would be liable to government action if they
treated victims of the violence without first obtaining
permission from the security forces.16 This order created
a dilemma for health professionals in Nepal;17 some
doctors continued to document human rights abuses and
torture by both sides. For taking this stance they were
supported by the Nepal Medical Association and other
physician’s advocacy groups around the world including
Physicians for Human Rights.
In great political change in April, 2006, the autocratic
regime of King Gyanendra was ousted and replaced with
a representative government; the earlier revolution in
1990 had limited some of the powers of the monarchy.
This change happened in response to widespread protests
around the country, which disrupted the fragile
701
Getty Images
Series
Nepalese protesters march against the autocratic regime of King Gyanendra
health-care infrastructure.18 Doctors and other members
of civil society including lawyers, journalists, engineers,
professors, teachers, and business people, were at the
forefront of this movement for the restoration of
parliament, and as in the earlier conflict, health
professionals were threatened because they treated
people injured during the protests.19 The Nepal Medical
Association and the Nepal Nursing Association, Nepal
Paramedic Association, and Nepal Ayurvedic Doctors’
Association called for treatment of victims and health
professionals responded promptly in providing this care.
The Tribhuvan University Teaching Hospital, Kathmandu
Model Hospital, Binayak Hospital, and many other
hospitals in Kathmandu provided treatment to the
victims. Health professionals associated with Physicians
for Social Responsibility in Nepal, the Nepali affiliate of
International Physicians for the Prevention of Nuclear
War, and several other health workers throughout the
country who were not affiliated with any organisation
participated in the movement for the restoration of peace
and democratic rights in Nepal.
Earlier that month, on April 8, 2006, the government of
Nepal detained seven physicians for peacefully protesting
the autocratic actions of the Nepali regime and defying
an imposed curfew. Dr Mathura Prasad Shrestha, the
president of Physicians for Social Responsibility Nepal
had been imprisoned for the previous 2 months on
frivolous charges. On April 10, 2006, 20 medical students
702
were arrested during a peaceful demonstration. Later
that night about 30 armed police, backed by the Royal
Nepalese Army, entered the hostel of the main medical
institution in Nepal, the Tribhuvan Institute Teaching
Hospital, and tortured and terrorised 20 medical students
and a doctor. On receiving this news I immediately
contacted my colleagues in different parts of the world
and apprised them of the developments in Nepal. The
outpouring of support for our Nepalese colleagues in
distress was instantaneous. In an expression of solidarity
more than 1400 physicians from more than 50 countries
signed our online petition demanding the immediate
release of our imprisoned colleagues.20 However, a few of
my Nepalese colleagues expressed serious reservations
about the role of these Nepalese doctors in the human
rights movement. They felt that the imprisoned doctors
had overstepped their boundaries and lost their scientific
neutrality.
The doctors who took part in the protests for democracy
used human-rights arguments to justify their participation
as necessary for the eradication of poverty and repression.
Many physician leaders have been at the forefront of this
ongoing political transformation in Nepal. The Maoist
rebels and the political parties drafted a historic new
Nepalese constitution in response to the mass uprising
against the monarchy. The new constitution was based
on equality, and enshrined the right to health as a
fundamental human right. Although the practical
realisation of these rights for millions of Nepalese people
is far in the future, this gesture might be the fruits of the
constant labour of health professionals in the
human-rights movement in Nepal.
One might question the role of physicians in a political
movement, but can health professionals in a poor,
undemocratic country like Nepal avoid politics? Can one
simply treat illness without taking into account the social
context in which it occurs? These are questions that need
to be tackled by physicians around the world wherever
humanitarianism is needed. The success of the revolution
in Nepal poses a dilemma far beyond that of Nepal. As
Vincanne Adams13 questioned: “Can medicine be
politicised and still be an objective science?”
I had witnessed a great deal of the Maoist insurgency
unfold in Nepal. Though once famous for its mountains,
Nepal had gained worldwide recognition because of
people disappearing. I had witnessed and painfully
documented both gross violations of human rights (such
as torture of women, children, and marginalised people)
and subtle erosion of human dignity (poor people unable
to bury their dead).10–13 A few months later, one of my
colleagues who had been freed reminded me that the
gesture of solidarity shown by doctors from around the
world had kept his spirits, and those of his colleagues,
high even during the most challenging times. Although
the act of witnessing seemed distant and aloof, and paled
by comparison with the personal sacrifices that these
doctors and most ordinary Nepalese made in these
www.thelancet.com Vol 370 August 25, 2007
Series
turbulent times. However, witnessing and sharing these
historic events in Nepal revealed what solidarity meant to
both doctors in need and their patients when health
professionals across the globe could find common cause
with their colleagues and fellow human beings in a faraway land.
What does it mean to bear witness?
The health-care provider acting as a witness serves the
patient first, not a political project, and in doing so,
honours the dignity of all people, and calls on others to
do the same. In this one goal, the actions of
humanitarians are apolitical, that is, they do not take
political sides in the conflict, and they always represent
and advocate for the victims. And yet, in the call to
others, the potential effects of bearing witness become
or could become political because they demand that the
methods of the political projects are changed to avoid
further affronts to the victim. The potential political
effects of humanitarian action lie in this dialectic, not
in the sentiment felt by the humanitarian provider. The
humanitarian health worker can be seen as an outsider
to both medicine as a purely technical skill, and to
society as a political project. They are probably
committed to freedom, both their own freedom and
that of their patients whose freedom is encumbered by
preventable unnecessary suffering. And the health
worker who both witnesses and experiences human
rights violations becomes an outsider to medicine as a
purely technical pursuit, because they can see the role
that health providers can have in instigating change for
the better.
We have few epidemiological instruments to measure
the extent of human-rights’ violations of individuals or of
communities.21 The standard hierarchy of evidence,
which places randomised trials and systematic reviews at
the top, does not apply in this context. The patients’
narrative might be given most importance in sudden
widespread violations of human rights.21 Narrative reveals
the sense, the sentimental feeling, and the humanity of
the experience. First-person narrative from the victims
and perpetrators can reveal the processes of life—the
humane and the inhumane—and these stories, in
combination with justice, can invite reconciliation and
contribute to political change.
Health researchers often feel that they have little to
contribute to humanitarian crises, because the immediate
needs of affected populations seem to relate to trauma or
neglected clinical programmes. However, researchers
and clinicians also have skills that advance the practice of
humanitarianism and human rights. The Iraq mortality
studies,22 responses to the Asian tsunami in 2004,23 and
documentation of systematic torture of refugees24 are
examples of humanitarian crises in which researchers
have documented the state of health and human rights
by scientific methods. Funding for health research related
to human rights and humanitarian crises is not always
available. States might not be interested in providing
www.thelancet.com Vol 370 August 25, 2007
support to neglected populations. This attitude, however,
should not be a barrier to research. The Cochrane
Collaboration is an example of how largely unfunded
research can have an effect on health and international
policy.25 Appropriate systematic review techniques can
overcome the criticisms of some states about the
inconsistencies of fact-finding missions and reports by
non-governmental organisations, and provide information about the state of human rights within certain
populations.26
Health practitioners are frequently among the first
witnesses to both subtle and gross humanitarian and
violations of human rights. Inadequate knowledge of
principles of human rights can contribute to inability or
unwillingness to bear witness in some repressive settings,
and physician tolerance and participation in human
rights abuses in others. Some people have advocated for
mechanisms to hold health-workers personally
responsible if they engage in human-rights’ violations.27
Health workers should be held accountable to the same
legal standards as other people, but bearing witness is a
further responsibility, and has risks in politically
ambiguous places.
Health practitioners around the world, in places where
human rights are routinely violated or human dignity is
eroded subtly, might relate to our testimonies. Our
responsibility as human beings and as health
professionals, whether physicians, nurses, logisticians or
researchers, is to bear witness honestly to the reality of
inhumanity, and to speak out against the moral dearth of
political inaction or complicity. What is good or right has
no moral neutrality when one is confronted with
attempted or actual genocide, wilful ignorance of
public-health threats, or the intimidation of caregivers.
Speaking out might not definitely save lives, but silence
certainly kills. Silence kills today, and it will kill tomorrow.
Our testimonials should encourage health professionals
to bear witness, and encourage others to stand in
solidarity with them. Knowledge of human rights
principals, documentation and bearing witness can be
the first step in addressing the social, political, and
economic forces that constrain human dignity.
Conflict of interest statement
We declare that we have no conflict of interest.
Acknowledgments
We thank Edward Mills for his contributions and reviews of this
manuscript.
References
1
Weiss TG. Principles, politics, and humanitarian action.
Ethics Int Aff 1999; 13: 1–32.
2
Red Cross: Iraq abuse tantamount to torture. http://www.msnbc.
msn.com/id/4944094/2004 (accessed June 25, 2007).
3
Médecins Sans Frontières. The crushing burden of rape. Sexual
violence in Darfur. http://www.artsenzondergrenzen.nl/usermedia/
files/Report%20Sexual%20Violence%20march%202005.pdf
(accessed Sept 18, 2006).
4
Moszynski P. Sudan arrests aid worker for “crimes against the
state”. BMJ 2005; 330: 1350.
5
Steiner H, Alston P. Civil society: human rights NGOs and other
groups. In: International human rights in context: law, politics,
morals. New York: Oxford University Press, 2000: 954–64.
703
Series
6
7
8
9
10
11
12
13
14
15
16
17
704
Shilts R. And the band played on: politics, people, and the aids
epidemic. New York: St. Martins Press, 1987.
Americans with Disabilities Act of 1990. http://www.usdoj.gov/crt/
ada/adahom1.htm (accessed June 25, 2007).
Hawkins V. Stealth conflicts: Africa’s world war in the DRC and
international consciousness. The Journal of Humanitarian
Assistance, 2004. http://www.jha.ac/articles/a126.htm (accessed
June 25, 2007).
Gourevitch P. Reversing the reversals of war. The New Yorker,
April 26, 1999.
Singh S, Sharma SP, Mills E, Poudel KC, Jimba M. Conflict induced
internal displacement in Nepal. Med Confl Surviv 2007; 23: 103–10.
Singh S, Bohler E, Dahal K, Mills E. The state of child health and
human rights in Nepal. PLoS Med 2006; 3: e203.
Maskey M. Practicing politics as medicine writ large in Nepal. Dev J
2004; 47: 122–130.
Adams V. Doctors for Democracy. Health professionals in the Nepal
revolution. Cambridge, UK: Cambridge University Press, 1998.
World Bank, Health Nutrition and Population, Nepal Country
Office. Assessment of impact of conflict on health service delivery
system for the rural population of Nepal, June, 2005. http://un.org.
np/reports/WB/2005/2005-jun-final-report-conflict-delivery-final.
pdf (accessed June 25, 2007).
Poudel K. Health a casualty of Maoist attacks in Nepal. Kathmandu:
One World South Asia, Apr 27, 2004. http://southasia.oneworld.net/
article/view/84604/1 (accessed May 27, 2007).
Anon. Curfew lifted in different districts of Nepal. Xinhua News,
April 30, 2006. http://english.peopledaily.com.cn/200604/30/
eng20060430_262322.html (accessed June 25, 2007).
Sharma GK, Osti B, Sharma B. Physicians persecuted for ethical
practice in Nepal. Lancet 2002; 359: 1519.
18
19
20
21
22
23
24
25
26
27
Pandey K Protests and bombing disrupt health care in Nepal. BMJ
2006; 332: 1052.
Pandey K. Doctors in Nepal take part in democracy protests. BMJ
2006; 332: 931.
Singh S, Arya N, Mills E, Holtz T, Westberg G. Free doctors and
medical students detained in Nepal. Lancet 2006; 367: 1730.
Robertson DW, Bedell R, Lavery JV, Upshur R. What kind of
evidence do we need to justify humanitarian medical aid? Lancet
2002; 360: 330–33.
Roberts L, Lafta R, Garfield R, Khudhairi J, Burnham G. Mortality
before and after the 2003 invasion of Iraq: cluster sample survey.
Lancet 2004; 364: 1857–64.
Thienkrua W, Cardozo BL, Chakkraband ML, et al. Symptoms of
posttraumatic stress disorder and depression among children in
tsunami-affected areas in southern Thailand. JAMA 2006; 296:
549–59.
Van Ommeren M, de Jong JT, Sharma B, et al. Psychiatric disorders
among tortured Bhutanese refugees in Nepal. Arch Gen Psychiatry
2001; 58: 475–82.
Clarke M, Langhorne P. Revisiting the Cochrane Collaboration.
Meeting the challenge of Archie Cochrane—and facing up to some
new ones. BMJ 2001; 323: 821.
Letter to Pierre Sane SGoAI, from Permanent Representative of
Republic of Indonesia to the UNHCR, Nov 2, 1994: UN Doc.
E/CN.4/1995/108. www.arabhumanrights.org/unconf/wchr/esc/ecn4-176-add1-95e.pdf (accessed June 25, 2007).
Grodin M. Medicine & Human Rights: a proposal for international
action. Hastings Centre Report 1993; 23: 8–12.
www.thelancet.com Vol 370 August 25, 2007
Viewpoint
Inconvenient truths about effective clinical teaching
Brendan M Reilly
I’ve been teaching clinical medicine for more than
30 years but it seems to be getting harder, not easier.
Conventional wisdom in the USA holds that the problem
is time and money (or, more precisely: time is money).
Hospitalised patients, discharged before doctors can get
to know them, are sicker and quicker today. Outpatient
teaching is no less awkward, 10-minute office visits and
outdated Medicare reimbursement rules gumming up
the works. Long overdue restrictions on resident work
hours won’t solve these problems.1
Too little time and money for clinical teaching betokens
lack of respect too. Most academic centres in the USA
don’t provide adequate support for clinician-educators’
salaries despite substantial government subsidies for
postgraduate education. This shortfall is not an oversight;
it is a calculated budgetary decision. Insult compounds
injury when physician-researchers openly disparage the
academic gravitas of physician-educators on the same
faculty.
This situation raises the obvious question: is clinical
teaching today not only more difficult but also less
effective? One might assume that our research-proud
profession would know the answer.2 In fact, despite
shocking indictments of the quality, safety, and equity of
US medical care,3–6 we know little about the effect of clinical
teaching on learners or patients, nor even how to measure
it.7,8 Worse, we don’t seem very concerned about this
situation. In 2006, four major medical journals (BMJ,
JAMA, Lancet, and New England Journal of Medicine) and
four medical education journals (Academic Medicine,
BMC Medical Education, Medical Education, and
Medical Teacher) published a total of one original outcomes
study of this kind (which found no correlation between
measures of teaching effectiveness and patients’ clinical
outcomes).9
Lacking evidence, I do what clinicians do when we
don’t have the data we need: I go with my gut instinct.
My gut tells me that clinical teaching today—my own
and others’—is less effective than it used to be and needs
to be. Among those who will disagree are many academic
leaders and quality gurus who don’t even acknowledge
the question. They maintain plausible deniability by
looking elsewhere: we need better systems, they say, not
better doctors. No doubt they are right about the
systems.
I propose that the decline of clinical teaching in our
training programmes is, like global warming, an
inconvenient truth. Even if we saw evidence as eerily
convincing as Al Gore’s pictures of melting polar
ice-caps,10,11 many in academic medicine would look the
other way. Rather than take remedial action, we will be
tempted to do the greenhouse-gas-shuffle: blame it on
random variation or transient aberration (anything but
www.thelancet.com Vol 370 August 25, 2007
ourselves) and hope the hurricanes and heat waves just
go away.
Doubly inconvenient would be to learn that fixes from
the past might not work in the present. For example, due
to digital information systems, clinical trainees inevitably
review patients’ laboratory data and diagnostic images
before they do a history or physical examination. This
change portends more than the devaluation of bedside
skills;12 it is nothing less than complete inversion of the
conventional diagnostic process. The good news is that
innovation in medical education eventually catches up
with advances in science and technology.13 The bad news
is that the pace of change is glacial.14,15 Worse, we know so
little about medicine’s informal curriculum (clinical
training) that it’s hard to know where to start.16–19 In this
spirit, I describe eight habits of exemplary clinical
teachers I have known and try to emulate still.
Lancet 2007; 370: 705–711
See Editorial page 630
Department of Medicine, Cook
County (Stroger) Hospital,
Rush Medical College, Chicago
IL 60612, USA (B M Reilly MD)
Correspondence to:
[email protected]
Think out loud
Making transparent to learners the teacher’s own clinical
reasoning is the most powerful predictor of learners’
satisfaction.20 This method is not the same as talking off
the top of one’s head, a habit common among ineffective
teachers. Instead, thinking out loud is highly disciplined
and strategic,21,22 with three primary purposes. First, it
communicates a general framework for solving the
clinical problem at hand. In the past, this was the main
challenge for clinical teachers: to extrapolate from the
particular patient to the general (all presentations like it)
and back again. But now even novice learners, armed
with expert guidelines and algorithms, can lend logic and
authority to their problem-solving strategy. Thus, teachers
today can afford to spend less time thinking out loud
about these things. This efficiency assumes that learners
read about their patients and apply well what they read,
not always a safe assumption.
Second, all clinicians struggle to translate the results of
published research into the care of each unique patient.
This population-to-person problem, the generic dilemma
of practising evidence-based medicine, needs not only
the skill to search and understand the published medical
literature but also the judgment to use it or not in
individual cases.23,24 This translational process, though
imperfectly understood, is the essence of what clinicians
do.25 For this reason, effective teachers do this translation
out loud, articulating it in detail for learners as well as
patients.
Third, effective teachers purposefully expose for
learners the ambiguity and ambivalence inherent to
clinical medicine by thinking out loud in the moment
(on the fly) as patients’ problems arise in real time. Such
spontaneity requires teachers to share extemporaneously
their own inchoate thoughts about what to ask, what to
705
Viewpoint
look for, what to do. Inevitably, some of these outspoken
ideas will seem, in retrospect, irrelevant or mistaken. But
learners need to see the teacher’s own problem-solving
journey—including fits and starts, blind alleys, and
missteps. Clinical medicine is messy, with many working
diagnoses disproved and therapeutic trials abandoned.
Effective teachers give learners a bird’s-eye view as they
struggle themselves to tidy up the mess.
The inconvenient truth is that thinking out loud needs
more than expertise and confidence; it also needs
humility, a virtue not encouraged widely enough in the
medical education hierarchy. If our profession is serious
about lifelong learning, we must recognise that learning
can’t happen without humility. Teachers who humbly
think out loud help to show the way.
Activate the learner
Experts agree that adult education is a tango: it takes two.
The dance will fail, no matter how expert the teacher, if
the learner is not actively, even passionately, engaged.
But clinical teachers typically lead teams of learners—in
the USA, groups of residents, interns, and
students—whose different skill levels need different
moves by the teacher. (By contrast, consider the
one-on-one mentoring deemed essential for all
researchers-in-training.) The group tango is made doubly
difficult by conditions on the dance floor: in today’s
hurly-burly hospital wards and clinics, getting the work
done—taking care of patients—must take priority over
teaching. Under such conditions, how do clinical teachers
activate learners?
Two steps are fundamental. First, effective teachers
insist on learners’ motivation as a precondition for their
activation. Unmotivated learners waste teachers’ time.
They don’t belong in a profession where lifelong learning
(indeed, love of learning) is an absolute requirement.
They should be encouraged to change careers (once
disability and fatigue have been excluded as the
underlying problem). Second, effective teachers synergise
learners’ needs with their patients’ needs. How? They
repeatedly pose two questions to their team of learners:
what do we know about this patient? What more must we
learn to provide them the best care? This agenda exploits
the fact that learners on different levels learn and help
patients in complementary ways. Whether it is the
student who elicits crucial new history or the senior
resident who develops the ultimate treatment plan, each
member of the team contributes by pulling their own
weight on their own level.
Beyond this basic two-step, the key to activating clinical
learners is the teacher’s style. The Socratic style would
seem ideal for medicine, guiding adult learners toward
self-discovery in dialogues orchestrated by the teacher.
But its tortuous process and delayed effect make the
Socratic style impractical in most clinical settings.
Alternatively, the autocratic style has many practical
advantages, especially when the patient’s clinical condi706
tion is dire and the right moves must be made right away.
However, this (shut-up-and-do-what-the-teacher-says)
approach fosters a culture where learners learn by
following orders, activated by fear of ridicule in the
present26 and reprisals (poor evaluations) in the future.
One needs look no further to appreciate why so many
practising clinicians rely on expert opinion, whether
evidence-based or not, and why they dread making
errors.
Most effective clinical teachers use the democratic style.
They assume that clinical learners mature most when
encouraged to think and act autonomously under
pressure. Here, the challenge for the teacher is knowing
when to stand back and when to jump in, giving learners
enough freedom to grow without hurting themselves
(and their patients). This balancing act is not for the faint
of heart: given too much autonomy, clinical learners
endanger patients in the present; given too little, they
might endanger them in the future. Thus, the democratic
style needs leadership as well as teaching skills. The need
for both explains why some teachers who perform
brilliantly in the classroom don’t do as well at the
bedside.
The inconvenient truth is that the success of the
democratic style is somewhat mysterious. William Penn
captured some of its nuance when he said: “Let the
people think they govern, and they will be governed.”27
Teaching democratically is all about activating learners’
initiative while protecting them from themselves. This
effort will succeed in medicine only if wise, watchful
teachers lead learners to “think they govern”.27
Listen smart
Effective clinicians listen carefully to their patients.
Effective teachers, who diagnose and treat learners in
parallel with diagnosing and treating their patients, also
listen carefully to learners.28 They tune in to learners’
acquisition, synthesis, and presentation of clinical data,
logic in clinical reasoning, patient-centredness when
making decisions, and grasp of the high standards of
medical professionalism. Listening to learners requires
insight and understanding beyond that needed to listen
to patients;29 for example, it needs a meta-analytical
understanding of what makes any clinician effective.
Such requirements explain why many effective clinicians
are not effective teachers: they don’t know how they do
what they do so well.22 Even master clinicians, when
listening to learners, might not know what to listen for.
Effective clinical teachers diagnose and treat two
general types of learning disorders: pathological conditions and developmental delays. Pathological conditions
need urgent attention because pathology in one domain
(eg, defective data synthesis) might metastasise to other
domains (eg, clinical reasoning); equally dangerous, one
learner’s pathology could infect other (typically, more
junior) learners on the same team. By contrast,
developmental delays are less urgent because all clinical
www.thelancet.com Vol 370 August 25, 2007
Viewpoint
learners attain some competencies more slowly than they
attain others. But these delays are important and should
not be ignored. So-called watchful waiting, often the best
option in clinical care, is rarely the best strategy in clinical
teaching. All clinical learners have room to grow, and the
teacher’s job is to help them grow. In the clinical
vernacular, effective teachers are interventionists. But
active intervention first requires active listening.
When listening to learners, teachers who lack
independent knowledge of learners’ patients will be less
effective. Why? Because teachers cannot listen smart
when they encounter a patient for the first time during a
learner’s presentation. Under these conditions, the teacher
can assess the internal validity of the presentation (does it
make sense?) but not its external validity (is it true?). Such
teachers, at best, will be inefficient; at worst, they will be
complicit in serious error.30 By contrast, hospitalists in the
USA tend to get high marks for inpatient teaching,
because their job is to assess patients independently.31
Similarly, tertiary referral centres might be less conducive
to effective teaching than some community hospitals
(where the clinicians already know their patients well).
No doubt it is hard work for teachers to assess
independently all of their learners’ patients in a timely
manner. Teachers who do this well complain of being
perpetual interns. The inconvenient truth is that personal
attention to detail is what is needed to teach clinical
medicine effectively. As Alfred North Whitehead noted in
The Aims of Education:
All practical teachers know that education is a patient
process of the mastery of details…There is no royal road
to learning…There is a proverb about the difficulty of
seeing the wood for the trees…The problem of education
is to make the pupil see the wood by means of the trees.32
Keep it simple
Recommending simplicity will seem disingenuous;
certainly medicine is not simple. But teaching medicine
as simple does not intend that teachers dumb it down,
make it simplistic; rather, simplicity exhorts them to
reduce their presentation, as chefs will do, boiling it
down to its hearty essence. Many clinical teachers don’t
do this, sometimes in deference to medicine’s complexity
(and sometimes to show off their own), but often because
they don’t appreciate its pedagogic power.
First, one must understand complexity well to express
it simply. Clinicians who can’t reduce to simple terms
what they think probably don’t understand what they
think well enough to apply to patients’ care. Effective
teachers keep the teaching simple because they know
that concise and clear expression improves
communication with patients, too. In boiling complexity
down for learners, they show learners how to boil it down
for patients.
Second, effective teachers address a specific scenario
by conveying general principles relevant to all situations
like it. For example, the principle of not letting the sun
www.thelancet.com Vol 370 August 25, 2007
set on a hot appendix is useful whether the particular
patient has appendicitis or not. Whitehead favoured this
approach to education:
The really useful training yields a comprehension of a
few general principles with a thorough grounding in the
way they apply to a variety of concrete details.32
To diagnose the cause of oedema, for example, only
two facts are pivotal: the patient’s jugular (central)
venous pressure and serum albumin level. This
principle is always useful, not because the two facts
always make the diagnosis but because, even when they
don’t (eg, in cases of vena cava obstruction), they always
point the way. Many such principles have been validated
in empirical studies to “help physicians…know what
clinical data are important to obtain.”33 Effective teachers
promulgate such rules because they give learners what
William Osler called “good methods and a proper point
of view.”34
Third, effective teachers recognise the difference
between scientific knowledge (which has intrinsic value)
and clinical knowledge (which has value only if applicable
to patient care). Ultimately, all clinicians must translate
complex clinical knowledge about their patient into one
simple decision: do this, or do that. Effective teachers
show them how.
The inconvenient truth is that keeping it simple is
complicated. Even the most effective teachers find it hard
to do consistently and well. Additionally, its reductive
technique sometimes annoys advanced learners who are
more interested in the exceptions than the rules. But this
point is where Osler’s “good methods” begin to pay
handsomely, where learning curves rise, for teachers as
well as learners. Getting there is a good thing, even if it
isn’t simple.
These first four habits comprise the acronym TALK
(table 1). But, just as clinical effectiveness is measured
more by what clinicians do than what they know, teaching
effectiveness is measured more by what teachers do than
what they say. Thus, the final four habits pertain to how
effective teachers “WALK the walk” (table 2), not how
they “TALK the talk”.
Wear gloves
Infection control has never been more important than it
is today, but wearing gloves addresses a broader issue
here: effective clinical teachers are hands-on role models.
This practice involves frequent physical interaction with
patients—demonstrating the clinical utility of physical
examination, the therapeutic value of touching, the
diverse benefits of bedside care.35–37 It also needs a
conscious effort to make real to learners the physical
experience of sick patients and the glorious relief good
doctoring can bring them. Most physicians-in-training
are young and healthy, unfamiliar with the travails of
being a patient; many have never felt excruciating pain,
profound weakness, or desperate dyspnoea. Thus, even
707
Viewpoint
Strategic goals
Educational challenges
Think out loud
Show learner the process, not
merely the outcome, of expert
reasoning
Articulate, in real time, pivotal steps Missteps inevitable
when making clinical decisions
Clinical challenges
Requires humility
Inconvenient truths
Activate the
learner
Promote learner’s initiative and
autonomy
Know when to stand back versus
when to assert clinical authority
Patient safety always the top
priority
Needs democratic leadership
skills
Listen smart
Efficiently assess validity of
learner’s presentation
Know what to listen for
Assess patient before
assessing learner
Requires mastery of patient’s
clinical details (teachers as
perpetual interns)
Keep it simple
Exemplify concise communication Use reductive general principles to
and rule-based decisionmaking
illuminate clinical complexity
Each patient is unique; some
don’t follow the rules
Easier said than done
Table 1: Teaching habits of effective clinical teachers—TALK the talk
the little things done by hands-on teachers can have great
effect on learners: feel the febrile patient’s sweat-soaked
back on early morning rounds; find her a fresh dry gown;
flip her pillow over to the cool side; see her close her eyes
in respite. These and other bedside ministrations have
been relegated to others today. But effective teachers
know, and show, that there is no better way for doctors to
connect with patients.
The importance of hands-on teaching has another
implication: the most effective clinical teachers are
practising clinicians. This point doesn’t mean that
researchers and administrators can’t teach; however,
they must do enough direct patient care to grow their
own clinical skills, not merely maintain them.38–40 In
many US centres today, academics practice only when
on-service (a few weeks per year); not infrequently, this
practice amounts to continuing medical education for
the attending physician, who unabashedly learns more
from the house staff and students than they do
from him. The notion that those who teach clinical
medicine need not practise it is absurd, a convenient
delusion that demeans the discipline and those training
to learn it.
The inconvenient truth is that hands-on clinical
teaching is largely unappreciated today, despite the effort
and expertise required. Well-intentioned reforms in
many US medical schools have created separate tracks
for academic promotion of clinician-educators but these
tracks are widely considered second-class. (Externally
funded researchers go first class, as they should.) Private
insurers and payers don’t reward teaching either, a
remarkable oversight in view of their alleged interest in
more effective clinical practice. In a very real sense,
hands-on clinical teaching has become its own reward, a
vestige of professional altruism that will survive only if
today’s teachers can pass the torch on to a new
generation—not a forgone conclusion.
Adapt, enthusiastically
As Osler said, “Medicine is a science of uncertainty and
an art of probability.”41 What the science predicts, however
confidently, might not happen; what clinicians do for
patients, however artfully, might not succeed. Thus,
despite the best laid plans, all clinicians must adapt to
the unexpected. Effective teachers seek these situations
because they present the greatest opportunities to learn
(and to help patients). Exploiting these opportunities is
not an easy thing to do, for several reasons.
Two demons haunt all clinicians: chance and fallibility.
Assessing the agency of chance is difficult42 but clinical
teachers should try, because clinical error is a different
kind of learning opportunity than bad luck. Both are
instructive but only error provides the opportunity to
learn the most difficult of all cognitive skills (when, and
why, to change one’s mind) and the most wrenching of
all clinical responsibilities (how, and to whom, to admit
Strategic goals
Educational challenges
Clinical challenges
Inconvenient truths
Wear gloves
Promote hands-on doctoring
Role-model unfashionable skills Continuing refinement of clinical
(physical exam) and
acumen to complement advances
countercultural behaviour
in science and technology
(nurses’ work)
Adapt,
enthusiastically
Embrace clinical uncertainty as Role-model aplomb and savoir
a valuable learning opportunity faire when unexpected clinical
events occur
Changing one’s mind; admitting
error; lack of evidence for many
clinical dilemmas
Link learning to
caring
Show, and expect of learners,
empathy and responsibility for
each patient
Role-model professionalism
and patient-centredness
Understand the patient’s illness as Medical consumerism (care as a
well as their disease
commodity) undermines medical
professionalism
Kindle kindness
Establish generosity (not
politeness) as the standard for
all clinical interactions
Give encouragement (hope) to
learners even when giving
critical feedback
Treat the disease as your enemy
but the patient as your friend
Bedside care undervalued and
inadequately rewarded
Managing clinical uncertainty
highly stressful yet poorly taught;
burn-out an occupational hazard
Unknown whether simple human
kindness is teachable
Table 2: Teaching habits of effective clinical teachers—WALK the walk
708
www.thelancet.com Vol 370 August 25, 2007
Viewpoint
mistakes).43Ironically, this learning opportunity could
explain why some teachers, always on the lookout for
teachable moments, find error when bad luck is equally
likely. Such attribution bias is dangerous; clinical errors
are frequent enough without inflating their number.4 All
clinicians accumulate guilt over the course of a
career—even when we deal with our mistakes
constructively, most of us incur a personal loss—and
increasing that guilt arbitrarily doesn’t help.44 Teachers
walk a tightrope here. Teetering between finding fault
and ruing randomness, their missteps have consequences
either way.
But adapting to the unexpected needs more than
hard-headed honesty about our errors and biases. It also
needs creativity, an ability to improvise when making
uncertain clinical judgments. Evidence-based medicine
zealots might disagree, but randomised trials and expert
guidelines will never address more than a fraction of all
conceivable eventualities in clinical medicine. In other
words, judicious improvisation will always be an integral
part of what clinicians do. But when (and how) do
clinicians learn this skill? Not in US medical schools,
according to Melvin Konner, the anthropologist who
wrote trenchantly about his experience as a medical
student:
Medical school [and] graduate school…have diametrically
opposite purposes. The graduate school must produce a
unique product: the student must…go as soon as possible
beyond what has been taught…The medical student
must on the contrary end by being as similar as possible
to every other medical student…according to a process
that leaves no room for originality. At the end of study,
all…medical students…should perform the same
examination, write the same assessment, and formulate
the same options for treatment.45
One might conclude that postgraduate medical
education also should produce carbon-copy exemplars of
some curricular ideal (although Konner doesn’t say as
much; he wrote only about medical school). But this
notion, seductive to many who want to standardise
clinical care, fundamentally misunderstands what
clinicians do and what effective teachers must teach. The
formal medical school curriculum, albeit bloated and
intimidating, is the easy part of medical education; the
hard part is learning how to decide what to do when no
one knows what to do, creatively using clinical judgment
to help the patient as best one can.46
The inconvenient truth is that clinicians learn to
manage uncertainty haphazardly, without formal
instruction, despite its manifest importance to patient
care. As a result, most clinicians don’t like surprise.
Enthusiasm for surprise—a sort of swashbuckling
eagerness to handle whatever happens next—allows
effective teachers to confront the unexpected head-on,
the only way to address it. Such brio is a lot to ask of any
teacher lifelong; many of us burn out, a problem we must
learn more about.47
www.thelancet.com Vol 370 August 25, 2007
Link learning to caring
Patient-centred teaching refers to teaching that is directly
and immediately relevant to each patient’s main clinical
problem. Whether sepsis or somatisation or a surgical
abdomen, this main problem determines what must be
taught and learned about a particular patient
today. Patient-centred teaching contrasts sharply with
teacher-centred teaching, an all too common practice
where clinical teachers teach what they know whether it
addresses the patient’s problem or not. Although
patient-centred teaching requires explicit prioritisation of
the patient’s problems, effective teachers take pains not
to prioritise disease (what the patient has) more highly
than illness (what the patient feels). This fundamental
tenet of clinical medicine has become countercultural in
many academic centres today.
But, whether the caring agenda is strictly technical
(disease-oriented) or more holistic (illness-oriented), its
defining characteristic is notable. Here, patient care
refers to what doctors do for patients, the services we
provide, whether brief counselling or major surgery.
Such things, of course, clinical teachers must teach. But
there’s the rub: thus defined, care is a thing—a product
provided by clinicians, received by patients, measured by
quality analysts, quantified by payers. This banal
definition, especially when combined with the mistaken
notion that well-trained physicians are interchangeable,
contributes to the increasingly popular idea that clinical
care is a commodity.
Francis Peabody took a different view when he observed
famously that “the secret of the care of the patient is in
caring for the patient.”48 In words that still resonate today,
he noted:
The physician who attempts to take care of a patient
while he neglects [the patient’s emotional life] is as
unscientific as the investigator who neglects to control
all the conditions that may affect his experiment. The
good physician knows his patient through and through,
and his knowledge is bought dearly. Time, sympathy
and understanding must be lavishly dispensed but the
reward is to be found in that personal bond which
forms the greatest satisfaction of the practice of
medicine.48
For Peabody, dying of cancer at age 47 when he wrote,
caring meant not only the things clinicians do for
patients but also the personal bond that gives them
meaning.
Of course, to link learning to such depth of caring
assumes that learners (and teachers) care so deeply. This
assumption is not true for all clinicians today; some of us
don’t care, not in the way Peabody meant. Worse, the
medical profession as a group has not tried to discover
who these careless clinicians are or how they got that
way. Ask patients what makes a good doctor and many
will give some variation on the same answer: a doctor
who cares about me. But the truth is that our profession
has not taken the trouble to study systematically what
709
Viewpoint
such caring means. Our negligence in this matter is not
merely inconvenient, it is deeply troubling, calling into
question how much our profession really cares about
patient care.
To their credit, august professional organisations have
defined competencies requisite for all clinicians,
including the 800-pound gorilla called professionalism.49–52
Its components no doubt make up much of what patients
look for in a doctor who cares. But defining
professionalism won’t make it happen; only the
professionals can do that. Fortunately, thousands do,
magnificently, every day. The unanswered question is
who will join their ranks, take their place? Who will
convince future generations that Peabody’s way of caring
is normative, not nostalgic? The inconvenient truth is
that no one knows the answers to these questions, and
not all of us care.
Kindle kindness
Ultimately, teaching is all about the learner, not the
teacher. Thus, effective clinical teachers aspire to a sort of
selflessness whose tangible expression is kindness to
learners, especially when assessing them (giving
feedback). Kindness makes even the toughest criticism
hopeful, empowering the learner by making learning
less oppressive. Not incidentally, patients appreciate
kindness too.
If, then, kindness makes patients more satisfied,
teachers more effective, and learners more receptive, we
should kindle it. But, what is it, exactly? It is not merely
politeness, as the philosopher André Comte-Sponville
has argued.53 He does not demean politeness, surely an
important thing:
All the world’s a stage, and living means acting…Not
being virtuous, we make a pretense of virtue; this is
called politeness. Not knowing how to love, we act as
though we did; this is called morality.53
But, in medicine today, this apparent pretence of virtue
is often confused with the real thing. The new
provider-customer model of the doctor-patient relationship abets this confusion, making little distinction
between the solicitude of a salesman and the benevolence
of a physician. Some learners today don’t know the
difference. Unfailingly polite, they pretend kindness, a
smarmy sham of the real thing.
The real thing, in the philosopher’s view, has more to
do with gentleness than politeness. On a mundane level,
physicians’ gentleness is tactical. Privileged to enter the
intimate lives of strangers, they use gentleness to enable
empathy and communication. But its deeper
relevance—to Peabody’s secret of caring, to the ideals of
medicine—shows itself when gentleness combines with
generosity.
Generosity invites us to give in the absence of love to the
very people we do not love, and to give them more the
more they need it, or the better equipped we are to help
710
them. Indeed when love cannot guide us because we do
not feel it, let us be guided by urgency and proximity.
This…is…generosity. Joined by justice, [generosity]
becomes equity. Coupled with compassion, it becomes
benevolence…But its most beautiful name is its secret,
an open secret that everyone knows: accompanied by
gentleness, it is called kindness.53
How, then, can teachers kindle kindness in medicine?
It can’t be done in a classroom. No doubt medical learners
are “good at reading what the environment expects of
them—and then meeting these expectations”.54 But
courses in medical ethics and medical humanities,
however valuable, can’t kindle kindness. The inconvenient
truth is that there is only one way: someone has to do it,
walk the walk for all to see. “Example is not the main
thing in influencing others,” Albert Schweitzer said, “it
is the only thing.”55
Demonstrating kindness doesn’t mean that clinical
teachers must be heroes or saints. In fact, what matters
are the little things, what Comte-Sponville calls “kindness
of manner”.53 Such simple human kindness is natural to
most physicians-in-training (in whom it must be
nurtured); unfortunately, it is not natural to all. Whether
kindness is teachable is a crucial question because,
without its spark, kindness in its nobler forms—altruism,
benevolence, equity—cannot be kindled. The ideals of
medicine, both in practice and in teaching, “begin…and
end…with the patient”.56
Some might think these habits set the bar too high.
Certainly they exceed my own reach more often than I
like. But there is much to recommend lofty goals,
especially today when the science of medicine is
soaring. Clinicians must catch up, not to compete with
the scientists but to become their equal partners
again.2,57 To do so, we can tap new sources of
energy—computer-simulated training, decision-support
systems, faculty development programmes—but we
must also prove their power and cost-effectiveness.
Better educational research is essential.58–60 Above all,
we must acknowledge a final inconvenient truth: “Our
will to take action is a renewable resource.”11
Temperatures are rising. We best take heed before the
tides rise too.61,62
Acknowledgments
I am grateful to Ian MacManus whose insightful review was most helpful.
References
1
Leach DC, Philibert I. High–quality learning for high-quality health
care: getting it right. JAMA 2006; 296: 1132–34.
2
Lenfant C. Clinical research to clinical practice—lost in translation?
N Engl J Med 2003; 349: 868–74.
3
McGlynn EA, Asch SM, Adams J, et al. The quality of health care
delivered to adults in the United States. N Engl J Med 2003; 348:
2635–45.
4
Kohn KT, Corrigan JM, Donaldson MS. To err is human:
building a safer health system. Washington, DC: National
Academy Press, 1999.
5
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL,
Pinder EL. The implications of regional variations in Medicare
spending. Part 1: The content, quality and accessibility of care.
Ann Intern Med 2003; 138: 273–87.
www.thelancet.com Vol 370 August 25, 2007
Viewpoint
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL,
Pinder EL. The implications of regional variations in Medicare
spending. Part 2: Health outcomes and satisfaction with care.
Ann Intern Med 2003; 138: 288–99.
Kilminster SM, Jolly BC. Effective supervision in clinical practice
settings: a literature review. Med Educ 2000; 34: 827–40.
Steinert Y, Mann K, Centeno A, et al. A systematic review of faculty
development initiatives designed to improve teaching effectiveness
in medical education: BEME Guide No. 8. Med Teacher 2006; 28:
497–526.
Mourad O, Redelmeier DA. Clinical teaching and clinical outcomes:
teaching capability and its association with patient outcomes.
Med Educ 2006; 40: 637–44.
Gore A. An Inconvenient Truth: the crisis of global warming. New
York: Rodale Books, 2006.
McManus IC, Mollon J, Duke OL, Vale JA. Changes in standard of
candidates taking the MRCP (UK) Part 1 examination, 1985 to 2002:
analysis of marker questions. BMC Med 2005; 3: 13.
Fred HL. Hyposkillia. Deficiency of clinical skills. Tex Heart Inst J
2005; 32: 255–56.
Ludmerer KM. Time to Heal. Oxford: Oxford University Press;
1999: 306.
Emanuel EJ. Changing premed requirements and the medical
curriculum. JAMA 2006; 296: 1128–31.
Cooke M, Irby DM, Sullivan W, Ludmerer KM. American medical
education 100 years after the Flexner report. N Engl J Med 2006; 355:
1339–44.
Mattern WD, Weinholtz D, Friedman CP. The attending physician
as teacher. N Engl J Med 1983; 308: 1129–32.
Irby DM. What clinical teachers in medicine need to know.
Acad Med 1994; 69: 333–42.
Weinholtz D, Edwards J. Teaching during rounds. A handbook for
attending physicians and residents. Baltimore: Johns Hopkins
University Press, 1992.
Hafferty FW. Professionalism—the next wave. N Engl J Med 2006;
355: 2151–52.
Smith C, Varkey A, Evans AT, Reilly BM. Evaluating the
performance of inpatient attending physicians: a new instrument
for today’s teaching hospitals. J Gen Intern Med 2004; 19: 766–72.
Ericsson KA, Simon HA. Protocol analysis: verbal reports as data,
revised edn. Cambridge MA: Bradford Books/MIT Press, 1993.
Ericsson KA, Charness N, Feltovich PJ, Hoffman RR, eds. The
Cambridge Handbook of Expertise and Expert Performance.
Cambridge UK: Cambridge University Press, 2006.
Shaneyfelt T, Baum K, Bell D, et al. Instruments for evaluating
education in evidence-based practice: a systematic review. JAMA
2006; 296: 1116–27.
Lucas B, Evans AT, Reilly BM, et al. The impact of evidence on
physicians’ inpatient treatment decisions. J Gen Intern Med 2004; 19:
402–409.
Reilly BM. The essence of EBM. BMJ 2004; 329: 991–92.
Brancati FL. The art of pimping. JAMA 1989; 262: 89–90.
Penn W. The peace of Europe: Some Fruits of Solitude and other
writings. London: JM Dent and Sons, 1930.
Aagard E, Teherani A, Irby DM. Effectiveness of the one-minute
preceptor model for diagnosing the patient and the learner: proof of
concept. Acad Med 2004; 79: 42–49.
Bowen JL. Educational strategies to promote clinical diagnostic
reasoning. N Engl J Med 2006; 355: 2217–25.
Gennis VM, Gennis MA. Supervision in the outpatient clinic:
effects on teaching and patient care. J Gen Intern Med 1993; 8:
378–80.
Haner KE, Wachter RM, McCulloch CE, Woo GA, Auerbach AD.
Effects of hospitalist attending physicians on trainee satisfaction
with teaching and with internal medicine rotations. Arch Intern Med
2004; 164: 1866–71.
Whitehead, AN. The aims of education. New York: Macmillan; 1929.
Wasson JH, Sox HC, Neff RK, Goldman L. Clinical prediction rules:
application and methodological standards. N Engl J Med 1985; 313:
793–99.
www.thelancet.com Vol 370 August 25, 2007
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
Osler W. On the need of a radical reform in our methods of
teaching senior students. Med News (New York) 1903; 82: 49–53.
Reilly BM. Physical examination in the care of medical inpatients:
an observational study. Lancet 2003; 362: 1100–05.
Linfors EW, Neelon FA. The case for bedside rounds. N Engl J Med
1980; 303: 1230–33.
Weinberg RB. The laying on of hands. Ann Intern Med 1992; 117:
83–84.
Wright SM, Kern DE, Kolodner K, Howard DM, Brancati FL.
Attributes of excellent attending physician role models.
N Engl J Med 1998; 339: 1986–93.
Skeff KM, Mutha S. Role models—guiding the future of medicine.
N Engl J Med 1998; 339: 2015–17.
Ericsson KA. Deliberate practice and the acquisition and
maintenance of expert performance in medicine and related
domains. Acad Med 2004; 79 (10 suppl): S70–81.
Bean WB. Sir William Osler: aphorisms from his bedside teachings
and writings. Springfield: Charles C Thomas, 1968.
Rothman KJ, ed. Causal inference. Chestnut Hill: Epidemiology
Resources, 1988.
Lazare A. Apology in medical practice: an emerging clinical skill.
JAMA 2006; 296: 1401–04.
Wu AW. Medical error: the second victim. BMJ 2000; 320: 726–27.
Konner M. Becoming a doctor: a journey of initiation in medical
school. New York: Penguin Books; 1988.
Groopman J. How doctors think. Boston: Houghton Mifflin, 2007:
234–59.
McManus IC, Keeling A, Paice E. Stress, burnout and doctors’
attitudes to work are determined by personality and learning style.
A twelve year longitudinal study of UK medical graduates.
BMC Medicine 2004; 2: 29.
Peabody FW. The care of the patient. JAMA 1927; 88: 877–82.
ACGME Outcome Project. Enhancing residency education through
outcomes assessment; general competencies, version 1.2. Chicago:
Accreditation Council for Graduate Medical Education, 1999.
General Medical Council. Good medical practice, 2006. http://www.
gmc–uk.org/guidance/good_medical_practice/index.asp. (accessed
Nov 10, 2006).
Arnold L, Stern DT. What is medical professionalism? In: Stern DT,
ed. Measuring medical professionalism. Oxford: Oxford University
Press, 2006.
Horton R, Gilmore I. The evolving doctor. Lancet 2006; 368:
1750–51.
Comte-Sponville A. A small treatise on the great virtues. New York:
Metropolitan Books, 2001.
Grady-Weliky TA, Kettyle CN, Hundert EM. The mentor–mentee
relationship in medical education. In: Wear D, Bickel J, eds.
Educating for professionalism. Iowa City: University of Iowa Press,
2000.
Maudsley RF. Content in context: medical education and society’s
needs. Acad Med 1999; 74: 143–45.
Osler W. Aequanimitas, with other addresses to medical students,
nurses and practitioners of medicine, 3rd edn. Philadelphia:
P Blakiston, 1932.
Arky RA. The family business—to educate. N Engl J Med 2006; 354:
1922–26.
Davis MH, Ponnamperuna GG. Medical education research at the
crossroads. Lancet 2006; 367: 377–78.
Torgerson CJ. Educational research and randomized trials.
Med Educ 2002; 36: 1002–03.
Schuwirth LWT, van der Vleuten CPM. Challenges for
educationalists. BMJ 2006; 333: 544–46.
McMichael AJ, Woodruff RE, Hales S. Climate change and human
health: present and future risks. Lancet 2006; 367: 859–69.
Haines A, Kovats RS, Campbell-Lendrum D, Corvalan C. Climate
change and human health: impacts, vulnerability and mitigation.
Lancet 2006; 367: 2101–09.
711
Case Report
Mysterious falls and a nasal voice
Avijit Bhandari, Firdaus Adenwalla
Lancet 2007; 370: 712
Neath Port Talbot Hospital,
Baglan Way, Port Talbot,
SA12 7BX, Wales
(A Bhandari MB,
F Adenwalla FRCP)
Correspondence to:
Dr Avijit Bhandari, Flat 3, Block C,
Staff Residences, Neath Port
Talbot Hospital, Baglan Way
Port Talbot, SA12 7BX Wales, UK
[email protected]
In December, 2003, a 77-year-old man was admitted to our
district general hospital, from a residential home. He had
been in the home for 6 h, having been transferred there
from a tertiary hospital, where he had spent 2 months.
His stay had been devoted to an unsuccessful investigation
of his recurrent falls, and deteriorating ability to walk.
The falls had started about 6 months before his admission,
transforming him from an independent man, the main
carer for his disabled wife, to one unable to leave his
house. His known diagnoses were ischaemic heart
disease, hypothyroidism, pernicious anaemia, cervical
spondylosis, and asbestosis; his prescribed medications
were glyceryl trinitrate spray, atenolol, thyroxine, vitamin
B₁₂ injections, ferrous sulphate, codeine phosphate, and
lansoprazole. Investigations in the tertiary hospital had
included screening blood tests, an autoantibody screen,
24-h ECG monitoring, CT of the head and thorax, and
endoscopy of the upper and lower gastrointestinal tracts.
These investigations had shown a mild normocytic
anaemia, and pleural plaques of asbestos—but nothing
unexpected, other than a high serum concentration of
p-ANCA. Prednisolone was prescribed, despite the
absence of clinical evidence of vasculitis, but proved to be
ineffective; the dose was being reduced at the time of
discharge. The patient had continued to fall while in
hospital, and had become able to walk for only about
5 m—with a wheeled walking frame. In the absence of a
diagnosis, he had received intensive physiotherapy and
occupational therapy, and had been discharged to the
residential home—where, after a further fall, the staff had
sent him to us.
Nerve ending
Vesicles containing ACh
Antibody
Some receptors blocked
or damaged by antibodies
Receptor on surface
of muscle fibre
Receptor stimulated by ACh
Figure: The cause of myasthenia gravis
Autoantibodies bind to acetylcholine receptors, reducing transmission at the neuromuscular junction.
712
On the patient’s admission to our hospital, the clinical
examination was normal, except for bilateral partial
ptosis, which the patient described as longstanding. We
then noticed that the patient’s voice became nasal towards
the end of a conversation; the change was not noticed by
him, but we concluded that he had palatal palsy, and
suspected myasthenia gravis. We took a blood sample,
requesting a test for acetylcholine-receptor antibodies.
While we awaited the result, the patient developed
dysphagia: videofluoroscopy revealed global pharyngeal
weakness. We had difficulties obtaining edrophonium
for a Tensilon test. We therefore prescribed
pyridostigmine, on a trial basis: within 24 h, the patient
was swallowing normally and walking independently; we
therefore continued to give him this drug. The
acetylcholine-receptor antibody test was strongly positive.
A subsequent serum p-ANCA concentration was normal.
When last seen, in May, 2006, the patient drove to our
clinic; he was able to walk without support, had no speech
impediment, and no difficulty swallowing, although his
ptosis remained. He had not fallen since he started taking
pyridostigmine.
The near-ubiquity of falls in elderly people1 should not
deter doctors from seeking the cause. The annual
incidence of myasthenia gravis in the UK is estimated at
1–2 in 100 000, but may be five times as high in elderly
people2,3—in whom the disorder is thought to be
substantially underdiagnosed.3 The mean time taken to
diagnose myasthenia gravis in people over 60 years of
age, after they develop symptoms, was estimated in 1996
as 4·5 months—almost twice the time taken to diagnose
the disorder in younger people.2 The classic presentation
of painless weakness on exertion can be overlooked in
elderly people, or attributed to other causes; symptoms
can remit and relapse; and doctors can omit to consider
the disease. Once the diagnosis is suspected, testing for
it is straightforward: the acetylcholine-receptor antibody
test is highly specific in patients with generalised
myasthenia gravis (80–90%), particularly in elderly
men.4 The diagnosis can also be made by use of the
Tensilon test, or single-fibre electromyography, on
which most people with myasthenia gravis have
abnormal jittering.
References
1
Blake AJ, Morgan K, Bendall MJ, et al. Falls by elderly people at home:
prevalence and associated factors. Age Ageing 1988; 17: 365–72.
2
Schon F, Drayson M, Thompson RA. Myasthenia gravis and elderly
people. Age Ageing 1996; 25: 56–58.
3
Vincent A, Clover L, Buckley C, Grimley Evans J, Rothwell PM; UK
Myasthenia Gravis Survey. Evidence of underdiagnosis of
myasthenia gravis in older people. J Neurol Neurosurg Psychiatry
2003; 74: 1105–08
4
Chua E, McLoughlin C, Sharman AK. Myasthenia gravis and
recurrent falls in the elderly patient. Age Ageing 2000; 29: 83–84.
www.thelancet.com Vol 370 August 25, 2007