Download minerva anestesiologica

MINERVA ANESTESIOLOGICA
ITALIAN JOURNAL OF ANESTHESIOLOGY AND ANALGESIA
MONTHLY JOURNAL FOUNDED IN 1935 BY A. M. DOGLIOTTI
OFFICIAL JOURNAL OF ITALIAN SOCIETY OF ANESTHESIOLOGY, ANALGESIA,
RESUSCITATION AND INTENSIVE CARE (S.I.A.A.R.T.I.)
Vol. 81
August 2015
No. 8
OFFICIAL JOURNAL OF ITALIAN SOCIETY OF ANESTHESIOLOGY, ANALGESIA,
RESUSCITATION AND INTENSIVE CARE (SIAARTI)
CONTENTS
825
837
Videolaryngoscopy: may the force be with you!
Surgical Pleth Index guided analgesia blunts the intraoperative sympathetic response to laparoscopic cholecystectomy
EDITORIALS
Kleine-Brueggeney M., Theiler L. G.
ORIGINAL ARTICLES
Colombo R., Raimondi F., Rech R., Castelli A., Fossali T.,
Marchi A., Borghi B., Corona A., Guzzetti S.
827
Measuring (and interpreting) the esophageal pressure:
a challenge for the intensivist
Grasso S., Cassano S.
846
830
Is esophageal pressure monitoring essential for the
most critically ill?
Berra L., Kacmarek R. M.
Indirect videolaryngoscopy using Macintosh blades in
patients with non-anticipated difficult airways results
in significantly lower forces exerted on teeth relative
to classic direct laryngoscopy: a randomized crossover
trial
Pieters B., Maassen R., Van Eig E., Maathuis B., Van
Den Dobbelsteen J., Van Zundert A.
832
How should I structure my Post-ICU Clinic? From early
goal rehabilitation to outpatient visits
Ranzani O. T., Jones C.
835
855
Near-infrared spectroscopy for monitoring brain oxygenation: to trust or not to trust?
Bruder N., Velly L.
Vol. 81 - No. 8
Esophageal pressure measurements under different
conditions of intrathoracic pressure. An in vitro study
of second generation balloon catheters
Mojoli F., Chiumello D., Pozzi M., Algieri I., Bianzina S.,
Luoni S., Volta C. A., Braschi A., Brochard L.
MINERVA ANESTESIOLOGICA
I
CONTENTS
865
Feasibility of Post-Intensive Care Unit Clinics:
an observational cohort study of two different
approaches
Dettling-Ihnenfeldt D. S., De Graaff A. E., Nollet F., Van
Der Schaaf M.
910
REVIEWS
Preoxygenation and general anesthesia: a review
Bouroche G., Bourgain J. L.
921
876
Assessment of cerebral oxygenation in neurocritical care patients: comparison of a new four wavelengths forehead regional saturation in oxygen sensor
(EQUANOX®) with brain tissue oxygenation. A prospective observational study
Esnault P., Boret H., Montcriol A., Carre E., Prunet B.,
Bordes J., Simon P., Joubert C., Dagain A., Kaiser E.,
Meaudre E.
Statin therapy in critically-ill patients with severe sepsis: a review and meta-analysis of randomized clinical
trials
Thomas G., Hraiech S., Loundou A., Truwit J., Kruger P.,
Mcauley D. F., Papazian L., Roch A.
931
LETTERS TO THE EDITOR
Association between steroid particle sizes and serious
complications during epidural injections
885
Moderate-degree acidosis is an independent determinant of postoperative bleeding in cardiac surgery
Ghobadifar M. A.
Ranucci M., Baryshnikova E., Simeone F., Ranucci M.,
Scolletta S.
932
894
Caldiroli D., Orena E. F.
OPRM1 receptor as new biomarker to help the prediction of post mastectomy pain and recurrence in breast
cancer
933
EXPERT OPINION
De Gregori M., Diatchenko L., Belfer I., Allegri M.
901
Epidural steroid injections: update on efficacy, safety,
and newer medications for injection
Kozlov N., Benzon H. T., Malik K.
Missed citations
Videolaryngoscopy offers us more than classic direct
laryngoscopy
Van Zundert A. A. J., Pieters B. M. A.
935
TOP 50 MINERVA ANESTESIOLOGICA
REVIEWERS
V O L U M E 8 1 · N o. 8 · A U G U S T 2 0 1 5
About the cover: the cover shows the trends of rSO2 and PbtO2 for patient n. 4 in the prospective,
observational, unblinded study about the assessment of cerebral oxygenation in neurocritical care
patients. Data show the inability of rSO2 to detect ongoing brain death. For more information, see
the article by Esnault P. et al. beginning on page 876.
II
MINERVA ANESTESIOLOGICA
August 2015

EDITORIAL
Videolaryngoscopy: may the force be with you!
M. KLEINE-BRUEGGENEY, L. G. THEILER
Department of Anaesthesiology and Pain Medicine, Inselspital, University of Bern and Bern University Hospital, Bern,
Switzerland
S
ince John Pacey, a surgeon, introduced the
GlideScope® into clinical practice in 2001,
videolaryngoscopes (VLS) have become increasingly successful. Similar to the use of ultrasound
guided techniques for vascular puncture and
nerve blocks, VLS have very quickly gained
popularity among anesthesiologists. They are
becoming more and more indispensable tools
for teaching purposes, for the management of
difficult airways and as documentation tools for
everyday cases. Many different VLS are available
and their number keeps steadily increasing. Prior
to marketing, all these devices lack evidence of
efficacy or safety. Hence, without academic guidance, the choice to use and to buy one particular
VLS will depend on marketing strategies of the
companies. The British Difficult Airway Society
has addressed this problem in an article that defines “a minimum level of evidence needed to make
a pragmatic decision about the purchase or selection
of an airway device”.1 In this issue of Minerva
Anestesiologica, Pieters et al. provide some of the
necessary evidence about efficacy and safety of
three VLS.2 From everyday clinical practice we
know that the force necessary to obtain a good
view of laryngeal structures is markedly decreased with VLS. This has also been shown by
Goto et al.3 Pieters and the study group led by
André van Zundert present more data enforcing
this knowledge. They confirm their previously
published finding that the force exerted on the
maxillary incisors is lower with the use of VLS
compared to the use of the Macintosh laryngoscope.2, 4, 5 We cannot directly deduct that the
Comment on p. 846.
Vol. 81 - No. 8
incidence of dental lesions is reduced by using
VLS, but it is difficult to study the incidence of
dental lesions because they occur in only about
1/2000 (0.05%) of anesthesia cases.6 The force
exerted on the teeth appears to be an acceptable
surrogate parameter. Importantly, those findings
apply to the non-difficult airway, not the nonanticipated difficult airway: the title of the study
might be misleading.
VLS can be divided into devices without a
guiding channel for the tracheal tube (such as
the three devices evaluated by Pieters et al.) and
devices with a guiding channel. Additionally,
VLS blades may resemble the standard Macintosh blade (e.g. the C-MAC® blades evaluated
in the study) or may feature a more pronounced
curve (e.g. the MacGrath® series 5 and the GlideScope® 7 evaluated by Pieters et al., or the CMAC “D-blade”). Curved blades are primarily
designed for the difficult airway and direct comparisons with Macintosh blades are difficult. The
more curved the blade, the more essential it is
to introduce a stylet into the tracheal tube for
guidance. If a stylet is not used, tracheal intubation will be more difficult, as shown by Pieters et
al. who did not use stylets in their study.2 Most
likely, this is why the GlideScope® seemed to
perform inferiorly.
Facing the emerging importance of VLS, a
crucial question becomes whether we should
abandon the 80-year old standard Macintosh
blade in favor of VLS. While superiority has
been claimed for VLS in the ICU setting 8 and
evidence shows that in normal airways, laryngoscopy becomes even easier when using videolaryngoscopes, there are important advantages
MINERVA ANESTESIOLOGICA
825
KLEINE-BRUEGGENEYVIDEOLARYNGOSCOPY
of direct laryngoscopy using the Macintosh
blades. The most obvious one is the fact that
one drop of blood or mucus may be sufficient to
completely obstruct the view obtained by videolaryngoscopes. Also, equipment failure remains
a problem.9 The Macintosh laryngoscope is a
simple, reliable tool that is difficult to break. It
is cheap, transportable, available in all sizes and
usable in all settings, even in the pre-hospital
setting in bright sunlight. Of note, VLS have so
far not been incorporated into difficult airway
algorithms, although this may change in the
near future.10 While VLS seem to be very valuable assets to the airway tool library, we risk losing our skills with two important techniques by
more and more using VLS: intubation with the
ubiquitously available Macintosh laryngoscope
and fibreoptic intubation. Several studies on
VLS in the simulated difficult airway situation
using manual inline stabilization have been conducted, mostly demonstrating a better visibility
of the vocal cords and some showing a higher intubation success rate with VLS compared to the
Macintosh laryngoscope.9, 11 Despite that, it is
also known that even with a good view obtained
by the VLS, there still might be problems to actually intubate the trachea (“you see that you
fail”).11 Therefore, alternative techniques like the
flexible fibreoptic intubation must continue to
be taught and used on a regular basis. To secure
the airway in the spontaneously breathing patient (awake intubation) remains the gold-standard in the management of the anticipated difficult airway, especially when difficult face-mask
ventilation is suspected, and should not be abandoned. Videolaryngoscopes are additions, not
replacements to our airway tool library. Their
role in securing patients’ airways is increasingly
being supported by evidence like the study by
Pieters et al. More evidence will have to follow
in the future, especially about the role of VLS in
the setting of difficult airway management.
References
 1. Pandit JJ, Popat MT, Cook TM, Wilkes AR, Groom P,
Cooke H et al. The Difficult Airway Society ‘ADEPT’ guidance on selecting airway devices: the basis of a strategy for
equipment evaluation. Anaesthesia 2011;66:726-37.
 2. Pieters B, Maassen R, van Eig E, Maathuis B, van Den
Dobbelsteen J, van Zundert A. Indirect videolaryngoscopy
using Macintosh blades in patients with non-anticipated
difficult airways results in significantly lower forces exerted
on teeth relative to classic direct laryngoscopy; a randomized crossover trial. Minerva Anestesiol 2015;81:846-54.
  3. Goto T, Koyama Y, Kondo T, Tsugawa Y, Hasegawa K. A
comparison of the force applied on oral structures during
intubation attempts between the Pentax-AWS airwayscope
and the Macintosh laryngoscope: a high-fidelity simulatorbased study. BMJ Open 2014;4:e006416.
  4. Lee RA, van Zundert AA, Maassen RL, Willems RJ, Beeke
LP, Schaaper JN et al. Forces applied to the maxillary incisors during video-assisted intubation. Anesth Analg
2009;108:187-91.
  5. Lee RA, van Zundert AA, Maassen RL, Wieringa PA. Forces applied to the maxillary incisors by video laryngoscopes
and the Macintosh laryngoscope. Acta Anaesthesiol Scand
2012;56:224-9.
  6. Newland MC, Ellis SJ, Peters KR, Simonson JA, Durham
TM, Ullrich FA et al. Dental injury associated with anesthesia: a report of 161,687 anesthetics given over 14 years. J
Clin Anesth 2007;19:339-45.
  7. Agro’ F, Doyle D, Vennari M. Use of Glidescope in adults:
an overview. Minerva Anestesiol 2015;81:342-51.
  8. De Jong A, Molinari N, Conseil M, Coisel Y, Pouzeratte Y,
Belafia F et al. Video laryngoscopy versus direct laryngoscopy for orotracheal intubation in the intensive care unit:
a systematic review and meta-analysis. Intensive Care Med
2014;40:629-39.
  9. Ilyas S, Symons J, Bradley W, Segal R, Taylor H, Lee K
et al. A prospective randomised controlled trial comparing tracheal intubation plus manual in-line stabilisation
of the cervical spine using the Macintosh laryngoscope
vs the McGrath Series 5 videolaryngoscope. Anaesthesia
2014;69:1345-50.
10. Frova G. Do videolaryngoscopes have a new role in the
SIAARTI difficult airway management algorithm? Minerva
Anestesiol 2010;76:637-40.
11. Byhahn C, Iber T, Zacharowski K, Weber CF, Ruesseler
M, Schalk R et al. Tracheal intubation using the mobile
C-MAC video laryngoscope or direct laryngoscopy for patients with a simulated difficult airway. Minerva Anestesiol
2010;76:577-83.
Conflicts of interest.—The authors are investigators on several randomized controlled trials of videolaryngoscopes and received grants from
the “Gottfried und Julia Bangerter-Rhyner Foundation”, from the “Fondation Latine des Voies Aériennes (FLAVA)” and the “Swiss Society
of Anaesthesiology and Resuscitation” for an ongoing study comparing videolaryngoscopes.
Received on November 15, 2014. - Accepted for publication on November 20, 2014. -Epub ahead of print on November 26, 2014.
Corresponding author: L. G. Theiler, Department of Anaesthesiology and Pain Medicine, University Hospital Bern, Inselspital CH-3010
Bern, Switzerland. E-mail: [email protected]
826
MINERVA ANESTESIOLOGICA
August 2015

EDITORIAL
Measuring (and interpreting)
the esophageal pressure:
a challenge for the intensivist
S. GRASSO, S. CASSANO
Dipartimento dell’Emergenza e Trapianti d’Organo (DETO), Sezione di Anestesiologia eRianimazione, Università degli
Studi Aldo Moro, Bari, Italy
I
n a mechanically ventilated passive patient
a portion of the positive pressure applied at
the airway opening (PAO) does not distend the
lung but works to “move” the chest wall. As a
consequence, the pleural pressure (PPL) increases
above its end-expiratory level, proportionally to
chest wall stiffness. What actually distends the
lung is the increase in trans-pulmonary pressure
(PL) above its end-expiratory level, i.e. PAO minus PPL.1 The straightforward implication is that
PL is always lower than the applied PAO.2 In patients with acute respiratory distress syndrome
(ARDS), in the attempt to minimize alveolar
hyperinflation, we limit the end-inspiratory airway opening plateau pressure (PAO,PLAT) to 30
cmH2O.3 In the most severe ARDS forms, in the
attempt to increase lung aeration, we use lungrecruiting maneuvers (LRM) followed by high
positive end-expiratory pressure (PEEP).4-6 Indeed, alveolar recruitment may change the clinical course of severe ARDS: a non “recruiting”
patient is candidate to “rescue” strategies, for
example extracorporeal membrane oxygenation
(ECMO).7 For PAO,PLAT, LRM and PEEP we
reason in terms of pressure applied at the airway
opening, assuming that the chest wall stiffness is
normal. Unfortunately, this is not the case in a
relevant portion of patients.8 Several pathologic
conditions, for example deformities, pleural efComment on p. 865.
Vol. 81 - No. 8
fusions and increased abdominal pressure may
impair chest wall compliance. If the chest wall
is stiff, a relevant portion of PAO is dissipated to
move it, generating higher PPL and lower PL.
When this happens, the PAO-based lung protective and/or lung-recruiting strategies often fail,
regardless the potential for alveolar recruitment.9
Studies have shown that titrating mechanical
ventilation on PL rather than on PAO significantly improves gas exchange and lung mechanics 10
and may reverse refractory hypoxemia.11
In the assisted ventilation modes, the patient
actively contracts his or her inspiratory muscles
and PL results by the interplay between the ventilator that pushes (positive PAO) and the patients
that pulls (negative PPL). Think to a patient with
mild ARDS non-invasively ventilated with a
pressure support of 15 cmH2O. If this patient
generates a substantial inspiratory effort (PPL
minus 20 cmH2O), the end-inspiratory PL will
be 35 cmH2O, a figure compatible with ventilator induced lung injury (VILI).12 Indeed, PL
is of paramount importance to estimate work
of breathing and patient-ventilator interactions
during assisted ventilation.13
Despite its importance, we rarely measure
PL in clinical practice or, even worst, take the PL
“concept” into account in our clinical reasoning. There are at least four reasons to explain this
paradox:
A) Measuring PPL is virtually impossible in the
MINERVA ANESTESIOLOGICA
827
GRASSO
MEASURING (AND INTERPRETING) THE ESOPHAGEAL PRESSURE
clinical setting. We have a surrogate, the esophageal pressure (PES), measured through a catheter
positioned in the lower esophageal third. Since
the esophageal lumen is a virtual space, to sense
the pressure acting on the esophageal wall, the
catheter tip must be inserted in an air filled balloon. The air volume put in the balloon needs
careful titration: if the balloon collapses on the
catheter tip, PES is underestimated. On the other
hand, if the balloon is overinflated, the stretch
on the balloon wall generates positive pressure
by itself, and PES is overestimated. In adjunct,
the PES reading is influenced by patient posture
and the esophageal muscular tone and contractions.1
B) The correct catheter positioning in the lower esophageal third requires expertise. There are
several methods to check the catheter position:14
some of them require active patients inspiration
against the occluded airway.15 Unfortunately few
ventilators are equipped with airway opening occlusion devices.
C) We lack of devices to measure PES. For a
correct reading PES should be showed together
with PAO, Flow and Volume, in real time. Few
ventilators and multi-parametric monitors are
equipped with an auxiliary port to measure PES.
D) After years of debates, we yet don’t know
how to interpret PES.16, 17 Some authors maintain that the absolute PES value is just the “right”
PPL value.10, 14, 18 Others do not trust on absolute PES and just trust on the PAO and PES swings
for partitioning the respiratory system elastance
(ERS) into its lung and chest wall components
(EL and EC, w).2 This is reasonable since: a) the
EL/ECW ratio defines the partitioning of PAO in
PPL and PL during positive pressure lung inflation; and b) a single “real” absolute PPL virtually
does not exist. In healthy subjects PPL is slightly
negative at functional residual capacity (FRC),
varies with gravity and body posture 1 and is occasionally frankly positive in the dependent lung
regions, for example in patients with abdominal
hypertension.14
In summary, while the experts recommend
measuring PL, several problems wait a solution. The paper by Mojoli et al. published in
the present issue of Minerva Anestesiologica is
a step in this direction.19 It is an elegant and
828
rigorous in vitro study testing six second-generation PES catheters at different balloon filling
volumes and surrounding pressures to define the
catheter-specific range of “appropriate” balloon
filling volumes. Rather surprisingly, it shows
that this range may be different from the one
recommended by catheters manufacturers. Most
importantly, the study establishes a “gold standard” in vitro approach to test PES catheters. We
urgently need preclinical and clinical studies like
this one.13 Our challenge is to quickly go from
theory to practice. Measuring and interpreting
PL is complex, requires considerable expertise
and technological improvements. Nevertheless,
we should wonder if it is challenging like measuring and interpreting the electrocardiogram.
Cardiologists won this challenge several years
ago. Will the intensivists do the same in the near
future?
References
  1. Hedenstierna G. Esophageal pressure: benefit and limitations. Minerva Anestesiol 2012;78:959-66.
  2. Gattinoni L, Chiumello D, Carlesso E, Valenza F. Benchto-bedside review: chest wall elastance in acute lung injury/
acute respiratory distress syndrome patients. Crit Care
2004;8:350-5.
  3. Network TA. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury
and the acute respiratory distress syndrome. The Acute
Respiratory Distress Syndrome Network. N Engl J Med
2000;342:1301-08.
  4. de Matos GF, Stanzani F, Passos RH, Fontana MF, Albaladejo R, Caserta RE et al. How large is the lung recruitability in
early acute respiratory distress syndrome: a prospective case
series of patients monitored by computed tomography. Crit
Care 2012;16:R4.
 5. Fan E, Wilcox ME, Brower RG, Stewart TE, Mehta S,
Lapinsky SE et al. Recruitment maneuvers for acute lung
injury: a systematic review. Am J Respir Crit Care Med
2008;178:1156-63.
  6. Gattinoni L, Caironi P, Cressoni M, Chiumello D, Ranieri
VM, Quintel M et al. Lung recruitment in patients with
the acute respiratory distress syndrome. N Engl J Med
2006;354:1775-86.
  7. Esan A, Hess DR, Raoof S, George L, Sessler CN: Severe
hypoxemic respiratory failure: part 1--ventilatory strategies.
Chest 2010;137:1203-16.
  8. Malbrain ML, Cheatham ML, Kirkpatrick A, Sugrue M,
Parr M, De Waele J et al. Results from the International
Conference of Experts on Intra-abdominal Hypertension
and Abdominal Compartment Syndrome. I. Definitions.
Intensive Care Med 2006;32:1722-32.
  9. Staffieri F, Stripoli T, De Monte V, Crovace A, Sacchi M,
De Michele M, Trerotoli P, Terragni P, Ranieri VM, Grasso
S: Physiological effects of an open lung ventilatory strategy titrated on elastance-derived end-inspiratory transpulmonary pressure: study in a pig model. Crit Care Med
2012;40:2124-31.
MINERVA ANESTESIOLOGICA
August 2015
MEASURING (AND INTERPRETING) THE ESOPHAGEAL PRESSUREGRASSO
10. Talmor D, Sarge T, Malhotra A, O’Donnell CR, Ritz R, Lisbon A et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med 2008;359:2095-104.
11. Grasso S, Terragni P, Birocco A, Urbino R, Del Sorbo L,
Filippini C et al. ECMO criteria for influenza A (H1N1)associated ARDS: role of transpulmonary pressure. Intensive Care Med 2012;38:395-403.
12. Slutsky AS, Ranieri VM: Ventilator-induced lung injury. N
Engl J Med 2014;370:980.
13. Akoumianaki E, Maggiore SM, Valenza F, Bellani G, Jubran A, Loring SH et al. The application of esophageal pressure measurement in patients with respiratory failure. Am J
Respir Crit Care Med 2014;189:520-31.
14. Loring SH, O’Donnell CR, Behazin N, Malhotra A, Sarge
T, Ritz R et al. Esophageal pressures in acute lung injury:
do they represent artifact or useful information about transpulmonary pressure, chest wall mechanics, and lung stress?
J Appl Physiol 2010;108:515-22.
15. Baydur A, Behrakis PK, Zin WA, Jaeger M, Milic-Emili J.
A simple method for assessing the validity of the esophageal
balloon technique. Am Rev Respir Dis 1982;126:788-91.
16. Hubmayr RD. Is there a place for esophageal manometry
in the care of patients with injured lungs? J Appl Physiol
2010;108:481-2.
17. Chiumello D, Cressoni M, Colombo A, Babini G, Brioni
M, Crimella F et al. The assessment of transpulmonary
pressure in mechanically ventilated ARDS patients. Intensive Care Med 2014;40:1670-8.
18. Talmor D, Sarge T, O’Donnell CR, Ritz R, Malhotra A,
Lisbon A et al. Esophageal and transpulmonary pressures in
acute respiratory failure. Crit Care Med 2006;34:1389-94.
19. Mojoli F, Chiumello D, Pozzi M, Algieri I, Bianzina S, Luoni S et al. Esophageal pressure measurements under different conditions of intrathoracic pressure. An in vitro study
of second generation balloon catheters. Minerva Anestesiol
2015;81:855-64.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on April 24, 2015. - Accepted for publication on April 25, 2015. - Epub ahead of print on April 30, 2015.
Corresponding author: S. Grasso,Università degli Studi di Bari “Aldo Moro”. Dipartimento dell’Emergenza e Trapianti d’Organo
(DETO),Sezione di Anestesiologia e Rianimazione, Azienda Ospedaliero-Universitaria, PoliclinicoPiazza Giulio Cesare 11, Bari, Italy.
E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
829

EDITORIAL
Is esophageal pressure monitoring
essential for the most critically ill?
L. BERRA 1, R. M. KACMAREK 1, 2,
1Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital and Harvard Medical
School, Boston, MA, USA; 2Department of Respiratory Care, Massachusetts General Hospital, Boston, MA, USA
T
he use of esophageal manometry has become
increasingly important for the management
of the most critically ill mechanically ventilated
patients.1 When performed properly, esophageal pressure measurement provides an estimate
of pleural pressure and its measurement has advanced our understanding of the pathophysiology of critically ill patients and has increasingly
assisted in the management of the most complex
patients. Esophageal pressure measurement is
useful in guiding the setting of positive end-expiratory pressure (PEEP) in severe adult respiratory
distress syndrome (ARDS),2 the determination
of maximum plateau pressure without undue risk
of induced lung injury in patients with altered
chest wall mechanics or marked obesity,3 and
for the determination of autoPEEP and work of
breathing in spontaneously breathing patients
receiving assisted ventilation. Scientific interest
in the exploration of possible applications of esophageal pressure guided ventilation has been recently highlighted by the PLUG working group.1
The use of esophageal manometry to describe
intrathoracic pressure changes during spontaneous breathing dates back to 1878.5 The estimation
of pleural pressure through esophageal manometry allows the determination of transpulmonary
pressures (end inspiratory plateau pressure minus
pleural pressure), i.e. the distending pressure of
the lungs.6 The assessment of how much pressure
is spent for the passive inflation of the thorax
is particularly helpful in any clinical condition
Comment on p. 855.
830
characterized by increased pleural pressure and/
or chest wall elastic abnormalities, including but
not limited to ARDS, obesity or surgical pneumoperitoneum. This differentiation allows the
clinician to better understand whether it is reasonably safe and feasible to increase airway pressures to better customize mechanical ventilation
in different pathophysiological settings.3 Specifically, to be able to identify those situations where
plateau pressure can exceed 28 cmH2O without
increased risk of lung injury. End expiratory
transpulmonary pressure has been proposed as
the ideal method of determining optimal PEEP.2
The goal, is to insure end expiratory transpulmonary pressure in positive and thus to avoid end
expiratory collapse. Additionally, the measurement of esophageal pressure has proven useful
in hemodynamic interpretation (as it allows the
calculation of transmural filling pressures). Monitoring of patient-ventilator synchrony, measurement of auto-PEEP during spontaneous breathing and estimation of work-of-breathing through
measurement of the pressure-time product has
been shown to be particularly useful in patients
difficult to wean from ventilator support.1
Since the introduction of esophageal manometry, there has been debate whether esophageal
pressure faithfully reflected pleural pressure. The
main issue with esophageal pressure measurements is its geometrically limited validity. As
pleural pressure is characterized by different values, following a gravity-vectored gradient, the
pleural pressure estimated through esophageal
manometry is valid only for structures in close
MINERVA ANESTESIOLOGICA
August 2015
ESOPHAGEAL PRESSURE MONITORING IN CRITICAL ILLNESSBERRA
proximity of the balloon itself. For the same reason, balloon positioning at different depths yields
different pleural pressure estimates. As a result positioning and functioning of the balloon is critical. Proper positioning is at the mid-esophageal
level identified by the presence of cardiac artifact
and a matching of esophageal pressure change
with airway pressure change during active inspiration determined by an occlusion test.6 Therefore, the esophageal pressure and airway pressure
change during inspiration and expiration should
be equivalent. Similarly during controlled ventilation, airway and esophageal pressure should
change equivalently. Additionally, the volume of
air used for balloon inflation influences the reliability of esophageal manometry. Balloon pressure
increases with balloon volume, as a result of distension of the balloon itself and the surrounding
structures, as shown in 1964 by Milic-Emili J et
al.4 These authors concluded that esophageal pressure best estimates pleural pressure at near-zero
balloon volume. More recently Walterspacher et
al.7 and Mojoli et al.8 published interesting papers
that address the issue of esophageal balloon filling volumes. Specifically, in this issue of Minerva
Anestesiologica, Mojoli et al. 8 studied in vitro the
minimum and maximum volume at which different catheters accurately measure the surrounding
pressure. It is interesting to note that the minimum volume to be injected for proper pressure
measurement was greater than recommended by
each of the four manufacturers of the catheters
tested, and the best balloon working volume did
vary considerably among catheters. Clinicians, in
addition, should be aware that the pressure-volume curve of the balloon changes after the first
inflation, suggesting that some inflation/deflation
cycles should be performed before placement.
The authors emphasize that in the case of an unsuccessful occlusion test evaluation of balloon
filling volume should occur before repositioning
of the balloon. According to their data, some balloons require working filling volumes of up to 7.5
mL considerably more that has been traditionally
recommended (0.5 to 1.0 mL).4 However, these
studies were performed exclusively in vitro and
on a single catheter per brand; adoption of the
suggested filling volumes is still debated. Furthermore, the findings of these two studies have yet to
be validated in an in vivo setting.
Esophageal manometry is an increasingly
important tool in the management of complex, refractory respiratory failure, allowing the
evaluation of the true amount of airway pressure
used for lung inflation and the minimum level
of PEEP required to avoid cyclic inflation and
deflation. It is also useful for the assessment of
autoPEEP, and work of breathing in the spontaneously breathing patient failing ventilator
discontinuation trials. However, esophageal balloon manometry is still flawed by methodological issued. Mojoli et al. have filled one of the gaps
that afflict measurement.8 Future clinical studies
however are needed to validate their results.
References
  1. Akoumianaki E, Maggiore SM, Valenza F, Bellani G, Jubran A, Loring SH et al. The application of esophageal pressure measurement in patients with respiratory failure. Am J
Respir Crit Care Med 2014;189:520-31.
  2. Talmor D, Sarge T, Malhotra A, O’Donnell CR, Ritz R, Lisbon A et al. Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med 2008;359:2095-104.
  3. Chiumello D, Carlesso E, Cadringher P, Caironi P, Valenza
F, Polli F et al. Lung stress and strain during mechanical
ventilation for acute respiratory distress syndrome. Am J
Respir Crit Care Med 2008;178:346-55.
  4. Milic-Emili J, Mead J, Turner Jm, Glauser Em. Improved
technique for estimating pleural pressure from esophageal
balloons. J Appl Physiol 1964;19:207-11.
  5. Luciani L. Archivio delle scienze mediche. Vol II. Torino:
Tipografo Vincenzo Bona; 1878. p. 186.
  6. Sarge T, Talmor D. Transpulmonary pressure: its role in preventing ventilator-induced lung injury. Minerva Anestesiol
2008;74:335-9.
  7. Walterspacher S, Isaak L, Guttmann J, Kabitz HJ, Schumann S. Assessing respiratory function depends on mechanical characteristics of balloon catheters. Respir Care
2014;59:1345-52.
 8. Mojoli F, Chiumello D, Pozzi M, Algieri I, Bianzina S,
Luoni S et al. Esophageal pressure measurements under
different conditions of intrathoracic pressure. An in vitro
study of second generation balloon catheters Minerva Anestesiol 2015;81:855-64.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on Febrary 27, 2015. - Accepted for publication on April 28, 2015. - Epub ahead of print on April 30, 2015.
Corresponding author: R. M. Kacmarek, Department of Respiratory and Department of Anesthesia, Massachusetts General Hospital and
Harvard Medical School, Warren Building 1225, 55 Fruit Street, Boston, MA 02114, USA. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
831

EDITORIAL
How should I structure my Post-ICU Clinic?
From early goal rehabilitation to outpatient visits
O. T. RANZANI 1, C. JONES
2, 3
1Respiratory Intensive Care Unit, Pulmonary Division, Heart Institute, Hospital das Clínicas, Faculdade de Medicina,
Universidade de São Paulo, São Paulo, Brazil; 2Critical Care Rehabilitation, Whiston Hospital, Prescot, UK; 3Institute
of Ageing and Chronic Disease, Faculty of Health and Life Sciences, University of Liverpool, Liverpool, UK
P
assing through a critically illness and facing
the Intensive Care Unit (ICU) scenario are
experiences that frequently brings uncomfortable physical, cognitive and psychological sequelae for patients and family members.1-3 In
contrast, a smaller proportion of patients deal
with surviving ICU as a victory in their life and,
so, a source of energy and a life changing opportunity.
Concurrently, the multidisciplinary staff involved in the care of critically ill patients also
receive many inputs during the workday apart
of technical issues: from happiness of recovering
patients and gratefulness of family members to
stressful situations, dealing with life threatening
conditions, death, and impotence feelings.4
Nowadays, the proportion of ICU survivors
is increasing and new challenges caused to the
health care system, health care workers and
population. Indeed, the suffering of recovering
patients and relatives can be high and negatively
affects their quality of life.1
Furthermore, the burden of ICU survivors is
enormous and needs special attention.5 In this
issue of Minerva Anestesiologica, Dettling-Ihnenfeldt et al.6 reported a pilot experience in conducting two Post-ICU Clinics in Netherlands.
Post-ICU clinics are specialized units with the
main aim to help ICU survivors and was first described in UK in 19857. Several centres around
the world have implemented these clinics; howComment on p. 865.
832
ever, there are questions around their implementation.6, 8, 9
What are the aims of the Post-ICU Clinic?
This is a crucial point in the topic and has
changed over time. Post-ICU Clinics can be
facilities for screening ICU survivors and their
families sometime after hospital discharge. The
rational is to identify patients who are not recovering or presented with new problems.6 PostICU Clinics can also be focused on physical and
psychological rehabilitation, manned by staff
specializing in the care of ICU survivors, who
have different demands from routine patients,
such as hallucinations related to their ICU stay.
However, the ideal goal of Post-ICU Clinics
should be an advanced facility that allows continuity of care. Indeed, initiatives beginning as
early as possible (sometimes inside the ICU)
could decrease the incidence of problems after
hospital discharge.10, 11 The context of each centre should be understood, since the cost-effectiveness of these models must be discussed. In a
national survey in the UK, financial constraints
was the main reason for not having a Post-ICU
Clinic.12 Other point raised by Dettling-Ihnenfeldt et al.6 is the existence of other services
within the health system, which already provide
rehabilitation and support, sometimes close to
the patients’ home or familiar to them because
of previous attendance. The ICU-rehabilitation
MINERVA ANESTESIOLOGICA
August 2015
HOW SHOULD I STRUCTURE MY POST-ICU CLINIC?RANZANI
team should be aware of all other possibilities of
referral and support. In this context, Post-ICU
Clinics could be a referral point for ICU survivors and caregivers.
Should we offer and refer to the Post-ICU
Clinic every ICU survivor or should
we identify high-risk patients? Should
family members be asked to visit?
In the study by Dettling-Ihnenfeldt et al.6
patients who were mechanical ventilated ≥48
h and were discharged to their own home were
invited to attend the Post-ICU Clinic. The authors reported that an average 15% of ICU patients, from the two hospitals, were invited to
the clinic. Although there is a rational for this
criteria, it can includes a myriad of patients,
from surgical patients to patients with acute
respiratory distress syndrome, one of the populations most affected and studied after ICU
discharge.13 Other studies included all patients
who stayed some time in ICU (i.e. 5 days) or included every patient who received high-level of
care.8 Although some risk factors for poor outcomes are known 14, there is no consensus on
how select patients, since the recovery process is
subject to the influence of individual characteristics.9, 12 Furthermore, as discussed by the authors, 18% of selected patients did not attended
because of having “no complaints”. This could
be a matter of concern, since it is possible they
did not attended because of avoidance behavior
as a symptoms of post traumatic stress disorder
and, so, are vulnerable to long-term problems.
There is no consensus whether active recruitment for this situation is advisable or whether
telephone follow-up could be an alternative.9
Certainly early contact with the patient and respective family whilst still hospitalized, can decrease the need for long-term follow-up and allows the recognition of those patients who may
need ongoing physiotherapy or psychological
support.
Another interesting point reported by the authors is how to care for family members and caregivers. The authors assessed the close relatives
and informal caregivers, finding high level of
Vol. 81 - No. 8
burden and stressful conditions in approximately 10-15%. This approach with close relatives
should be integrated into the design of Post-ICU
Clinics because they are fundamental actors in
the process of care and suffering.7 Although not
reported in this study, family members of patients who died in ICU should also be evaluated,
similar to palliative care teams supporting relatives in the grief period. Whether close relatives
should visit separately from patients is a question
that still needs to be addressed.
What staff should compose
the Post-ICU Clinic?
Another question about Post-ICU Clinics is
what professions or services should be available.
In the study by Dettling-Ihnenfeldt et al.,6 each
centre have different profile. One centre was
nurses-led, a common finding in the literature 7-9
and in the other centre, nurses, physiotherapists,
psychologists and physicians were present. We
believe that a multidisciplinary team is advisable
and the inclusion of staff from the ICU is desirable.3 Indeed, patients and family members may
have close relationship to the ICU staff who attended them during critical illness and this could
enhance the coping in stressful situations. Furthermore, the ICU staff will have a chance to
face different challenges, observe good results, or
feed back to other staff on practices that have
had a positive or negative impact on the patients’
recovery. Indeed, outcomes from critically ill patients should be assessed sometime after ICU
discharge.
Challenges and promising approaches
As reported by the authors,6 patients from
hospitals of different levels of care deserves
care after critical illness. When planning the
introduction of a Post-ICU Clinic, health care
workers and policy makers should consider all
the above aspects. The risks and features of the
patients following ICU admission is dynamic
over time,3, 15 ICU and hospital discharge are
moments of great anxiety and carry important
care level changes. Therefore, focusing on each
moment of this time frame is essential: the im-
MINERVA ANESTESIOLOGICA
833
RANZANI
HOW SHOULD I STRUCTURE MY POST-ICU CLINIC?
plementation of light sedation and early rehabilitation,16 ICU diaries,10 manualise rehabilitation programmes such as the ICU Recovery
Manual,17 home-based programmes 18 and new
approaches to optimize an effective communication between teams from ICU and ward, ward
and outpatient services are fundamental.
The introduction of a Post-ICU Clinic is desirable and might be very useful, but has several
challenges.9, 12, 19 Based on the balance between
benefits and costs, we should discuss whether
it is better to guarantee inpatient rehabilitation
and discharge facilities first and only then introduce a Post-ICU Clinic; or whether is suitable
conduct both strategies at the same time.
References
  1. Jones C. What’s new on the post-ICU burden for patients
and relatives? Intensive Care Med 2013;39:1832-5.
  2. Herridge MS. Recovery and long-term outcome in acute
respiratory distress syndrome. Crit Care Clin 2011;27:685704.
  3. Flaatten H. Follow-up after intensive care: Another role for
the intensivist? Acta Anaesthesiol Scand 2005;49:919-21.
 4. Azoulay E, Herridge M. Understanding ICU staff burnout: The show must go on. Am J Respir Crit Care Med
2011;184:1099-100.
  5. Prescott HC, Langa KM, Liu V, Escobar GJ, Iwashyna TJ.
Increased 1-year healthcare use in survivors of severe sepsis.
Am J Respir Crit Care Med 2014;190:62-9.
  6. Dettling-Ihnenfeldt DS, de Graaff AE, Nollet F, van der
Schaaf M. Feasibility of Post-Intensive Care Unit Clinics:
an observational cohort study of two different approaches.
Minerva Anestesiol 2015;81:865-75.
 7.Griffiths RD, Jones C. Seven lessons from 20 years of
follow-up of intensive care unit survivors. Curr Opin Crit
Care 2007;13:508-13.
 8.Cuthbertson BH, Rattray J, Campbell MK, Gager M,
Roughton S, Smith A et al. The PRaCTICaL study of nurse
led, intensive care follow-up programmes for improving
long term outcomes from critical illness: a pragmatic randomised controlled trial. BMJ 2009;339:b3723.
  9. Modrykamien AM. The ICU follow-up clinic: a new paradigm for intensivists. Respir Care. 2012;57:764-72.
10. Jones C, Bäckman C, Capuzzo M, Egerod I, Flaatten H,
Granja C et al. Intensive care diaries reduce new onset post
traumatic stress disorder following critical illness: a randomised, controlled trial. Crit Care 2010;14.
11. Herridge M, Cox C. Linking ICU practice to long-term
outcome: Fostering a longitudinal vision for ICU-acquired
morbidity. Am J Respir Crit Care Med 2012;186:299-300.
12. Griffiths JA, Barber VS, Cuthbertson BH, Young JD. A national survey of intensive care follow-up clinics. Anaesthesia
2006;61:950-5.
13. Dowdy DW, Eid MP, Sedrakyan A, Mendez-Tellez PA, Pronovost PJ, Herridge MS et al. Quality of life in adult survivors of critical illness: A systematic review of the literature.
Intensive Care Med 2005;31:611-20.
14. Jones C, Bäckman C, Capuzzo M, Flaatten H, Rylander C,
Griffiths RD. Precipitants of post-traumatic stress disorder
following intensive care: A hypothesis generating study of
diversity in care. Intensive Care Med 2007;33:978-85.
15. Ranzani OT, Zampieri FG, Park M, Salluh JI. Long-term
mortality after critical care: what is the starting point? Crit
Care 2013;17:191.
16. Park M, Pires-Neto RC, Nassar Junior AP. Awaking, exercising, sitting, walking and extubating: moving on the paradigms for mechanically ventilated patients. Rev Bras Ter
Intensiva 2014;26:203-4.
17. Jones C, Skirrow P, Griffiths RD, Humphris GH, Ingleby S,
Eddleston J et al. Rehabilitation after critical illness: a randomized, controlled trial. Crit Care Med 2003;31:2456-61.
18. Elliott D, McKinley S, Alison J, Aitken LM, King M, Leslie
GD et al. Health-related quality of life and physical recovery
after a critical illness: a multi-centre randomised controlled
trial of a home-based physical rehabilitation program. Crit
Care 2011;15:R142.
19. Herridge MS. The challenge of designing a post-critical illness rehabilitation intervention. Critical Care 2011;15.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on November 24, 2014. - Accepted for publication on December 3, 2014. - Epub ahead of print on December 5, 2014.
Corresponding author: O. T. Ranzani, LIM/09 – Laboratorio de Pneumologia, Faculdade de Medicina da USP, Av. Dr. Arnaldo, 455 Laboratório de Pneumologia, 2º andar, sala 2144, Cerqueira César 01246903 - Sao Paulo, SP, Brasil. E-mail: [email protected]
834
MINERVA ANESTESIOLOGICA
August 2015

EDITORIAL
Near-infrared spectroscopy for monitoring
brain oxygenation: to trust or not to trust?
N. BRUDER, L. VELLY
Department of Anesthesiology and Intensive Care, CHU Timone, Aix-Marseille University, Marseille, France
N
ear-infrared spectroscopy (NIRS) is increasingly used in anesthesia to monitor brain
regional saturation in oxygen (rSO2). In cardiac
surgical patients, rSO2 monitoring has shown
interesting results suggesting an improvement
in brain morbidity in some situations.1, 2 Large
changes in brain oxygenation may be expected
during cardiac surgery due to cannula malposition or during carotid endarterectomy after carotid artery clamping, for example. Still, definitive
demonstration of the usefulness of this monitoring technique for improving patient outcome is
lacking. In the neurointensive care unit, the main
objective is to avoid or limit cerebral ischemia, especially after traumatic brain injury. Changes in
brain oxygenation may guide our treatments for
this purpose. However, the amplitude of brain oxygenation changes is usually small, needing accurate monitoring methods. Spurious brain oxygenation values may lead to inappropriate therapeutic
interventions and inadequate patient management. Observational studies using brain interstitial tissue pressure in oxygen (PbtO2) have shown
promising results. Compared to historical control
groups, patients managed with PbtO2-guided
management had better outcomes.3, 4 However,
this monitoring is invasive, limiting its use only
to sedated severe brain injured patients. It would
certainly be a real monitoring breakthrough if
intensivists could obtain reliable information on
brain oxygenation using a non-invasive device like
NIRS, as it is used in the operating room.
Comment on p. 876.
Vol. 81 - No. 8
In the past, a few studies have shown important limitations of NIRS to monitor brain
oxygenation.5 In normal subjects NIRS monitoring can be adversely affected by extra-cranial
contamination of the signal after routine physiological tests, like Valsalva maneuver, hyperventilation and head-up tilt.6 It was confirmed
in volunteers with no difference between three
NIRS devices.7 In anesthetized patients in
beach-chair position for arthroscopic shoulder
surgery, there was no agreement between SjO2
and rSO2 values, and rSO2 was not reliable in
detecting a low SjO2. In anesthetized children,
NIRS devices demonstrated a correlation with
SjO2, but the agreement between rSO2 and SjO2
was poor.8 In brain-injured patients, Ter Minassian et al. used a carbon dioxide (CO2) and
blood pressure challenge with norepinephrine to
compare changes in rSO2 and SjO2. They demonstrated that changes in rSO2 were positively
correlated with changes in SjO2 during a CO2
challenge but negatively during the blood pressure challenge. This suggested that extracerebral
vasoconstriction influenced rSO2 values making
it unreliable for clinical monitoring in the ICU.
One can argue that NIRS not only measures venous blood oxygenation but both arterial and
venous saturation. In fact, NIRS devices use a
fixed proportion of arterial (30%), capillary and
venous (70%) blood within brain tissue. This is
also a limitation in brain-injured patients because this proportion may change over time.
Recently, new devices have appeared on the
market and all manufacturers claim the algo-
MINERVA ANESTESIOLOGICA
835
BRUDER
NEAR-INFRARED SPECTROSCOPY FOR MONITORING BRAIN OXYGENATION
rithms have been improved in order to limit extracerebral contamination of the signal. Thus, it
is important to test these devices to assess their
value for clinical monitoring. The study by Esnault et al. published in this issue of Minerva
Anestesiologica compared one new NIRS device
with brain tissue oxygen pressure in oxygen
(PbtO2).9 Strength of this study was to use an
accurate measurement method of regional cerebral oxygenation in the same area than rSO2
measurement. Of course, one can argue again
that PbtO2 and rSO2 are not the same measures
of brain oxygenation. However, the results were
straightforward: like the other monitors (INVOS™ 10, CEROX™ 11) rSO2 measured by the
EQUANOX™ monitor could not detect most
ischemic episodes detected by PbtO2.
This does not mean that NIRS monitoring is
useless. For example, NIRS has been validated as
an acceptable method for continuous assessment
of cerebral autoregulation.12 However, only
changes over time were taken into account and
not absolute values of brain oxygenation. At this
time, we have to follow recent guidelines telling
us NIRS is not currently indicated for routine
monitoring of neurointensive care patients.13
Undoubtedly, NIRS will still be used for research
purposes, but at this time we have to take the information provided by NIRS devices cautiously.
References
  1. Murkin JM, Adams SJ, Novick RJ, Quantz M, Bainbridge
D, Iglesias I et al. Monitoring brain oxygen saturation during coronary bypass surgery: a randomized, prospective
study. Anesth Analg 2007;104:51-8.
  2. Zheng F, Sheinberg R, Yee MS, Ono M, Zheng Y, Hogue
CW. Cerebral near-infrared spectroscopy monitoring and
neurologic outcomes in adult cardiac surgery patients: a systematic review. Anesth Analg 2013;116:663-76.
 3.Spiotta A, Stiefel M, Gracias V, Garuffe A, KoFKe W,
Maloney-wilensky E et al. Brain tissue oxygen–directed
management and outcome in patients with severe traumatic
brain injury. J Neurosurg 2010;113:571-80.
  4. Stiefel M, Spiotta A, Gracias V, Garuffe A, Guillamondegui
O, Maloney-Wilensky E et al. Reduced mortality rate in patients with severe traumatic brain injury treated with brain
tissue oxygen monitoring. J Neurosurg 2005;103:805-11.
 5. Messerer M, Daniel RT, Oddo M. Neuromonitoring after major neurosurgical procedures. Minerva Anestesiol
2012;78:810-22.
 6. Canova D, Roatta S, Bosone D, Micieli G. Inconsistent
detection of changes in cerebral blood volume by near infrared spectroscopy in standard clinical tests. J Appl Physiol
2011;110:1646-55.
  7. Davie SN, Grocott HP. Impact of extracranial contamination on regional cerebral oxygen saturation: A comparison
of three cerebral oximetry technologies. Anesthesiology
2012;116:834-40.
  8. Nagdyman N, Ewert P, Peters B, Miera O, Fleck T, Berger
F. Comparison of different near-infrared spectroscopic cerebral oxygenation indices with central venous and jugular venous oxygenation saturation in children. Paediatr Anaesth
2008;18:160-6.
  9. Esnault P, Boret H, Montcriol A, Carre E, Prunet B, Bordes
J et al. Assessment of cerebral oxygenation in neurocritical
care patients: comparison of a new four wavelengths forehead regional saturation in oxygen sensor (equanox®)
with brain tissue oxygenation. A prospective observational
study. Minerva Anestesiol 2015;81:876-84.
10. Leal-Noval SR, Cayuela A, Arellano-Orden V, Marin-Caballos A, Padilla V, Ferrandiz-Millon C et al. Invasive and
noninvasive assessment of cerebral oxygenation in patients
with severe traumatic brain injury. Intensive Care Med
2010;36:1309-17.
11. Rosenthal G, Furmanov A, Itshayek E, Shoshan Y, Singh V.
Assessment of a noninvasive cerebral oxygenation monitor
in patients with severe traumatic brain injury. J Neurosurg
2014;120:901-7.
12. Zweifel C, Castellani G, Czosnyka M, Carrera E, Brady
KM, Kirkpatrick PJ et al. Continuous assessment of cerebral
autoregulation with near-infrared spectroscopy in adults after subarachnoid hemorrhage. Stroke 2010;41:1963-8.
13. Oddo M, Bosel J. Monitoring of brain and systemic oxygenation in neurocritical care patients. Neurocrit Care
2014;21:S103-20.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on January 24, 2015. - Accepted for publication on January 28, 2015. - Epub ahead of print on January 30, 2015.
Corresponding author: N Bruder, Department of Anesthesiology and Intensive Care, CHU Timone, Aix-Marseille University, 265 rue
St-Pierre, 13385 Marseille, France. E-mail: [email protected]
836
MINERVA ANESTESIOLOGICA
August 2015

O R I G I N A L A RT I C L E
Surgical Pleth Index guided analgesia blunts
the intraoperative sympathetic response
to laparoscopic cholecystectomy
R. COLOMBO 1, F. RAIMONDI 2, R. RECH 1, A. CASTELLI 1, T. FOSSALI 1
A. MARCHI 2, B. BORGHI 1, A. CORONA 1, S. GUZZETTI 3
1Anesthesiology and Intensive Care Unit, Azienda Ospedaliera Luigi Sacco, Polo Universitario, University of Milan,
Milan, Italy;2Anesthesiology and Intensive Care Unit, Istituto Clinico Humanitas IRCCS, Rozzano, Italy; 3Emergency
Department, Azienda Ospedaliera Luigi Sacco, Polo Universitario, University of Milan, Milano, Italy
ABSTRACT
Background. Surgical noxious stimuli generate a stress response with an increased sympathetic activity, potentially
affecting the perioperative outcome. Surgical Pleth Index (SPI), derived from the pulse plethysmogram, has been
proposed as a tool to assess nociception-antinociception balance. The relationship between SPI and autonomic nervous system (ANS) during general anesthesia is poorly understood and it is doubtful if SPI-guided analgesia may offer
advantages over the standard clinical practice. The study was designed to evaluate if SPI-guided analgesia leads to a
lower sympathetic modulation compared with standard clinical practice.
Methods. Electrocardiographic wave, non-invasive blood pressure and SPI were recorded in ASA I-II patients undergoing elective laparoscopic cholecystectomy, randomized to receive SPI-guided analgesia or standard analgesia.
Hemodynamic parameters, SPI, mean and variance of heart rate, low (LF) and high frequency (HF) spectral components of heart rate variability were measured at four time points: (T0) baseline, (T1) after induction of general
anesthesia, (T2) after pneumoperitoneum insufflation and (T3) after pneumoperitoneum withdrawal.
Results. SPI, hemodynamic and ANS parameters changed significantly in both groups during the study period
(P<0.0001). At T2 SPI and markers of sympathetic modulation were significantly lower in SPI group (mean [SD]
SPI 38.1 [15.3] vs. 48.1 [16.2] normalized units, P<0.05; LF 38 [8.6] vs. 56.2 [20.6] normalized units, P<0.01; LF/
HF 1.01 [1.1] vs. 2.68 [2.07], P<0.01). There was no difference in remifentanil consumption, recovery time from
anesthesia, or postoperative pain and complications.
Conclusion. SPI-guided analgesia led to a more stable sympathetic modulation but didn’t seem to offer clinically relevant advantages over the standard clinical practice for laparoscopic cholecystectomy. (Minerva Anestesiol 2015;81:837-45)
Key words: Autonomic nervous system - Nociception - Heart Rate.
A
cute painful stimulation induces modification of the autonomic nervous system
(ANS) modulation through activation of the
sympathetic nerve activity directed to the heart
and vessels.1, 2 During general anesthesia the
conscious perception of pain is abolished, thus
the assessment of nociception is difficult. Traditionally, patients’ movements or autonomic
mediated reflexes like tachycardia, hypertension
and tearing are used by anesthesiologists to as-
Vol. 81 - No. 8
sess the adequacy of antinociception. Unfortunately, these clinical signs have been proven to
carry low specificity and sensitivity as indicators
of nociception in anaesthetized patients.3-6 Recently the Surgical Pleth Index (SPI), a novel
pulse plethysmograph-derived index has been
proposed as a tool to predict the nociceptionanti-nociception balance during general anesthesia.7 The SPI seems to be related to the remifentanil effect-site concentration, is unrelated to the
MINERVA ANESTESIOLOGICA
837
COLOMBO
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESS
propofol effect-site concentration during general
anaesthesia,8, 9 and it predicts the probability of
movements at skin incision more accurately than
commonly used clinical parameters.7 High SPI
values are considered indicators of a prevalence
of the nociception over the antinociception.
To date, it is unclear if a SPI-guided analgesia
offers some advantages in terms of ANS activation during general anesthesia. Direct measurement of ANS activity is not feasible in a clinical
context. Heart rate variability (HRV) analysis
in the frequency domain provides an indirect
measurement of sympathetic and vagal modulation directed to the heart.10, 11 We conducted a
prospective randomized controlled trial to investigate if SPI-guided analgesia results in a lower
sympathetic activity measured by means of the
HRV analysis compared to a standard clinical
practice in patients undergoing laparoscopic
cholecystectomy.
Materials and methods
Patients selection
This is a single centre, prospective, randomized, double blinded study, conducted at the
Luigi Sacco Hospital in Milan, Italy, from February 2010 to March 2011. Eligible participants
were adults aged between 18 and 50 years old,
American Society of Anaesthesiologists status I
and II who were scheduled for elective laparoscopic cholecystectomy under general anesthesia.
Exclusion criteria were comprised of: a history of
diabetes, hypertension requiring pharmacological treatment, chronic cardiovascular or neurological disease, obesity, history of alcohol or drug
abuse, arrhythmic cardiac disease (atrial fibrillation, atrial flutter, ectopic beats >5% of normal
sinus beats), endocrine pathology (thyroidal or
adrenal), peripheral neuropathy, and any known
autonomic dysfunction. The study was previously approved by the local Ethics Committee and
all patients provided written informed consent.
Study protocol
The eligible patients were randomized into
one of the two study groups: SPI-guided analge-
838
sia group (SPI group) and standard practice analgesia group (control group). All patients were
asked to fast eight hours preoperatively, and no
vagolytic drugs (i.e. atropine, scopolamine) or
benzodiazepines were administered as premedication. Ten milligrams of oxycodone was administered orally 1 hour before the surgery. Upon
arrival at the operating theatre, patients were
placed in a supine position and connected to a
S/5 Advance Monitor provided with SPI (General Electric, Helsinki, Finland). Monitoring consisted of continuous electrocardiogram (ECG),
non-invasive blood pressure every 2.5 min, pulse
oximetry, respiratory gas analysis, end-tidal carbon dioxide (EtCO2), electroencephalographic
rest (RE) and state (SE) entropy, pulse photoplethysmography, cutaneous temperature at the
tenar eminence of the left hand, and SPI. The
SPI ranges from 0 to 100 and may be calculated
as SPI=100-(0.33xPBInorm + 0.67xPPGAnorm),9
where PBI and PPGA are the normalized pulse
beat interval and pulse plethysmographic amplitude, derived from the photo plethysmographic
probe. Normalization algorithm is a property
of the manufacturer and it is undisclosed. All
data were recorded continuously and were exported to a laptop computer provided with the
S/5 Collect software and real time SPI measurement (General Electric, Helsinki, Finland). All
variables were analysed at four main time points:
1) T0, baseline before induction of general anesthesia; 2) T1, after the tracheal intubation and
five min after starting of mechanical ventilation;
3) T2, five min after the insufflation of pneumoperitoneum; 4) T3, five min after the removal
of pneumoperitoneum.
Induction and maintenance of anesthesia
In all patients, anesthesia was induced with
propofol and remifentanil via target-controlled
infusion pumps (Asena Alaris; Cardinal Health,
Basingstoke, United Kingdom). The pharmacokinetic models used were Schnider’s and
Minto’s for propofol and remifentanil respectively. The predicted effect-site concentration of
propofol (CEprop) initially was set at 4 µg·mL-1
and remifentanil (CEremi) at 4 ng·mL-1. After loss
of consciousness, oxygen was given by facemask
MINERVA ANESTESIOLOGICA
August 2015
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESSCOLOMBO
and muscle relaxation was induced with cis-atracurium bolus (0.2 mg x kg-1). After tracheal tube
placement, mechanical ventilation was started at
15 breaths min-1 with a volume targeted to get
an EtCO2 between 35-38 mmHg. The respiratory rate was increased if the EtCO2 exceeded 38
mmHg with a tidal volume of 10 mL x kg-1 and/
or an airway plateau pressure exceeding 35 cmH2O. A respiratory rate <15 breaths x min-1 was
not allowed during the study period.
The CEprop was adjusted by step of 1 µg x
mL-1 every min to maintain an electroencephalographic state entropy (SE) level between 40-60
(the minimum allowed CEprop was 3 µg x mL-1).
A rescue dose of cis-atracurium (0.08 mg x kg-1)
was administrated 40 min after the induction of
general anesthesia if needed. In the SPI group,
the remifentanil infusion rate was adjusted by
step of 1 ng x mL-1 every 1 min to maintain SPI
below 50 (the allowed range CEremi was 2-15 ng x
mL-1). In the control group the SPI was unavailable to the anesthesiologist while it continued to
be acquired by the researchers to a laptop computer. In the control group, the anesthesiologist
adjusted the remifentanil infusion rate based on
the clinical signs of inadequate analgesia (systolic blood pressure above +20% from the baseline
or mean blood pressure >100 mmHg, heart rate
>90 bpm, coughing, chewing, grimacing, midriasis, tears) or excessive analgesia (heart rate <50
bpm, systolic blood pressure below -20% from
the baseline or mean blood pressure <55 mmHg
in absence of active bleeding). A pneumoperitoneum was surgically induced and maintained
between 12-14 mmHg.
Severe bradycardia (<45 bpm) was treated
with atropine 0.01 mg x kg-1 i.v. and hypotension (systolic blood pressure <90 mmHg or
mean blood pressure <50 mmHg) unresponsive
to fluid challenge of ringer’s acetate 500 mL was
treated with ephedrine boluses 5 mg i.v. The patients were considered drop-outs and the data
were excluded from the analysis if atropine or
vasopressors were needed at any time before the
conclusion on the study protocol, blood loss
was >200 mL, laparoscopy was converted into
laparotomic surgery, or cutaneous temperature
dropped >1 °C from baseline.
Non-invasive blood pressure was measured
Vol. 81 - No. 8
every 2.5 min. Cutaneous temperature, SE, RE,
and SPI values represented the mean of values
recorded every 15 seconds during the ECG acquisition. The heart rate is displayed as R-to-R
mean interval of ECG (msec). Beats per minute
are calculated with the equation: HR=(60/
RRmean) x 103.
After the withdrawal of the gallbladder, tramadol 100 mg and acethaminophen 1 g were
administered intravenously. Propofol and
remifentanil were stopped at the time of trocar
withdrawal.
Recovery time was calculated as the time
elapsed from the stop of propofol and remifentanil infusion to tracheal extubation.
Postoperative pain was measured with Numerical Rating Scale (NRS) 6 and 24 hours after
the surgery.
Heart rate variability analysis
At each study time point, the ECG was sampled at 300 Hz for HRV analysis according to
the recommendations of the European Society
of Cardiology Task Force.12 HRV analysis was
performed offline by a semi-automatic tachogram identifier (R-to-R interval identification
made by a derivative/threshold method taken
from the ECG).13 After detecting the QRS complex on the ECG and locating the R-apex, the
interval between two consecutive R was computed as heart period. Postoperatively, the tachogram was extracted from the recorded ECG and
reviewed by an investigator to avoid erroneous
detections or missed beats. If isolated ectopic
beats occurred, they were removed and linearly
interpolated using the closest values unaffected
by ectopic beats.
The power spectrum was estimated using an
autoregressive model.10 Sequences of 300 consecutive heart beats were selected inside each
experimental step. The mean and the variance
of heart period are expressed in msec and ms2
respectively. Autoregressive spectral density
was factorized into components each of them
characterized by a central frequency. A spectral
component was labeled as LF if its central frequency was between 0.04 and 0.15 Hz, while
it was classified as HF if its central frequency
MINERVA ANESTESIOLOGICA
839
COLOMBO
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESS
was between 0.15 and 0.5 Hz.12 The LF and HF
powers are defined as the sum of the powers of
all LF and HF spectral components respectively.
The HF power was considered to represent respiration-driven vagal heart rate modulation.10, 14
The LF may reflect vasomotor activity, which is
an indirect index of sympathetic nerve activity
and partially affected by parasympathetic activity.15-17 Spectral values were expressed in normalized units (nu) to remove the effect of changes
in total power spectrum densities on LF and HF
components. Normalization consists of dividing the power of a given spectral component by
the total power minus the power below 0.04 Hz
(Very Low Frequency spectral component), and
multiplying the ratio by 100. The ratio of the LF
power to the HF (LF/HF) was considered an indicator of the balance between sympathetic and
vagal modulation directed to the heart.10
The investigator who analysed the heart rate
variability data was blinded to SPI value.
Outcomes
Primary endpoint: to evaluate if SPI-guided
analgesia leads to a lower sympathetic modulation compared to standard clinical practice during laparoscopic cholecystectomy.
Secondary endpoints: to identify if the two
groups have a difference in intraoperative hemodynamic variables, consumption of remifentanil,
recovery time from anesthesia or postoperative
pain.
Statistics
Randomization was based on computer generated list balanced in block of six. Sample size
was calculated with PS Power and Sample Size
Calculator Software for Windows (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize). There is a wide range in size of ANS
variables reported in literature. A previous study
found that opiate administration to healthy volunteers during a paced spontaneous breathing
trial led to a significant reduction of LF/HF (1.6
vs. 0.92, P=0.013).18 A minimum of 30 subjects
in each study group was required to detect a difference of 0.8 (SD 1.1) in the means of LF/HF
840
between groups at the time of maximal visceral
stimulation (pneumoperitoneum insufflation),
with a probability of 0.8 and alpha error of 0.05
with two tails test for unpaired groups.
Continuous variables are expressed as mean
±SD, or median (25-75th percentiles) when not
normally distributed. The D’Agostino-Pearson’s
test was used to check for normal distribution.
Student’s t test for independent samples was used
to compare the means (or Mann-Whitney U
when not normally distributed) between groups
of non-repeated measures. Repeated measures
were analysed with two way ANOVA followed
by Bonferroni’s post hoc test for the comparison between groups. The categorical variables
differences in proportions were compared using
the χ2 test or Fisher’s exact test, as appropriate.
The recovery time from general anesthesia were
analysed using Kaplan–Meier log-rank survival
analysis (calculating the cumulative probability
of patients remaining intubated after discontinuation of the anesthetic drugs).
Two tailed P values <0.05 were considered statistically significant. GraphPad Prism 5 was used
for statistical analysis.
Results
During the study, 64 patients were recruited.
Two patients in the SPI-group and one patient
in the control group were excluded because of
intraoperatively atropine administration. One
patient in the control group was excluded because of conversion to open laparotomy. A total
of 60 patients (30 in each group) were analysed
(Figure 1).
Patient characteristics are summarized in
Table I. Hemodynamic variations and ANS
variables during the study protocol are summarized in Figure 2. The SPI showed significant
variation during the study protocol with group
(P=0.009) and time (P<0.0001) accounting for
the total variance. During the maximal visceral
stimulation corresponding to the induction
of pneumoperitoneum, SPI was significantly
higher in the control group than in SPI group
(48.1±16.2 vs. 38.1±15.3, P<0.05). Heart rate,
systolic and mean arterial pressures, and all HRV
indexes changed significantly during the study
MINERVA ANESTESIOLOGICA
August 2015
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESSCOLOMBO
Figure 1.—Enrolment diagram of the studied sample.
Table I.—Characteristics of the studied patients.
Age (yrs)
BMI
M/F
ASA I/II
Control group
(N.=30)
SPI group
(N.=30)
P
49.9 (11.4)
26.18 (4.19)
14/16
19/11
46.6 (12.2)
26.2 (3.6)
13/17
19/11
0.28
0.79
1
1
Data are expressed as mean (SD) or number of subjects.
protocol, with time accounting for the total
variance (P<0.0001). At time of the induction
of pneumoperitoneum the power of variability in the low frequency spectrum and the LF/
HF ratio were significantly higher in the control group than in SPI group (LF 56.2±20.6 nu
vs. 38±18.6 nu, P<0.01; LF/HF 2.68±2.07 vs.
1.01±1.1, P<0.01). The heart rate, expressed
as R-to-R interval on ECG, was also different
between groups (control group 0.944±0.139
Vol. 81 - No. 8
s vs. SPI group 1.045±0.139 s, P<0.05). There
were no significant differences in the CEremi between control and SPI group during study protocol (T1 3.84±0.51 vs. 4.03±1.05 ng x mL-1,
T2 4.61 ±1.69 vs. 5.17±1.23 ng x mL-1, and T3
4.23 ±2.18 vs. 4.56 ±1.65 ng x mL-1). There was
no significant differences in the total remifentanil consumption between control and SPI
group (1056.1±440.5 μg vs. 1036.7±420.6 μg,
P=0.5) or indexed remifentanil consumption
MINERVA ANESTESIOLOGICA
841
COLOMBO
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESS
Figure 2.—Grey columns represent control group, white columns represent the SPI group. *P<0.05, **P<0.01 between groups.
SPI: Surgical Pleth Index; LF: low frequency component of power spectrum of HRV expressed in normalized units; HF: high
frequency component of power spectrum of HRV expressed in normalized units; R-to-R: mean interval between QRS peaks on
ECG expressed in sec; SAP: systolic arterial pressure; MAP: mean arterial pressure.
ent between control and SPI groups at 6 hours
(NRS 3.5±2.1 vs. 3.2±2.2, P=0.6) and 24 hours
(NRS 2.3±1.8 vs. 1.5±1.6, P=0.17) after the
end of surgery. No other complications were
reported during the postoperative period in the
recovery room or during the first 24 hours in the
surgical ward.
Discussion
Figure 3.—Kaplan-Meier curves represent the percentage of
remaining intubated patients after stopping the infusion of
remifentanil and propofol (P=0.66).
(0.189±0.049 vs. 0.199±0.055 μg x kg-1 x min,
P=0.55). No differences were found in the recovery time from anesthesia (Figure 3). Postoperative nociception was not significantly differ-
842
Our findings contribute significantly to understanding of the autonomic nervous system
modulation changes that occur during general
anesthesia for laparoscopic surgery. Although
it is commonly believed that laparoscopic surgery is a “minimally invasive” surgery, actually
it is “maximally perturbing” surgery because of
the changes in intra-abdominal pressure, intrathoracic pressure, diaphragm shift, right and
left atrial pressure, vascular resistances, cardiac
MINERVA ANESTESIOLOGICA
August 2015
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESSCOLOMBO
output and carbon dioxide overload. Moreover, it is well known that nociception induces
a sympathetic activation 2, 19 generating a stress
response of increased heart rate, blood pressure,
catecholamine and corticosteroid release, and it
has been supposed to influence postoperative
outcome.20, 21 Unfortunely, it is difficult to assess
the surgical stress during general anesthesia. Previously we found that changes in SPI correlate
with changes in ANS modulation during general anesthesia.22 Nowadays it was unknown if
titrating analgesia according to a low SPI might
reduce the intraoperative autonomic stress. This
study found that SPI-guided analgesia aimed at
maintain SPI below 50 leads to lower sympathetic activity than the standard clinical practice.
Some Authors found that SPI-guided analgesia,
with the index kept between 20 and 50 compared to standard clinical practice, resulted in
lower remifentanil and propofol consumption,
more stable hemodynamic, faster recovery but
with contradictory results on incidence of unwanted events.23, 24 This study did not find a
difference in remifentanil consumption, hemodynamic, or recovery time from anesthesia between groups. This might be due to the different
surgery type and patient population.
There has been uncertainty about neurally mediated hemodynamic distress elicited by pneumoperitoneum during laparoscopic surgery. An
anecdotal vagal hypertone or sympathetic hyperactivity caused by insufflation of pneumoperitoneum has been alternatively advocated as a cause
of hemodynamic changes.25-29 This study provides
additional evidence that pneumoperitoneum insufflation increases the sympathetic modulation
during a standard clinical practice anesthesia, as
stated by high LF/HF ratio due to high LF spectral component whereas the vagal component remained unchanged in the control group.
The relationship between SPI and LF/HF may
be explained by considering the physiologic bases
of SPI. It is calculated from the pulse plethysmographic amplitude that reflects pulsatile changes
of arteriolar blood volume into the tissue. This is
related to the arteriolar wall distensibility 30 and
influenced both by intravascular volume status
and by sympathetic-mediated vasoconstriction
of the arterioles. It is well demonstrated that
Vol. 81 - No. 8
different types of sympathetic stimuli (painful, orthostatic, lower body negative pressure)
increase the firing rates of sympathetic muscular fibers resulting in vessels constriction.2, 31-34
These changes ultimately affect the PPGA and
SPI. Furthermore, SPI is affected by several confounding factors (atropine administration, level
of sedation, spinal anesthesia, intravascular volume status and patient’s position),35-38 all influencing ANS activity. We believe that this point
is the core of the misunderstanding about pulse
plethysmographic-derived indexes as a measure
of nociception-antinociception balance: during general anesthesia, altered homeostasis and
autonomic response can be induced by several
factors, not just nociceptive. The results of this
study suggest that SPI reflects the sympathetic
mediated vasoconstriction. Other monitoring
systems based on HRV analysis have been proposed to measure adequacy of anti-nociception
under general anesthesia. The Autonomic Nociceptive Index (ANI) 39 is derived from the automatic analysis of the HRV using a wavelets
transformation algorithm in order to keep only
HF variations, and then the oscillations’ amplitude are normalized and displayed as absolute
number. ANI is based on the assumption that
nociception causes a reduction of vagal activity
that is reciprocal to increased sympathetic activity. This is not true in all cases. Vagal and sympathetic modulation can show quite different
oscillations in amplitude and are not always reciprocal to each other. For example, in this study
we found a stable HF component and a marked
different LF component after pneumoperitoneum insufflation. Thus, any heart rate variability
or pulse plethysmographic based index should
be considered for their physiological meanings:
they measure only the modulation of ANS on
cardiovascular system in response to a multitude
of afferences, not only nociception.
Changes in HRV during general anesthesia
likely result from the interaction of hypnosis,
surgical stimulation, antinociception, and direct
cardiovascular effects of drugs.19 In this study
we kept a stable level of hypnosis measured with
electroencephalographic state entropy in order
to rule out the effect of the depth of hypnosis on
SPI and ANS activity.
MINERVA ANESTESIOLOGICA
843
COLOMBO
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESS
Commonly, it is believed that high surgical
stress is equal to high blood pressure and high
heart rate, and vice-versa in the case of low surgical stress. This is true in the majority of cases,
but these clinical modifications are late signs of
unbalanced or overbalanced sympathetic outflow. Assessing the ANS modulation by means of
heart rate variability or pulse plethysmographic
analysis might give anesthesiologists more detailed information about the ANS activity under general anesthesia. If these indexes could be
complementary and could help to better titrate
the hypnotic and analgesics drugs during anesthesia should be an interesting subject for future research.
This study has few limitations. We studied
almost healthy subjects. Among these patients
we did not found clinically relevant advantages
over the standard clinical practice. It should be
further investigated if SPI-guided analgesia can
allow the anesthesiologists to achieve better control of intraoperative stress, more stable hemodynamic, and prevent perioperative morbidity
in moderate-to-high risk patients or in high risk
surgery. Although there is no evidence of the optimal SPI value to be maintained during general
anesthesia, this study chose values below 50 in
the SPI group. Unlike other studies,23, 24 this
study did not set a lower limit of SPI in the SPI
group. Although there was no difference between
groups in remifentanil consumption or CEremi at
the three intraoperative study points, mean SPI
was lower than 50 in both groups. At the time
of abdominal distension by pneumoperitoneum,
SPI increased significantly only in the control
group. We hypothesize that SPI-guided analgesia
may help the anesthesiologist to better titrate analgesia during major surgical stress and decrease
opioid infusion during less stressful phases.
Conclusions
In conclusion, this study found that pneumoperitoneum insufflation induces a sympathetic activation of ANS and SPI guided analgesia reduces the sympathetic response during
laparoscopic cholecystectomy. However, there
are no clinically relevant advantages over standard clinical practice in healthy subjects. Further-
844
more, these data suggest that SPI seems to reflect
sympathetic cardiovascular modulation under
general anesthesia.
Key messages
—— Pneumoperitoneum represents a
stressful stimulus inducing a marked sympathetic activation.
—— A target controlled opioid infusion
based on Surgical Pleth Index values below
50 reduces the sympathetic response during
laparoscopic cholecystectomy.
—— Surgical Pleth Index appears to reflect
changes in sympathetic modulation during
general anesthesia for laparoscopic surgery.
References
 1.Terkelsen AJ, Molgaard H, Hansen J, Andersen OK,
Jensen TS. Acute pain increases heart rate: differential
mechanisms during rest and mental stress. Auton Neurosci
2005;121:101-9.
  2. Burton AR, Birznieks I, Bolton PS, Henderson LA, Macefield VG. Effects of deep and superficial experimentally induced acute pain on muscle sympathetic nerve activity in
human subjects. J Physiol 2009;587:183-93.
  3. Buhre W, Rossaint R. Perioperative management and monitoring in anesthesia. Lancet 2003;362:1839-46.
  4. Bruhn J, Myles PS, Sneyd R, Struys MM. Depth of anesthesia monitoring: what’s available, what’s validated and
what’s next? Br J Anesth 2006;97:85-94.
 5. Heier T, Steen PA. Assessment of anesthesia depth. Acta
Anaesthesiol Scand 1996;40:1087-100.
  6. Rantanen M, Yli-Hankala A, van Gils M, Yppärilä-Wolters
H, Takala P, Huiku M et al. Novel multiparameter approach
for measurement of nociception at skin incision during
general anesthesia. Br J Anesth 2006;96:367-76.
  7. Ahonen J, Jokela R, Uutela K, Huiku M. Surgical stress
index reflects surgical stress in gynecological laparoscopic
day-case surgery. Br J Anesth 2007;98:456-61.
  8. Huiku M, Uutela K, van Gils M, Korhonen I, Kymäläinen
M, Meriläinen P et al. Assessment of surgical stress during
general anesthesia. Br J Anaesth 2007;98:447-55.
 9. Struys MM, Vanpeteghem C, Huiku M, Uutela K, Blyert NB, Mortier EP. Changes in a surgical stress index in
response to standardized pain stimuli during propofolremifentanil infusion. Br J Anaesth 2007;99:359-67.
10. Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R,
Pizzinelli P et al. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 1986;59:17893.
11. Montano N, Ruscone TG, Porta A, Lombardi F, Pagani M,
Malliani A. Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during
graded orthostatic tilt. Circulation 1994;90:1826-31.
12.Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the
European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation
1996;93:1043-65.
MINERVA ANESTESIOLOGICA
August 2015
SURGICAL PLETH INDEX HELPS TO PREVENT SYMPATHETIC STRESSCOLOMBO
13.Porta A, Baselli G, Caiani E, Malliani A, Lombardi F,
Cerutti S. Quantifying electrocardiogram RT-RR variability interactions. Med Biol Eng Comput 1998;36:27-34.
14. Akselrod S, Gordon D, Ubel FA, Shannon DC, Berger
AC, Cohen RJ. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular
control. Science 1981;213:220-2.
15. Hopf HB, Skyschally A, Heusch G, Peters J. Low-frequency spectral power of heart rate variability is not a specific
marker of cardiac sympathetic modulation. Anesthesiology
1995;82:609-19.
16. Goldsmith RL, Bigger JT, Steinman RC, Fleiss JL. Comparison of 24-hour parasympathetic activity in endurancetrained and untrained young men. J Am Coll Cardiol
1992;20:552-8.
17. Cook JR, Bigger JT, Kleiger RE, Fleiss JL, Steinman RC,
Rolnitzky LM. Effect of atenolol and diltiazem on heart
period variability in normal persons. J Am Coll Cardiol
1991;17:480-4.
18. Vettorello M, Colombo R, De Grandis CE, Costantini E,
Raimondi F. Effect of fentanyl on heart rate variability during spontaneous and paced breathing in healthy volunteers.
Acta Anaesthesiol Scand 2008;52:1064-70.
19. Jeanne M, Logier R, De Jonckheere J, Tavernier B. Heart
rate variability during total intravenous anesthesia: effects of
nociception and analgesia. Auton Neurosci 2009;147:91-6.
20. Parker SD, Breslow MJ, Frank SM, Rosenfeld BA, Norris EJ, Christopherson R et al. Catecholamine and cortisol responses to lower extremity revascularization: correlation with outcome variables. Perioperative Ischemia
Randomized Anesthesia Trial Study Group. Crit Care Med
1995;23:1954-61.
21. Monk TG, Ding Y, White PF. Total intravenous anesthesia:
effects of opioid versus hypnotic supplementation on autonomic responses and recovery. Anesth Analg 1992;75:798804.
22. Colombo R, Raimondi F, Corona A, Rivetti I, Pagani F,
Porta VD et al. Comparison of the Surgical Pleth Index
with autonomic nervous system modulation on cardiac
activity during general anesthesia: A randomised cross-over
study. Eur J Anesthesiol 2014;31:76-84.
23. Chen X, Thee C, Gruenewald M, Wnent J, Illies C, Hoecker J et al. Comparison of surgical stress index-guided analgesia with standard clinical practice during routine general
anesthesia: a pilot study. Anesthesiology 2010;112:117583.
24. Bergmann I, Gohner A, Crozier TA, et al. Surgical pleth
index-guided remifentanil administration reduces remifentanil and propofol consumption and shortens recovery
times in outpatient anesthesia. Br J Anesth 2013;110:6228.
25. Baltayian S. A brief review: anesthesia for robotic prostatectomy. J Robotic Surg 2008;2:59-66.
26.Gainsburg DM. Anesthetic concerns for robotic-assisted
laparoscopic radical prostatectomy. Minerva Anestesiol
2012;78:596-604.
27.O’Leary E, Hubbard K, Tormey W, Cunningham AJ.
Laparoscopic cholecystectomy: hemodynamic and neuroendocrine responses after pneumoperitoneum and changes in position. Br J Anesth 1996;76:640-4.
28. Di Iorio C, Cafiero T, Di Minno RM. The effects of pneumoperitoneum and head-up position on heart rate variability and QT interval dispersion during laparoscopic cholecystectomy. Minerva Anestesiol 2010;76:882-9.
29. Bickel A, Yahalom M, Roguin N, Frankel R, Breslava J, Ivry
S et al. Power spectral analysis of heart rate variability during positive pressure pneumoperitoneum: the significance
of increased cardiac sympathetic expression. Surg Endosc
2002;16:1341-4.
30. Burton AC. Relation of structure to function of the tissues
of the wall of blood vessels. Physiol Rev 1954;34:619-42.
31. Furlan R, Porta A, Costa F, Tank J, Baker L, Schiavi Ret
al. Oscillatory patterns in sympathetic neural discharge and
cardiovascular variables during orthostatic stimulus. Circulation 2000;101:886-92.
32. Cooke WH, Rickards CA, Ryan KL, Kuusela TA, Convertino VA. Muscle sympathetic nerve activity during intense
lower body negative pressure to presyncope in humans. The
J Physiol 2009;587:4987-99.
33. Fu Q, Iwase S, Niimi Y, Kamiya A, Kawanokuchi J, Cui J et
al. Effects of lower body positive pressure on muscle sympathetic nerve activity response to head-up tilt. Am J Physiol
Regul Integr Comp Physiol 2001;281:R778-85.
34. Joyner MJ, Shepherd JT, Seals DR. Sustained increases in
sympathetic outflow during prolonged lower body negative
pressure in humans. J Appl Physiol 1990;68:1004-9.
35. Bonhomme V, Uutela K, Hans G, Maquoi I, Born JD,
Brichant JF et al. Comparison of the surgical Pleth Index
with hemodynamic variables to assess nociception-antinociception balance during general anesthesia. Br J Anesth
2011;106:101-11.
36.Ilies C, Gruenewald M, Ludwigs J, Thee C, Höcker J,
Hanss R et al. Evaluation of the surgical stress index during
spinal and general anesthesia. Br J Anesth 2010;105:533-7.
37. Ilies C, Ludwigs J, Gruenewald M, Thee C, Hanf J, Hanss
R et al. The effect of posture and anesthetic technique on
the surgical pleth index. Anesthesia 2012;67:508-13.
38. Höcker J1, Broch O, Gräsner JT, Gruenewald M, Ilies C,
Steinfath M et al. Surgical stress index in response to pacemaker stimulation or atropine. Br J Anesth 2010;105:1504.
39. Logier R, Jeanne M, De Jonckheere J, Dassonneville A,
Delecroix M, Tavernier B. PhysioDoloris: a monitoring
device for analgesia/nociception balance evaluation using
heart rate variability analysis. Conf Proc IEEE Eng Med
Biol Soc 2010;2010:1194-7.
Acknowledgements.—We would like to thank Prof. A. Porta, Ph. D. (Dipartimento di Scienze Biomediche per la Salute, University of
Milan, Italy) for providing us with the software for HRV analysis.
Funding.—No financial support was needed for this study.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on October 22, 2013. - Accepted for publication on November 4, 2014. - Epub ahead of print on November 6, 2014.
Corresponding author: R. Colombo, Anesthesiology and Intensive Care Unit, Azienda Ospedaliera Luigi Sacco, Polo Universitario, Via G.
B. Grassi 74, 20157 Milan, Italy. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
845

O R I G I N A L A RT I C L E
Indirect videolaryngoscopy using Macintosh
blades in patients with non-anticipated
difficult airways results in significantly lower
forces exerted on teeth relative to classic direct
laryngoscopy: a randomized crossover trial
B. PIETERS 1, 2, R. MAASSEN 3, E. VAN EIG 4, B. MAATHUIS 4,
J. VAN DEN DOBBELSTEEN 4, A. VAN ZUNDERT 2, 5, 6
1Department of Anesthesiology, Intensive Care and Pain Therapy, Catharina Hospital Eindhoven, Eindhoven, The
Netherlands; 2Department of Anesthesiology, Maastricht University Hospital, Maastricht, The Netherlands; 3Department
of Anesthesiology, Laurentius Hospital Roermond, Roermond, The Netherlands; 4Delft University of Technology,
Delft, The Netherlands; 5Department of Anesthesiology, Faculty of Medicine, University Ghent Hospital, Ghent,
Belgium; 6Department of Anesthesiology and Perioperative Medicine, University of Queensland; 7School of Medicine,
Royal Brisbane and Women’s Hospital, Herston Brisbane, Australia
ABSTRACT
Background. Videolaryngoscopy has proven advantageous over direct laryngoscopy for a variety of outcome variables, most importantly, making laryngoscopy more successful. We tested whether three videolaryngoscopes (VLS),
McGrath® series 5 (Aircraft Medical Ltd, Edinburgh, UK), C-MAC® (Karl Storz, Tuttlingen, Germany) and GlideScope® Cobalt (Verathon Medical, Bothell, WA, USA) exert reduced forces on maxillary incisors and lower teeth,
and compared them with a classic Macintosh MAC 3 laryngoscope blade during laryngoscopy.
Methods. In this randomized crossover trial, we included 141 patients (ASA I-III) with non-anticipated difficult airways. They were randomly allocated to undergo direct laryngoscopy and videolaryngoscopy performed with one of
three VLS. Primary outcome was the magnitude of forces applied to the maxillary incisors during laryngoscopy. Secondary outcomes were the frequency with which forces were applied, and the magnitude of forces applied to the lower teeth.
Results. Forces applied to the maxillary incisors during direct classic laryngoscopy were on average higher than
forces applied during videolaryngoscopy. Among the VLS the average force applied was significantly lower for the
C-MAC® as compared to the McGrath® and the GlideScope® VLS. The frequency with which a force was applied
to the maxillary incisors was significantly lower for the C-MAC®, compared to the other VLS and classic Macintosh
laryngoscope. The number of cases in which force was applied to the lower teeth was smallest for the McGrath VLS.
Conclusion. Forces exerted on maxillary incisors are lower using video-assisted Macintosh blade laryngoscopy compared to classic direct laryngoscopy. The number and magnitude of forces applied to maxillary incisors also differ
substantially between different VLS. (Minerva Anestesiol 2015;81:846-54)
Key words: Laryngoscopes - Pressure - Dentition.
D
uring endotracheal intubation the anesthesiologist has to avoid using the maxillary incisors as a fulcrum to lever the soft tissues
Comment in p. 825.
846
upwards. This may otherwise result in dental
trauma. It is obvious that contact with teeth and
– even worse – the incidence of accidental dental
trauma, are directly related to the difficulty of
the intubation.1
MINERVA ANESTESIOLOGICA
August 2015
FORCES ON TEETH DURING VIDEOLARYNGOSCOPYPIETERS
Videolaryngoscopy has proven advantageous
over direct laryngoscopy using a classic Macintosh blade, for improved viewing of the glottis,
with subsequent more successful intubations,
and a shorter effective airway time both in patients with non-anticipated difficult and anticipated difficult airways.2, 3 Previously, it has been
demonstrated that the forces applied to the patient’s maxillary incisors are reduced when using
a VLS, compared to a classic Macintosh laryngoscope. However, in these studies only one type
of VLS (V-MAC®, Karl Storz, Tuttlingen, Germany) or Airtraq® (Prodol Meditec, Guecho,
Spain)) was used.4, 5 In the other, different types
of VLS than the ones used in this study were investigated, and forces on lower teeth were not
investigated.6
We tested whether three different brands of
VLS, i.e. McGrath® series 5 (Aircraft Medical
Ltd, Edinburgh, UK), C-MAC® (Karl Storz,
Tuttlingen, Germany) and GlideScope® Cobalt
(Verathon Medical, Bothell, WA, USA) apply
similar reduced forces on both maxillary incisors
and lower teeth, and compared them with a classic Macintosh MAC 3 laryngoscope.
The primary outcome was the magnitude of
forces applied to the maxillary incisors.
Secondary outcomes were the frequency with
which forces were applied and the magnitude of
the forces applied to the lower teeth.
Materials and methods
After obtaining Institutional Review Board
approval (Catharina Hospital Eindhoven, The
Netherlands (Chairman: dr. R. Grouls), registration M12-1217), CCMO registration
(NL39915.06012) and registration at Clinical
Trials (NCT01599312), written informed patient consent was obtained. In total, 150 consecutive patients (ASA I-III) with non-anticipated
difficult airways, undergoing a variety of surgical
interventions, for which endotracheal intubation
was indicated, were enrolled in this randomized
crossover trial between May and September
2012 at the Catharina Hospital Eindhoven. The
patients were randomly allocated, using computer-generated tables, to receive direct laryngoscopy (classic Macintosh laryngoscope [MAC
Vol. 81 - No. 8
3 blade]) and one of three VLS: McGrath®, CMAC® or GlideScope®. The C-MAC® was used
with indirect (monitor) viewing throughout the
experiment. Exclusion criteria were no informed
patient consent, patients younger than 18 years,
patients requiring other than size 3 blade Macintosh laryngoscope, patients with preoperative
predictors of a difficult airway (Mallampati score
IV, restricted neck movement, thyromental distance <65 mm, interincisor/interdental distance
<35 mm), patients with ASA class ≥IV, emergency surgery, surgery of head and/of throat, locoregional anaesthesia, patients fasted <6 hours,
bad dentition, dental crowns and/or fixed partial
denture. Patients were enrolled during the preanesthetic visit of the patient (performed by anesthesiologists not involved in this study). During this visit, patient characteristics mentioned
above (with respect to exclusion criteria) were
recorded.
When the patients arrived in the OR, they
were connected to standard monitoring devices. Patients were placed in the supine position with the patient’s head resting on a pillow
8 cm in height. After three minutes of oxygen
administration, IV induction of anesthesia was
performed with 1 μg/kg fentanyl, 3 mg/kg propofol, and rocuronium 0.7 mg/kg, followed by
manual ventilation via a facemask using sevoflurane in oxygen. The laryngoscope was inserted at
least 90 seconds after completion of induction,
and only when the anesthesiologist considered
the level of anesthesia adequate for intubation
and the end-tidal sevoflurane concentration was
≥1.5%. Muscle paralysis was measured using the
train-of-four-ratio.
The patients had the classic Macintosh laryngoscope placed in their mouth. The anesthesiologist subsequently determined the best possible
view of the glottic opening and an endotracheal
tube (7.5 mm endotracheal tube for men, 7.0
mm for women) was brought into position in
front of the glottic opening. After the Macintosh
laryngoscope was removed, one of three VLS
was placed in their mouth and the procedure
was repeated. Actual intubation was only performed with the VLS. Three anesthesiologists,
familiar with the tested VLS (minimum 100
uses on actual patients) and classical intubation,
MINERVA ANESTESIOLOGICA
847
PIETERS
FORCES ON TEETH DURING VIDEOLARYNGOSCOPY
participated in the study. They were specifically instructed in avoiding any contact with the
teeth. We chose to instruct anesthesiologists in
this way because we could not blind them for the
sensors placed on the laryngoscope blade. The
participating anesthesiologists did know the purpose of the study and therefore may have taken
extra care in avoiding dental trauma.
Effective airway time was measured as the
time between picking up the endotracheal tube
and placing it in front of the glottic opening (in
case of the classic Macintosh laryngoscope) or
insertion between the vocal cords (in case of the
VLS). Each approach towards the glottic opening was counted as an intubation attempt. After
two unsuccessful attempts a stylet was inserted
into the endotracheal tube to facilitate intubation.
Patient characteristics (i.e. weight, age, gender, height, BMI), specific metrics of intubation
difficulty (i.e. Mallampati grade, thyromental
distance, neck movement, mouth opening, dentition, Cormack-Lehane grade, number of intubation attempts), and any complication were
recorded
The method used to measure the forces applied to the teeth (dependent parameter) for the
duration of insertion of each laryngoscope was
the same as used by Lee et al., using Flexiforce®
sensors (A201-25, Tekscan, MA, USA). These
were fixed to the blade of the laryngoscope
at the possible area of contact with the teeth.
Three sensors were mounted on top of the blade
(the site of the blade directed at the maxillary
incisors during intubation) along its length and
three at the bottom of the blade (the site of
the blade directed towards the tongue during
intubation) (Figure 1).4, 6 This configuration
completely covers the possible contact points
between the blade and the teeth. The sensors
are rated to 110 N. Calibration was performed
by applying a known mass (from 1 to 12 kg, in
steps of 1 kg) using a flat-headed screwdriver (as
geometrical approximation of the contact with
the teeth) to the sensors mounted upon the
blade. The sensors were invariant to the contact
point of the applied load. Each of the sensors
was individually calibrated in situ. The sensors
were cleaned and sterilized between each use by
848
Figure 1.—Flexiforce® sensors (A201-25, Tekscan, MA, USA)
fixed to the blade of the laryngoscope.
submersion of the blade in chlorhexidine 0.5%
in alcohol 70%. The sensors were assumed to
work reliably, as the forces were an order of
magnitude lower than required to permanently
damage the sensor. This assumption was tested
by calibration tests at the end of the experiments, whereby none of the sensors indicated
degradation.
Data acquisition was achieved with a National Instruments® (DAQ6009 (National Instruments Corporation, TX, USA) card at 500
Hz, using Labview® 7.0 (National Instruments
Corporation, TX, USA) on a laptop computer
(Hewlett Packard Company, CA). Peak forces
were subsequently noted for each of the patients.
Because drift and noise in the sensors could result in very low but incorrectly detected contact
forces between the teeth and the blade we used
a threshold of 2N to determine whether a force
was applied to the teeth or not. All registrations
of forces applied to the maxillary incisors or lower teeth were registered on the laptop computer
in such a way that these were completely blinded
for the anesthesiologist who intubated the patient’s trachea. The results of these measurements
were only available to an engineer (JVD) of the
Technical University Delft, The Netherlands.
Prior to the experiment we performed a power
analysis to estimate the required sample size of
patients required for finding differences between
the magnitudes of used forces. We were interested in finding a difference between the laryngoscopes of at least 20N as, based on earlier results
by Lee et al., we assumed that such a difference
would have a significant clinical value.4, 6 Fur-
MINERVA ANESTESIOLOGICA
August 2015
FORCES ON TEETH DURING VIDEOLARYNGOSCOPYPIETERS
thermore, we used a standard deviation of 20N
as an estimate of the population variability, the
latter being based on the variability found in our
previous study.4 Using these assumptions, the
power analysis showed that an appropriate sample size to achieve adequate power was at least
22 measurements per laryngoscope. The actual
sample size used in the experiment was twice
as large for each of the VLS to overcome possible exclusions for technical failure, while each
insertion with a VLS was also paralleled with a
measurement with the classic laryngoscope. For
the cases in which forces (>2N) were applied, we
determined whether the magnitude of force differed among the laryngoscopes.
Non parametric testing was used to compared
medians across groups using the Median Test
for k samples (IBM SPSS 19 [IBM Corporation, NY 10589, USA]) We made six comparisons between groups for each depended variable
(force, lower teeth, maxillary incisors and time).
Corrections were made for multiple hypothesis
tested using Bonferroni (i.e.=alpha/k=0.05/6).
Results
All 150 patients were successfully intubated in
the study. However, in nine patients in the VLS
group and in one patient in the direct laryngoscopy group, registration of forces was not realized due to technical problems.
The actual sample size was larger than initially calculated by the power analysis, because
during the study it was not directly clear how
many measurements were lost due to the technical problems. Therefore the study included
the results of 141 patients (63 males and 78 females), all intubated with one of the three VLS.
Patient characteristics are given in Table I, results
obtained during laryngoscopy in Table II. No
differences were seen between the groups regarding patient characteristics, the objective metrics
Table I.—Patient characteristics.
Classic
N.=141
Male: female ratio, N.
Age; years
Height; cm
Weight; kg
BMI; kg.m-2
ASA class I:II:III, N.
Mallampati score I: II: III: IV, N.
Thyromental distance; mm
Interincisor/interdental distance; mm
Denture: double: single: none, N.
Adequate neck movement; yes/no, N.
63:78
53.8±15.5
171.5±8.9
81.6±19.3
27.7 ±6.3
45:87:9
70:59:12:0
83.4±10.5
43.4±6.6
104:9:28
141:0
McGrath®
N.=48
22:26
55.5±15.0
172.4 ±9.4
80.1±15.6
27.0±6.2
16:31:1
25:20:3:0
84.3±9.9
44.0±6.4
34:1:13
48:0
C-MAC®
N.=47
GlideScope®
N.=46
21:26
51.9±15.0
171.3±9.0
80.4±20.1
27.4±7.1
15:28:4
21:20:6:0
84.1±8.4
43.5±6.1
36:4:7
47:0
20:26
54.2±16.6
171.0±8.5
84.5±21.9
28.9±7.4
14:28:4
24:19:3:0
82.4±3.5
43.2±7.2
34:4:8
46:0
Values are numbers, or mean±SD.
Table II.—Results obtained during laryngoscopy using three videolaryngoscopes (McGrath®,C-MAC®, GlideScope®) and the
classic Macintosh laryngoscope.
Cormack Lehane grade
I:II:III:IV; N.
Effective airway time, s
Intubation attempts;
1:2:3:4, N.
Intubation failures
Use of sylet; yes:no; N.
Classic
N.=141
McGrath®
N.=48
C-MAC®
N.=47
GlideScope®
N.=46
77:52:10:2*
(55:37:7:1)
22±13
N/A
47:1:0:0
(98:2:0:0)
28±12
7:11:18:12
(15:23:37:25)
0
30:18
(63:37)
43:4:0:0
(91:9:0:0)
11±6*
36:10:1:0*
(77:21:2:0)
0
1:46 *
(2:98)
41:5:0:0
(89:11:0:0)
27±12
6:12:19:9
(13:26:41:20)
0
28:18
(61:39)
N/A
N/A
P value
<0.05
<0.001
<0.001
NS
<0.001
Values are numbers (%), or mean±SD. NS: not significant; N/A: not applicable.
*Significant difference between * and the other groups.
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
849
PIETERS
FORCES ON TEETH DURING VIDEOLARYNGOSCOPY
Figure 2.—Peak force for the maxillary incisors per laryngoscope, averaged across all measurements.
Figure 3.—Peak force for the lower teeth per laryngoscope, averaged across all measurements.
of intubation difficulty, and Cormack-Lehane
grade, although the latter differed significantly
between direct and indirect laryngoscopy.
Figures 2, 3 display the peak force per laryngoscope for the maxillary incisors and lower
teeth, respectively, averaged across all measurements. Figure 2 shows that the force applied
850
with the classic laryngoscope (30.2±33.9 N) was
on average higher than the force applied with the
VLS (16.2±17.4 N, McGrath® (NS); 9.31±11.3
N, GlideScope® (P<0.004)) and that the average force applied was lowest for the C-MAC®
VLS (1.18±4.73 N (P<0.001)). Post Hoc testing
revealed that the average force applied for the C-
MINERVA ANESTESIOLOGICA
August 2015
FORCES ON TEETH DURING VIDEOLARYNGOSCOPYPIETERS
Table III.—Forces (Newton) applied per laryngoscope for the maxillary incisors (across all measurements).
Median
IQR
Classic
N.=141
McGrath
N.=48
C-MAC
N.=47
GlideScope
N.=46
P-Value
17.1
54.5*
16.5
24.6
-0.4
2.2
5.0
16.0
<0.001
*Significant difference between * and the other groups.
MAC® VLS is significantly lower than the average forces found for the McGrath® (P<0.001)
and the GlideScope® VLS (P=0.001). Table III
displays the median and IQR for the forces applied to the maxillary incisors, across all measurements. Figure 3 proves that hardly any force
was exerted on the lower teeth with any of the
laryngoscopes. The differences between median
forces for the classic laryngoscope (2.5N [IQR
7.5N]) and the C-MAC® VLS (2.6N [IQR
8.2N]) were very small. For the GlideScope®
(<0N [IQR 4.3N]) and McGrath® VLS (<0N
[IQR 1.1N]), median forces were negligible.
The results displayed in Figures 2, 3 also incorporate the cases in which no contact was made
between the teeth and the blade, effectively lowering the average peak force. To determine whether
the applied force differs among the various laryngoscopes we selected the cases in which contact
was made between the teeth and the blade, and
in which force (>2N) was applied to the maxillary incisors, before computing the averages. The
results are displayed in Figure 4. Figure 4 shows
that the peak force applied to the maxillary incisors with the classic laryngoscope (44.6±30.8
N) was on average higher than the force applied
with the VLS and that the average force was lowest for the C-MAC® VLS (11.5±8.1 N).
Post hoc testing revealed that there was a
significant difference between the average peak
force found for the classic Macintosh laryngoscope and the average values found for the VLS.
The difference between the C-MAC® and the
other VLS did, however, not reach significance.
For the lower teeth (Figure 5), the peak force
found for the different laryngoscopes was low,
the median being 1.5 N. The mean forces applied to the lower teeth were significantly lower
for the McGrath® compared to the classic Macintosh laryngoscope (P<0.001). Of the other
Figure 4.—Average peak force for the maxillary incisors per laryngoscope. Averages are based on the selection of cases in which
contact was made between teeth and blade.
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
851
PIETERS
FORCES ON TEETH DURING VIDEOLARYNGOSCOPY
Figure 5.—Average peak force for the lower teeth per laryngoscope. Averages are based on the selection of cases in which contact
was made between teeth and blade (>2N).
averages, none were significantly different from
each other.
Tables IV, V summarize the proportion of cases where contact was made with the upper and
lower teeth, respectively.
The mean effective airway time was 8.6±5.6 s
for the patients in the direct laryngoscopy group,
compared to 18±11 s in the McGrath® group
(P<0.001); 23±16 s in the GlideScope® group
(P<0.001) and 11±10 s in the C-MAC® group
(NS). In the latter group significantly more first
attempt successful intubations were realized compared to the other two VLS groups. In only 2/47
(4%) of the patients in the C-MAC group, a stylet
was needed to intubate the trachea, which was significantly lower than in the McGrath (29/48, i.e.
60%) and GlideScope® (28/46, i.e. 61%) groups.
Discussion
This study confirms that when using videolaryngoscopy, the forces applied to the patient’s
maxillary incisors are reduced when compared
with the classic Macintosh laryngoscope, while
there are no differences in the forces exerted on
the lower teeth.
The number of contacts made with the maxillary insicors was lowest with the C-MAC® VLS,
whereas there were no differences in number of
contacts between the other VLS studied and the
classic laryngoscope. Average peak forces were
lowest when using the C-MAC®. The difference
in average peak force between the C-MAC®
and the other VLS was not statistically significant. However, when unexpectedly, intubation
Table IV.—Proportion of cases in which contact was made between the maxillary incisors and the blade and a force was applied
(>2N).
Macintosh laryngoscope
McGrath® videolaryngoscope
C-MAC® videolaryngoscope
GlideScope® videolaryngoscope
Total
N
Contact
Percentage
P-value
141
48
47
46
282
93
34
9
28
164
66
71
19*
61
58
P<0.001
force was applied (>2N).
*Significant difference between * and the other groups.
852
MINERVA ANESTESIOLOGICA
August 2015
FORCES ON TEETH DURING VIDEOLARYNGOSCOPYPIETERS
Table V.—Proportion of cases in which contact was made between the lower teeth and the blade and a force was applied (>2N).
Macintosh laryngoscope
McGrath® videolaryngoscope
C-MAC® videolaryngoscope
GlideScope® videolaryngoscope
Total
N.
Contact
141
48
47
46
282
73
4
24
14
115
Percentage
52
8*
51
30
41
P value
<0.001
*Significant difference between * and the other groups.
was more difficult than anticipated (Cormack
Lehane grade III/IV), and for example the
BURP maneuver had to be used or the help of
an assistant was required, this did not result in
differences between VLS concerning forces applied to the maxillary incisors.
When intubation could only be realized after
multiple attempts, differences between numbers
of contacts at each intubation attempt did exist. The trend seemed to be that providers made
more contact at subsequent attempts. However,
there was no clear increase in magnitude of forces applied to teeth.
The design of the C-MAC® laryngoscope blade
is similar to a classic Macintosh blade, but differs
from the blades of the McGrath® and GlideScope®
VLS. The fact that a VLS with a Macintosh blade
creates more room for intubation compared to
other VLS, makes intubation not only easier
(more room to direct/manipulate the tube and
less use of stylets), but also faster (shorter effective
airway time because the laryngeal and pharyngeal
axes are more in one line compared to a VLS with
a curved blade, making the route, which the tube
has to follow, more direct) and with less frequent
contacts with maxillary incisors.
When using a classic direct laryngoscope with
a Macintosh blade, all three axes (laryngeal,
pharyngeal and especially oral) need to be positioned in one line. Often this requires the anesthesiologist to use more force and so frequently
more force is applied to the maxillary incisors.
The results of this study show a large difference between forces applied during direct and
indirect laryngoscopy, with the mean force applied to the maxillary incisors being lowest for
all measurements in the C-MAC® group (i.e.
1.18±4.73 N), which is comparable to the results of Lee et al.4, 6
Patients were not intubated using the classic
Vol. 81 - No. 8
laryngoscope, and this could have affected the
effective airway time realized by direct laryngoscopy. The results on the forces obtained with the
classic Macintosh laryngoscope may have been
greater than the ones recorded. Actually passing the tube through the vocal cords could have
turned out to be more difficult than just positioning it in front of the vocal cords, requiring
more force to be used.
Limitations of the study include: 1) the restricted number of different brands of VLS in
this study, since the amount of available brands
of VLS seems to increase daily; 2) this study only
included patients with non-anticipated difficult
airways, and hence nothing can be deduced for
patients with difficult airways. One could argue
if patients with a Mallampati Score III can be
considered as having a non-anticipated difficult
airway. Indeed, this is often associated with a difficult airway. However, sometimes even a patient
with a Mallampati I turns out to be difficult to
intubate because of a high-anterior postioned
glottis. The combination of exclusion criteria
mentioned earlier and a cut-off value of Mallampati III is valid to exclude the majority of patients
with a difficult airway. This was confirmed by the
fact that all patients were successfully intubated;
3) we considered 2N as the lower range for determination of any force, which means that the
actual number of contacts might be higher; and
4) actual intubation was only performed with the
VLS, and hence intubation time could have been
influenced for direct laryngoscopy. Also, additional pressure may be applied and a higher number of contacts with the teeth may be experienced
during actual passing of the ETT between the vocal cords. This could have resulted in more contacts with teeth and greater force being applied to
the teeth for the classic Macintosh laryngoscope;
5) a stylet was inserted into the endotracheal
MINERVA ANESTESIOLOGICA
853
PIETERS
FORCES ON TEETH DURING VIDEOLARYNGOSCOPY
tube to facilitate intubation only after two unsuccessful attempts. This may in part explain why
a large proportion of the intubations required
3-4 attempts in the McGrath® group (28/46 or
61% from Table II) and the GlideScope® group
(30/48 or 63% from Table II) even though the
glottic views were good with both the McGrath®
and GlideScope® (mostly with Cormack Lehane
grade I). In addition, this may also explain why
the number of contacts made with the maxillary
incisors was lowest with the C-MAC® compared
with the GlideScope® and McGrath® because of
few attempts for tracheal intubation.
We deliberately chose to only insert a stylet after two unsuccessful attempts, although
the manufacturers of both GlideScope® and
McGrath® recommend to always use a stylet. In
our opinion, a stylet poses an extra risk of (oral)
trauma to the patient.7
We proved that even without a stylet intubation with GlideScope® (39%) and McGrath®
(40%) is possible. Also, when we would have chosen to always use a stylet with the GlideScope®
and McGrath®, we also should have always used
one with the C-MAC®, since the tube is easier
to direct, making intubation more easily. Using
a stylet with all VLS, as mentioned above, poses
an extra risk to the patient. Also, this would have
made intubating using the C-MAC® far more
easy, possibly influencing outcome.
Conclusions
Our study demonstrates that indirect videoassisted laryngoscopy results in significantly lower forces applied to maxillary incisors relative to
classical direct laryngoscopy in patients with no
expected difficulty for intubation, and that there
are differences among the VLS used. The number of contacts and forces applied to maxillary
incisors significantly differs between the tested
VLS. Forces applied to lower teeth seem clinically insignificant.
We strongly recommend considering using
VLS in patients with poor dentition, dental
crowns and/or fixed partial denture, needing intubation. When choosing a videolaryngoscope
for this category of patients, the anesthesiologist
has to be aware of differences between VLS concerning risk of dental trauma.
Key messages
—— Compared to classic direct intubation,
intubating with VLS results in reduced forces
exerted on the maxillary incisors.
—— VLS differ in the magnitude of forces
applied to maxillary incisors during intubation. Forces are lowest when VLS with
Macintosh shaped blades are used.
—— Forces applied to lower teeth seem
clinically insignificant.
References
  1. Givol N, Gershtansky Y, Halamish-Shani T, Taicher S, Perel
A, Segal E. Perianesthetic dental injuries: analysis of incident reports. J Clin Anesth 2004;16:173-6.
  2. van Zundert A, Maassen R, Lee R, Willems R, Timmerman
M, Siemonsma M et al. A Macintosh laryngoscope blade
for videolaryngoscopy reduces stylet use in patients with
normal airways. Anesth Analg 2009;109:825-31.
  3. Maassen R, Lee R, Hermans B, Marcus M, van Zundert A.
A comparison of three videolaryngoscopes: the Macintosh
laryngoscope blade reduces, but does not replace, routine
stylet use for intubation in morbidly obese patients. Anesth
Analg 2009;109:1560-5.
 4. Lee RA, van Zundert AAJ, Maassen RLJG, Willems RJ,
Beeke LP, Schaaper JN et al. Forces applied to the maxillary incisors during videoassisted intubation. Anesth Analg
2009;109:187-91.
  5. Hindman BJ, Santoni BG, Puttlitz CM, From RP, Todd
MM. Laryngoscope Force and Cervical Spine Motion during Intubation with Macintosh and Airtraq Laryngoscopes.
Anesthesiology 2014;121:260-71
  6. Lee RA, van Zundert AAJ, Maassen RLJG, Wieringa PA.
Forces applied to the maxillary incisors by video laryngoscopes and the Macintosh laryngoscope. Acta Anaesthesiol
Scand 2012;56:224-9.
  7. Van Zundert A, Pieters B, van Zundert T, Gatt S. Avoiding palatopharyngeal trauma during videolaryngoscopy:
do not forget the ‘blind spots’. Acta Anaesthesiol Scand.
2012;56:532-4.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Funding.—None.
Received on April 2, 2014. - Accepted for publication on October 8, 2014. - Epub ahead of print on October 14, 2014.
Corresponding author: A. van Zundert, Department of Anesthesiology and Perioperative Medicine, University of Queensland, School of
Medicine, Royal Brisbane and Women’s Hospital, Herston Brisbane Qld 4029 Australia. E-mail: [email protected]
854
MINERVA ANESTESIOLOGICA
August 2015

O R I G I N A L A RT I C L E
Esophageal pressure measurements under
different conditions of intrathoracic pressure.
An in vitro study of second generation balloon catheters
F. MOJOLI 1, D. CHIUMELLO 2, M. POZZI 1, I. ALGIERI 3, S. BIANZINA
S. LUONI 3, C. A. VOLTA 4, A. BRASCHI 1, L. BROCHARD 5
1
1Sezione di Anestesia Rianimazione e Terapia Antalgica, Dipartimento di Scienze Clinico-chirurgiche, Diagnostiche e
Pediatriche, SC Anestesia e Rianimazione 1, Università degli Studi di Pavia, Fondazione IRCCS Policlinico S. Matteo,
Pavia, Italy; 2Dipartimento di Anestesia, Rianimazione (Intensiva e Subintensiva) e Terapia del Dolore, Fondazione
IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milano, Italy; 3Dipartimento di Fisiopatologia Medico-Chirurgica
e dei Trapianti, Università degli Studi di Milano, Milano, Italy; 4Dipartimento di Scienze Morfologiche, Medicina
Sperimentale e Chirurgia, Sezione di Anestesia e Terapia Intensiva, Università di Ferrara, Arcispedale Sant’ Anna,
Ferrara, Italy; 5Keenan Research Centre and Interdepartmental Division of Critical Care, Critical Care Department,
University of Toronto, St. Michael’s Hospital, Toronto, Ontario, Canada
ABSTRACT
Background. The aim of this study was to evaluate in vitro the accuracy of second generation esophageal catheters
at different surrounding pressures and filling volumes and to suggest appropriate catheter management in clinical
practice.
Methods. Six different esophageal catheters were placed in an experimental chamber at four chamber pressures (0,
10, 20 and 30 cmH2O) and at filling volumes ranging from 0 to 10 mL. The working volume was defined as the
volume range between the maximum (Vmax) and minimum (Vmin) volumes achieving acceptable accuracy (defined by a balloon transmural pressure ±1 cmH2O). Accuracy was evaluated for a standard volume of 0.5 mL and for
volumes recommended by manufacturers. Data are shown as median and interquartile range.
Results. In the four conditions of chamber pressure Vmin, Vmax and working volume were 1.0 (0.5, 1.5), 5.3 (3.8,
7.1), and 3.5 (2.9, 6.1) mL. Increasing chamber pressure increased Vmin (rho=0.9; P<0.0001), that reached 2.0 mL
(1.6-2.0) at 30 cmH2O. Vmax and working volumes differed among catheters, whereas Vmin did not. By injecting
0.5 mL and the minimum recommended volume by manufacturer, balloon transmural pressure was <-1 cmH2O in
71% and 53% of cases, it was negatively related to chamber pressure (rho=-0.97 and -0.71; P<0.0001) and reached
values of -10.4 (-12.4, -9.7) and -9.8 (-10.6, -3.4) at 30 cmH2O.
Conclusion. Measuring positive esophageal pressures needs higher injected volumes than usually recommended.
The range of appropriate filling volumes is catheter-specific. Both absolute values and respiratory changes of esophageal pressure can be underestimated by an underfilled balloon. (Minerva Anestesiol 2015;81:855-64)
Key words: Esophagus - Catheters - Lung - Pressure - Pleura.
I
n 1878 Luciani 1 was the first to publish esophageal pressure measurements and in 1949
Buytendijk 2 proposed an esophageal pressure
measurement technique with a balloon in the
Comment in p. 827 and p. 830.
Vol. 81 - No. 8
esophagus to estimate pleural pressure. Subsequently, several studies demonstrated the possibility of applying esophageal pressure as an
adequate surrogate of pleural pressure in the
clinical practice.3 The most widely applied technique consists of an esophageal balloon filled
MINERVA ANESTESIOLOGICA
855
MOJOLI
ESOPHAGEAL PRESSURE MEASUREMENTS
with air, connected to a long thin catheter for
pressure transmission.3, 4 The length, diameter
and filling volume of these esophageal balloon
catheters significantly affected the accuracy of
esophageal pressure measurements.5-7 In particular, based on the experience on spontaneous
breathing patients, a very small filling volume
of about 0.5 mL was considered mandatory to
avoid both balloon overdistention and esophageal artifacts.5, 8 This first generation of catheters
was mainly applied in research, but scarcely in
the clinical practice.
In the last years clinicians have reassessed this
technique, mainly in the light of several studies showing the utility of esophageal pressure
monitoring for correct management of patients
with acute respiratory distress syndrome,9-12 and
due to the availability of a second generation
of esophageal balloon catheters.3, 13 The technique is now considered to estimate the role of
the chest wall in airway pressures or to adjust
ventilator pressures based on the absolute values
measured in the thorax with the balloon. Technical information concerning these new catheters
is still very limited.3 A recent study found that
pressure recorded may be influenced by artifacts
caused by material adhesion and may depend on
the balloon design and filling volumes, underlining the risk of balloon overfilling, which might
lead to overestimation of esophageal pressure.14
We also reported that, under positive pressure
conditions, esophageal pressure can be under-
Figure 1.—Esophageal balloon catheters.
Esophageal balloon’s material is polyethylene for Cooper, NutriVent, SmartCath and SmartCathG and latex for Marquat and
Microtek catheters. Balloon’s length is 100 mm for Microtek, NutriVent, SmartCath and SmartCathG, 95 mm for Cooper and 80
mm for Marquat. Catheter’s diameter (Fr) and length (cm) are 5/85 (Cooper), 6/78 (Marquat), 8/100 (Microtek), 14/91 (NutriVent), 8/101 (SmartCath) and 16/114 (SmartCathG).
856
MINERVA ANESTESIOLOGICA
August 2015
ESOPHAGEAL PRESSURE MEASUREMENTSMOJOLI
transmitted when too low inflation volumes are
used.15
The aim of this in vitro study was thus to
evaluate the accuracy of common commercial
esophageal balloon catheters at different filling
volumes and surrounding pressures, in order to
suggest clinicians how to properly manage this
second generation catheters under different intrathoracic conditions.
Materials and methods
Measurements
Esophageal catheters
Six different commercially available esophageal balloon catheters were evaluated (Figure 1):
SmartCath (Avea SmartCath Carefusion, San
Diego, CA, USA), SmartCathG (Avea SmartCath nasogastric pressure, Carefusion, San Diego, CA, USA), Nutrivent (NutriVent, Sidam,
San Giacomo Roncole, Mirandola, Modena,
Italy), Marquat (Marquat Genie Biomedical,
Boissy-Saint-Léger Cedex, France), Cooper
(Cooper Surgical, Trumbull, CT, USA) and Microtek (Microtek, Zutphen, The Netherlands).
SmartCath, SmartCathG, Marquat, Cooper
and Microtek are equipped with an esophageal
balloon. Nutrivent is equipped with both an esophageal and a gastric balloon. SmartCathG and
Nutrivent have also an internal line for feeding
and stomach aspiration. Inflating air volumes
recommended by the manufacturer are 1-2 mL
(Cooper), 0.5-3 mL (Marquat), 4 mL (NutriVent) and 0.5-2.5 mL (SmartCath and SmartCathG). To our knowledge, no manufacturer’s
indications are available for Microtek, therefore
we considered 0.5-1.0 mL as previously reported
for this catheter.16
For each catheter type, five samples were
visually inspected and fully inflated at ambient
pressure, in order to exclude different behavior
among samples due to manufacturing defects.
Experimental setup
The experimental setup (Figure 2) consisted
of an experimental chamber and a data acquisition system (OptiVent, Sidam, Italy). The experimental chamber consisted of a rigid plastic
Vol. 81 - No. 8
cylinder (inner volume 2 liters) connected to a
100 mL syringe and to the data acquisition system by a 3-way stopcock and a 120 cm tube
line. The esophageal balloon catheters were
introduced into the experimental chamber
through a small opening, that was subsequently
sealed. Esophageal catheters were connected to
a 10 mL syringe for balloon inflation and to the
data acquisition system by a 3-way stopcock and
a 80 cm tube line.
Simultaneous measurement of chamber and
balloon pressure were provided by pressure sensors (26PCAFA6G, Honeywell Micro Switch
Sensing and Control, USA) located in the OptiVent system; data were sampled at 100Hz and
processed on this dedicated data acquisition system.
Figure 2.—Experimental setup.
The experimental chamber consisted in a rigid plastic cylinder
(C) in which four levels of inner pressure were obtained by the
injection and pressurization of different volumes of air by a 100
mL syringe (S1). Inflation of the balloon (B) was obtained by
injection of air in the esophageal catheter by a 10 mL syringe
(S2). The experimental chamber and the esophageal catheter
were connected to the syringes and to a pressure monitor (M)
by 3-way stopcocks (T) and tube lines (L).
MINERVA ANESTESIOLOGICA
857
MOJOLI
ESOPHAGEAL PRESSURE MEASUREMENTS
Experimental protocol
To simulate a wide range of “pleural” pressures surrounding the esophageal balloons, the
pressure in the chamber was set at 4 different
levels: 0, 10, 20 and 30 cmH2O. The chamber
pressure was returned to ambient by disconnecting the 100 mL syringe. Positive values of
chamber pressure were obtained by the injection
and pressurization of different volumes of air in
the cylinder by the 100 mL syringe. Different
esophageal catheters were introduced in the experimental chamber in a random order. At each
level of chamber pressure, the esophageal balloon catheters were progressively inflated by 0.5
mL steps, from 0 up to a maximum inflating air
volume of 10 mL or a maximum of balloon pressure of 50 cmH2O, whatever came first. For each
step, before the injection of the desired volume
by the 10 mL syringe, the esophageal balloon
catheter was deflated with generation of negative
pressure and then equilibrated at zero pressure
by disconnecting the 10 mL syringe for at least
10 seconds. After that, the catheter was inflated
up to the maximum level in order to distend the
balloon’s wall and then deflated to the desired
one. Since the increase of balloon volume could,
in theory, slightly modify the chamber pressure,
this was rechecked and eventually adjusted at the
desired level.
In all the tested conditions the balloon and
chamber pressures were obtained in duplicate.
Balloon transmural pressure was calculated
as the difference between the balloon and the
chamber pressure. Acceptable accuracy was defined by a transmural pressure between -1 and
+1 cmH2O. For each esophageal balloon catheter the working volume (Vworking) was defined
by the difference between the maximum (Vmax)
and minimum (Vmin) injected air volume
above zero which provided an acceptable accuracy. The balloon underfilling, appropriate filling
and overfilling were defined by all the injected
air volumes respectively below Vmin, between
Vmin and Vmax and above Vmax. The overlap
range of appropriate volumes (Voverlap) was defined by the maximum value of Vmin and the
minimum value of Vmax observed at the four
different levels of chamber pressure. Because the
858
esophagus can be considered a liquid milieu,
Vmin and Vmax were systematically verified by
submerging all the catheters under a 10, 20 and
30 cm water column.
A single catheter (NutriVent, Sidam, Italy)
was also studied at negative (-10 cmH2O) chamber pressure; in this case, after the catheter had
been equilibrated with ambient pressure, both
volume injection into and volume removal from
the catheter were evaluated, up to 10 mL or 50
cmH2O or down to -10 mL or -50 cmH2O, respectively.
Statistical analysis
Data are given as median (interquartile ranges). Agreement and correlation between measurements obtained in the experimental chamber
and those obtained in the water column were
evaluated by Bland-Altman analysis and Spearman’s test. Vmin, Vmax and Vworking were
compared among catheters by Kruskal Wallis test. Correlation between Vmin, Vmax and
chamber pressure was evaluated by Spearman’s
test. Balloon transmural pressures associated to
a standard volume of 0.5 mL and to volumes
recommended by manufacturers were computed
and correlation with chamber pressure evaluated
by Spearman’s test.
MedCalc®, version 13.0.6.0 (MedCalc Software, Belgium) was used for statistics. A P value<0.05 was considered significant.
Results
Considering both balloon and chamber pressure, the difference between two repeated measurements was always smaller than 0.2 cmH2O.
Appropriate injected volumes were determined for each esophageal balloon catheter in all
conditions (Figure 3). Considering all catheters
and conditions the median and interquartile
range for the Vmin, Vmax and Vworking were
1.0 (0.5, 1.5), 5.3 (3.8, 7.1) and 3.5 (2.9, 6.1)
mL, respectively. In Table I are shown Vmin,
Vmax, Vworking and Voverlap for each type
of esophageal balloon catheter. Vmax differed
among catheters (P=0.002), whereas Vmin did
not. Marquat and NutriVent had the highest
MINERVA ANESTESIOLOGICA
August 2015
ESOPHAGEAL PRESSURE MEASUREMENTSMOJOLI
50
NutriVent
40
pressure (cmH2O)
pressure (cmH2O)
50
30
20
10
Marquat
40
30
20
10
0
0
-10
-10
-1
0
1
2
3
4
5
6
7
8
9
-1
10
0
1
50
pressure (cmH2O)
pressure (cmH2O)
SmartCath
40
30
20
10
4
5
6
7
8
9
10
8
9
10
8
9
10
SmartCathG
40
30
20
10
0
0
-10
-10
-1
0
1
2
3
4
5
6
7
8
9
-1
10
0
1
50
50
pressure (cmH2O)
Microtek
40
2
3
4
5
6
7
injected volume (ml)
injected volume (ml)
pressure (cmH2O)
3
injected volume (ml)
injected volume (ml)
50
2
30
20
10
Cooper
40
30
20
10
0
0
-10
-10
-1
0
1
2
3
4
5
6
7
8
9
10
-1
0
1
2
3
4
5
6
7
injected volume (ml)
injected volume (ml)
Figure 3.—Balloon pressure as a function of the volume injected in the esophageal catheters, at four levels of chamber pressure.
Plain lines: balloon pressure, circles: Vmin, squares: Vmax. Dashed horizontal lines refer to the four tested values of chamber pressure: 0, 10, 20 and 30 cmH2O.
Vmax, while Cooper the lowest one (P<0.05).
The working volumes differed among catheters
(P=0.001): the greatest were observed with Marquat and NutriVent (P<0.05 vs. all the other
Vol. 81 - No. 8
catheters), whereas Cooper showed the smallest
one (P<0.05 vs. all the other catheters).
By increasing chamber pressure, Vmin significantly increased (rho=0.88, CI 0.75-0.95;
MINERVA ANESTESIOLOGICA
859
MOJOLI
ESOPHAGEAL PRESSURE MEASUREMENTS
Table I.—Appropriate and recommended filling volumes in the tested esophageal balloon catheters.
Cooper
Marquat
Microtek
NutriVent
SmartCath
SmartCathG
Vmin
(mL)
Vmax
(mL)
Vworking
(mL)
Voverlap
(mL)
Vrec
(mL)
0.8 (0.4-1.1)
1.3 (0.8-1.6)
2.0 (1.1-2.6)
1.3 (0.8-1.6)
1.3 (0.8-1.6)
1.0 (0.8-1.1)
1.8 (1.4-2.1)#
8.3 (7.5-9.3)§
5.3 (4.0-6.1)
7.5 (6.8-8.1)§
4.3 (4.0-4.6)
4.8 (3.8-5.5)
1.0 (1.0-1.0)#
7.5 (7.3-7.6)§
3.3 (2.9-3.5)
6.3 (6.0-6.5)§
3.0 (3.0-3.3)
3.5 (3.0-4.1)
/
2.0-7.5
/
2.0-6.0
2.0-4.0
1.5-3.0
1.0-2.0
0.5-3.0
0.5-1.0
4.0
0.5-2.5
0.5-2.5
Vmin and Vmax: minimum and maximum volumes providing appropriate balloon filling.; Vworking: Vmax-Vmin difference; Voverlap: overlap
range of appropriate volumes; Vrec: volumes recommended by manufacturers.
§P<0.05 vs. all others catheters.
#P<0.05 vs. Marquat, Microtek, Nutrivent, SmartCath, SmartCathG.
°P<0.05 vs. Marquat, Microtek, Nutrivent.
^P<0.05 vs. Cooper, SmartCath, SmartCathG.
p<0.0001), being 0.5 mL (0.5-0.5) at 0 cmH2O, 1.0 mL (1.0-1.0) at 10 cmH2O, 1.5 mL
(1.1- 1.5) at 20 cmH2O and 2.0 mL (1.6-2.0)
at 30 cmH2O. Marquat, NutriVent, SmartCath
and SmartCathG showed an overlap among the
appropriate volumes observed at the four different chamber pressures (Figure 3); thus, for these
four catheters it was possible to identify a range
of volumes (Voverlap) providing acceptable accuracy throughout the 0-30 cmH2O external
pressure range (Table I).
Values of Vmin and Vmax obtained in the water
column agreed (mean difference -0.1±0.2 mL and
0.1±0.4 mL, respectively) and were strictly correlated (rho=0.98, CI 0.94-0.99, P<0.0001 and
rho=0.99, CI 0.96-0.99, P<0.0001, respectively)
with those obtained in the experimental chamber.
In 15/24 cases (63%), the volume ranges recommended by manufacturers included not appropriate values. The minimum recommended volumes
of Marquat, Microtek, SmartCath and SmartCathG were associated with balloon underfilling
balloon transmural pressure (cmH2O)
2
SmartCathG
SmartCath
0
Marquat
Cooper
-2
NutriVent
-4
Microtek
-6
-8
-10
-12
-14
-16
-5
0
5
10
15
20
25
30
35
chamber pressure (cmH2O)
Figure 4.—Balloon transmural pressure as a function of chamber pressure (injected volume 0.5 mL).
860
MINERVA ANESTESIOLOGICA
August 2015
ESOPHAGEAL PRESSURE MEASUREMENTSMOJOLI
in the 10-30 cmH2O range of chamber pressure.
For Cooper, the minimum and maximum recommended volumes were associated with underfilling at 30 cmH2O and with overfilling in the 0-10
cmH2O range of chamber pressure, respectively.
The single volume recommended for NutriVent
was appropriate in all the conditions tested.
In case of appropriate filling, balloon transmural pressure was -0.2 cmH2O (-0.6, -0.0).
Balloon transmural pressures associated to a
standard volume of 0.5 mL and to the minimum volume recommended by manufacturers
were lower than -1 cmH2O in many cases (71%
and 53%, respectively) and were highly correlated with the level of chamber pressure (rho=0.97 and -0.71; P<0.0001), reaching values of
-10.4 (-12.4, -9.7) and -9.8 (-10.6, -3.4) at 30
cmH2O, respectively (Figure 4).
balloon transmural pressure (cmH2O)
15
Discussion
The major findings of this study are: 1) all the
esophageal balloon catheters showed a good accuracy in transmitting the external pressure when
inflated with appropriate filling volumes; 2) un-
a
b
a1
10
5
In case of negative surrounding chamber
pressure, the equilibration of the catheter with
ambient pressure led to balloon overinflation
(positive transballoon pressure) and therefore to
overestimation of the external pressure, despite
no volume injected in the catheter (Figure 5).
The removal of some air volume from the catheter was effective in restoring the appropriate
filling of the balloon and allowing an accurate
transmission of the external pressure, with negligible balloon transmural pressure.
a3
a2
0
b2
-5
b3
b1
-10
removal
injection
-15
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
volume (ml)
Figure 5.—Balloon transmural pressure as a function of the removed/injected volume at negative vs. positive surrounding pressure.
The NutriVent catheter was tested both at negative (-10 cmH2O, part “a” of the figure) and at positive surrounding pressure (+10
cmH20, part “b” of the figure). Part a. In case of negative external pressure, when the catheter is equilibrated with ambient pressure,
transmural pressure becomes positive, thus the balloon is already overinflated without any volume injection (a1). Some volume
should be removed from the catheter in order to reduce the balloon volume below the value at which hyperinflation occurs, thus
decreasing the transmural pressure to zero (a2). Further volume removal generates negative transmural pressure with underestimation of external pressure (a3). Part b. In case of positive external pressure, when the catheter is equilibrated with ambient pressure,
transmural pressure becomes negative and the balloon volume is zero (b1). Some volume needs to be injected (and compressed)
in the catheter to increase the inner pressure up to the external one and to a transmural pressure of zero (b2). Any further volume
injection translates into an increase of balloon volume, without any increase of balloon pressure, until the overfilling volume is
reached (b3): at this point, balloon pressure rapidly increases due to the elastic recoil generated by its walls.
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
861
MOJOLI
ESOPHAGEAL PRESSURE MEASUREMENTS
der positive pressure conditions, the minimum
appropriate filling volume was greater than 0.5
mL in most cases and was directly dependent on
the external pressure, and 3) the range of appropriate filling volumes differed among the esophageal balloon catheters and in most cases did not
completely correspond with manufacturer’s recommendations.
The most common indications for measuring
esophageal pressure are the assessment of elastic
properties of the lung, the optimization of the
ventilator setting like the settings of positive endexpiratory pressure (PEEP) and the measurement
of work of breathing.3
Reliability of esophageal pressure measurements depends on a proper positioning in the
esophagus,4, 5, 7, 17, 18 but also on the amount of
volume injected in the balloon catheter.6 An underfilled balloon cannot properly transmit the
surrounding pressure,8 whereas an overfilled and
overdistended one can generate a significant recoil pressure by itself and overestimate.3, 5-7 The
injected volumes required to avoid both the underfilling and overfilling of the balloon, i.e. the
range of appropriate volumes, depend on both
balloon and catheter length, diameter and compliance.5-7 Thus, appropriate filling volumes may
significantly differ among the various esophageal
catheters available on the market.3, 8
In the present study, appropriate filling volumes were identified for all the catheters at all
the chamber pressure tested. Importantly, these
volumes did not differ when a liquid instead of
a gas environment for the balloon catheters was
set up. The minimum volume to avoid underfilling (Vmin) was very similar among the esophageal balloon catheters. Median value of Vmin was
1 mL, but accordingly to our preliminary report
15 it clearly rose with the increase of the external
pressure and reached values as high as 3 mL. Our
experimental observations are only apparently in
contrast with the assumption that a minimal volume, i.e. more or less 0.5 mL, is necessary for the
balloon to accurately transmit pressure.8 When a
certain amount of volume is injected in the catheter by a syringe, this leads to a corresponding
inflation of the balloon only in the case of atmospheric pressure surrounding the balloon. Instead,
when this pressure is positive, the injection of a
862
volume of air actually translates into a pressurization of the gas contained in the catheter, until the
inner pressure reaches the one surrounding the
balloon. At this point, the balloon’s transmural
pressure is no more negative and the balloon starts
inflating (Figure 4). Therefore, the more positive
the external pressure, the higher the volume that
has to be injected in the catheter to avoid the underfilling of the balloon.
Balloon overinflation can also occur. At low
filling volumes, esophageal balloons show a nonelastic behaviour, i.e. they can increase their volume without generation of elastic recoil pressure.3,
7 Above a critical value, any further air volume
injection produces a steep increase of pressure,
i.e. the balloon starts being overdistended, thus
leading to a progressive and significant increase
of the balloon transmural pressure with overestimation of the real external pressure. Despite
very similar balloon lengths, the volume at which
overinflation started (Vmax) varied significantly
among the different catheters, ranging from 1
to 7.5 mL at ambient pressure. Recently, it was
observed that a sustained overinflation of an esophageal balloon can stretch its wall and increase
the volume at which overdistention occurs, thus
slightly prolonging the range of appropriate volumes.14 Higher volumes of overinflation lead to
larger working volumes and allow an overlap of
the appropriate volumes observed at different external pressure levels. In 4 of the 6 catheters tested
we could identify a range of volumes (Voverlap)
that gave accurate transmission of the surrounding pressures in the whole 0-30 cmH2O range.
The possibility to inject in the esophageal catheter
the same volume for almost all intrathoracic conditions simplifies the clinical practice and allows
good accuracy also in case of great respiratory
variations of pleural pressure. Interestingly, these
volume ranges only partially corresponded to the
volumes suggested by literature and by manufacturers. In particular, the conventional and widely
accepted volume of 0.5 mL and the minimum
recommended volume by single manufacturers
led to balloon underfilling in more than 70% and
50% of the cases, respectively.
Conventional small volumes (0.5-1.0 mL)
seem to be appropriate for measuring near-atmospheric intrathoracic pressures in upright sponta-
MINERVA ANESTESIOLOGICA
August 2015
ESOPHAGEAL PRESSURE MEASUREMENTSMOJOLI
neously breathing subjects, but are prone to underestimate the more positive pleural pressures
generally observed in ICU patients undergoing
positive pressure mechanical ventilation while lying in supine position. Moreover, the more positive the external pressure, the greater the underestimation: this means that end-inspiratory values
of esophageal pressure will be more affected than
end-expiratory ones. Therefore, both absolute values and respiratory oscillations of esophageal pressure may be under-transmitted by an underfilled
balloon.
Any time esophageal pressure is considered
for clinical or research purposes, accuracy of the
measurement should be checked by performing
an airway occlusion test 3, 4 Our data suggest that,
in case of incomplete transmission of pleural pressure to the esophageal balloon during this test, the
risk of balloon underfilling should be considered
before attempting the repositioning of the catheter. Accordingly, an injected volume among those
in vitro validated as appropriate under positive
pressure conditions should be tried out. Alternatively, the volume could be in vivo titrated up to
the actual Vmin, that is the volume at which balloon pressure stabilizes and the balloon pressurevolume relationship becomes roughly flat.
When negative intrathoracic pressures were
simulated, the risk of balloon overfilling also for
small injected volumes (or even no injection at
all) was clearly demonstrated. In theory, in case
of negative surrounding pressures, the removal of
some volume from the esophageal catheter could
restore the appropriate balloon filling volume and
the accuracy of esophageal pressure measurement
(Figure 4). Anyway, it seems more reasonable to
avoid equilibrating the esophageal catheter with
ambient pressure during patient’s inspiratory efforts that may lead to negative intrathoracic pressures. For this purpose, the patient could be asked
for a respiratory pause or a Valsalva maneuver.
Study limitations
This study has some limitations. Our experimental setup, conceived to evaluate the accuracy
of esophageal balloon catheters in transmitting
the surrounding pressure, was clearly a simplistic
model. Actually, the difference between pleural
Vol. 81 - No. 8
and esophageal balloon pressures have been attributed to both the elastic properties of the balloon
and the elastance of the esophageal wall.8, 19-21 We
decided not to simulate the esophageal reaction
to balloon inflation because it was considered not
technically feasible and reliable. Moreover, we
applied a constant and uniform pressure in our
experimental chamber, but pleural pressure is
not homogeneous along the lung surface 18 and a
vertical gradient is normally present.22 Therefore,
beneficial effects on esophageal pressure measurements provided by titrating balloon filling volume
up to an appropriate value as suggested by our in
vitro experience should be confirmed by further
in vivo studies.
Conclusions
In conclusion, all the esophageal catheters tested in this in vitro study allow accurate transmission of the external pressure when properly filled,
being the range of appropriate filling volumes
different among catheters. The minimum volume
that has to be injected is greater than those usually recommended and is directly dependent on
the external pressure surrounding the balloons.
Under and overfilling of esophageal balloons can
lead to significant errors in the assessment of both
absolute values and respiratory changes of esophageal pressure.
Key messages
—— The range of appropriate filling volumes differs among the second-generation
esophageal balloon catheters.
—— Under and overfilling of esophageal
balloons can lead to significant errors in the
assessment of esophageal pressure-based parameters.
—— Under positive pressure conditions the
minimum volume that has to be injected may
be greater than those usually recommended.
—— In case of unsatisfactory occlusion
test, before attempting catheter repositioning
consider the possibility of balloon underfilling and try to achieve the appropriate volume range by further volume injection.
MINERVA ANESTESIOLOGICA
863
MOJOLI
ESOPHAGEAL PRESSURE MEASUREMENTS
References
  1. Zin WA, Milic-Emili J. Esophageal pressure measurement.
In: Physiologic basis of respiratory disease, first edition.
Hamilton, Canada: BC Decker Inc; 2005. p. 639-47.
  2. Buytendijk HJ. Oesophagusdruck en Longelasticiteit [Dissertatie]. The Netherlands: University of Groningen; 1949.
  3. Akoumianaki E, Maggiore SM, Valenza F, Bellani G, Jubran A, Loring SH et al. The application of esophageal pressure measurement in patients with respiratory failure. Am J
Respir Crit Care Med 2014;189:520-31.
  4. Baydur A, Cha EJ, Sassoon CS. Validation of esophageal
balloon technique at different lung volumes and postures. J
Appl Physiol 1987;62:315-21.
  5. Milic-Emili J, Mead J, Turner JM, Glauser EM. Improved
technique for estimating pleural pressure from esophageal
balloons. J Appl Physiol 1964;19:207-11.
  6. Petit JM, Milic-Emili G. Measurement of endoesophageal
pressure. J Appl Physiol 1958;13:481-5.
  7. Mead J, McIlroy MB, Selverstone NJ, Kriete BC. Measurement of intraesophageal pressure. J Appl Physiol
1955;7:491-5.
  8. Hedenstierna G. Esophageal pressure: benefit and limitations. Minerva Anestesiol 2012;78:959-66.
  9. Chiumello D, Carlesso E, Cadringher P, Caironi P, Valenza
F, Polli F et al. Lung stress and strain during mechanical
ventilation for acute respiratory distress syndrome. Am J
Respir Crit Care Med 2008;178:346-55.
10. Grasso S, Terragni P, Birocco A, Urbino R, Del Sorbo L,
Filippini C et al. ECMO criteria for influenza A (H1N1)associated ARDS: role of transpulmonary pressure. Intensive Care Med 2012;38:395-403.
11. Talmor D, Sarge T, O’Donnell CR, Ritz R, Malhotra A,
Lisbon A et al. Esophageal and transpulmonary pressures in
acute respiratory failure. Crit Care Med 2006;34:1389-94.
12. Terragni PP, Filippini C, Slutsky AS, Birocco A, Tenaglia T,
Grasso S et al. Accuracy of plateau pressure and stress index
to identify injurious ventilation in patients with acute respiratory distress syndrome. Anesthesiology 2013;119:880-9.
13. Chiumello D, Gallazzi E, Marino A, Berto V, Mietto C,
Cesana B et al. A validation study of a new nasogastric polyfunctional catheter. Intensive Care Med 2011;37:791-5.
14. Walterspacher S, Isaak L, Guttmann J, Kabitz HJ, Schumann S. Assessing respiratory function depends on mechanical characteristics of balloon catheters. Respir Care
2013;59:1345-52.
15. Mojoli F, Volta C, Pozzi M, Currò I, Milic-Emili J, Torriglia
F et al. The use of an oesophageal balloon to measure pleural pressure: effects of different inflation volumes. Intensive
Care Med 2011;37:S194.
16. Clarenbach CF, Camen G, Sievi NA, Wyss C, Stradling JR,
Kohler M. Effect of simulated obstructive hypopnea and
apnea on thoracic aortic wall transmural pressures. J Appl
Physiol 2013;115:613-17.
17. Higgs BD, Behrakis PK, Bevan DR, Milic-Emili J. Measurement of pleural pressure with esophageal balloon in anesthetized humans. Anesthesiology 1983;59:340-3.
18. Milic-Emili J, Mead J, Turner JM. Topography of esophageal pressure as a function of posture in man. J Appl Physiol
1964;19:212-6.
19.Knowles JH, Hong SK, Rahn H. Possible errors using
esophageal balloon in determination of pressure-volume
characteristics of the lung and thoracic cage. J Appl Physiol
1959;14:525-30.
20. Milic-Emili G, Petit JM. Relationship between endoesophageal and intrathoracic pressures in dog. J Appl Physiol
1959;14:535-7.
21. Hedenstierna G, Jarnberg PO, Torsell L, Gottlieb I. Esophageal elastance in anesthetized humans. J Appl Physiol
Respir Environ Exerc Physiol 1983;54:1374-8.
22. Pelosi P, Goldner M, McKibben A, Adams A, Eccher G,
Caironi P et al. Recruitment and derecruitment during acute respiratory failure: an experimental study. Am J
Respir Crit Care Med 2001;164:122-30.
Acknowledgements.—Catherine Klersy (Servizio di Biometria e Statistica, IRCCS Fondazione Policlinico S. Matteo, Pavia) for statistics.
Conflicts of interest.—F. Mojoli participated in the development of Nutrivent, and he is involved in a University research spin-off and he
received a honorarium as member of the advisory board. He did not receive any payment for writing or contributing to the report.
D. Chiumello participated in the development of Nutrivent, and he is involved in a University research spin-off and he received a honorarium as member of the advisory board. He did not receive any payment for writing or contributing to the report.
A. Braschi participated in the development of Nutrivent, and he is involved in a University research spin-off and he received a honorarium
as member of the advisory board. He did not receive any payment for writing or contributing to the report.
Received on November 12, 2014. - Accepted for publication on January 27, 2015. - Epub ahead of print on January 30, 2015.
Corresponding author: D. Chiumello, Dipartimento di Anestesia, Rianimazione e Terapia del dolore, Fondazione IRCCS Cà GrandaOspedale Maggiore Policlinico, Via F. Sforza 35, Milano, Italy. E-mail: [email protected].
864
MINERVA ANESTESIOLOGICA
August 2015

O R I G I N A L A RT I C L E
Feasibility of Post-Intensive Care Unit Clinics:
an observational cohort study
of two different approaches
D. S. DETTLING-IHNENFELDT 1, A. E. DE GRAAFF 2, F. NOLLET 1, M. VAN DER SCHAAF 1
1Department of Rehabilitation, Academic Medical Center, University of Amsterdam, Amsterdam, The
Netherlands; 2Department of Intensive Care, Tergooi Hilversum, The Netherlands
ABSTRACT
Background. Post-ICU clinics have been advocated to reduce long-term physical and psychological impairments
among ICU survivors. A format for optimal structure, timing, and care content has not yet been established. We
developed and implemented two post-ICU clinics in different hospital settings and evaluated the feasibility.
Methods. In this prospective cohort study ICU-survivors of a university hospital (AMC) and a general hospital
(TG), who were mechanically ventilated ≥2 days and discharged to their homes, were invited to the post-ICU clinic
one month after hospital discharge (AMC) or three months after ICU discharge (TG). Feasibility was evaluated as 1)
the number of eligible ICU-survivors and the proportion that attended; 2) the prevalence of ICU-related abnormalities, that required referral for further treatment; and 3) patient satisfaction.
Results. Forty-five of 629 AMC-patients and 70 of 142 TG-patients were eligible for the post-ICU clinic. Of
these, 49% and 67% respectively, visited the outpatient clinic (P=0.026). The majority of all screened patients had
functional restrictions, and 68% required referral for further diagnosis and treatment. Patient satisfaction was high.
Conclusion. This study provides valuable information to support the implementation of post-ICU clinics. The use
of validated screening instruments facilitates the identification of patients with need for further treatment. Early inhospital screening and recruiting patients at highest risk for adverse outcome could be a more targeted approach to
achieve greater benefit. (Minerva Anestesiol 2015;81:865-75)
Key words: Intensive care - Rehabilitation - Feasibility studies - Outpatient clinics, hospital.
E
ach year about 80,000 adults are admitted
to intensive care units (ICUs) in the Netherlands and, due to progress in critical care,
survival rates have increased. A significant proportion of survivors have long-term physical,
psychological, and cognitive impairments that
negatively affect daily function, employment,
and health related quality of life (HRQOL).1-5
The complexity and magnitude of these ICU-related consequences, recently denoted as Post Intensive Care Syndrome (PICS), require comprehensive multidisciplinary care.6 However, ICU
Comment in p. 832.
Vol. 81 - No. 8
survivors experience inadequate and disjointed
multidisciplinary care after hospital discharge,
with inconsistent service provision.7 Particularly
for ICU survivors who are discharged to their
homes, PICS may not be reliably and promptly
recognized, resulting in incomplete or late referral to the appropriate care.
Post-ICU clinics have been advocated to manage ICU-related problems in survivors,8-10 but
to date such clinics are scarce, their organization varies, and their optimal structure, timing,
and care content has not been established yet.
Furthermore, there is no direction or consensus
on how to implement it. We developed and im-
MINERVA ANESTESIOLOGICA
865
DETTLING-IHNENFELDT
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICS
plemented a post-ICU clinic in a large tertiary
university hospital and in a general hospital in
the Netherlands, based on the recommendations
from the National Institute for Health and Clinical Excellence (NICE) guidelines.11 The aim
of this post-ICU clinic according to the NICE
guideline, is to ‘screen and detect’ patients “that
recover at a slower rate than anticipated, and to
identify patients that has developed new physical and/or psychological morbidity, that was not
previously identified”, and to initiate tailored
treatment, if required. We assumed that patients
of a university hospital would have more complex and serious illnesses, resulting in longer
ICU and hospital stays and a greater need for
the post-ICU clinic than ICU survivors of a general hospital would have. Also, we expected that
the programmatic evaluation of both approaches
would provide important practical information
for further improvement of post-ICU clinics.
The purpose of this study was to evaluate the
feasibility of the post-ICU clinic in each hospital setting by determining: 1) the number of
eligible ICU survivors and the proportion that
attends the post-ICU clinic; 2) the prevalence
of physical and psychological impairments and
functional restrictions that require referral for
further diagnosis and treatment; and 3) the level
of patient satisfaction with the post-ICU clinic.
Materials and methods
Study design, setting and participants
This prospective cohort study was undertaken in the Academic Medical Center (AMC) in
Amsterdam and the Tergooi (TG) in Hilversum.
The AMC is a 1000-bed university hospital with
a 34-bed mixed medical and surgical closedformat ICU and TG is a 633-bed general hospital with a 10-bed mixed medical and surgical
closed-format ICU.
All consecutive adult (age ≥18 years) ICU patients admitted in a 20 months period (20102012) in the AMC and TG were screened for
participation in the study. Critically ill patients
mechanically ventilated for ≥2 days and discharged to their homes from the hospital were
considered eligible. Patients with insufficient
866
knowledge of the Dutch language or comorbidity that would impair completing questionnaires
or visiting the outpatient clinic (e.g. mental retardation, mechanical ventilation at home, terminal disease, etc.) were excluded. Considering
their limited physical resilience and mobility
range, AMC patients living beyond the service
area of the hospital (a travel distance of more
than 15 km) were also excluded. The Ethical
Review Board of the AMC waived the need for
informed consent because of the non-interventional nature of this study.
Description of the post-ICU clinics
The main purposes of our post-ICU clinics
were 1) to screen patients for physical and psychological impairments, functional restrictions,
and HRQOL; 2) to identify care giver strain
and symptoms of post-traumatic stress disorder
(PTSD) in close relatives; 3) to refer patients or
close relatives with unanticipated ICU-related
sequelae for further treatment; and 4) to inform
patients and their close relatives about short and
long-term ICU-related consequences. As part
of the standard clinical care, in both hospitals
physical therapy interventions started early on
the ICU, and were continued on the wards until
hospital discharge.
Although the purposes of the post-ICU clinic
were similar in the two hospitals, the approach
differed with respect to the involved professionals, the timing, and the used satisfaction questionnaires because of the differences in existing
care approaches. In the AMC, the post-ICU
clinic was implemented in May 2010 by the department of rehabilitation medicine and carried
out by a senior physical therapist specialized in
critical care. Patients and relatives were invited
one month after hospital discharge to enable
early identification of ICU-related problems and
referral to other practitioners. In October 2010,
a similar post-ICU clinic was implemented in
the general hospital under the responsibility of
the ICU, led by nursing staff. Patients and their
relatives were invited three months after ICU
discharge to evaluate recovery and initiate additional care if necessary.
In both hospitals, two weeks prior to the visit
MINERVA ANESTESIOLOGICA
August 2015
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICSDETTLING-IHNENFELDT
to the post-ICU clinic, patients completed the
Dutch-translated and validated versions of the
following questionnaires:
1) The Medical Outcomes Study 36-item
Short Form Health Survey (SF-36) was used to
measure HRQOL.12, 13 The domains “physical
function” and “role limitation due to physical
problems” were used to screen patients for physical impairments and functional restrictions.
2) The Hospital Anxiety and Depression Scale
(HADS) was uses to assess symptoms of anxiety
and depression.14-16 Sub-scale scores (0-21) of ≥8
indicate anxiety or depressive symptoms.16, 17
3) The 10-item Trauma Screening Questionnaire (TSQ) was used to identify patients at risk
for PTSD. A cut-off score of 6 has found to be
optimal.18, 19
4) Health care utilization, a self-composed list
on which patients had to mark all health care
professionals that they received care from.
Information on socio-demographics (age,
height, body weight, living situation, and employment [number of working hours] before and
after the ICU admission) was recorded, and clinical data were retrieved from medical records.
In close relatives, PTSD and potential care
giving concerns were assessed before their visit to
the post-ICU clinic, using the TSQ and Caregiver Strain Index (CSI). The CSI is a validated 13
item questionnaire with a score of >7 indicating
a higher risk for strain.20 In addition, socio-demographic characteristics (age, gender, relationship to the patient, work situation before and
after patient’s hospital admission) were recorded.
At the post-ICU clinic, a shortened version of
the International Classification of Functioning,
Disability, and Health (ICF) checklist (Part 1a:
impairments of body functions and Part 2: activity limitations and participation restrictions),
and the Malnutrition Universal Screening Tool
(MUST) were completed for the screening of
ICU-related impairments or functional restrictions. The ICF checklist of the World Health
Organization is a practical tool to identify and
classify information on the functioning and disability of an individual.21 The MUST is used in
hospitals to identify patients at risk for undernourishment.22 A MUST score of 2 implies a
high risk of malnutrition.23
Vol. 81 - No. 8
Patients with clinically meaningful ICUrelated physical or psychological impairments
according to the questionnaires and screening
tools were referred for further treatment. Close
relatives with high burden of care and/or symptoms of PTSD were advised to contact their general practitioner for support or additional care.
AMC patients were asked to rate their satisfaction with the post-ICU clinic on an ordinal
5-item scale (4=very satisfied, 3=satisfied, 2=nor
satisfied/nor dissatisfied, 1=dissatisfied, 0=very
dissatisfied). At TG, a 10-point scale, with higher scores indicating more satisfaction (1=very
dissatisfied, 10=very satisfied) was used.
ICU survivors who declined the invitation to
visit the post-ICU clinic were contacted by phone
to ascertain their reasons for not attending.
Feasibility
The feasibility of each post-ICU clinic was
evaluated as: 1) the number of eligible ICU
survivors and the proportion that attended one
month after hospital discharge for AMC, and
three months after ICU discharge for TG; 2) the
prevalence of physical and psychological impairments, and functional restrictions that required
referral for further diagnosis and treatment; and
3) patient satisfaction.
Statistical analysis
Demographic and clinical data of ICU patients, including age, gender, Body Mass Index
(BMI), Acute Physiology and Chronic Health
Evaluation II (APACHE II), ICU admission
diagnosis, duration of mechanical ventilation,
length of ICU and hospital stay, and working
hours prior to ICU admission, were compared
both between the two hospital settings, and between patients that attended the post-ICU clinic
and patients who declined or who were lived beyond the service area of the hospital. Also, physical and psychological impairments, and functional restrictions after hospital discharge were
compared between the two hospitals, as well
as the demographic data of close relatives, care
giver strain, and post-traumatic stress. Comparisons were performed with the independent t-test
MINERVA ANESTESIOLOGICA
867
DETTLING-IHNENFELDT
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICS
(continuous data), the Mann-Whitney U-test
(ordinal data), and the Chi-square test (categorical and dichotomous data) or Fisher’s exact test.
A P-value less than 0.05 was considered statistically significant. For descriptive analysis, mean
values were presented with standard deviation
(±SD), median values with interquartile range
(IQR) and proportions with percentages and total numbers. All analyses were performed in IBM
SPSS Statistics 19.0 (SPSS Inc., Chicago, USA).
Results
Number of eligible patients
Of the patients who were mechanically ventilated ≥2 days (AMC: N.=629, TG: N.=142),
45 patients from the AMC and 70 from TG met
the inclusion criteria and were discharged home
(Figure 1). The baseline characteristics of the 60
AMC-patients who were excluded because of
travel distance >15 km, did not differ from those
of included AMC-patients, except for age (median [IQR] age 57 [42-66.8] in excluded patients
vs. 61 [54-69.5] in included patients, P=0.041).
Of all eligible patients, 22 AMC patients
(49%) and 47 TG patients (67%) visited the
post-ICU clinic (P=0.026). AMC patients visited the outpatient clinic at a mean of 5.5 weeks
(SD 2.6) after hospital discharge and TG patients
at 12.1 weeks (SD 3.7) after ICU discharge.
Reasons for not attending (N.=23) the postICU clinic in the AMC were: good health (N.=2)
or participation in a cardiac rehabilitation program (N.=8), no need (N.=6), health-related difficulties (N.=4), and re-admission to the hospital
(N.=2); one patient could not be traced. At TG,
the reasons for not attending (N.=23) were: impossible to contact (N.=12), health-related difficulties (N.=6), and no need (N.=5).
The demographic and clinical data of the
study participants in the two hospitals were comparable (Table I), except for gender (42% men in
AMC vs. 76% in TG, P=0.008). Patient characteristics between ICU survivors who visited the
post-ICU clinic and who declined did not differ
(P>0.05) (Table II). Three AMC patients and
one TG patient were excluded from the analysis
because of incomplete questionnaires (Figure 1).
Functional status and referral
There were no statistical differences in functional status of ICU survivors in the two hospitals (Table III). Out of all SF-36 domain scores,
only “bodily pain” and “mental health” scores
were comparable to the normative values of
the general Dutch population. All other SF-36
Table I.—Demographic and clinical data of intensive care unit aftercare patients.
Patient characteristics
AMC
N. of patients
Age (years), mean (SD)
Gender (male), N. (%)
BMI, median (IQR)
Obesity (BMI>30), N. (%)
ICU admission diagnosis, N. (%)
Medical
Elective surgery
Non-elective surgery
APACHE II Score, median (IQR)
Duration of mechanical ventilation (days), median (IQR)
LOS in ICU (days), median (IQR)
LOS in hospital after ICU-discharge (days), median (IQR)
Remunerative employment, N. (%)
Working hours per week, median (IQR)
Voluntary work, N. (%)
Hours per week, range
19
58.1 (14.2)
8 (42.1)
24.9 (22.5-31.2)
5 (26.3)
16 (84.2)
1 (5.3)
2 (10.5)
21 (16-24)
6 (4-7)
7 (5-8)
11 (8-14)
6 (31.6)
36 (27-40)
2 (10.5)
3-5
Tergooi
46
63.4 (11.5)
35 (76.1)
27.4 (23.5-29.8)
10 (21.7)
36 (78.3)
4 (8.7)
6 (13)
18 (15-24.5)
4.15 (2.9-7.3)
6.85 (5-10)
12 (6.8-17.3)
13 (28.3)
36 (31-39)
8 (17.4)
2-20
P
0.118
0.008
0.478
0.077
0.891
0.155
0.840
0.817
0.160
0.485
IQR: interquartile range (25th and 75th percentile); N.: number; BMI: Body Mass Index (kg/m2); ICU: intensive care unit; APACHE II: Acute
Physiology and Chronic Health Evaluation; LOS: length of stay.
868
MINERVA ANESTESIOLOGICA
August 2015
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICSDETTLING-IHNENFELDT
Table II.—Comparison of patient characteristics between intensive care unit survivors who visited the post-ICU clinic, and
who did not.
Patient characteristics
Participants
post-ICU clinic
Non-participants
post-ICU clinic
N. of patients
Age (years), mean (SD)
Gender (male), N. (%)
BMI, median (IQR)
ICU admission diagnosis, N. (%)
Medical
Elective surgery
Non-elective surgery
APACHE II score, median (IQR)
Duration of mechanical ventilation (days), median (IQR)
LOS in ICU (days), median (IQR)
LOS in hospital after ICU-discharge (days), median (IQR)
Remunerative employment, N. (%)
Working hours per week, median (IQR)
Voluntary work, N. (%)
Hours per week, range
65
64.8 (14.7)
43 (66.2)
26.8 (23.3-30.2)
46
61.8 (12.5)
27 (58.7)
28.4 (25.3-23.9)
52 (80.0)
5 (7.7)
8 (12.3)
18 (15.5-24)
5 (3-7)
6.9 (5-10)
12 (7.5-16.5)
22 (33.8)
36 (29.5-40)
10 (15.4)
2-20
35 (76.1)
3 (6.5)
8 (17.4)
19 (15-23.25)
8 (3-10)
8.6 (4.7-12.4)
11 (6-20)
NO
NO
P
0.247
0.423
0.151
0.799
0.838
0.078
0.481
0.808
-
IQR: interquartile range (25th and 75th percentile); N.: number; BMI: Body Mass Index (kg/m2); ICU: intensive care unit; APACHE II: Acute
Physiology and Chronic Health Evaluation; LOS: length of stay; NO: not available.
Figure 1.—Study patient recruitment diagram.
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
869
DETTLING-IHNENFELDT
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICS
Table III.—Functional status at 4 weeks after hospital discharge and 3 months after intensive care unit discharge.
Screening instrument
SF36 (range 0-100, normative value)a, median (IQR)
Physical function (76.8)
Role physical (70.8)
Bodily pain (70.8)
General health (65)
Vitality (69)
Social function (82.7)
Role emotional (82.4)
Mental health (75.6)
TSQ (0-10)
Patients with risk for PTSD (sum score ≥ 6)
HADS sub-scale scores (0-21)
Patients with indication for anxiety (score ≥ 8)
Patients with indication for depression (score ≥ 8)
Patients with signs of malnutrition (MUST≥2)
Returned to work
Resumed voluntary work
Hours per week, range
AMC (N.=19)
Tergooi (N.=46)
P
30 (18.8-56.3)
12.5 (0-25)
67.3 (55.1-100)
55 (20-67.5)
40 (33.8-61.3)
62.5 (37.5-78.1)
66.7 (25-100)
76 (59-84)
50 (25-75)
0 (0-25)
67.3 (44.9-81.6)
50 (35-61.3)
55 (35-66.3)
62.5 (37.5-87.5)
66.6 (33.3-100)
76 (55-88)
0.131
0.772
0.455
0.822
0.308
0.635
0.987
0.811
3 (15.8%)
9 (19.6%)
0.196
6 (31.6%)
5 (26.3%)
9 (42.1%)
0 of 6
1 (5.3%)
5
13 (28.3%)
8 (17.4%)
15 (32.6%)
0 of 13
3 (6.5%)
4-5
0.080
0.493
0.680
1.000
0.848
SF36: Medical Outcomes Study 36-Item Short-Form Health Survey; IQR: interquartile range; TSQ: Trauma Screening Questionnaire; PTSD: Post
Traumatic Stress Disorder; HADS: Hospital Anxiety and Depression Scale; MUST: Malnutrition Universal Screening Tool.
aHigher scores representing better functioning; SF-36 normative data for the general Dutch population.
domain scores were lower in our study population. Figure 2 shows the ten most frequently
reported impairments and restrictions according
to the ICF checklist. For these, predominantly
physical impairments, 47% of the AMC patients
and 54% of the TG patients received physical
therapy. Signs of PTSD were found in 19% of
the 65 participants and symptoms of anxiety
and depression were present in 29% and 20%,
respectively. Of the patients with psychological
Figure 2.—Ten most reported impairments and restrictions on the International Classification of Functioning, Disability and
Health (ICF) checklist (AMC N.=19/TG N.=46).
870
MINERVA ANESTESIOLOGICA
August 2015
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICSDETTLING-IHNENFELDT
symptoms, 41% received treatment from a psychologist, psychiatrist, or social worker. Thirtyseven percent of all patients had signs of malnutrition of which 16% received treatment from a
dietician.
No patient had returned to paid work, but
four out of ten patients had resumed the volunteer work they did prior to ICU admission.
In the AMC post-ICU clinic, 15 patients were
referred to other specialties: physical therapist
(N.=6), psychologist/psychiatrist (N.=3), and
dietician (N.=8). In addition, 2 patients with
physical impairments were referred back to their
physical therapist with specific training instructions. In TG, 27 patients were referred to a physical therapist (N.=8), psychologist/psychiatrist
(N.=7), and dietician (N.=12), and 12 patients
were referred back to their physical therapist
with additional training instructions.
Close relatives
In the AMC 11 and in TG 41 close relatives
completed the questionnaires. The median age
of the close relatives and the reported hours per
week spent on care were significantly higher for
the TG than for the AMC patients (median age:
58 vs. 63 [P=0.023]; care assignment: 3.8 vs. 10
hours per week [P=0.025]). Burden of care was
high (CSI≥7) in 9% of the close relatives in the
AMC and 18% in the TG. Signs of PTSD were
not found in the AMC population, but were
present in 16% of close relatives of TG patients.
High scores on TSQ and CSI were particularly
found in partners and children.
Patient satisfaction
AMC patients were “very satisfied” (65%) and
“satisfied” (35%) with the post-ICU clinic. For
the TG patients, the median (IQR) satisfaction
score was 8 (8-9).
Discussion
The results of our study show that post-ICU
clinics are useful in identifying potential problems in daily functioning in survivors of critical
illness. Although, the number of eligible patients
and the follow-up rates of ICU survivors were
limited in our study, the programmatic evaluation of the two post-ICU clinics in a university
and a general hospital provides important information on different formats of post-ICU clinics.
We found that a limited proportion of the
AMC survivors were discharged to their homes,
17%, versus 49% of the TG survivors. This can
be explained by the fact that a tertiary university
hospital has a large catchment area and many patients were transferred back to a general hospital
near their residential area early after specialized
medical treatment. Additionally, the majority
Table IV.—Demographic data of close relatives.
Demographic characteristics
N. of close relatives
Age (years), median (IQR)
Gender (female), N. (%)
Relationship to patient, N. (%)
Partner
Family (other than partner)
No family
Remunerative employment, N. (%)
Working hours/week baseline, median (IQR)
Working hours/week after ICU discharge, median (IQR)
Relatives working less hours after ICU, N. (%)
Care assignment, N. (%)
Hours/week, median (IQR)
CSI≥7
TSQ≥6
AMC
Tergooi
P
11
58 (52-59)
7 (63.6)
41
63 (57-69)
30 (73.2)
0.023
0.535
9 (81.8)
1 (9.1)
1 (9.1)
5 (45.5)
36 (26-40)
32 (10-40)
1 (9.1)
8 (72.7)
3.8 (1-5)
1 (9.1)
0
32 (78)
7 (17.1)
2 (4.9)
15 (36.6)
32 (21.3-40)
29 (16-40)
4 (26.7)
19 (55.8)
10 (3-18)
7 (17.9) N.=39
6 (15.8) N.=38
0.576
0.436
0.158
0.329
0.781
0.025
1.000
0.529
CSI: Caregiver Strain Index (sum score ≥7 indicating higher risk for strain); IQR: interquartile range; no: number; TSQ: Trauma Screening Questionnaire (sum score ≥ 6 indicating risk for post-traumatic stress disease).
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
871
DETTLING-IHNENFELDT
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICS
of AMC-patients lived beyond the service area
of the hospital, which resulted also in a small
number of eligible patients for the post-ICU
clinic.
Of all eligible patients 49% of the AMC and
67% of the TG patients visited the outpatient
clinic. This result was in line with comparable
studies from Schandl 9 and Cutler 24 with attendance rates of 66% and 32% three months
after ICU discharge. Ten eligible AMC patients
(22%) did not attend the post-ICU clinic, because they had no complaints, or were already
involved in a rehabilitation program. These
patients may be assumed to have received care
tailored to their individual needs. However, of
greater concern was the group of eligible patients
(13% at AMC and 23% at TG) that reported no
need for ICU follow-up, because these patients
may be suffering serious psychological problems,
such as anxiety or PTSD, and consequently
avoiding hospital contact.
With respect to the limited physical and psychological resilience of ICU survivors, also the
timing of the post-ICU clinic seems important. One month after hospital discharge, 29%
of the eligible AMC patients were not able to
visit the outpatient clinic due to poor health.
For that reason, 16% of the AMC patients visited the post-ICU clinic later, between 5 and
12 weeks after hospital discharge. Of the TG
patients, only 9% did not attend the post-ICU
clinic at three months after ICU discharge because of poor health. Based on these findings,
we suppose that a post-ICU clinic three months
after ICU discharge would be more feasible for
ICU-survivors. With the purpose of screening
in mind, this timing enables patients to reflect
on their recovery process and to determine any
physical, mental or cognitive impairments, that
were not present during hospital stay. This is also
in line with the recommendation in the NICE
guideline.11, 25 However, early identification of
functional impairments is essential to initiate
rehabilitation treatment as soon as possible, to
improve recovery and to prevent chronic complaints.1, 19, 26 In addition, several studies emphasize the importance of early support because of
the many difficulties patients face shortly after
hospital discharge.27, 28 Therefore, early in-hospi-
872
tal screening and risk stratification, followed by
an assessment in a post-ICU clinic could be a
more targeted approach to identify ICU patients
at highest risk for adverse outcome.
We expected that patients of a university hospital (AMC) would have more complex and serious illnesses, resulting in longer ICU and hospital stays and a greater need for the post-ICU
clinic than ICU survivors of a general hospital
(TG). However, we found no significant differences between patient and clinical characteristics, functional status, and indications for
referral between the two hospitals. These results should be taken cautiously, given the small
study-population and the several differences in
approach of the post-ICU clinics. Given the fact
that TG patients visited the clinic two months
later than AMC patients, we expected that they
would have been further in their recovery process and perceive better HRQOL. Nevertheless,
this was not the case in our study. Although,
it is not possible to draw firm conclusions on
the basis of the small numbers involved, these
results show that patients experienced similar
restrictions and HRQOL, regardless of type of
ICU, severity of illness, demographic characteristics and time after ICU discharge. Despite the
exclusion of patients who were transferred to a
long-term care facility, we found severe physical
and psychological impairments and functional
restrictions in the majority of screened patients,
both one month and three months after ICU
discharge. This finding is in line with previous
studies2, 8, 10, 29, 30 and underlines the importance
of a continued chain of care for ICU survivors
who are discharged home.
As expected, our study also demonstrated that
close relatives, such as the partners and children
of ICU survivors, were at risk for PTSD or a high
burden of care, which is in line with other studies that found a high prevalence of anxiety and
depressive conditions.31, 32 Therefore, we recommend the assessment of close relatives for symptoms of stress, anxiety, depression, and care giver
strain as a regular part of the post-ICU clinic.
With the screening instruments used in this
study, problems in different health domains were
identified. Subsequently, patients were referred
to different health care providers based on the
MINERVA ANESTESIOLOGICA
August 2015
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICSDETTLING-IHNENFELDT
available cut-off scores. The validity of these criteria for referral purposes was not assessed in this
study and should be evaluated in future research.
Besides the identification of ICU-related
problems and referral to other professionals,
providing information about the ICU period
and the physical and psychological recovery was
an integral part of the post-ICU clinic. A better understanding of what had happened and
how to recognize and to deal with ICU-related
consequences appears to be very valuable for patients and their close relatives, and contributed
to an overall high satisfaction with the post-ICU
clinic. It is not convincing that the use of different questionnaires has affected this result, because on both instruments the responses ranged
between the highest scores. Nevertheless, the
5-item scale might be more feasible, because 5
different response options are more easily to rate
than 10.
Our study has some limitations that should
be considered in the interpretation of the results.
Due to the small number of patients and the selection of ICU survivors who were discharged to
their homes, our study results might be biased
and not generalizable to other ICU survivors.
Furthermore, a significant proportion of eligible
patients was unable or not willing to participate in follow-up. Developing strategies to offer
post-ICU clinics to all ICU survivors who might
benefit, would be an important improvement in
this. Additionally, an appropriate comparison
between the two post-ICU clinic approaches
was limited, because of the differences in hospital settings. Finally, this study does not provide
long-term outcome data of the patients to control for efficacy and treatment success. In future
work, patient relevant outcome data should be
collected in longitudinal studies to evaluate the
effectiveness of post-ICU clinics and to further
improve the care of ICU survivors after hospital
discharge. Despite the restrictions of this study,
it provides useful information to support the
implementation of post-ICU clinics.
Based on the results of this study and current
developments we recommend the following longitudinal approach to improve the care for ICU
survivors. First of all, a systematic early screening for patients at risk for post ICU physical and
Vol. 81 - No. 8
psychological impairments should be performed
repeatedly during hospital admission. With this,
patients can be referred directly to outpatient
rehabilitation services or be scheduled for a follow-up in a post-ICU clinic. Three months after
ICU discharge, patients and their next relatives
should be invited to administer questionnaires,
to screen for remaining or new ICU-related
problems in daily functioning, and to evaluate
whether the offered care is sufficient or that additional care is needed. The use of computerized questionnaires and tele-medicine applications could be useful to facilitate this process.
With the use of electronic surveys, it would be
possible to assess more patients, also those who
are physically or practically unable to return to
a clinic, or those with avoidant behaviors. We
assume that this could improve the recruitment
of ICU survivors. Furthermore, appropriate use
of technology may also enhance the cost-effectiveness of post-ICU clinics, another concern of
post-ICU clinics. Tele-medicine as an alternative
to face-to-face consultation serves clients, clinicians, and systems by minimizing the barriers
of distance, time, and costs.33, 34 Based on this
“electronic” reassessment, patients and relatives
with symptoms of PICS could be invited for a
post-ICU clinic visit or could be directly referred
for further diagnosis and targeted treatment. Additionally, we suppose that the development of
a network with predefined arrangements with
other allied health and medical specialists will
improve the continuity of care. With such an integrated, stepped care program, the continuum
of care for the transition from the hospital can be
ensured for critically ill patients at risk for PICS.
Conclusions
Post-ICU clinics are important to facilitate
the continuity of care for critically ill patients
who are discharged to their homes. Validated
screening instruments should be used for the
identification of physical and psychological impairments and for referral to medical and allied
health professionals. To increase the proportion
of patients that can take advantage of post-ICU
services, an early in-hospital screening to identify patients at risk for long-term ICU-related
MINERVA ANESTESIOLOGICA
873
DETTLING-IHNENFELDT
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICS
sequelae would be a more targeted approach. In
addition, future research should investigate the
feasibility of computerized screening and telemedicine interventions to improve the benefit of
post-ICU clinics.
Key messages
—— The pragmatic evaluation of post-ICU
clinics in two different settings provides useful information to support the implementation of post-ICU clinics.
—— A post-ICU clinic is feasible for the
identification of patients with ICU-related
sequelae, and to refer these patients for further diagnosis and treatment.
—— Early in-hospital risk stratification,
followed by a post-ICU clinic evaluation 3
months after ICU discharge might be a beneficial approach to improve the continuum
of targeted care after critical illness.
References
  1. Cuthbertson BH, Roughton S, Jenkinson D, Maclennan
G, Vale L. Quality of life in the five years after intensive
care: a cohort study. Crit Care 2010;14:R6.
 2. Desai SV, Law TJ, Needham DM. Long-term complications of critical care. Crit Care Med 2011;39:371-9.
  3. Griffiths RD, Hall JB. Intensive care unit-acquired weakness. Crit Care Med 2010;38:779-87.
  4. Herridge MS, Tansey CM, Matte A, Tomlinson G, DiazGranados N, Cooper A et al. Functional disability 5 years
after acute respiratory distress syndrome. N Engl J Med
2011;364:1293-304.
  5. van der Schaaf M, Beelen A, Dongelmans DA, Vroom MB,
Nollet F. Poor functional recovery after a critical illness: a
longitudinal study. J Rehabil Med 2009;41:1041-8.
  6. Needham DM, Davidson J, Cohen H, Hopkins RO, Weinert C, Wunsch H et al. Improving long-term outcomes after
discharge from intensive care unit: report from a stakeholders’ conference. Crit Care Med 2012;40:502-9.
 7. Griffiths J, Hatch RA, Bishop J, Morgan K, Jenkinson
C, Cuthbertson BH et al. An exploration of social and
economic outcome and associated health-related quality
of life after critical illness in general intensive care unit survivors: a 12-month follow-up study. Crit Care 2013;17:
R100.
  8. Cuthbertson BH. The PRaCTICal study of nurse led, intensive care follow-up programmes for improving long term
oucomes from critical illness: a pragmatic randomised controlled trial. Br Med J 2009;339:b3723.
  9. Schandl A.R. Screening and treatment of problems after intensive care: A descriptive study of mutidisciplinary followup. Intensive Crit Care Nursing 2011;27:94-101.
10. Williams TA, Leslie GD. Beyond the walls: a review of ICU
clinics and their impact on patient outcomes after leaving
hospital. Aust Crit Care 2008;21:6-17.
11. NICE. Rehabilitation after critical illness NICE Clinical
874
Guideline 83. 2009. London, UK: National Institute for
Health and Clinical Excellence. Ref Type: Report.
12. Aaronson NK, Muller M, Cohen PD, Essink-Bot ML, Fekkes M, Sanderman R et al. Translation, validation, and norming of the Dutch language version of the SF-36 Health
Survey in community and chronic disease populations. J
Clin Epidemiol 1998;51:1055-68.
13. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health
Survey Manual and Interpretation Guide. Boston, MA:
New England Medical Center, The Health Institute; 1993.
14. Myhren H, Ekeberg O, Toien K, Karlsson S, Stokland O.
Posttraumatic stress, anxiety and depression symptoms in
patients during the first year post intensive care unit discharge. Crit Care 2010;14:R14.
15. Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens
AE, Van Hemert AM. A validation study of the Hospital
Anxiety and Depression Scale (HADS) in different groups
of Dutch subjects. Psychol Med 1997;27:363-70.
16. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361-70.
17. Crawford JR, Henry JD, Crombie C, Taylor EP. Normative
data for the HADS from a large non-clinical sample. Br J
Clin Psychol 2001;40(Pt 4):429-34.
18. Brewin CR, Rose S, Andrews B, Green J, Tata P, McEvedy
C et al. Brief screening instrument for post-traumatic stress
disorder. Br J Psychiatry 2002;181:158-62.
19. Dekkers AM, Olff M, Naring GW. Identifying persons at
risk for PTSD after trauma with TSQ in the Netherlands.
Community Ment Health J 2010;46:20-5.
20. Robinson BC. Validation of a Caregiver Strain Index. J Gerontol 1983;38:344-8.
21. World Health Organization. The International Classification of Functioning, Disability and Health (ICF). 2001.
Geneva. Ref Type: Report.
22. Elia M. Screening for Malnutrition: A Multidisciplinary
Responsibility. Development and Use of the “Malnutrition
Universal Screening Tool” (“MUST”) for Adults. 2013. Ref
Type: Report
23. Stratton RJ, Hackston A, Longmore D, Dixon R, Price S,
Stroud M et al. Malnutrition in hospital outpatients and
inpatients: prevalence, concurrent validity and ease of use
of the ‘malnutrition universal screening tool’ (‘MUST’) for
adults. Br J Nutr 2004;92:799-808.
24. Cutler L, Brightmore K, Colqhoun V, Dunstan J, Gay M.
Developing and evaluating critical care follow-up. Nurs
Crit Care 2003;8:116-25.
25. Tan T, Brett SJ, Stokes T. Rehabilitation after critical illness:
summary of NICE guidance. Br Med J 2009;338:b822.
26. Connolly B, Denehy L, Brett S, Elliott D, Hart N. Exercise
rehabilitation following hospital discharge in survivors of
critical illness: an integrative review. Crit Care 2012;16:226.
27. Deacon K.S. Re-building life after ICU: A qualitative study
of the patients’ perspective. Intensive Crit Care Nursing
2012;28:114-22.
28. Prinjha S, Field K, Rowan K. What patients think about
ICU follow-up services: a qualitative study. Crit Care
2009;13:R46.
29. Jones C, Griffiths RD, Slater T, Benjamin KS, Wilson S.
Significant cognitive dysfunction in non-delirious patients
identified during and persisting following critical illness.
Intensive Care Med 2006;32:923-6.
30. Iwashyna TJ, Ely EW, Smith DM, Langa KM. Long-term
cognitive impairment and functional disability among survivors of severe sepsis. JAMA 2010;304:1787-94.
31. Paul F, Rattray J. Short- and long-term impact of critical illness on relatives: literature review. J Adv Nurs 2008;62:27692.
32. Davidson JE, Jones C, Bienvenu OJ. Family response to
critical illness: postintensive care syndrome-family. Crit
Care Med 2012;40:618-24.
33. Brennan DM, Mawson S, Brownsell S. Telerehabilitation:
MINERVA ANESTESIOLOGICA
August 2015
FEASIBILITY OF POST-INTENSIVE CARE UNIT CLINICSDETTLING-IHNENFELDT
enabling the remote delivery of healthcare, rehabilitation, and self management. Stud Health Technol Inform
2009;145:231-48.
34. McCue M, Fairman A, Pramuka M. Enhancing quality of
life through telerehabilitation. Phys Med Rehabil Clin N
Am 2010;21:195-205.
Funding.—The study in the AMC is financially supported by the Academic Medical Center in Amsterdam, the Netherlands. In TG the
study is funded by the health insurance company AGIS and the Intensive Care Unit Tergooi. The study sponsors had no involvement in
the execution of the study, the analysis of the data or in writing the paper.
Acknowledgements.—We wish to thank Annemarie Heining-Korteweg, Hanneke Oonk, Edwin de Jong, Margreet Sanders, and Marja
Vilijn (Tergooi Hilversum), and Dennis Gommers and Mirjam van Groen (AMC) for their assistance with the implementation of the
post-ICU clinic and with the data collection.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on March 20, 2014. - Accepted for publication on October 27, 2014. - Epub ahead of print on October 30, 2014.
Corresponding author: D. Dettling-Ihnenfeldt, Department of Rehabilitation, Academic Medical Center, PO Box 22660, 1100 DD
Amsterdam, The Netherlands. Room: A01-348. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
875

O R I G I N A L A RT I C L E
Assessment of cerebral oxygenation
in neurocritical care patients:
comparison of a new four wavelengths
forehead regional saturation in oxygen sensor
(EQUANOX®) with brain tissue oxygenation.
A prospective observational study
P. ESNAULT 1, H. BORET 1, A. MONTCRIOL 1, E. CARRE 2, B. PRUNET 1, J. BORDES 1
P. SIMON 1, C. JOUBERT 3, A. DAGAIN 3, E. KAISER 1, E. MEAUDRE 1
1Intensive
Care Unit, Sainte Anne Military Teaching Hospital, Toulon, France; 2Military Biomedical Research Institute,
Brétigny sur Orge, France; 3Department of Neurosurgery, Sainte Anne Military Teaching Hospital, Toulon, France
ABSTRACT
Background. Because of restricted information given by monitoring solely intracranial pressure and cerebral perfusion pressure, assessment of the cerebral oxygenation in neurocritical care patients would be of interest. The aim
of this study was to determinate the correlation between the non-invasive measure regional saturation in oxygen
(rSO2) with a third generation NIRS monitor and an invasive measure of brain tissue oxygenation tension (PbtO2).
Methods. We conducted a prospective, observational, unblinded study including neurocritical care patients requiring a PbtO2 monitoring. Concomitant measurements of rSO2 were performed with a four wavelengths forehead
sensor (EQUANOX Advance®) of the EQUANOX® 7600 System. We determined the correlation between rSO2
and PbtO2 and the ability of the rSO2 to detect ischemic episodes defined by a PbtO2 less than 15 mmHg. The rSO2
ischemic threshold was 60%.
Results. During 2 months, 8 consecutives patients, including 275 measurements, were studied. There was no correlation between rSO2 and PbtO2 (r=0.016 [-0.103-0.134], r2=0.0003, P=0.8). On the 86 ischemic episodes detected
by PbtO2, only 13 were also detected by rSO2. ROC curve showed the inability for rSO2 to detect cerebral hypoxia
episodes (AUC=0.54).
Conclusion. rSO2 cannot be used as a substitute for PbtO2 to monitor cerebral oxygenation in neurocritical care
patients. (Minerva Anestesiol 2015;81:876-84)
Key words: Brain - Cell respiration - Hypoxia, brain - Spectroscopy, near-infrared
I
t has been established that monitoring of cerebral perfusion pressure (CPP) is the cornerstone parameters in brain-injured patients in
neurocritical care unit.1. However, intracranial
pressure (ICP) and CPP monitoring alone are
Comment in p. 835.
876
unable to indicate whether the chosen threshold is efficient to maintain an optimal cerebral
oxygenation.2 Moreover, deleterious hypocapnia
remains undetected by the sole measurement of
CPP. For these reasons, advanced cerebral monitoring has been used increasingly to measure
brain oxygenation and/or cerebral blood flow
MINERVA ANESTESIOLOGICA
August 2015
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTSESNAULT
(CBF).2 Some monitors are invasive (brain tissue
oxygen tension [PbtO2]) catheter or jugular bulb
venous oxygen saturation [SjvO2]), and others are
non-invasive as the regional saturation in oxygen
(rSO2) measured with near-infrared spectroscopy
(NIRS). NIRS is a recent attractive monitor, easy
to use, accessible anywhere and provides continuous real-time assessment of cerebral oxygenation.
However, a recent study suggested that rSO2
measured with the INVOS 5100® (NIRS monitor, Somanetics INVOS® system, Somanetics
Inc., MI, USA) is poorly correlated with PbtO2
and did not represent a substitute for PbtO2 values after traumatic brain injury (TBI).3 This system only used two wavelengths of near-infrared
light to measure rSO2. The NONIN Medical®
society has developed a new sensor with 4 wavelengths and 2 emitters (EQUANOX Advance®,
Nonin Medical®, Inc, Plymouth, MN, USA) in
order to determine more accurately cerebral saturation. The main objective of this study was to
compare this third generation NIRS monitoring
with measure of cerebral oxygenation by PbtO2
in brain-injured patients.
Material and methods
Study design and patients selection
After approval of the institutional review
board, we conducted a prospective, observational, single-center study at the Sainte Anne
Military Teaching Hospital of Toulon (France).
Between November 2011 and January 2012,
brain-injured patients admitted to the neurocritical care unit and requiring advanced multimodal
neuromonitoring (ICP and PbtO2 at least) were
eligible. Inclusion criteria were patients with age
over 18 and a Glasgow Coma Scale (GCS) Score
of ≤9 on admission with severe injury requiring
multimodal monitoring. Patients suffering from
forehead or skull lesions, and those with deficient signals measured by NIRS or PbtO2 were
excluded from the study. Informed consent was
obtained from patient’s family and relatives. Intracranial pressure was monitored continuously
by using an intraparenchymal sensor (Codman®,
Randolph, MA, USA). Patients were sedated, intubated and mechanically ventilated in accord-
Vol. 81 - No. 8
ance with the international guidelines and with
the following objectives: head elevation (30°),
PaO2>85 mmmHg, PaCO2 between 35 mmHg
and 45 mmHg, natremia between 140 mEq/L
and 145 mEq/L, targeted temperature <37.5
°C, ICP <20 mmHg, and CPP>60 mmHg initially then threshold adapted daily to PbtO2. If
ICP increased, the protocol included moderate
hypocapnia (PaCO2 between 30 mmHg and
35 mmHg), increase of sedation with propofol,
muscle paralysis with cisatracurium and moderate hypothermia (33-35 °C). Third line treatment included barbiturates and/or unilateral
craniectomy.1, 4-8
Regional saturation in oxigen monitoring (rSO2)
rSO2 was measured using the EQUANOX®
7600 monitor (Nonin Medical®, Inc, Plymouth, MN, USA) associated with its specific
sensor (EQUANOX Advance®) placed on the
skin in the frontal region. The sensor was placed
ipsilaterally to PbtO2 probe in all patients. The
EQUANOX® system uses NIRS technology to
allow estimation of grey matter oxygen. rSO2 is
a composite value based on the expected relative proportion of arterial (30%), capillary and
venous (70%) blood within brain tissue. The
EQUANOX® uses dual emitters alternately creating pairs of reflected light paths through surface tissue to the shallow receiver and through
the cerebral cortex to the far receiver. The algorithm is supposed to remove surface effects
that modulate light amplitude, resulting in a
measurement of rSO2 of cortical grey matter in
the anterior and middle cerebral arteries territories. The EQUANOX Advance® sensor uses 4
wavelengths ranging from 730 nm to 880 nm
to measure equilibrium between oxy- and desoxyhemoglobin. The manufacturer claims that
doubling of the LED light sources allows the
device to compare rSO2 measurements in adjacent tissues to confirm accuracy of a measurement. A recent study confirmed this hypothesis.
The EQUANOX® system demonstrated less
degree of extracranial contamination than the
INVOS® system.9 The ischemic threshold from
manufacturer and given by some studies is under 60%.10
MINERVA ANESTESIOLOGICA
877
ESNAULT
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTS
The EQUANOX® system did not play a role in
the medical management of the patients.
Brain tissue oxygen tension monitoring (PbtO2)
A PbtO2 probe (LICOX®, Integra Neurosciences, USA) was inserted into the frontal lobe
of the most injured hemisphere through the same
screw of ICP probe (Bolt system, IM2, Integra
Neurosciences, USA). The correct location of the
monitor was confirmed by CT-scan and manipulation of oxygen inspired fraction (FiO2). As recommended, values were taken into account only
after an initial in vivo equilibration period of 2
hours. In accordance with preliminary clinical
studies, the ischemic threshold is 15 mmHg and
below.11, 12
Data collected
Patient characteristics, nature of brain injury,
CT-scan classifications according to the National
Institutes of Health Traumatic Coma Data Bank
(TCDB), surgical procedures and Glasgow Outcome Scale (GOS) were collected.13, 14
Every 3 hours during all the monitoring
phase, the following variables were simultaneously recorded: rSO2 (%), PbtO2 (mmHg),
CPP (mmHg), ICP (mmHg), End-tidal CO2
(EtCO2) (mmHg), SpO2, FiO2 and bladder
temperature (°C). Temperature is particularly
relevant because, at identical partial pressure in
oxygen, hemoglobin saturation is physiologically
increased by hypothermia.15 All PbtO2 and rSO2
values were adjusted to patient bladder temperature. PbtO2 values were adjusted by means of a
computer (LICOX® system). rSO2 values were
adjusted a posteriori with equations given elsewhere.15 PbtO2 and rSO2 values were paired
measurements simultaneously obtained from the
software of the both monitors.
Endpoints
The primary endpoint of this study was to assess a correlation between rSO2 and PbtO2. Secondary endpoints were to assess a correlation between rSO2 and the others data recorded (CPP,
ICP, SpO2, FiO2 and EtCO2), and to establish
the ability of rSO2 to detect ischemic events.
878
Statistical analysis
Statistical analysis was performed with SPSS
version 15.0 (SPSS Inc., Chicago, IL, USA). Continuous data were reported as the mean±standard
deviation or median with interquartile ranges
(25-75th)] when not normally distributed. Nominal variables are reported as numbers and proportions (%). A univariate analysis was conducted
using the χ2 test or Fisher’s exact test to compare
categorical variables and the Mann-Whitney test
or Student’s t-tests to compare groups for continuous variables (respectively for comparison of
medians and comparison of means). The correlation between rSO2 and relevant parameters were
evaluated by Pearson correlation test. A receiveroperator characteristic (ROC) curve analysis was
performed to evaluate the ability of ipsilateral
rSO2 to detect ischemic events. We determined 3
levels of cerebral hypoxia: moderate (defined by
a PbtO2≤15 mmHg), severe (PbtO2≤10 mmHg)
and critical (PbtO2≤5 mmHg). The sensitivity,
specificity, predictive values of rSO2 thresholds
were determined for each level of brain hypoxia.
For all tests, P<0.05 was considered statistically
significant.
Results
Patient characteristics and brain hypoxia episodes
We included consecutively 8 brain-injured
patients (5 male and 3 female). No patients were
excluded. Patient characteristics are presented
in Table I. The mean age was 49.6±15.6 years,
the median Simplified Acute Physiology Score
(SAPS II) was 42,34-62 and the median GCS on
admission was 6 (range 3-9). The median patient
temperature during monitoring was 35.1 (34.536) °C. The type of surgical intervention (if applicable) and the location of monitors are detailed in Table I. The brain tissue oxygen probe
was placed in the most-injured hemisphere. In
this cohort, no complication was attributed to
the PbtO2 probe (no intracranial hemorrhage or
infection). The PbtO2 probe of patient number
4 was placed ipsilaterally to the internal carotid
artery dissection. The PbtO2 probe for patient
number 7 was inserted in the edema surround-
MINERVA ANESTESIOLOGICA
August 2015
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTSESNAULT
GCS on
admission
SAH
8
2
M
66
73
TBI
3
3
4
M
F
39
44
5
6
M
F
51
53
34
TBI
42 Right ICA
dissection
73
TBI
34
TBI
3
6
7
8
M
F
69
20
58
34
5
6
ICH
TBI
6
9
Fisher IV
Median frontal lobe hematoma
Right frontal lobe SDH
Left parieto-temporal lobes petechia
Necrosis of the right fronto-parietotemporal lobes (ACA and MCA territories).
Right temporal lobe hematoma (10x4 cm)
Left temporal lobe petechia
Right temporal lobe contusion
Left frontal lobe hematoma (4x5 cm)
Right basal nuclei contusion
5
3
5
3
2
Licox monitor
side
Equanox monitor
side
GOS before
departure of ICU
Pathology
41
ICP monitor side
SAPS II
55
Neurosurgery
Age
M
TCDB score
Sex
1
CT-scan
Patient
Table I.—Patient characteristics.
EVD
Rt
Rt
Rt
4
Craniotomy and
Lt
craniectomy
No
Lt
Craniectomy (day 2) Rt
Lt
Lt
1
Lt
Rt
Lt
Rt
3
1
Craniectomy
Lt
Craniectomy (day 6) Rt
Lt
Rt
Lt
Rt
1
1
Rt
Rt
Rt
Rt
1
5
Craniotomy
No
Rt
Rt
SAH: subarachnoid hemorrhage. ICA: internal carotid artery. ACA: anterior cerebral artery. MCA: middle cerebral artery. ICH: intracerebral
hemorrhage. ER: emergency room. GCS: Glasgow Coma Scale. GOS: Glasgow Outcome Scale (1: death, 2: vegetative state, 3: severely disable, 4:
moderately disable, 5: no disability). ICU: Intensive Care Unit. SAPS: simplified acute physiology score. EVD: external ventricular drainage. SDH:
subdural hematoma. ICP: intracranial pressure. TBI: traumatic brain injury. Rt: right. Lt: left.
ing the evacuated hematoma. The PbtO2 probe
of the others patient was placed in normal appearing tissue.
Two hundred and seventy five EQUANOX®
measurements were recorded. The median duration of NIRS monitoring was 4 days (3.5-5.5
days) and the median measurement per patient
was 30 (21-46). The median signal quality was
100% (range 75-100). No complications as skin
irritation due to the EQUANOX® sensor were
observed.
During the study, 86 moderate cerebral hypoxia episodes were detected by the PbtO2
probe. The EQUANOX® system was able to
detect 74 moderate cerebral hypoxia episodes.
EQUANOX® and PbtO2 detected only 13 moderate cerebral hypoxia episodes simultaneously.
Patients with good outcomes (GOS=3, 4, and
5) had significantly better PbtO2 values than patients with bad outcomes (GOS=1, and 2): 24.2
(20.6-29.7) vs. 17.4 (12.7-24.5) (P<0.0001).
rSO2 values were also significantly better in patients with good outcomes: 78 (73-81) vs. 65
(55-71) (P<0.0001).
Correlation between rSO2 and PbtO2
Vol. 81 - No. 8
No correlation was found between the rSO2
and the PbtO2 (r=0.016 [-0.103 – 0,134],
r2=0.0003, P=0.8) (Figure 1). Figure 2 showed
the care of the patient number 4. When the patient evolved into brain death with a PbtO2 falling to 0 mmHg, rSO2 was still showing normal
values with no variation.
Concerning the temporal correlation of rSO2
and the PbtO2 changes over time, the values followed the same trends in only 58.9% of cases.
Correlation between rSO2 and others variables
rSO2 values were also indirectly related to ICP
(r=-0.54 [-0.62 – -0.45], r2=0.29, P<0.0001).
rSO2 values were weakly directly correlated to
CPP (r=0.21 [0.09-0.32], r2=0.05, P=0.0004).
Correlations with the others data recorded are
reported in Table II.
Detection of cerebral hypoxia with rSO2
The ROC curve demonstrated that rSO2 had
low accuracy to detect moderate cerebral hypoxia (PbtO2≤15 mmHg) with an area under curve
(AUC) with 95% confidence interval (95%
CI) of 0.54 (0.43-0.65) (Figure 3). The recom-
MINERVA ANESTESIOLOGICA
879
ESNAULT
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTS
Figure 1.—Graphic showing the absence of correlation between PbtO2 and rSO2 (pooled values).
Figure 2.—Trends of rSO2 and PbtO2 for patient n°4. Data showing the inability of rSO2 to detect ongoing brain death.
Table II.—Correlation between rSO2 and physiological parameters values
Parameter
SpO2
FiO2
ICP
CPP
EtCO2
880
95% CI
Coefficient
of correlation
Lower Bound
Upper Bound
0.1
- 0.17
- 0.54
0.21
0.21
- 0.02
- 0.28
- 0.62
0.09
0.09
0.22
- 0.05
- 0.45
0.32
0.32
MINERVA ANESTESIOLOGICA
P Value
0.1
0.005
< 0.001
< 0.001
< 0.001
August 2015
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTSESNAULT
Figure 3.—ROC curve constructed with 5 rSO2 thresholds (50%, 55%, 60%, 65% and 70%) to detect moderate cerebral hypoxia
(PbtO2≤15). AUC=0.54.
Table III.—Sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values of the optimal rSO2 threshold.
PbtO2<15 mmHg
rSO2<70%
rSO2<65%
rSO2<60%
rSO2<55%
rSO2<50%
PbtO2<10 mmHg
rSO2<70%
rSO2<65%
rSO2<60%
rSO2<55%
rSO2<50%
PbtO2<5 mmHg
rSO2<70%
rSO2<65%
rSO2<60%
rSO2<55%
rSO2<50%
Sensitivity
Specificity
PPV
NPV
LR +
LR -
78.2 (68.4 – 85.6)
39.1 (29.5-49.6)
14.9 (8.9-23.9)
10.3 (5.5-18.5)
4.6 (1.8-11.2)
49.5 (42.4-56.6)
59 (51.9-65.8)
67.6 (60.6-73.8)
75.5 (68.9-81.1)
90.4 (85.4-93.9)
41.7 (34.4-49.4)
30.6 (22.8-39.7)
17.6 (10.6-27.8)
16.4 (8.9-28.3)
18.2 (7.3-38.5)
83 (75-88.9)
67.7 (60.2-74.4)
63.2 (56.3-69.6)
64.6 (58-70.6)
67.2 (61.2-72.7)
1.55
0.95
0.46
0.42
0.48
0.44
1.03
1.26
1.19
1.06
93.6 (82.8-97.8)
46.8 (33.3-60.8)
24.4 (14.2-38.7)
20 (10.9-33.8)
8.9 (3.5-20.7)
47.8 (41.4-54.3)
63.2 (56.7-69.2)
72.6 (66.5-78)
80 (74.4-84.7)
92.2 (88-95)
27 (20.8-34.3)
20.8 (14.1-29.4)
14.9 (8.5-24.7)
16.4 (8.9-28.3)
18.2 (7.3-38.5)
97.3 (92.4-99.1)
85.2 (79.1-89.8)
83.1 (77.3-87.6)
83.6 (78.2-87.9)
83.8 (78.8-87.8)
1.79
1.27
0.89
1
1.14
0.13
0.84
1.04
1
0.99
100 (90.6-100)
51.4 (36-66.6)
27 (15.4-43)
24.3 (13.4-40.1)
10.8 (4.3-24.7)
47.1 (40.8-53.4)
61.3 (55- 67.3)
73.1 (67.1-78.3)
80.7 (75.2-85.2)
92.4 (88.4-95.2)
22.7 (16.9-29.7)
17.1 (11.2-25.2)
13.5 (7.5-23.1)
16.4 (8.9-28.3)
18.2 (7.3-38.5)
100 (96.7-100)
89 (83.3-92.9)
86.6 (81.2-90.6)
87.3 (82.2-91.1)
87 (82.3-90.6)
1.89
1.34
1
1.26
1.43
0
0.79
0.99
0.94
0.97
Values are expressed as percentages with 95% confidence intervals. The likelihood ratios for a positive test (LR+) and negative test (LR-) are also
shown. A value of rSO2<70% was found to be the optimal threshold for prediction a PbtO2<15 mmHg. A value of rSO2<60% was found to be a
better threshold for prediction a PbtO2<10 mmHg.
mended threshold of 60% had only a positive
predictive value (PPV) of 17.6%, a sensitivity of
14.9%, a specificity of 67.6% and a likelihood
ratio for a positive test (LR+) very low of 0.46.
We repeated the calculations for a rSO2 threshold of 50%, 55%, 65% and 70%. We found that
the best rSO2 threshold to predict a PbtO2≤15
mmHg was 70%, with a PPV of 41.7%, a sen-
Vol. 81 - No. 8
sitivity of 78.2% and LR+ of 1.55 (Table III).
By investigating the accuracy of rSO2 for
detecting severe cerebral hypoxia (PbtO2≤10
mmHg), we observed a minimal improvement
in the AUC (0.64 [0.51-0.76], P<0.0001) and
sensitivity (20%), and an LR+ of 0,89 for a
threshold of rSO2<60% (Table III).
The accuracy of rSO2 for detecting critical cer-
MINERVA ANESTESIOLOGICA
881
ESNAULT
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTS
ebral hypoxia (PbtO2≤5 mmHg) was moderately improved with an AUC of 0.67 (0.54-0.78)
(P<0.0001) and sensitivity of 27%, and an LR+
of 1 for a threshold of rSO2<60% (Table III).
Discussion
To our knowledge, this study is the first assessing the validity of a third generation NIRS compared to PbtO2. Our results highlighted an absence of correlation between EQUANOX® and
PbtO2 measurements. This study also showed
the very low ability of rSO2 to detect patients
with moderate cerebral hypoxia. When using a
threshold of 60%, there was a high rate of falsenegative (37%) and false-positive (22%) results,
which could lead to miss true ischemic episode
or treat unnecessary false ischemic episode. Severe and critical cerebral hypoxias were better
detected by EQUANOX® than moderate cerebral hypoxia (AUC=0.64 and 0.67 respectively).
This study confirmed limitations of rSO2
to detect cerebral hypoxia in neurocritical care
patients, as shown previously, in TBI and subarachnoid hemorrhage (SAH) patients.3, 16, 17 In
the study by Naidech et al. using the INVOS
5100® to monitor 6 SAH patients, the authors
were unable to find a value that was associated
with cerebral hypoxia.17 Likewise, information
given by PbtO2 and rSO2 (measured by the
INVOS 5100®) were compared in 22 TBI patients.3 A direct poor correlation between PbtO2
and rSO2 was demonstrated (r=0.179), but the
ability of rSO2 to identify patients with intracerebral hypoxia was moderate (AUC=0.62). Recently, Rosenthal et al. demonstrated that PbtO2
was not significantly related to rSO2 measured
by the CerOx® 3110 monitor (Ornim Medical
Ltd.) (r=0.014), indicating that these two measures of brain oxygenation are not correlated.16
Although both PbtO2 and rSO2 measure cerebral tissue oxygenation, they differ in the type of
data that they provide. PbtO2 reflects the balance
between oxygen delivery and oxygen consumption. PbtO2 is primarily correlated with CBF
(CPP), arterial oxygen tension (FiO2, PaO2),
cerebral oxygen diffusion, and hemoglobin levels.18-21 In contrast, rSO2 depends primarily by
the degree of oxygen extraction determined by
882
the balance between oxygen delivery and utilization.16 In fact, rSO2 is supposed to measure
the hemoglobin saturation in 3 different compartments (venous, arterial, and capillar), but
because cerebral blood volume is largely represented by venous blood, rSO2 may predominantly represent the hemoglobin saturation in
the venous bed.22, 23 Thereby, it is not surprising
that, like in our study, rSO2 is correlated with
FiO2, PaO2, CPP and/or ICP.17, 24-26
Although rSO2 was not correlated to PbtO2,
its use has demonstrated its interest in some
situations, especially in operating room during carotid endarterectomy or pediatric cardiac
surgery.27, 28 In our study, rSO2 is well correlated with ICP, CPP and PaCO2 (through the
EtCO2), three essential parameters in braininjured patients. rSO2 could be interesting in
situations where invasive monitoring of ICP or
PbtO2 is not available: elective surgery, prehospital care, unspecialized ICU or coagulopathy. In
these specific conditions, rSO2 can bring useful
information concerning cerebral oxygenation.
NIRS monitoring might maybe also replace the
SjvO2 monitoring in neurocritical care patients.
In fact, a recent study showed a good correlation between the SjvO2 and the NIRS measure
in 18 TBI patients.16 However, NIRS monitoring is noninvasive, with an excellent signal quality (100% in our study) and devoid of complications. In contrary, the continuous measure
of SjvO2 entails technical difficulties like the
availability of good-quality signal, the need for
frequent recalibration and the risks of vascular
puncture.29
The recommended ischemic threshold by
NONIN® company for rSO2 is 60%. We found
that this threshold was not efficient to detect
moderate cerebral hypoxia and weakly efficient
to detect severe and critical cerebral hypoxia.
We found that the optimal threshold of rSO2
was 70% to detect moderate cerebral hypoxia.
However, this value would have a high sensitivity but a low specificity, making it of little clinical value. Moreover, it is necessary to define an
ischemic threshold according to the patient’s
core temperature when using NIRS with an absolute rSO2 value. Indeed, it is physiologically
established that hemoglobin saturations depends
MINERVA ANESTESIOLOGICA
August 2015
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTSESNAULT
on temperature with a left deviation of the dissociation curve during hypothermia.15 Contrary
to the others study comparing rSO2 and PbtO2
measurements, we performed this correction,
but despite that, the results confirm the low ability of NIRS to detect brain hypoxia.
Study limitations
The present study has several limitations. First,
even though PbtO2 is largely used in neurocritical
care unit, it is not yet a recognized gold-standard
technique for cerebral oxygenation monitoring. It
is however useful in detecting ischemic injury and
has a prognosis value when PbtO2 drops below 10
and 5 mmHg.2 Second, we studied only a limited
number of patients, probably limiting the power of our study and the realization of sub group
analysis. Third, we included different pathologies with different pathophysiology. However, it
is representing the real life, where monitoring is
not only used for a specific condition. Fourth, unlike of other studies, which the PbtO2 probe was
placed into the less injured cerebral hemisphere,
in the present study the probe was placed in the
most injured hemisphere. This could render important differences when comparing the reported
results with those obtained by other authors.3, 16
Fifth, the most of GOS departure of ICU were
1, meaning that most of patient died at ICU or
promptly after. We recognize that this could represent a selected patient population with a poor
prognosis or impaired cerebral autoregulation,
but we believe this impacts both equally PbtO2
and rSO2 values. Finally, we recognize that the
comparison of these monitors is problematic, as
each monitors measures a distinct physiological
parameter: rSO2 is monitoring a volume in the
grey matter, when PbtO2 is monitoring a surface
(13 mm2) in the white matter, with different CBF
in these regions.
Conclusions
Despite limitations, these results emphasize
the low ability of rSO2 to detect cerebral hypoxia
compared to PbtO2. Even using a third generation NIRS monitoring, rSO2 cannot be used a
substitute for PbtO2 after brain injury. However,
Vol. 81 - No. 8
further prospective studies investigating the relationship between invasive and non-invasive assessment of cerebral hypoxia, are needed.
Key messages
—— This study is the first to assess a correlation between a third generation NIRS and
PbtO2.
—— Our results highlighted an absence
of correlation between EQUANOX® and
PbtO2 measurements.
—— This study also showed the very low
ability of rSO2 to detect patients with cerebral hypoxia.
—— rSO2 values were also indirectly related
to ICP and weakly directly correlated to CPP.
References
  1. Bratton S, Chestnut R, Ghajar J, McConnell Hammond
F, Harris O, Hartl R et al. Guidelines for the management
of severe traumatic brain injury. Cerebral perfusion thresholds. J Neurotrauma 2007;24:59-64.
  2. Andrews PJD, Citerio G, Longhi L, Polderman K, Sahuquillo J, Vajkoczy P, Neuro-Intensive Care and Emergency
Medicine (NICEM) Section of the European Society of Intensive Care Medicine (2008) NICEM consensus on neurological monitoring in acute neurological disease. Int Care
Med 34:1362-70.
 3.Leal-Noval SR, Cayuela A, Arellano-Orden V, MarínCaballos A, Padilla V, Ferrándiz-Millón C et al. Invasive
and noninvasive assessment of cerebral oxygenation in
patients with severe traumatic brain injury. Int Care Med
2010;36:1309-17.
  4. Bratton S, Chestnut R, Ghajar J, McConnell Hammond F,
Harris O, Hartl R et al. Guidelines for the management of
severe traumatic brain injury. Blood pressure and oxygenation. J Neurotrauma 2007;24:7-13.
  5. Bratton S, Chestnut R, Ghajar J, McConnell Hammond
F, Harris O, Hartl R et al. Guidelines for the management
of severe traumatic brain injury. Hyperosmolar therapy. J
Neurotrauma 2007;24:14-20.
  6. Bratton S, Chestnut R, Ghajar J, McConnell Hammond F,
Harris O, Hartl R et al. Guidelines for the management of
severe traumatic brain injury. Intracranial pressure thresholds. J Neurotrauma 2007;24:55-8.
  7. Bratton S, Chestnut R, Ghajar J, McConnell Hammond F,
Harris O, Hartl R et al. Guidelines for the management of
severe traumatic brain injury. Brain oxygen monitoring and
thresholds. J Neurotrauma 2007;24:65-70.
  8. Bratton S, Chestnut R, Ghajar J, McConnell Hammond F,
Harris O, Hartl R et al. Guidelines for the management of
severe traumatic brain injury. Hyperventilation. J Neurotrauma 2007;24:87-90.
  9. Davie SN, Grocott HP. Impact of extracranial contamination on regional cerebral oxygen saturation. A comparison
of three cerebral oxymetry technologies. Anesthesiology
2012;116:834-40.
10. Tobias JD. Cerebral oxygenation monitoring: near-infrared
spectroscopy. Expert Rev Med Dev 2006;3:235-43.
MINERVA ANESTESIOLOGICA
883
ESNAULT
ASSESSMENT OF CEREBRAL OXYGENATION IN NEUROCRITICAL CARE PATIENTS
11. Valadka AB, Gopinath SP, Contant CF, Uzura M, Robertson CS. Relationship of brain tissue pO2 to outcome after
severe head injury. Crit Care Med 1698;26:1576-81.
12.Van Santbrink H, Maas AIR, Avezaat CJJ. Continuous
monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery 1996;38:2132.
13. Marshall LF, Marshall SB, Klauber MR, van Berkum Clark
M, Eisenberg HM, Jane JA et al. A new classification of
head injury based on computerized tomography. J Neurosurg 1991;75:S14-20.
14. Jennett B, Bond M. Assessment of outcome after severe
brain damage. Lancet 1975;1:480-4.
15. Tremey B, Vigué B. Changes in blood gases with temperature: implications for clinical practice. Ann Fr Anesth Rea
2004;23:474-81.
16. Rosenthal G, Furmanov A, Itshayek E, Shoshan Y, Singh V.
Assessment of a noninvasive cerebral oxygenation monitor
in patients with severe traumatic brain injury. J Neurosurg
2014;120:901-7.
17. Naidech AM, Bendok BR, Ault ML, Bleck TP. Monitoring with the Somanetics INVOS 5100C After Aneurysmal
Subarachnoid Hemorrhage. Neurocrit Care 2008;9:32631.
18. Narotam PK, Morrison JF, Nathoo N. Brain tissue monitoring in traumatic brain injury and major trauma: outcome analysis of a brain tissue oxygen-directed therapy. J
Neurosurg 2009;111:672-82.
19. Rose JC, Neill TA, Hemphill JC. Continuous monitoring
of the microcirculation in neurocritical care: an update of
brain tissue oxygenation. Curr Opin Crit Care 2003;12:97102.
20. Rosenthal G, Hemphill JC III, Sorani M, Martin C, Morabito D, Obrist WD et al. Brain tissue tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care
Med 2008;36:1917-24.
21.Leal-Noval SR, Rincon-Ferrari MD, Marin-Niebla A,
Cayuela A, Arellano-Orden V, Marin-Caballos A et al.
Transfusion of erythrocyte concentrates produces a variable
increment on cerebral oxygenation in patients with severe
traumatic brain injury: a preliminary study. Intensive Care
Med 2006;32:1733-40.
22. Smith M, Elwell C. Near-infrared spectroscopy: shedding
light on the injured brain. Anesth Analg 2009;108:10557.
23.Watzman HM, Kurth CD, Montenegro LM, Rome J,
Steven JM, Nicolson SC. Arterial and venous contributions to near-infrared cerebral oximetry. Anesthesiology
2000;93:947-53.
24. Stoneham MD, Lodi O, de Beer TC, Sear JW. Increased
oxygen administration improves cerebral oxygenation in
patients undergoing awake carotid surgery. Anesth Analg
2008;107:1670-5.
25. Skhirtladze K, Birkenberg B, Mora B, Moritz A, Ince I, Ankersmit HJ et al. Cerebral desaturation during cardiac arrest: its relation to arrest duration and left ventricular pump
function. Crit Care Med 2009;37:471-5.
26. Matsumoto S, Nakahara I, Higashi T, Iwamuro Y, Watanabe Y, Takahashi K et al. Near-infrared spectroscopy in
carotid artery stenting predicts cerebral hyperperfusion syndrome. Neurology 2009;28:1512-8.
27. Samra SK, Dy EA, Welch K, Dorje P, Zelenock GB, Stanley
JC. Evaluation of a cerebral oximeter as a monitor of cerebral ischemia during carotid endarterectomy. Anesthesiology 2000;93:964-70.
28. Zulueta JL, Vida VL, Perisinotto E, Pittarello D, Stellin
G. Role of intraoperative regional oxygen saturation using
near infrared spectroscopy in the prediction of low output syndrome after pediatric heart surgery. J Cardio Surg
2013;28:446-52.
29.Kiening KL, Unterberg AW, Bardt TF, Schneider GH,
Lanksch WR. Monitoring of cerebral oxygenation in patients with severe head injuries: brain tissue PO2 versus
jugular vein oxygen saturation. J Neurosurg 1996;85:7517.
PE, HB, EC contributed to the study concept and design. PE, HB, AM, JB, CJ, and AD contributed to the acquisition of data. PE, HB,
BP, and PS contributed to the analysis and interpretation of data. PE, HB, BP, EK, and EM contributed to the drafting of manuscript and
critically revising the manuscript for important intellectual content.
Acknowledgments.—The authors thank Mrs. Tina Tari for the language revision.
Funding.—This study was supported solely by internal departmental funds. NONIN™ loaned the EQUANOX device and the sensors to
the ICU while it was being evaluated. No other financial support, honorary fees, research personnel, consulting fees or statistical support
was provided. The authors collected the data in a database of their own design and maintained control at all times. They were responsible
for analyzing the data as well as the decision to submit the data for publication.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on September 21, 2014. - Accepted for publication on November 19, 2014. - Epub ahead of print on November 21, 2014.
Corresponding author: P. Esnault, Service de Réanimation, Brûlés, Hôpital d’Instruction des Armées Sainte Anne, Boulevard Sainte Anne,
83000 Toulon, France. E-mail: [email protected]
884
MINERVA ANESTESIOLOGICA
August 2015

O R I G I N A L A RT I C L E
Moderate-degree acidosis
is an independent determinant
of postoperative bleeding in cardiac surgery
M. RANUCCI 1, E. BARYSHNIKOVA 1, F. SIMEONE 2, M. RANUCCI 1, S. SCOLLETTA 2
1Department of Cardiothoracic and Vascular Anesthesia and ICU, IRCCS Policlinico San Donato, San Donato Milanese,
Milan, Italy; 2Department of Medical Biotechnologies, Anesthesia and Intensive Care, University of Siena, Siena, Italy
ABSTRACT
Background. Acidosis is a well-known factor leading to coagulopathy. It has been widely explored as a risk factor for
severe bleeding in trauma patients. However, no information with respect to acidosis as a determinant of postoperative bleeding in cardiac surgery patients exists. The aim of this study was to investigate the role of acidosis and hyperlactatemia (HL) in determining postoperative bleeding and need for surgical revision in cardiac surgery patients.
Methods. We carried out a retrospective analysis on 4521 patients receiving cardiac operations in two institutions.
For each patient the preoperative data and operative profile was available. Arterial blood gas analysis data at the arrival in the intensive care unit were analyzed to investigate the association between acidosis (pH<7.35), HL (>4.0
mMol/L) and postoperative bleeding and surgical revision rate.
Results. After correction for the potential confounders, both acidosis (P=0.001) and HL (P=0.001) were significantly associated with the amount of postoperative bleeding. HL was an independent risk factor for postoperative
bleeding even in absence of acidosis. Overall, surgical revision rate was 5.6% in patients with HL and no acidosis;
7.7% in patients with acidosis and HL, and 7.2% in patients with acidosis and no HL. All these values are significantly (P=0.001) higher than the ones in patients without acidosis/HL (2%).
Conclusions. Even a moderate degree of postoperative acidosis is associated with a greater postoperative bleeding
and surgical revision rate in cardiac surgery patients. Correction of acidosis with bicarbonate does not lead to an
improvement of the postoperative bleeding asset. (Minerva Anestesiol 2015;81:885-93)
Key words: Cardiac surgical procedures - Hemorrhage - Acidosis - Lactates.
B
lood acidosis is a recognized factor underlying coagulopathy and bleeding in different conditions. The main clinical environment
where a low blood pH has been found as a determinant of severe bleeding is the traumatic
coagulopathy, where coagulopathy, acidosis, and
hypothermia concur in increasing the patients’
risk of ongoing bleeding and death.1-4
Cardiac surgery-associated coagulopathy in
patients undergoing cardiac surgery with cardiopulmonary bypass (CPB) is a complex syndrome, including dilution/consumption of coagulation factors, thrombocytopenia, platelet
Vol. 81 - No. 8
dysfunction, decreased fibrinogen levels, hyperfibrinolysis, and residual heparin/protamine effects.5
During cardiac operations, different mechanisms may lead to acidosis: ischemia-reperfusion
of the heart and lung during CPB; cooling-rewarming associated ischemia-reperfusion; inadequate oxygen delivery during CPB; low cardiac
output before and after CPB. All these factors
are determinants of metabolic acidosis, whose
marker is an increased lactate (LAC) formation. As a matter of fact, hyperlactatemia (HL)
is common during CPB and after cardiac sur-
MINERVA ANESTESIOLOGICA
885
RANUCCI
ACIDOSIS AND BLEEDING IN CARDIAC SURGERY
gery,6-8 and is the most typical mechanism leading to bicarbonate buffer consumption and metabolic acidosis. Conversely, considering that all
the patients are mechanically ventilated, other
forms of acidosis (respiratory acidosis) are rare
in cardiac surgery patients during and early after
surgery.
Despite the relatively common finding of HL
and metabolic acidosis immediately after cardiac surgery, and the equally common finding of
postoperative bleeding in the first 12 hours after
surgery, no reports have investigated the possible
role of acidosis as a determinant of postoperative
bleeding.
The present study is a retrospective analysis exploring the hypothesis that postoperative
acidosis may be an independent risk factor for
postoperative bleeding in patients undergone
cardiac surgery with CPB.
Materials and methods
Study design
Two-center, retrospective cohort study based
on data prospectively collected at the two participating institutions (IRCCS Policlinico San
Donato and University Hospital of Siena). The
study was approved by the Local Ethics Committee of the IRCCS Policlinico San Donato,
and the need for an informed consent from the
patients was waived. At the hospital admission
all the patients gave written approval to the
treatment of their data in an anonymous form
and for scientific purposes.
Patients
We analyzed data routinely collected in the
institutional databases of the two participating
institutions from January 2010 through December 2013. The databases include all the patients
receiving a cardiac surgery operation, with the
exclusion of transplant operations. Both institutions are routinely using the same database,
where specific fields contain the blood gas analysis values at the arrival in the Intensive Care Unit
(ICU) immediately after surgery. Patients aged
<18 years, patients operated without CPB, and
886
patients with missing data in the fields of interest
were excluded from the study population. The
final study population included 4251 patients
(3824 at the IRCCS Policlinico San Donato and
427 at the University Hospital of Siena).
Data collection and definitions
For each patient, the following data were
collected and were available: 1) preoperative:
demographics; left ventricular ejection fraction
(%); preoperative HCT (%); recent (within 30
days prior to surgery) myocardial infarction;
congestive heart failure; active endocarditis; unstable angina; serum creatinine value (mg/dL);
chronic obstructive pulmonary disease; diabetes
(on medication); previous cerebrovascular accident; previous cardiac surgery; urgent/emergent
procedure; use of anticoagulants (warfarin), lowmolecular weight heparin, anti-platelet agents
(aspirin and/or thienopyridines). 2) Operative:
type of operation; CPB duration (minutes). 3)
Postoperative: moderate-high dose cathecolamines at the arrival in the ICU (defined as dopamine >5 µg/kg/min or epinephrine/norepinephrine >0.03 µg/kg/min); postoperative bleeding
(drain blood collected in the first 12 postoperative hours); rate of surgical revision due to bleeding; arterial blood gas analysis data at the arrival
in the ICU (pH, PaCO2, HCO3 -, LAC concentration [mMol/L]).
CPB and surgery
CPB was generally conducted under moderate (32 °C) hypothermia and alpha-stat management, unless for specific operations. The
priming volume consisted of a mixture of 80%
gelatin and 20% tromethamine (THAM) solution (4236 patients) or crystalloid solution (15
patients).
The cardioplegic arrest was achieved with antegrade administration of cold crystalloid cardioplegia and, in a minority of the cases, with
antegrade cold blood cardioplegia. Anticoagulation was achieved with unfractionated heparin
according to the standard institution protocols,
and heparin reversal was achieved with adequate
doses of protamine sulphate. All the patients re-
MINERVA ANESTESIOLOGICA
August 2015
ACIDOSIS AND BLEEDING IN CARDIAC SURGERYRANUCCI
ceived tranexamic acid at a dose of 15 mg/kg
before CPB and 15 mg/kg after protamine administration. No patient received aprotinin during the period of observation.
Statistical analysis
All data are presented as number with percentage for categorical variables, mean with standard
deviation for continuous, normally distributed
variables, and median with interquartile range
for non-normally distributed variables. Normality of distribution was checked with the
Kolgomorov-Smirnov test. The association between pH values, LAC values, and postoperative bleeding was investigated using polynomial
regression analyses. Postoperative acidosis was
defined as a pH value <7.35; the pH values distribution was analyzed according to deciles of
distribution, with the first decile arbitrarily settled at a level of pH<7.35. HL was defined as
a blood LAC value >4.0 mMol/L. LAC values
distribution was analyzed according to deciles
of distribution, with the last decile corresponding to a LAC value >4.0 mMol/L. Differences in
postoperative bleeding at the various deciles of
pH and LAC distribution were analyzed with an
Analysis of Variance (ANOVA) with post-hoc
Bonferroni’s test and adjustment for multiple
comparisons.
Multivariable analyses included regression
analyses having postoperative bleeding as the
dependent variable, the values of pH and LAC
at the arrival in the ICU, and a number of possible confounders included as covariates. To be
considered as a possible covariate, pre and intraoperative factors should be univariately associated with postoperative bleeding at a P level <0.1;
intercorrelation was checked; a multivariable regression model with stepwise forward procedure
was applied.
No patient with missing data was admitted to
the study.
All tests were two-sided. A P value <0.05
was considered to be significant for all statistical tests. Statistical calculations were performed
using computerized statistical programs (SPSS
13.0, Chicago, IL, and GraphPad Prism 6, San
Diego, CA, USA).
Vol. 81 - No. 8
Results
The general characteristics of our patient
population are shown in Table I. At a univariate analysis, there was a significant (P=0.001)
negative association between the pH values and
postoperative bleeding, according to a polynomial (cubic) regression analysis. The association
between the pH values and postoperative bleeding according to the decile distribution of the pH
values is shown in Figure 1. An ANOVA analysis
of the between-groups difference demonstrated
that only the value of postoperative bleeding in
patients with a pH value <7.35 (540±460 mL) at
Table I.—Demographics, risk profile, operative details and
acid-base balance at the arrival in the intensive care unit
(N.=4251). Data are number (%) or median (interquartile range).
Variable
Age (years)
Weight (kg)
Gender male
Left ventricular ejection fraction (%)
Previous (30 days) myocardial infarction
Unstable angina
Congestive heart failure
Active endocarditis
Serum creatinine (mg/dL)
Previous stroke
Diabetes on medication
Chronic obstructive pulmonary disease
Preoperative hematocrit (%)
Redo surgery
Anti-thrombotic therapy
Warfarin
Low molecular weight heparin
Antiplatelets
Isolated coronary surgery
Isolated valve surgery
Double valve or valve+coronary surgery
Ascending aorta
Others
Non-elective procedure
Cardiopulmonary bypass duration (min)
Moderate-high dose of cathecolamines
Acid-base balance at the arrival in intensive
care unit
pH
PaCO2 (mmHg)
HCO3 – (mEq/L)
Lactates (mMol/L)
Acidosis (pH<7.35)
Hyperlactatemia (lactates >4.0 mMol/L)
Postoperative bleeding (mL/12 h)
MINERVA ANESTESIOLOGICA
Data
69 (60-76)
74 (65-82)
2848 (67)
55 (49-61)
217 (5.1)
93 (2.2)
393 (8.7)
95 (2.1)
1.0 (0.8-1.2)
127 (3.0)
823 (18.2)
375 (8.3)
39 (36-42)
316 (7.0)
50 (1.1)
624 (13.8)
881 (19.5)
1447 (34.0)
1343 (31.6)
1062 (25.0)
57 (1.4)
392 (8.0)
466 (10.3)
80 (60-111)
563 (13.2)
7.5 (7.46-7.53)
34 (31-37.3)
25.4 (23.7-27.1)
1.4 (1.0-2.1)
156 (3.4)
436 (9.6)
350 (200-525)
887
RANUCCI
ACIDOSIS AND BLEEDING IN CARDIAC SURGERY
Figure 1.—Postoperative bleeding as a function of pH at the arrival in the intensive care unit. Dashed lines are 95% confidence
intervals of the cubic spline function.
the arrival in the ICU was significantly (P=0.001)
higher than the value registered at the other
deciles of distribution (overall, 412±304 mL).
Different additional factors (listed in Table
I) were investigated as possible determinants
of postoperative bleeding. Significant associations were found with gender male, serum creatinine values, redo operations, preoperative
use of unfractionated heparin, double valve or
valve+coronary surgery, and CPB duration. Patients receiving moderate-high dose cathecolamines at the arrival in the ICU had a significantly
(P=0.001) larger bleeding than those who did
not receive this treatment (498±392 mL/12 h
vs. 411±295 mL/12 h), a significantly (P=0.001)
lower pH (7.44±0.08 vs. 7.50±0.06) and a significantly (P=0.001) higher level of arterial blood
LACs (3.7±2.r mMol/L vs. 1.7±1.3 mMol/L).
A multivariable regression analysis (Table II)
confirmed a pH value <7.35 as an independent
determinant of postoperative bleeding, after correction for the other confounders, including the
use of moderate-high dose cathecolamines.
The LAC values demonstrated a significant (P=0.001) association with postoperative
bleeding according to a polynomial (cubic)
regression analysis. At the analysis per deciles
of distribution (Figure 2), only patients in the
upper decile (HL, LAC value at the arrival in
the ICU>4.0 mMol/L) demonstrated a significantly higher bleeding rate (513±417 mL)
with respect to all the other deciles (overall,
408±297 mL).
At the multivariable analysis, HL remained
independently associated with postoperative
bleeding (Table III) after correction for the other
Table II.­—Multivariable analysis for pH and other postoperative bleeding determinants
Factor
Gender male
CPB duration (min)
Moderate-high dose cathecolamines
pH at the arrival in intensive care unit <7.35
Constant
Regression coefficient
130
0.88
50
61
252
95% CI
P value
109-150
0.64-1.13
20.8-78.5
5.4-117
0.001
0.001
0.001
0.032
CI: confidence interval.
888
MINERVA ANESTESIOLOGICA
August 2015
ACIDOSIS AND BLEEDING IN CARDIAC SURGERYRANUCCI
Figure 2.—Postoperative bleeding as a function of arterial blood lactates at the arrival in the intensive care unit. Dashed lines are
95% confidence intervals of the cubic spline function.
confounders, including the use of moderatehigh dose cathecolamines.
This analysis was conducted separately from
the previous one, due to intercorrelation between pH and LAC values.
To investigate the relative role of acidosis and
HL in the determinism of postoperative bleeding, a sensitivity analysis was undertaken. The
patients were divided into 4 groups: 1) no acidosis and no HL (N.=3740); 2) no acidosis and
HL (compensated metabolic acidosis), N.=355;
3) acidosis and no HL (respiratory acidosis),
N.=74; 4) acidosis and HL (decompensated
metabolic acidosis) N.=82.
Postoperative bleeding in the 4 groups is depicted in Figure 3. Overall, there was a significantly
(P=0.001) higher bleeding in the groups with
compensated (493±393 mL/12 h) and decompen-
sated (576±489 mL/12h) metabolic acidosis vs. no
acidosis/HL (406±293 mL/12 h). Patients with
respiratory acidosis had a higher bleeding, but due
to the limited number (1.7% of the patient population) the difference did not reach statistical significance. However, a linear relationship between
PaCO2 at the arrival in the ICU and postoperative
bleeding was found significant (P=0.001).
Surgical revision due to bleeding was significantly (P=0.001) higher in patients with any
kind of acidosis and HL without acidosis (7.2%
and 7% respectively) with respect to patients
without acidosis and HL (2%)
Discussion
Postoperative bleeding in cardiac surgery patients is a multi-factorial event. The bleeding
Table III.—Multivariable analysis for lactate value and other postoperative bleeding determinants.
Factor
Gender male
CPB duration (min)
Moderate-high dose cathecolamine
Lactate level at the arrival in the ICU >4.0 mMol/L
Constant
Regression coefficient
130
0.83
37
63
254
95% CI
P value
110-151
0.58-1.08
7.1-67.3
26-99
0.001
0.001
0.015
0.001
CI: confidence interval; ICU: intensive care unit.
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
889
RANUCCI
ACIDOSIS AND BLEEDING IN CARDIAC SURGERY
Figure 3.—Postoperative bleeding in patients with and without acidosis and hyperlactatemia.
pattern is usually ascribed to residual heparin effects; coagulation factors consumption/dilution;
low levels of fibrinogen; thrombocytopenia and/
or poor platelet function; hyperfybrinolysis.5
Additionally, there is the possibility of surgical
sources of bleeding. The great part of the existing
literature is based on the diagnosis and treatment
of the above disturbances. However, it should
be noticed that the prerequisite for an effective
hemostasis is a well-balanced combination of
normal temperature, acid-base balance, and calcium concentration. These last factors have been
strongly emphasized in other settings, like the
trauma patient, where severe acidosis and hypothermia are commonly found.
In our study, we have found that: 1) even a
moderate degree of acidosis (pH 7.35) is significantly associated with increased postoperative
bleeding in cardiac surgery patients; 2) arterial
blood LAC values exceeding 4.0 mMol/L are associated with increased postoperative bleeding,
even if the pH is normal; and 3) the effects of
acidosis and HL on postoperative bleeding are
clinically relevant, leading to a 3-4 times higher
rate of surgical revision to control bleeding.
890
Other determinants of bleeding have been
identified in our study. These are patient-related
(gender male, serum creatinine levels), drug-related (preoperative use of heparin) and procedurerelated (CPB duration, complex surgery, redo surgery) factors. The above factors are well-known
determinants of bleeding, and after adjustment
for these confounders, both acidosis and HL remained independently associated with postoperative bleeding. Additionally, the use of moderatehigh degree cathecolamines was found associated
with postoperative bleeding, acidosis, and HL.
The effects of acidosis on the different coagulation steps have been elucidated by various
animal studies. Severe acidosis (pH=7.1) moderately inhibits the initiation phase of thrombin
generation, and greatly inhibits the propagation
phase.9 This indicates an inhibitory effect on factors V, VIII, IX and X. Other studies 10, 11 highlighted that severe acidosis induces a decrease
in fibrinogen levels and platelet count. Platelet
activation is downregulated by extracellular acidosis.12 Finally, fibrinolysis is enhanced by acidosis.13, 14 Overall, the acidosis-induced coagulopathy involves thrombin generation, clot firmness
MINERVA ANESTESIOLOGICA
August 2015
ACIDOSIS AND BLEEDING IN CARDIAC SURGERYRANUCCI
(resulting from the interaction between platelets
and fibrinogen), and hyperfibrinolysis.
Cardiac surgery-induced coagulopathy has
a similar pattern in patients receiving surgery
with CPB.5 Thrombin generation is usually impaired by coagulation factors dilution and consumption, even if this mechanism rarely leads
to severe bleeding.15 Low fibrinogen levels are
common after major cardiac operations, and
have been associated with an increased bleeding
tendency.16, 17 Thrombocytopenia and decreased
platelet activation are common,18 as well as hyperfibrinolysis.19
In our patient series, a moderate degree of
acidosis is associated with a significantly higher
postoperative bleeding. Given the fact that the
mechanisms of acidosis-induced coagulopathy
resemble those of the cardiac-surgery induced
coagulopathy, a synergistic mechanism could
be hypothesized. The retrospective nature of
our study does not allow us to differentiate the
specific impact of acidosis on the different coagulation steps. This would be feasible, as done
in animal experiments,10 by routinely running
viscoelastic coagulation tests after surgery. In
viscoelastic tests an indirect information on the
various dynamic steps of the coagulation process
is available,20 and could be correlated with the
different pH values.
Our study includes an additional piece of information. Patients with HL have a significant
higher bleeding tendency even if the pH is not
in the acidotic range. This information needs an
interpretation based on the clinical environment
of HL in cardiac surgery.
The values of arterial blood LACs observed in
our study have been measured immediately at the
arrival in the ICU. Therefore, those with values
>4.0 mMol/L experienced an “early HL”, that is
the marker of intraoperative LAC formation and
is associated with bad outcomes.8 This type of
HL is usually “type B”, deriving from an inadequate oxygen delivery to the peripheral organs
(which may occur during or after CPB). In both
cases, lactic acid is formed as an end-product of
the anaerobic metabolism, and buffering of lactic
acid leads to LAC formation and consumption of
bicarbonates, with a pattern of metabolic acidosis. In presence of this condition, the physicians
Vol. 81 - No. 8
in charge for intraoperative assistance usually
introduce a number of measures to increase the
oxygen delivery; simultaneously, it is common to
correct the pH values intraoperatively (and especially during CPB) with variable doses of bicarbonates or other buffering agents. Unfortunately, we
are lacking data on the pharmacologic measures
applied during and after CPB. Conversely, we
have the information related to the use of cathecolamines at the arrival in the ICU. Metabolic acidosis and HL are very likely to be ascribed to an
intraoperative poor oxygen delivery, due to a low
cardiac output. Of notice, the use of moderatehigh degree of cathecolamines is strongly associated with both these conditions, being a marker
of the therapeutic approach to a cardiac-related
low output syndrome. However, even after correction for the use of cathecolamines, both acidosis and HL maintains an independent association
with postoperative bleeding.
The final pattern is that of a “corrected metabolic acidosis” where the pH goes back to normal, unless it is possible that the hemodynamic
conditions leading to HL continue to deteriorate. The process of buffering acid lactic leads to
CO2 formation, and the relationship between
arterial PaCO2 and postoperative bleeding is
confirmative of this interpretation.
In light of this observation, the interpretation
of the conditions leading to HL with normal
pH (about 9% of our patient population) is that
these patients experienced an exposure to metabolic acidosis during surgery, whose effects on
the pH were subsequently corrected.
Despite this, these patients demonstrate an
increased bleeding tendency, without significant
differences with respect to those with both acidosis and HL. Therefore, correction of the acidosis with bicarbonates does not seem to limit
the acidosis-induced impairment of coagulation
and hemostasis.
From this perspective, it is useful to underline that previous animal studies on the effects
of bicarbonate correction of acidosis gave similar
results. Martini et al.10 performed a study based
on viscoelastic tests in acidotic swines, before
and after correction of the acidosis with bicarbonate infusion. After recovery of a normal (7.4)
pH value, there was no recovery of the coagula-
MINERVA ANESTESIOLOGICA
891
RANUCCI
ACIDOSIS AND BLEEDING IN CARDIAC SURGERY
tion function in terms of clotting time and clot
firmness. The same results were obtained using
alternative measures of buffering.11 Additionally,
fibrinogen levels and platelet count remained
at depleted levels after pH neutralization. The
authors hypothesize that this depletion of substrates is the main reason for acidosis-induced
coagulation disturbance.
From the clinical perspective, this piece of information leads to the concept that, in presence
of intra or postoperative acidosis, neutralization
of pH by bicarbonates or other buffers is not effective for preventing acidosis-induced bleeding.
However, the clinicians should be aware that the
effects of acidosis and HL are not clinically irrelevant, leading to a considerable increase in
surgical revision rate, that is, per se, a cause of
morbidity and mortality.21 Possible measures to
treat acidosis-related bleeding should therefore
rely more on substrate supplementation (basically, fibrinogen and platelets).
Key messages
—— In cardiac surgery patients, a moderate degree of early postoperative acidosis
(pH<7.35) is associated with an increased
postoperative bleeding.
—— Early HL (blood lactates >4 mMol/L)
is associated with an increased postoperative
bleeding.
—— HL is associated with increased postoperative bleeding even in absence of acidosis.
—— Intraoperative acidosis/HL are the
most likely determinants of early acidosis/
HL at the arrival in the intensive care unit.
References
Study limitations
There are limitations in our study. The first
one is the retrospective nature of the data analysis, that is based on our institutional databases.
This allowed us to include a large series of patients, but inevitably led to some missing information: basically, we are lacking data on intraoperative pH, LAC levels, and administration
of bicarbonates. Additionally, we do not have
available data on viscoelastic tests after surgery,
platelet counts, and fibrinogen levels at the arrival in the ICU. This does not allow us to clearly
demonstrate on which part(s) of the coagulation
system acidosis exerts the most detrimental effects. Due to these factors, the interpretation of
our data is at least in part speculative.
Nevertheless, our study adds an additional interpretation to the factors leading to postoperative bleeding and surgical revision rate in cardiac
surgery, which certainly deserves further studies
to be fully elucidated.
Conclusions
The final interpretation of our data should be
that, besides a number of well recognized factors,
892
moderate acidosis and HL play an independent
role in determining postoperative bleeding in
cardiac surgery patients.
  1. Hess JR, Brohi K, Dutton RP, Hauser CJ, Holcomb JB,
Kluger Y et al. The coagulopathy of trauma: a review of
mechanisms. J Trauma 2008;65:748-54.
  2. Kashuk JL, Moore EE, Sawyer M, Wohlauer M, Pezold M,
Barnett C et al. Primary fibrinolysis is integral in the pathogenesis of the acute coagulopathy of trauma. Ann Surg
2010;252:434-42.
  3. Moore EE, Thomas G. Orr memorial lecture. Staged laparotomy for the hypothermia, acidosis, and coagulopathy
syndrome. Am J Surg 1996;172:405-10.
 4. Niles SE, McLaughlin DF, Perkins JG, Wade CE, Li Y,
Spinella PC et al. Increased mortality associated with the
early coagulopathy of trauma in combat casualties. J Trauma 2008;64:1459-63.
  5. Ranucci M. Hemostatic and thrombotic issues in cardiac
surgery. Semin Thromb Hemost 2015;41:84-90.
  6. Demers P, Elkouri S, Martineau R, Couturier A, Cartier R.
Outcome with high blood lactate levels during cardiopulmonary bypass in adult cardiac surgery. Ann Thorac Surg
2000;70:2082-6.
  7. Ranucci M, Isgrò G, Carlucci C, De La Torre T, Enginoli
S, Frigiola A. Central venous oxygen saturation and blood
lactate levels during cardiopulmonary bypass are associated
with outcome after pediatric cardiac surgery. Crit Care
2010;14:R149.
  8. Maillet JM, Le Besnerais P, Cantoni M, Nataf P, Ruffenach
A, Lessana A et al. Frequency, risk factors, and outcome of hyperlactatemia after cardiac surgery. Chest
2003;123:1361-6.
 9. Martini WZ, Pusateri AE, Uscilowicz JM, Delgado AV,
Holcomb JB. Independent contributions of hypothermia and acidosis to coagulopathy in swine. J Trauma
2005;58:1002-9.
10. Martini WZ, Dubick MA, Pusateri AE, Park MS, Ryan KL,
Holcomb JB. Does bicarbonate correct coagulation function impaired by acidosis in swine? J Trauma 2006;61:99106.
MINERVA ANESTESIOLOGICA
August 2015
ACIDOSIS AND BLEEDING IN CARDIAC SURGERYRANUCCI
11. Martini WZ, Dubick MA, Wade CE, Holcomb JB. Evaluation of tris-hydroxymethylaminomethane on reversing
coagulation abnormalities caused by acidosis in pigs. Crit
Care Med 2007;35:1568-74.
12. Etulain J, Negrotto S, Carestia A, Pozner RG, Romaniuk
MA, D’Atri LP et al. Acidosis downregulates platelet hemostatic functions and promotes neutrophil proinflammatory responses mediated by platelets. Thromb Haemost
2012;107:99-110.
13. Martini WZ, Holcomb JB. Acidosis and coagulopathy: the
differential effects on fibrinogen synthesis and breakdown
in pigs. Ann Surg 2007;246:831-5.
14. Dirkmann D, Radü-Berlemann J, Görlinger K, Peters J.
Recombinant tissue-type plasminogen activator-evoked
hyperfibrinolysis is enhanced by acidosis and inhibited by
hypothermia but still can be blocked by tranexamic acid. J
Trauma Acute Care Surg 2013,74:482-8.
15. Ternström L, Radulovic V, Karlsson M, Baghaei F, Hyllner
M, Bylock A et al. Plasma activity of individual coagulation factors, hemodilution and blood loss after cardiac
surgery: a prospective observational study. Thromb Res
2010;126:e128-33.
16. Pillai RC, Fraser JF, Ziegenfuss M, Bhaskar B. The influence
of circulating levels of fibrinogen and perioperative coagula-
tion parameters on predicting postoperative blood loss in
cardiac surgery: a prospective observational study. J Card
Surg 2014;29:189-95.
17. Kindo M, Hoang Minh T, Gerelli S, Perrier S, Meyer N,
Schaeffer M et al. Plasma fibrinogen level on admission to
the intensive care unit is a powerful predictor of postoperative bleeding after cardiac surgery with cardiopulmonary
bypass. Thromb Res 2014;134:360-8.
18. Greilich PE, Brouse CF, Beckham J, Jessen ME, Martin EJ,
Carr ME. Reductions in platelet contractile force correlate
with duration of cardiopulmonary bypass and blood loss
in patients undergoing cardiac surgery. Thromb Res
2002;105:523-9.
19.Teufelsbauer H, Proidl S, Havel M, Vukovich T. Early
activation of hemostasis during cardiopulmonary bypass:
evidence for thrombin mediated hyperfibrinolysis. Thromb
Haemost 1992;68:250-2.
20. Martini WZ. Coagulopathy by hypothermia and acidosis:
mechanisms of thrombin generation and fibrinogen availability. J Trauma 2009;67:202-9.
21. Karthik S, Grayson AD, McCarron EE, Pullam DM, Desmond MJ. Reexploration for bleeding after coronary artery
bypass surgery: risk factors, outcomes, and the effects of
time delay. Ann Thorac Surg 2004;78:527-34.
Acknowledgments.—M. R conceived the experimental design, analyzed the data, and wrote the draft manuscript; E. B collected the data
and participated in data interpretation; F. S collected the data and reviewed the manuscript; M. R (Matteo Ranucci) participated in data
analysis and interpretation; S. S analyzed the data and reviewed the manuscript.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript.
Received on March 21, 2015. - Accepted for publication on May, 21 2015. - Epub ahead of print on May 26, 2015.
Corresponding author: M. Ranucci, Department of Anesthesia and Intensive Care, IRCCS Policlinico San Donato, Via Morandi 30,
20097 San Donato Milanese, Milan, Italy. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
893

E X PE RT O P I N I O N
OPRM1 receptor as new biomarker
to help the prediction of post mastectomy pain
and recurrence in breast cancer
M. DE GREGORI 1, 2, L. DIATCHENKO 2, 3, I. BELFER 2, 4, M. ALLEGRI 2, 5, 6
1Pain Therapy Service, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy; 2SIMPAR (Study in Multidisciplinary
Pain Research) group, Rome, Italy; 3McGill University, Montréal, Canada; 4Department of Anesthesiology, University
of Pittsburgh, Pittsburgh, PA, USA; 5Department of Surgical Sciences, University of Parma, Parma, Italy; 6Anesthesia,
Intensive care and Pain Service, Azienda Ospedaliero-Universitaria di Parma, Parma, Italy
ABSTRACT
Breast cancer is the most common type of cancer among women worldwide. Short-term postsurgical recovery is
complicated by many factors, including imbalanced inflammatory and immune response, acute pain associated with
functional impairment, and chronic postmastectomy pain (CPMP), developed by about 25-60% of patients. Opioids, most common drugs used for treatment of cancer pain, are immunosuppressive, and therefore, they might directly and/or indirectly influence long-term cancer recurrence. Moreover, they also produce endocrinopathy, which
consists primarily of hypothalamic-pituitary-gonadal axis or hypothalamic-pituitary-adrenal axis dysfunction. The
interindividual variability in both CPMP and opioid response is believed to be largely underlined by genetic variability in the gene locus for μ-opioid receptor (OPRM1) that modulates opioid pharmacodynamics. For this reason,
OPRM1 genotype may play a key role both in short-term postmastectomy outcome and in long-term follow-up, becoming a new biomarker for breast cancer recurrence in patients suffering from chronic postmastectomy pain managed by opioid therapy. Hence OPRM1 might be used in near future to customize the opioid therapy, avoiding not
only opioid side effects but also the disease progression. In this review we evaluate the literature state of the art on this
topic and possible steps towards obtaining the safest individualized postmastectomy analgesic therapy. Therefore, a
personalized pain treatment strategy might be useful to both manage pain and control cancer disease progression.
(Minerva Anestesiol 2015;81:894-900)
Key words: Genetics - Pain - Breast neoplasms - Mastectomy.
B
reast cancer (BC) is one of the most common
types of cancer among women worldwide
with high social direct and indirect costs.1 More
than 205,000 BC diagnoses are made only in
USA per year, with 40,230 annual deaths.2 Surgery is one step of multidisciplinary approach to
cure BC and to ensure oncologic radicalness and
preservation of femininity through oncoplastic
techniques.3 Mastectomy with reconstructive
surgery provides restored body symmetry to patients, improving their satisfaction and quality
of life.4
894
However, women after breast surgery are susceptible to severe postoperative pain and chronic
postmastectomy pain (CPMP), regardless of
the type of surgery.5 CPMP includes phantom
breast pain, scar pain (“iron bra”), or neuroma
pain caused by damage to cutaneous nerves entrapped in scar tissue. Moreover, pain associated
with breast reconstruction following mastectomy could be related to capsule formation, to
the compression of lateral and medial pectoral
nerves under the pectoralis muscle and to the
detachment of the serratus anterior muscle re-
MINERVA ANESTESIOLOGICA
August 2015
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
sulting from pressure exerted on the muscle by
the implant.6
Wallace et al. reported that over half of the
patients who have undergone mastectomy with
reconstruction suffered from pain for at least one
year, compared to one-third after mastectomy
alone (53% vs. 30%).7 In a recent meta-analysis
the prevalence of pain patients with a probable/
definite neuropathic component has been estimated approximately 67%.8
There is still debate about the clinical determinants of CPMP: young age and lymphadenectomy
are major risk factors,9, 10 however, it is possible that
neural (brain morphology, activity and connectivity) or psychosocial factors may also predispose to
CPMP.11 Genetic factors also have been reported
as influencing the susceptibility to CPMP.12
In turn, pain worsens quality of life, increasing emotional stress and disability;13 it represents
an important social and clinical concern, also because of its complex causes (including psychosocial factors such as anxiety and catastrophizing)
and manifestations.11, 13
Uncontrolled perioperative pain can severely
affect patients’ long-term outcome both as it
could act as important immunodepressant factor
and as it is one of the most important risk factors
of chronic postoperative pain. Hence, an adequate pain perioperative treatment is mandatory
to reduce immunodepression pain related and
the risk of chronic persistent pain. At biological level, pain is a mediator of surgery-induced
immune suppression, and then anesthesia and
analgesia techniques, by reducing postoperative
pain, would also impact the long-term postoperative outcome.14
Opioids, alone or associated with non-opioid
drugs, such as NSAIDS and acetaminophen, are
currently the main used drugs to control postoperative and chronic pain (both nociceptive
and neuropathic) in these patients, even if both
paravertebral blocks and intrawound continuous
infusion of local anestetics could be promising
in guaranteeing a better control of acute and
chronic pain.15
There is a substantial variability in acute and
chronic response to opioids, whose sources have
not yet been fully elucidated at molecular, neurophysiologic and clinical level.
Vol. 81 - No. 8
DE GREGORI
Opioids are known to be immunosuppressive,
although their actions are complicated, often indirect, and not well understood.16 Some authors
argue that they could both directly and indirectly
influence cancer growth, facilitating recurrence
through modulation of cell proliferation and/
or apoptosis;17 these drugs may also modulate
neoplastic cell proliferation and/or apoptosis,
suppressing immune function affected by surgical stress and volatile anesthetics.18, 19 Immunomodulation is due to the interaction between
opioid receptors and several molecules involved
in the complex immune response, such as transcription factors and receptors of both myeloid
and lymphoid cells.20 They also decrease the
concentrations of circulating natural killer (NK)
cells, and can have a dose-dependent effect on
NK cell cytotoxicity.21
Gupta and collaborators showed that morphine stimulates human microvascular endothelial cell proliferation and angiogenesis in vitro and
in vivo, and at clinically relevant doses, it may
promote tumor neovascularization.22 At high
concentrations, the disruption of the endothelial barrier occurs, and this might be harmful in
angiogenesis-dependent cancers. Nevertheless,
further studies examining the direct association
between opioids and cancer are warranted.
The mu-opioid receptor (OPRM1) is the
main site of action for opioid analgesia.23 Gene
encoding this receptor, OPRM1, has a number
of functional variants and, as demonstrated by
Klepstad et al., its polymorphisms could be involved in modulating the morphine clinical efficacy in cancer pain.24 The most common and well
known genetic variant is the single nucleotide
polymorphism (SNP) A118G (rs1799971), with
30% frequency of minor allele G in Caucasians
and 7% in African Americans (HapMap data;
http://hapmap.ncbi.nlm.nih.gov/). It seems that
this variant influences the immunomodulation
caused by exogenous opiates, with a protective effect of G allele;25 moreover, it decreased
response to a social stressor by means of suppression of the hypothalamic–pituitary–adrenal
(HPA) axis activation.26 Larger studies, aimed to
delineate the effect of AG and GG genotypes on
immunosuppression and endocrine response, are
needed.
MINERVA ANESTESIOLOGICA
895
DE GREGORI
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
As also immunosuppression and endocrine
alterations are related to OPRM1,27, 28 a deeper
knowledge about the effects of the genetic variation in this receptor might be useful not only in
predicting acute and chronic postoperative pain,
but also in optimizing the analgesic therapy with
a long-term perspective, that is BC recurrence.
As cancer patients homozygous for the G allele of A118G variant required higher doses of
oral morphine for long-term treatment,24 the
genotype might be considered as a potential
biomarker of therapeutic intervention in oncologic surgery setting.
Therefore, pharmacogenetics, by detecting the
patients who need to receive more morphine to
achieve the same level of analgesia, might identify the patients at major risk to develop cancer progression due to morphine treatment. It
would be suitable, for this group of patients, to
administer analgesic drugs without immunosuppressive effects as good alternative to opioids.
Material and methods and results
Analyses of OPRM1 polymorphisms associated with
opioid response and BC recurrence
Firstly, we analyzed the literature data about
OPRM1 gene, its polymorphisms, and the interindividual variability of opioid pharmacodynamics. By using the key words: “OPRM1,
polymorphisms, opioids”, we found 122 results
on PubMed database (http://www.ncbi.nlm.nih.
gov/pubmed), from 2002 to 2014: 15 clinical
trials and 28 reviews.
We found some consistent data that genetic
variants of OPRM1, the primary site of action
for the most commonly used opioids, play a key
role in the interindividual variability of opioid
pharmacodynamics.29 One of the most frequent
polymorphisms reported corresponds to the
mentioned above SNP rs1799971, in exon 1,
which causes nucleotide substitution from an
adenine to guanine at position 118 (A118G),
in turn producing the amino acid exchange at
position 40 of the corresponding protein from
asparagine to aspartic acid (N40D), and the
consequent loss of a N-glycosylation site in the
extracellular region of the receptor.30 This variant has been already associated with decreased
potency of morphine and morphine-6-glucuronide,31 and could act as a potential marker to
predict adequate opioid dosages in individualized pain treatment,32 nevertheless, in postoperative setting, its role is not always confirmed.33, 34
Another functional OPRM1 SNP, rs563649,
located within a structurally conserved internal
ribosome entry site (IRES) in the 5-UTR of a
recently discovered exon 13-containing OPRM1
isoforms (MOR-1K), affects both mRNA levels
and translation efficiency (Figure 1).35-37 This
SNP has been already associated with individual
variations in pain perception and differences in
morphine responses.36
By using the keywords: “mu opioid receptor,
breast cancer”, we found 25 papers. Bortsov et
al. reported the first study examining the association between genetic polymorphisms and the
function of opioid pathways and cancer survival: they suggested a protective effect of A118G
polymorphism in patients with invasive BC.38
Women with at least one copy of G allele may
experience significant reduction in BC-specific
mortality, however, further research is needed to
verify this hypothesis.
Lastly, we focused on opioid-induced endocrinopathy (by using the keyword “opioid-in-
Figure 1.—Genomic structure of the human OPRM1 gene is based on multispecies genome alignments created by comparative
genomes analysis,36 current NCBI database, UCSC genome browser and published data.36, 37 Exons and introns are shown by
vertical and horizontal boxes, respectively. Shaded boxes represent constitutive exons.
896
MINERVA ANESTESIOLOGICA
August 2015
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
duced endocrinopathy” we found 16 items) that
is one of the most common, even if least often
diagnosed, consequences of prolonged opioid
therapy.28 Opioids inhibit the luteinizing hormone (LH) and follicle-stimulating hormone
(FSH).28, 39 Hormone substitution could be
indicated to treat symptoms, and the decrease/
termination of opioid treatment reverses endocrine dysfunction.40 Nevertheless, the inhibition
of the hormones by opioids might be an advantage for the BC patients: antiestrogens and aromatase inhibitors are clinically used to arrest the
estrogen-dependent BC recurrence.41 Therefore,
to optimize the postmastectomy outcome, physicians should realize the challenges associated
with opioids,42 recognizing the balance between
their positive effects and the possible risks for
each patient, also defined by the genetic background.
Discussion
Even if opioids continue to be one of the most
commonly prescribed medications (together
with non steroidal anti-inflammatory drugs),
there are still important open questions related
to their efficacy, safety and predictability of patient’s response.32, 34
OPRM1 polymorphism and CPMP could be
evaluated as possible prognostic biological and
clinical factors in BC patients. Research on opioid effects on cancer is still an emerging field, and
combination of basic and clinical new knowledge would provide insights for postmastectomy
patient care. Even though in animal models
morphine-based analgesia reduces the number
of pulmonary metastases after surgery,43 in humans recent evidence suggests that opioid-based
perioperative analgesia is associated with reduced recurrence-free survival, when compared
DE GREGORI
with regional techniques, such as paravertebral
blocks and intrawound continuous infusion of
local anesthetics, for breast surgery.44
Future clinical and basic studies that seek to
identify the functional genetic variants within
OPRM1 locus, and associated molecular mechanisms, will result in a better understanding of
individual responses to opioid therapy and ultimately to the development of new diagnostic
tools.29 As high interindividual variability of opioid response is mostly due to OPRM1 polymorphisms, and since opioids, mediating immunosuppression might facilitate tumor recurrence,45
the OPRM1 genotype could play a key role also
in long-term follow-up in target BC patients,
assuming a great importance as new biomarker
for breast cancer recurrence, overall in patients
suffering from CPMP under opioid therapy.
In fact, OPRM1 genotype could be related not
only to interindividual safety and effectiveness of
opioid therapy, but also to a different immune
and endocrine response to the same opioid.
Then, in managing BC pain, the OPRM1 genetic analysis might contribute to improve both
postoperative life quality and life expectancy,
by also reducing BC recurrence and mortality
(Table I).23, 32, 37, 46,47 A deeper expertise of the
mechanism of action of opioids would be pivotal,46 together with a better characterization of
the variables that differentiate patients at major
risk to develop CPMP from those whose acute
postoperative pain resolves.11
OPRM1, playing a role in immunosuppression and endocrinopathy, could be useful not
only in predicting acute and chronic postoperative pain, but also in optimizing the analgesic
therapy with a long-term perspective, that is BC
recurrence.
In the near future, genetics-based opioid pain
therapy could solve CPMP symptoms, also be-
Table I.—Effects of A118G polymorphism in different clinical settings.
A118G effect
Setting
Increased morphine requirements
Immunomodulation
Suppression of the hypothalamic–pituitary–adrenal (HPA) axis activation
Increased amount of morphine and highest pain scores
Protective effect
Vol. 81 - No. 8
Malignant disease
Opiate users
Health volunteers
Post-hysterectomy pain
Breast cancer survival
MINERVA ANESTESIOLOGICA
Authors
Klepstad P et al.23
Hajj A et al. 46
Mao J 47
Sia AT et al.32
Bortsov AV et al. 37
897
DE GREGORI
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
Table II.—Publications by the authors about the role of OPRM1 variability, in chronologic order.
First author
Journal, year
Dever 48
AIDS 2014
Kolesnikov 49
De Gregori 33
Mura 50
Bortsov 37
De Gregori 51
Diatchenko 28
Serohijos 52
Gris 34
Gris 53
Shabalina 35
Max 54
Title
Differential expression of the alternatively spliced OPRM1 isoform μ-opioid receptor1K in HIV-infected individuals
Mol Pain 2013
Chronic pain after lower abdominal surgery: do catechol-O-methyl transferase/opioid
receptor μ-1 polymorphisms contribute?
Eur J Clin Pharmacol 2013 Genetic variability at COMT but not at OPRM1 and UGT2B7 loci
modulates morphine analgesic response in acute postoperative pain
J Pain Res 2013
Consequences of the 118A&gt;G polymorphism in the OPRM1 gene: translation from
bench to bedside?
Anesthesiology 2012
μ-Opioid receptor gene A118G polymorphism predicts survival in patients with breast
cancer.
Metab Brain Dis 2012
Morphine metabolism, transport and brain disposition.
Eur J Pain Suppl 2011
Elucidation of mu-Opioid Gene Structure: How Genetics Can Help Predict Responses
to Opioids
Structure 2011
Structural basis for μ-opioid receptor binding and activation
Mol Pain 2010
A novel alternatively spliced isoform of the mu-opioid receptor: functional antagonism.
Methods Mol Biol 2010
Molecular assays for characterization of alternatively spliced isoforms of the u opioid
receptor (MOR)
Hum Mol Genet 2009
Expansion of the human mu-opioid receptor gene architecture: novel functional variants
Mol Pain 2006
A clinical genetic method to identify mechanisms by which pain causes depression and
anxiety.
coming a treatment for the oncological disease
at reasonable cost. A multidisciplinary approach
would obtain this innovative role, through a
network involving different professionals ranging from clinicians (surgeons and anesthesiologists) to pharmacologists and biologists, with a
“from bench to bedside” approach.47 The analysis might be feasible on a large scale because the
sequencing cost per base is falling dramatically”.
The authors of this review already reported,
in 12 publications from 2009 to 2014, the importance of OPRM1 variability (used as search
criteria in PubMed: each of the 4 surnames followed by OPRM1, A118G, or mu opioid receptor) (Table II).28, 33-35, 37, 48-54
Conclusions
Development of translational approaches for
BC pain therapy treatment involving genotypic
analysis of OPRM1 gene locus, would contribute to better-customized short-term and longterm postoperative BC outcomes.
These new biomarkers of opioid pharmacodynamics will help to stratify patients’ risk of
CPMP and cancer recurrence, increasing the
role of pain therapy all the way from the patient
care focused on pain relief to cancer recovery.
898
Key messages
—— Postmastectomy patients often report
severe acute postoperative pain and chronic
postmastectomy pain, which are usually
treated with opioids.
—— Breast cancer recurrence is potentially
related to opioid consumption, as opioids
modulate neoplastic cell proliferation and/
or apoptosis and suppress immune function,
already affected by surgical stress and volatile
anesthetics.
—— OPRM1 variability, influencing pharmacodynamics, can potentially help predicting short- and long-term postoperative outcomes, such as opioid requirement, analgesic
response to opioids, and breast cancer recurrence.
References
  1. Kasper M, Jaks V, Fiaschi M, Toftgård R. Hedgehog signalling in breast cancer. Carcinogenesis 2009;30:903-11.
  2. Hornberger J, Alvarado MD, Rebecca C, Gutierrez HR, Yu
TM, Gradishar WJ. Clinical validity/utility, change in practice patterns, and economic implications of risk stratifiers to
predict outcomes for early-stage breast cancer: a systematic
review. J Natl Cancer Inst 2012;104:1068-79.
 3.Goldfarb Y, Ben-Eliyahu S. Surgery as a risk factor for
breast cancer recurrence and metastasis: mediating mecha-
MINERVA ANESTESIOLOGICA
August 2015
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
nisms and clinical prophylactic approaches. Breast Dis
2006-2007;26:99-114.
 4.Eltahir Y, Werners LL, Dreise MM, van Emmichoven
IA, Jansen L, Werker PM et al. Quality-of-life outcomes
between mastectomy alone and breast reconstruction:
comparison of patient-reported BREAST-Q and other
health-related quality-of-life measures. Plast Reconstr Surg
2013;132:201e-209e.
 5. Belfer I, Schreiber KL, Shaffer JR, Shnol H, Blaney K,
Morando A et al. Persistent postmastectomy pain in breast
cancer survivors: analysis of clinical, demographic, and psychosocial factors. J Pain 2013;14:1185-95.
  6. Nguyen JT, Buchanan IA, Patel PP, Aljinovic N, Lee BT.
Intercostal neuroma as a source of pain after aesthetic and
reconstructive breast implant surgery. J Plast Reconstr Aesthet Surg 2012;65:1199-1203.
  7. Wallace JE, Sayler C, McDowell NG, Moseley HS. The role
of stereotactic biopsy in assessment of nonpalpable breast
lesions. Am J Surg 1996;171:471-3.
  8. Haroutiunian S, Nikolajsen L, Finnerup NB, Jensen TS.
The neuropathic component in persistent postsurgical pain:
a systematic literature review. Pain 2013;154:95-102.
  9. Andersen KG, Kehlet H. Persistent pain after breast cancer
treatment: a critical review of risk factors and strategies for
prevention. J Pain 2011;12:725-46.
10. Alves Nogueira Fabro E, Bergmann A, do Amaral E Silva
B, Padula Ribeiro AC, de Souza Abrahão K, da Costa Leite
Ferreira MG et al. Postmastectomy pain syndrome: incidence and risks. Breast 2012;21:321-5.
11. Schreiber KL, Martel MO, Shnol H, Shaffer JR, Greco C,
Viray N et al. Persistent pain in postmastectomy patients:
comparison of psychophysical, medical, surgical, and psychosocial characteristics between patients with and without
pain. Pain 2013;154:660-8.
12. Nissenbaum J, Devor M, Seltzer Z, Gebauer M, Michaelis
M, Tal M, et al. Susceptibility to chronic pain following
nerve injury is genetically affected by CACNG2. Genome
Res. 2010;20:1180-90.
13. Macrae WA. Chronic post-surgical pain: 10 years on. Br J
Anaesth 2008;101:77-86.
14. Valkanoff TA, Kline-Simon AH, Sterling S, Campbell C,
Von Korff M. Functional disability among chronic pain
patients receiving long-term opioid treatment. J Soc Work
Disabil Rehabil 2012;11:128-42.
15. Andreae MH, Andreae DA. Local anaesthetics and regional
anaesthesia for preventing chronic pain after surgery. Cochrane Database Syst Rev 2012 Oct 17;10:CD007105.
16. Feng Y, He X, Yang Y, Chao D, Lazarus LH, Xia Y. Current
research on opioid receptor function. Curr Drug Targets
2012;13:230-46.
17. Afsharimani B, Cabot P, Parat MO. Morphine and tumor
growth and metastasis. Cancer Metastasis Rev 2011;30:22538.
18. Franchi S, Panerai AE, Sacerdote P. Buprenorphine ameliorates the effect of surgery on hypothalamus-pituitaryadrenal axis, natural killer cell activity and metastatic colonization in rats in comparison with morphine or fentanyl
treatment. Brain Behav Immun 2007;21:767-74.
19. Mojadadi S, Jamali A, Khansarinejad B, Soleimanjahi H,
Bamdad T. Acute morphine administration reduces cellmediated immunity and induces reactivation of latent herpes simplex virus type 1 in BALB/c mice. Cell Mol Immunol 2009;6:111-6.
20. Sacerdote P, Franchi S, Panerai AE. Non-analgesic effects
of opioids: mechanisms and potential clinical relevance
of opioid-induced immunodepression. Curr Pharm Des
2012;18:6034-42.
21. Lennon FE, Moss J, Singleton PA. The μ-opioid receptor in
cancer progression: is there a direct effect? Anesthesiology
2012;116:940-5.
22. Gupta K, Kshirsagar S, Chang L, Schwartz R, Law PY, Yee
Vol. 81 - No. 8
DE GREGORI
D et al. Morphine stimulates angiogenesis by activating
proangiogenic and survival-promoting signaling and promotes breast tumor growth. Cancer Res 2002;62:4491-8.
23. Lötsch J, Geisslinger G. Are mu-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol
Med 2005;11:82-9.
24. Klepstad P, Rakvåg TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC et al. The 118 A > G polymorphism in the
human mu-opioid receptor gene may increase morphine
requirements in patients with pain caused by malignant disease. Acta Anaesthesiol Scand 2004;48:1232-9.
25. Sharad S, Gupta AK, Singh RA, Kapoor M, Kapur S. Correlation of circulatory immunoglobulin levels with Mu opiate receptor allele. Indian J Biochem Biophys. 2007;44(5):394-400.
26. Chong RY1, Oswald L, Yang X, Uhart M, Lin PI, Wand
GS. The mu-opioid receptor polymorphism A118G predicts cortisol responses to naloxone and stress. Neuropsychopharmacology 2006;31:204-11.
27. Roy S, Barke RA, Loh HH. MU-opioid receptor-knockout
mice: role of mu-opioid receptor in morphine mediated immune functions. Brain Res Mol Brain Res 1998;61:190-4.
28. Colameco S, Coren JS. Opioid-induced endocrinopathy. J
Am Osteopath Assoc 2009;109:20-5.
29. Diatchenko L, Robinson JE, Maixner W. Elucidation of
mu-Opioid Gene Structure: How Genetics Can Help Predict Responses to Opioids. Eur J Pain Suppl 2011;5:433-8.
30. Bond C, LaForge KS, Tian M, Melia D, Zhang S, Borg L
et al. Single-nucleotide polymorphism in the human mu
opioid receptor gene alters beta-endorphin binding and activity: possible implications for opiate addiction. Proc Natl
Acad Sci USA 1998;95:9608-13.
31.Lötsch J, Skarke C, Liefhold J, Geisslinger G. Genetic
predictors of the clinical response to opioid analgesics:
clinical utility and future perspectives. Clin Pharmacokinet
2004;43:983-1013.
32. Allegri M, De Gregori M, Niebel T, Minella C, Tinelli C,
Govoni S et al. Pharmacogenetics and postoperative pain: a
new approach to improve acute pain management. Minerva
Anestesiol 2010;76:937-44.
33. Sia AT, Lim Y, Lim EC, Ocampo CE, Lim WY, Cheong P
et al. Influence of mu-opioid receptor variant on morphine
use and self-rated pain following abdominal hysterectomy. J
Pain 2013;14:1045-52.
34. De Gregori M, Garbin G, De Gregori S, Minella CE, Bugada D, Lisa A et al. Genetic variability at COMT but not at
OPRM1 and UGT2B7 loci modulates morphine analgesic
response in acute postoperative pain. Eur J Clin Pharmacol
2013;69:1651-8.
35. Gris P, Gauthier J, Cheng P, Gibson DG, Gris D, Laur O
et al. A novel alternatively spliced isoform of the mu-opioid
receptor: functional antagonism. Mol Pain 2010;6:33.
36. Shabalina SA, Zaykin DV, Gris P, Ogurtsov AY, Gauthier J,
Shibata K et al. Expansion of the human mu-opioid receptor gene architecture: novel functional variants. Hum Mol
Genet 2009;18:1037-51.
37. Xu J, Xu M, Hurd YL, Pasternak GW, Pan YX. Isolation
and characterization of new exon 11-associated N-terminal
splice variants of the human mu opioid receptor gene. J
Neurochem. 2009;108:962-72.
38. Bortsov AV, Millikan RC, Belfer I, Boortz-Marx RL, Arora
H, McLean SA. μ-Opioid receptor gene A118G polymorphism predicts survival in patients with breast cancer. Anesthesiology 2012;116:896-902.
39. Craft RM, Mogil JS, Aloisi AM. Sex differences in pain
and analgesia: the role of gonadal hormones. Eur J Pain
2004;8:397-411.
40. Rhodin A, Stridsberg M, Gordh T. Opioid endocrinopathy:
a clinical problem in patients with chronic pain and longterm oral opioid treatment. Clin J Pain 2010;26:374-80.
41.Caldas C, Tannock IF. Breast cancer: Tamoxifen--when
more might be better. Nat Rev Clin Oncol 2013;10:125-6.
MINERVA ANESTESIOLOGICA
899
DE GREGORI
OPRM1 GENOTYPE: A NEW BIOMARKER TO PREDICT MASTECTOMY OUTCOME
42. Stein C. Opioids, sensory systems and chronic pain. Eur J
Pharmacol 2013;716:179-87.
43. Page GG, McDonald JS, Ben-Eliyahu S. Pre-operative versus postoperative administration of morphine: impact on
the neuroendocrine, behavioural, and metastatic-enhancing
effects of surgery. Br J Anaesth 1998;81:216-23.
44. Forget P1, Vandenhende J, Berliere M, Machiels JP, Nussbaum B, Legrand C et al. Do intraoperative analgesics influence breast cancer recurrence after mastectomy? A retrospective analysis. Anesth Analg 2010;110:1630-5.
45. Olson BM, McNeel DG. Antigen loss and tumor-mediated
immunosuppression facilitate tumor recurrence. Expert
Rev Vaccines 2012;11:1315-7.
46. Hajj A, Khabbaz L, Laplanche JL, Peoc’h K. Pharmacogenetics of opiates in clinical practice: the visible tip of the
iceberg. Pharmacogenomics 2013;14:575-85.
47. Mao J. Translational pain research: achievements and challenges. J Pain 2009;10:1001-11.
48. Dever SM1, Costin BN, Xu R, El-Hage N, Balinang J, Samoshkin A et al. Differential expression of the alternatively
spliced OPRM1 isoform μ-opioid receptor-1K in HIVinfected individuals. AIDS. 2014;28:19-30.
49. Kolesnikov Y1, Gabovits B, Levin A, Veske A, Qin L, Dai
F, et al. Chronic pain after lower abdominal surgery: do
catechol-O-methyl transferase/opioid receptor μ-1 polymorphisms contribute? Mol Pain. 2013;9:19.
50. Mura E, Govoni S, Racchi M, Carossa V, Ranzani GN, Allegri M et al. Consequences of the 118A>G polymorphism
in the OPRM1 gene: translation from bench to bedside? J
Pain Res. 2013;6:331-53.
51. De Gregori S, De Gregori M, Ranzani GN, Allegri M,
Minella C, Regazzi M. Morphine metabolism, transport
and brain disposition. Metab Brain Dis. 2012;27:1-5.
52.Serohijos AW, Yin S, Ding F, Gauthier J, Gibson DG,
Maixner W et al. Structural basis for μ-opioid receptor
binding and activation. Structure 2011;19:1683-90.
53. Gris P, Cheng P, Pierson J, Maixner W, Diatchenko L. Molecular assays for characterization of alternatively spliced
isoforms of the u opioid receptor (MOR). Methods Mol
Biol. 2010;617:421-35.
54. Max MB, Wu T, Atlas SJ, Edwards RR, Haythornthwaite
JA, Bollettino AF et al. A clinical genetic method to identify
mechanisms by which pain causes depression and anxiety.
Mol Pain 2006;2:14.
The authors received comments on a draft of this paper from the following members of the SIMPAR group: Dr. Adele Sgarella (Pavia,
Italy), Dr. Virginia Gallo (Pavia, Italy), Dr. Lorenzo Cobianchi (Pavia, Italy), Dr. Christian Compagnone (Parma, Italy).
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on May 20, 2014. - Accepted for publication on October 8, 2014. - Epub ahead of print on October 10, 2014.
Corresponding author: M. De Gregori, Fondazione IRCCS Policlinico San Matteo, Viale Golgi 19, 27100 Pavia, Italy.
E-mail: [email protected]
900
MINERVA ANESTESIOLOGICA
August 2015

E X PE RT O P I N I O N
Epidural steroid injections: update on efficacy,
safety, and newer medications for injection
N. KOZLOV, H. T. BENZON, K. MALIK
Department of Anesthesiology, Northwestern University Feinberg School of Medicine, Chicago, IL, USA
ABSTRACT
The best evidence for epidural injection appears to be in the setting of radicular pain with epidural steroid and nonsteroid injections more efficacious than non-epidural injections. Studies showed the efficacy of non-particulate steroid
to approach the efficacy of particulate steroid and very limited comparisons demonstrated no significant difference
between epidural steroid and epidural non-steroid (local anesthetic) injection. Preliminary studies evaluating epidural
injection of disease modifying anti-rheumatic drugs such etanercept and tocilizumab showed conflicting results and
had significant limitations. Randomized studies support better efficacy of transforaminal injection due to greater
incidence of ventral epidural spread of injectate when compared to interlaminar injection. Thus, the transforaminal
approach is recommended when unilateral radicular pain is limited to one nerve root. However, the transforaminal
approach is associated with greater incidence of central nervous system injury, including paraplegia, attributed to
embolization of the particulate steroid. Recent studies showed that non-particulate steroids potentially last as long
as particulate steroids. Therefore non-particulate steroid should be used in initial transforaminal epidural injection.
Future studies should look into the role of adjunct diagnostic aids, including digital subtraction angiography, in detecting intravascular injection and the ideal site of needle placement, whether it is the safe triangle or the triangle of
Kambin. Finally, the role of epidural disease -modifying antirheumatic drugs in the management of back pain needs
to be better elucidated. (Minerva Anestesiol 2015;81:901-9)
Key words: Injections, epidural - Injections, epidural, complications - Steroids.
R
ecent publications on epidural steroid injections (ESIs) focused on epidural route,
epidural steroid injectate composition, optimal
steroid dose, cost effectiveness, impact on healthcare utilization, prevention of surgery, safety, and
alternatives to steroid as the injectate. There have
been reports of central nervous system (CNS)
injuries after transforaminal (TF) ESIs prompting a Working Group, under the sponsorship
of the Food and Drug Administration (FDA)
Safe Use Initiative, to issue recommendations
to improve the safety of ESIs. Non-particulate
steroids were recently shown to be as effective
and as long lasting as particulate steroids. The efficacy of disease modifying anti-rheumatic drugs
(DMARDs) has been studied as an alternative to
Vol. 81 - No. 8
steroid. In this review, we will update the reader
on these topics.
We performed a Medline/PubMed search on
the topics discussed in our review, including the
following: 1) safety of ESIs; 2) ESIs and need
for surgery; and 3) tumor necrosis factor alpha
inhibitors. In view of word limitations, the list
of publications related to these topics can be requested from one of the authors (HTB).
Injection route
Differences in epidural injection route, vertebral level, control group, injectate composition
and spinal pathology make drawing conclusions
regarding the efficacy of epidural steroid injec-
MINERVA ANESTESIOLOGICA
901
KOZLOV
EPIDURAL STEROID INJECTIONS
tions difficult. Several studies have attempted to
compare the ESI routes including interlaminar
(IL), transforaminal (TF), and caudal with mixed
findings. Each route has its advantages and disadvantages in terms of proximity to pain pathology and the potential risks of dural puncture,
intravascular injection or neurologic injury.1
Some of the studies on the efficacy of the
TF compared with the IL approach were retrospective or retrospective case-control studies.2-5
Other studies compared different injectates in
the transforaminal technique, i.e. methylprednisolone-bupivacaine versus saline,6 bupivacaine
with betamethasone versus bupivacaine alone,7
or triamcinolone versus saline;8 TL injection versus trigger point injection;9 or, ganglionic versus
preganglionic TF injection.10
Several studies directly compared the TF with
the IL approach in a prospective randomized
manner (Table I). One study showed epidural
perineural injections to be more effective than
conventional posterior epidural injections or
paravertebral local anesthetic.8 In this study,
the transforaminal technique is not the typical
TF technique that we use today. In their TF approach, the authors inserted their needle contralaterally, passing through the ligamentum flavum
with the needle tip ending at the lateral aspect
of the anterior epidural space.8 Another study
compared nerve root injection with IL injections
and noted no difference between the two techniques.11 In this study, the glucocorticoid cortivazol was used, a steroid that is not used in the
United States or Canada.
In a randomized study, patients with S1 radiculopathy from herniated nucleus pulposus
received epidural steroid injections via TF, caudal or IL route with the TF approach resulting
in better pain relief than either IL or caudal injection route.12 Furthermore, the patients who
were observed to have ventral epidural spread on
fluoroscopy during injection corresponded to a
better outcome and this was most common in
patients who received TF epidural steroid injection. These findings were confirmed in two studies.13, 14 In a study of patients with axial back
Table I.—Results of controlled studies comparing transforaminal from interlaminar epidural steroid injections.
Study
Kraemer et
Study design; Subjects
al.8
1997
Ackerman & Ahmad 12 2007
Lee et al.13 2009
Gharibo et al.14 2011
Candido et al.15 2008
Rados et al.16 2011
R; 93 patients
Intervention
Perineural vs. IL vs.
paravertebral local
anesthetic
R, evaluator blinded;
90 with L5-S1 disc
herniation and radicular
pain
C, 40 mg TA + 19 mL
PF saline;
IL, 40 mg TA + 4 mL
NS; TF, 40 mg TA + 4
mL NS
R, evaluator blinded;
TF, 20 mg TA +
192 patients with HNP lidocaine per side; IL,
or SS
40 mg TA + lidocaine
R, DB; 38 patients
TF, 40 mg TA + B; IL,
40 mg TA + B
R; 60 patients with
unilateral radiculopathy
from HNP or
degenerated disc
R, 64 patients with
unilateral radiculopathy
from HNP
TF, 80 mg MP + NS +
lidocaine; Parasagittal
IL, 80 mg MP + NS +
lidocaine
TF, 40 mg MP +
lidocaine;
IL, 80 mg MP +
lidocaine
Results
Comments
Better results with
perineural compared to
IL; both groups better
than paravertebral
At 24 weeks, pain relief
was greatest for TF
followed by IL and C
Doses not stated;
perineural technique not
classic TF approach
At 4m, there was greater
relief with TF than with
IL in patients with SS
Better pain relief with
TF injection
Patients with ventral
epidural spread reported
greater pain relief
TF injections were done
bilaterally
No difference in terms
of disability, function,
depression, and opioid
use
No difference in VAS
scores between TF and
parasagittal IL groups
up to 6 months
No difference in pain
Steroid dose in TF was
relief or function at 6m half the dose compared
to IL approach
B: bupivacaine; C: caudal; D: dexamethasone; DB: double-blind; HNP: herniated nucleus polpusos; IL: interlaminar; MP: methylprednisolone;
NS: normal saline; PF: preservative free; R: randomized; TA: triamcinolone; TF: transforaminal; SS: spinal stenosis.
Study by Kolsi et al.11 is not included since the steroid used, cortivazol, is not commonly used for epidural steroid injections.
902
MINERVA ANESTESIOLOGICA
August 2015
EPIDURAL STEROID INJECTIONSKOZLOV
pain from herniated nucleus pulposus or spinal
stenosis, either IL or bilateral TF epidural steroid injections were performed.13 Patients who
received TF epidural steroid injections had better outcomes than those who received IL injection for up to four months follow-up. In the
other study, patients with radicular leg pain who
received TF steroid injections had better pain
relief than those who had IL injections.14 However, the improvements in secondary outcomes
including the Oswestry Disability Index, Depression Scale, and walking tolerance were not
different between the two techniques.
In contrast, a study found no difference in analgesia between TF and parasagittal IL epidural
steroid injections for up to 6 months.15 The IL
approach was associated with better spread of the
contrast in the anterior epidural space. Additionally, the lack of difference in efficacy between the
TF and IL approach was noted in another trial.16
In this study, different doses were given in the
two approaches: 80 mg methylprednisolone in
the IL approach and 40 mg methylprednisolone
in the TF approach.
Overall, the results of randomized studies are
not uniform (Table I). However, there is moderate support for the TF route given its association with a greater incidence of ventral epidural
spread. This must be balanced with an increased
risk profile when compared to the IL or caudal
approach.
Steroid injectate composition,
optimal dose of steroid
The type of steroid and the optimal steroid
dose has been the subject of recent research.
When comparing different types of steroids, particulate to non-particulate, the results again are
mixed.1 An older study comparing non-particulate with particulate steroid in interlaminar injections showed better efficacy of the particulate
steroid.17 A recent study where equipotent doses
of methylprednisolone (80 mg) and dexamethasone (15 mg) were used, the authors noted
that dexamethasone approached the safety and
effectiveness of methylprednisolone although
there was a non-significant trend toward less
pain relief and shorter duration of action of the
Vol. 81 - No. 8
non-particulate steroid.18 For transforaminal injections, older studies showed less efficacy and
shorter duration of pain relief of non-particulate
steroids.19, 20 One randomized study showed increased benefit with respect to pain reduction
after TF injection of the particulate steroid triamcinolone compared to dexamethasone.21
More recent studies, however, showed the efficacy of non-particulate steroids to approach
that of the particulate steroids in transforaminal
injections. A retrospective study of patients with
cervical radiculopathy showed no difference in
pain score between patients receiving non-particulate versus participate steroid.22 However,
the dose of dexamethasone and triamcinolone
used in this study were not equipotent, possibly
confounding the results. Another retrospective
observational study found the non-particulate
steroid dexamethasone to be non-inferior to particulate steroids triamcinolone or betamethasone
in terms of pain relief and functional improvement.23 The study however was retrospective and
follow-up was only for two months. Moreover,
the non-particulate dexamethasone was injected
after 2010 while the particulate steroids triamcinolone and betamethasone were injected from
2006 to 2010. A more recent randomized multicenter trial showed lumbar TF dexamethasone
to be as effective as triamcinolone at 6 months
follow-up although there was a slight increase
in the number of patients requiring three injections.24 It now appears that the efficacy of the
non-particulate steroid closely approximates the
efficacy of particulate steroid steroids in transforaminal injections. This is important because of
the potential neurologic injuries resulting from
TF particulate steroid injection.
Studies attempted to address the optimal steroid dose. A study 25 randomized patients with
newly exacerbated lumbar radicular pain to receive one lumbar interlaminar ESI with either
40mg or 80mg of methylprednisolone. Comparable improvements in VAS scores were observed
in both groups at both two weeks and three
months post injection. Additionally a non-statistically significant reduction in post-injection
flares, flushing and hyperglycemia was observed
among patients who received low dose versus
high dose steroid. Another study looked at the
MINERVA ANESTESIOLOGICA
903
KOZLOV
EPIDURAL STEROID INJECTIONS
dose-response after two TF ESIs with 5 mg, 10
mg, 20 mg or 40 mg of triamcinolone, given one
week apart.26 The number of patients reporting
considerable pain relief was significantly lower at
one week after the first epidural injection among
the patients who received the 5 mg dose. However, pain scores improved in all the groups at
one week after the second epidural injection.
Given these findings the authors recommended
a minimal effective dose of 10 mg triamcinolone
for immediate pain relief in patients with lumbar
radicular pain.26 Although the number of studies
is small, the literature appears to favor the ability of a reduced steroid dose to provide similar
pain relief with possibly less steroid associated
side effect. The reduced dose has implications
in patients with diabetes mellitus because of the
hyperglycemic effect of steroids.27 As these findings are contrary to usual dose used in clinical
practice, additional larger studies are needed.
Epidural steroid versus non-steroid injections
A recent systematic review and meta-analysis
evaluated the “control” injections, mainly local anesthetic, in randomized control trials.28
The authors determined whether epidural nonsteroid injections constituted treatment or true
placebo in comparison with non-epidural injections. Indirect comparisons suggested that epidural non-steroid injections were more likely
than non-epidural injections to attain a positive outcome and provide greater reduction of
pain. In very limited direct comparisons, no
significant difference in outcome was observed
between epidural non-steroid and non-epidural
injections. The authors concluded that epidural
non-steroid injections may provide benefit beyond placebo but that further research is needed
to determine the potential role of epidural nonsteroid injections in patients with lumbar pain in
whom steroid use may be considered high risk.
The efficacy of local anesthetic as control was
recently highlighted in a recent publication that
showed no significant difference between lidocaine and steroid in lumbar spinal stenosis.29
The lack of superiority of steroid in the study
may also be due to the etiology of the back pain,
i.e. canal stenosis.
904
Epidural steroid injections
and the need for surgery
In a study of patients who were considered to
be operative candidates, 55 patients were randomized to receive a selective nerve root injection
of either bupivacaine with or without betamethasone.7 Twenty of 28 patients who received the
steroid injection decided not to have surgery. In
contrast, only 9 of 27 patients who were injected
with bupivacaine opted not to have the operation (P<0.004). A follow-up of the 29 patients
who did not have surgery showed that 21 still
did not have surgery at 5 years.30 There was no
difference between the patients who received
bupivacaine (1/9) and those who had the bupivacaine and dexamethasone (3/12, P=0.422).
Interestingly, there was no difference between
the patients who had HNP (1/7) and those with
spinal stenosis (3/14). The authors concluded
that the relief by the ESI was long enough for the
symptoms including pain to resolve naturally. A
subgroup analysis of a randomized controlled
trial noted that at one year follow-up, TF methylprednisolone-bupivacaine injection prevented
operations for contained herniations but not in
the patients with extruded discs.31
Two retrospective studies showed a surgerysparing effect of the ESIs. In one study, 26 of 30
patients had rapid regression of their pain after
periradicular injection of triamcinolone, 60% of
which had permanent resolution of their pain
allowing avoidance of surgery for an average
of 16 months.32 In another study, 51 of 90 patients with sciatica from lumbar disc herniation
avoided surgery after TF epidural triamcinolone.33 Studies with need for surgery as a primary
outcome demonstrated the strongest evidence
for epidural steroid injection, correlating with
a reduced incidence of surgical intervention.
However, the benefit appears to be greatest in
the short-term rather than in the long-term.
The limited number of patients in one
study,7, 30 the subgroup analysis nature in another study,31 and the retrospective nature in
two other publications 32, 33 make one hesitate to
conclude a surgery-sparing effect of ESIs. Also,
other studies where the incidence of surgery was
analyzed as a secondary end point, showed no
MINERVA ANESTESIOLOGICA
August 2015
EPIDURAL STEROID INJECTIONSKOZLOV
difference between epidural steroid and placebo.34-36
Safety of epidural steroid injections
A cost utility analysis was performed using
direct payment data for 480 patients over a 2
year period from four previously published randomized control trials assessing the clinical effectiveness of caudal epidural injections with or
without steroid for pain related to lumbar disc
herniation, lumbar degenerative disc, lumbar
spinal stenosis and post lumbar surgery syndrome.37 All patients showed clinical improvement and positive cost utility analysis results.
The cost per one-year quality-adjusted-life-year
was $2,172.50 for all patients and $1,966.03 for
patients judged to be successful. Although the
study showed a better cost utility of managing
chronic lumbar pain with caudal epidural injections than non-interventional therapy, this study
was limited to a single center. Another study assessed cost effectiveness by looking at mean quality- adjusted-life-year gains of patients receiving
epidural steroid injections in an outpatient setting in England.38 Epidural steroid injections
were noted to provide short-term but cost-effective means of managing chronic back pain.
Paraplegia has been reported after TF ESIs,
with the cause ascribed to embolism of the particulate steroid. Animal studies showed death or
cerebral hemorrhage after injection of particulate steroid into the vertebral artery of pigs 39 or
into the cerebral artery of rats.40 In contrast,
these complications were not noted after nonparticulate steroids. The routes of embolism include the segmental radicular artery, ascending
cervical and deep cervical arteries, artery of Adamkiewicz, and lateral sacral arteries which anastomose with the sacral radicular arteries and the
ansa communications at the conus. A Working
Group, under the sponsorship of the Food and
Drug Administration, made recommendations
to decrease the complications. Some of their recommendations include the use of image guidance and injection of contrast for ESIs, interlaminar over the transforaminal approach in cervical
ESIs, transforaminal over interlaminar approach
in lumbar ESIs, the preferred use of non-particulate steroid in cervical transforaminal injections
and the initial use of non-particulate steroids in
lumbar TF ESIs.41
To better detect intravascular placement of
the needle in lumbar TF injection of particulate
A
B
Cost effectiveness and healthcare utilization
Figure 1.—Fluoroscopy images showing epidural contrast pattern with no visible intravascular injection after injection of contrast
(A) and simultaneous epidural and clear intravascular pattern with DSA. From Lee MH et al.42
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
905
KOZLOV
EPIDURAL STEROID INJECTIONS
steroid, digital subtraction angiography (DSA)
has been recommended. As concomitant epidural spread is common in intravascular injections, the use of DSA improves the fluoroscopic
image enhancing the detection of intravascular
injection (Figure 1).42 Retrospective 43 as well as
prospective studies 42, 44 showed superiority of
DSA over aspiration in detecting intravascular
injection. In one study, the sensitivity of DSA
was 60% compared to 25% with aspiration (Table II).42 The DSA however is not foolproof, is
expensive and increases radiation exposure.
TF injection at the triangle of Kambin, where
the needle tip is at the postero-inferior intervertebral disc space, decreases the incidence of intravascular injection compared to injection at the “safe
triangle” (2 of 50 versus 6 of 50).45 The efficacy appears not to be decreased and there is less contact
with the nerve root. The less vascular nature of the
inferior portion of the foramen was supported in a
retrospective review of thoracic and lumbar spinal
angiograms. In the study, 113 radiculomedullary
arteries were clearly identified. In eighty seven percent the radiculomedullary arteries (artery of Adamkiewicz) were noted in the upper third, in 9%
they were noted in the middle third, and only 2%
were located in the lower third of the foramen.46
A retrospective cohort study found increasing
number of lumbar epidural steroid injections
was associated with an increasing risk for vertebral body fractures by a factor of 1.21.47 Al-
though modest, the risk was statistically significant. However, it should be noted that the study
did not control for other variables that may also
contribute to facture risk such as smoking, level
of exercise, or Body Mass Index.
Radiation exposure to both the patient and
clinician is a concern with fluoroscopy-guided
ESI. A prospective, randomized single blind
clinical study by found no statistical difference
in terms of pain score or Oswestry Disability Index in patients receiving ultrasound-guided versus fluoroscopy-guided caudal epidural steroid
injection for unilateral lumbar radicular pain.48
Although this study has a number of limitations
in terms of blinding, practitioner experience, patient number and selection, it highlights potential benefits of ultrasound guidance that can be
explored in larger trials.
Epidural anti-inflammatory
non-steroid injections
The role of inflammation in the causation of
low back and radicular pain led to the use of
disease modifying anti-rheumatic drugs for this
condition. Of these drugs, only etanercept and
tocilizumab were investigated epidurally (Table
III). Two trials showed greater efficacy of epidural etanercept compared to placebo. A study
of 24 patients showed better efficacy of three
doses of etanercept (2, 4, 6 mg) compared to sa-
Table II.—Incidence of Intravascular Injection with and without digital subtraction angiography*.
Study
McLean et al.43
Contrast injection under real-time
fluoroscopy
DSA
Lee et al.42
Contrast injection under real-time
fluoroscopy after aspiration
DSA
Hong et al.44
Vascular uptake on real-time fluoroscopy
Positive vascular injection on DSA
Number of subjects or
Number of positive
injections
intravascular injections
Sensitivity
Specificity
25%
60%
100%
100%
71%
100%
(Subjects)
67
67
87 injections
249 injections
12 (17.9%)
22 (32.8%)
5
12
22 of 31
31 of 31
Modified from Benzon HT. Use of digital subtraction angiography in epidural steroid injections: PRO. American Society of Regional Anesthesia
(ASRA) Newsletter 2014: p. 9, 19-20.
P value between contrast injection and DSA, McLean et al. study: 0.0471.43
P value between real-time fluoroscopy and DSA, Hong et al. study: 0.007.44
906
MINERVA ANESTESIOLOGICA
August 2015
EPIDURAL STEROID INJECTIONSKOZLOV
Table III.—Results of studies on epidural injections of disease modifying anti-inflammatory drugs.
Study, drug
Study design
Cohen et al.48, etanercept
Freeman et al.50, etanercept
Ohtori et al.51, etanercept
Cohen et al.52, etanercept
Ohtori et al.53, tocilizumab
Intervention
Double-blind
placebo-controlled,
dose-response; 24
patients
Randomized,
double-blind,
placebo-controlled, 49
patients
Prospective,
randomized; 80 patients
Transforaminal epidural Superiority of 0.5 mg
etanercept (0.5, 2.5,
etanercept over placebo
12.5) vs. placebo
in terms of leg pain,
Oswestry
Transforaminal epidural Etanercept more
etanercept (10 mg)
effective for low back
and lidocaine vs.
and leg pain, and leg
dexamethasone (3.3
numbness
mg) and lidocaine
Multicenter,
Transforaminal epidural Epidural steroid
3-group, randomized
methylprednisolone (60 resulted in better
placebo-controlled trial; mg) vs. etanercept (4
relief of leg pain and
84 patients
mg) vs. saline
improvement in
functional capacity than
etanercept
Prospective, 60 patients Transforaminal epidural Tocilizumab more
tocilizumab (80 mg)
effective in relieving
and lidocaine vs.
back and leg pain and
dexamethasone (3.3
numbness
mg) and lidocaine
line at one month.49 Another trial of 49 patients
showed efficacy of etanercept over placebo of
up to six months but only with the lowest dose
of etanercept (0.5 mg).50 Two trials compared
transforaminal epidural etanercept with steroid. One trial of 80 patients showed superiority of etanercept over dexamethasone in terms
of relief of leg pain and numbness at 4 weeks.51
However, another study showed less efficacy of
etanercept compared to methylprednisolone in
terms of relief of leg pain or in functional outcome, although the results were not statistically
significant.52 The single study on transforaminal
80 mg tocilizumab showed it to be more effective than 3.3 mg dexamethasone.53
Conclusions
In this review, we discussed current issues associated with ESI. The TF route appears to be
more effective than the IL route, however, this
must be balanced with the higher risk of vascular injection or neurologic injury with this technique. More recent studies demonstrated the
efficacy and duration of TF non-particulate steroid to approach that of particulate steroids. This
Vol. 81 - No. 8
Results
Transforaminal
Improvement in
injection of 2, 4, or 6
etanercept but not in
mg etanercept vs. saline saline
Comments
Small number
of patients (24),
short-term follow-up
Lowest dose of
etanercept noted to be
the most effective
Dose of dexamethasone
low; follow-up short (1
month)
One month follow-up;
dexamethasone dose
low (3.3 mg)
equal efficacy will probably result in increased
use of non-particulate steroid TF injections in
unilateral radicular pain, thereby reducing the
incidence of CNS injury. Studies evaluating TF
injection at the triangle of Kambin, instead of
the “safe triangle” should be pursued. The evidence appears to favor a steroid ceiling effect so
for patients in whom steroid use is high risk, a
lower steroid dose, non-particulate steroid or
non-steroid epidural injection should be considered. Studies that look at the need for surgery as
a primary outcome showed a reduced need for
surgery in the short term. Research studies on
epidural anti-inflammatory non-steroid injections are preliminary and recommendations on
their future clinical applicability cannot be made
at this time.
Key Messages
—— There is moderate support for the
greater efficacy of transforaminal injection
when compared with interlaminar and caudal injection approaches, due to greater in-
MINERVA ANESTESIOLOGICA
907
KOZLOV
EPIDURAL STEROID INJECTIONS
cidence of ventral epidural spread, but this
must be balanced with an increased risk of
neurologic injury.
—— Use of reduced steroid dose, nonparticulate steroid or non-steroid epidural
injection may significantly reduce steroid associated complications and side effects such
as embolic neurologic injury and hyperglycemia while potentially providing similar pain
relief.
—— Epidural steroid injections may potentially reduce healthcare costs by reducing
healthcare utilization and/or the need for
surgical intervention.
—— Future areas of research should focus on improving the safety and efficacy of
epidural injections by examining the role of
diagnostic aids (digital subtraction angiography), needle position (safe triangle versus
triangle of Kambin), injectate (non-particulate steroid or alternative to steroid injection such as anti-inflammatory non-steroid
injection), and new approaches (ultrasound
versus fluoroscopic guidance).
References
  1. Cohen SP, Bicket MC, Jamison D, Wilkinson I, Rathmell
JP. Epidural steroids: A comprehensive review. Reg Anesth
Pain Med 2013;38:175-200.
  2. Schaufele M, Hatch L, Jones W. Interlaminar versus transforaminal epidural injections for the treatment of symptomatic lumbar intervertebral disc herniations. Pain Physician 2006;9:361-6.
  3. Lee J, Moon J, Lee S. Comparison of effectiveness according to different approaches of epidural steroid injection in
lumbosacral herniated disk and spinal stenosis. J Back Musculoskelet Rehabil 2009;34:83-9.
  4. Mendoza-Lattes S, Weiss A, Found E,Zimmerman B, Gao
Y. Comparable effectiveness of caudal vs. trans-foraminal
epidural steroid injections. Iowa Orthop J 2009;29:91-6.
  5. Smith CC, Booker T, Schaufele M, Weiss P. Interlaminar
versus transforaminal epidural steroid injections for the
treatment of symptomatic lumbar spinal stenosis. Pain Med
2010;11:1511-15.
 6.Karppinen J, Malmivaara A, Kurunlahti M, Kyllonen
E, Pienimaki T, Nieminen P et al. Periradicular infiltration for sciatica: A randomized controlled trial. Spine
2001;26:1059-67  7. Riew KD, Yin Y, Gilula L, Bridwell KH, Lenke LG, Lauryssen C, Goette K. The effect of nerve-root injections on the
need for operative treatment of lumbar radicular pain. A
prospective, randomized, controlled, double-blind study. J
Bone Joint Surg Am 2000;82-A:1589-93.
  8. Kraemer J, Ludwig J, Bickert U, Owczarek V, Traupe M.
Lumbar epidural perineural injection: a new technique. Eur
Spine J 1997;6:357-61.
908
  9. Vad VB, Bhat AL, Lutz GE, Cammisa F. Transforaminal
epidural steroid injections in lumbosacral radiculopathy: A
prospective randomized study. Spine 2002;27:11-6.
10. Jeong HS, Lee JW, Kim SH, Myung JS, Kim JH, Kang HS.
Effectiveness of transforaminal epidural steroid injection by
using a preganglionic approach: A prospective randomized
controlled study. Radiology 2007;245:584-90.
11. Kolsi I, Delecrin J, Berthelot J, Thomas L, Prost A, Maugars
Y. Efficacy of nerve root versus interspinous injections of
glucocorticoids in the treatment of disk-related sciatica. A
pilot, prospective, randomized, double-blind study. Joint
Bone Spine 2000;67:113-8.
12. Ackerman 3rd W, Ahmad M. The efficacy of lumbar epidural steroid injections in patients with lumbar disc herniations. Anesth Analg 2007;104:1217-22.
13. Lee J, An J, Lee S. Comparison of the effectiveness of interlaminar and bilateral transforaminal epidural steroid injections in treatment of patients with lumbosacral disc herniation and spinal stenosis. Clin J Pain 2009;25:206-10.
14. Gharibo C, Varlotta G, Rhame E, Bendo JA, Perloff MD.
Interlaminar versus transforaminal epidural steroids for the
treatment of subacute lumbar radicular pain: a randomized, blinded, prospective outcome study. Pain Physician
2011;14:499-511.
15. Candido K, Raghavendra M, Chinthagada M, Badiee S,
Trepashko DW. A prospective evaluation of iodinated contrast flow patterns with fluoroscopically guided lumbar epidural steroid injections: the lateral parasagittal interlaminar
epidural approach versus the transforaminal epidural approach. Anesth Analg 2008;106:638-44.
16. Rados I, Sakic K, Fingler M, Kapural L. Efficacy of interlaminar vs. transforaminal epidural steroid injection for the
treatment of chronic unilateral radicular pain: prospective,
randomized study. Pain Med 2011;12:1316-21.
17. Noe CE, Haynsworth RF Jr. Comparison of epidural depomedrol vs. aqueous betamethasone in patients with low
back pain. Pain Pract 2003;3:222-5.
18. Kim D, Brown J. Efficacy and safety of lumbar epidural
dexamethasone versus methylprednisolone in the treatment
of lumbar radiculopathy: a comparison of soluble versus
particulate steroids. Clin J Pain 2011;27:518-22.
19. Dreyfuss P, Baker R, Bogduk N. Comparative effectiveness of cervical transforaminal injections with particulate
and nonparticulate corticosteroid preparations for cervical
radicular pain. Pain Med 2006;7:237-42.
20. Lee JW, Park KW, Chung SK, Yeom JS, Kim KJ, Kim HJ,
Kang HS. Cervical transforaminal epidural steroid injection for the management of cervical radiculopathy: a comparative study of particulate versus non-particulate steroids.
Skeletal Radiol 2009;38:1077-82.
21. Park C, Lee S, Kim B. Comparison of the effectiveness of
lumbar transforaminal epidural injection with particulate
and non-particulate corticosteroids in lumbar radiating
pain. Pain Med 2010;11:1654-8.
22. Shakir A, Ma V, Mehta B. Comparison of pain score reduction using triamcinolone versus dexamethasone in cervical
transforaminal epidural steroid. Am J Phys Med Rehabil
2013;92:768-75.
23. El-Yahchouchi C, Geske JR, Carter RE, Diehn FE, Wald
JT, Murthy NS et al. The noninferiority of the nonparticulate steroid dexamethasone vs the particulate steroids betamethasone and triamcinolone in lumbar transforaminal
epidural steroid injections. Pain Med 2013;14:1650-7.
24. Kennedy DJ, Plastaras C, Casey E, Visco CJ, Rittenberg
JD, Conrad B et al. Comparative effectiveness of lumbar
transforaminal epidural steroid injections with particulate
versus nonparticulate corticosteroids for lumbar radicular
pain due to intervertebral disc herniation: A prospective,
randomized, double-blind trial. Pain Med 2014;15:548-55.
25. Owlia M, Salimzadeh A, Alishiri G, Haghighi A. Comparison of two doses of coticosteroid in epidural steroid
MINERVA ANESTESIOLOGICA
August 2015
EPIDURAL STEROID INJECTIONSKOZLOV
injection for lumbar radicular pain. Singapore Med J
2007;48:241-5
26. Kang SS, Hwang BM, Son HJ, Cheong IY, Lee SJ, Lee SH,
Chung TY. The dosages of corticosteroid in transforaminal
epidural steroid injections for lumbar radicular pain due to
a herniated disc. Pain Physician 2011;14:361-70.
27. Kim WH, Sim WS, Shin BS, Lee CJ, Jin HS, Lee JY et al.
Effects of two different doses of epidural steroid on blood
glucose levels and pain control in patients on diabetes mellitus. Pain Physician 2013;16:557-68.
28. Bicket MC, Gupta A, Brown CH, Cohen SP. Epidural
injections for spinal pain: a systematic review and metaanalysis evaluating the “control” injections in randomized
controlled trials. Anesthesiology 2013;119:907-31.
29. Friedly JL, Comstock BA, Turner JA, Heagerty PJ, Deyo
RA, Sullivan SD et al. A randomized trial of epidural gluococorticoid injections for spinal stenosis. N Engl J Med
2014;371:11-21.
30. Riew D, Park JB, Choo YS, Gilula L, Patel A, Lenke LG et
al. Nerve root blocks in the treatment of lumbar radicular
pain. A minimum five-year follow-up. J Bone Joint Surg
Am 2006;88:1722-5.
31.Karppinen J, Ohinmaa A, Malmivaara A, Kurunlahti
M, Kyllönen E, Pienimäki T et al. Cost effectiveness of
periradicular infiltration for sciatica: subgroup analysis
of a randomized controlled trial. Spine (Phila Pa 1976)
2001;26:2587-95.
32. Narozny M, Zanetti M, Boos N. Therapeutic efficacy of
selective nerve root blocks in the treatment of lumbar radicular leg pain. Swiss Med Wkly 2001;131:75-80.
33. Manson NA, McKeon MD, Abraham EP. Transforaminal
epidural steroid injections prevent the need for surgery in
patients with sciatica secondary to lumbar disc herniation: a
retrospective case series. Can J Surg 2013;56:89-96.
34. Wilson-MacDonald J, Burt G, Griffin D, Glynn C. Epidural steroid injection for nerve root compression. A randomized, controlled trial. J Bone Joint Surg Br 2005;87:352-5.
35. Arden NK, Price RC, Reading I, Stubbing J, Hazelgrove J,
Dunne C et al. WEST Study Group. A multicentre randomized controlled trial of epidural corticosteroid injections for
sciatica: the WEST study. Rheumatology 2005;44:1399406.
36.Tafazal S, Ng L, Chaudhary N, Sell P. Corticosteroids in
periradicular infiltration for radicular pain: a randomized
double-blind controlled trial. One year results and subgroup analysis. Eur J Spine 2009;18:1220-5.
37. Manchikanti L, Falco FJ, Pampati V, Cash KA, Benyamin
RM, Hirsch JA. Cost utility analysis of caudal epidural injections in the treatment of lumbar disc herniation, axial or
discogenic low back pain, central spinal stenosis, and post
lumbar surgery syndrome. Pain Physician 2013;16:E12943.
38. Whynes DK, McCahon RA, Ravenscroft A, Hardman J.
Cost effectiveness of epidural steroid injections to manage
chronic lower back pain. BMC Anesthesiol 2012;12:26.
39. Okubadejo GO, Talcott MR, Schmidt RE, Sharma A, Patel
AA, Mackey RB et al. Perils of intravascular methylprednisolone injection into the vertebral artery. J Bone Joint Surg
Am 2008;90:1932-8.
40. Dawley JD, Moeller-Bertram T, Wallace MS, Patel PM. Intra-arterial injection in the rat brain. Evaluation of steroids
used for transforaminal epidurals. Spine 2009;34:1638-43.
41. Rathmell JR, Benzon HT, Dreyfuss P, Huntoon M, Wallace M, Baker R et al. Safeguards to prevent neurological
complications after epidural steroid injections: Consensus
opinions from a multidisciplinary working group and national organizations. Anesthesiology 2015;122:974-84.
42. Lee MH, Yang KS, Kim YH, Jung HD, Lim SJ, Moon DE.
Accuracy of live fluoroscopy to detect intravascular injection during lumbar transforaminal epidural injections. Korean J Pain 2010;23:18-23.
43. McLean JP, Sigler JD, Plastaras CT, Garvan CW, Rittenberg JD. The rate of detection of intravascular injection
in cervical transforaminal epidural steroid injections with
and without digital subtraction angiography. PM R 2009;
1:636-42.
44. Hong JH, Huh B, Shin HH. Comparison between digital
subtraction angiography and real-time fluoroscopy to detect intravascular injection during lumbar transforaminal
epidural injections. Reg Anesth Pain Med 2014;39:329-32.
45. Park KD, Lee J, Jee H, Park Y. Kambin triangle versus the
supraneural approach for the treatment of lumbar radicular
pain. Am J Phys Med Rehabil 2012;91:1039-50.
46. Murthy NS, Maus TP, Behrns CL. Intraforaminal location
of the great radiculomedullary artery (artery of Adamkiewicz): a retrospective review. Pain Med 2010;11:1756-64.
47. Mandel S, Schilling J, Peterson E, Rao S, Sanders W. A
retrospective analysis of vertebral body fractures following
epidural steroid injections. J Bone Joint Surg Am 2013;95:
961-4.
48. Park Y, Lee JH, Park KD, Ahn JK, Park J, Jee H. Ultrasound-guided vs. fluoroscopy-guided caudal epidural steroid injection for the treatment of unilateral lower lumbar
radicular pain: a prospective, randomized, single-blind
clinical study. Am J Phys Med Rehabil 2013;92:575-86.
49. Cohen SP, Bogduk N, Dragovich A, Buckenmaier CC 3rd,
Griffith S, Kurihara C et al. Randomized, double-blind,
placebo-controlled, dose response, and preclinical safety
study of transforaminal epidural etanercept for the treatment of sciatica. Anesthesiology 2009;110:1116-26.
50. Freeman BJ, Ludbrook GL, Hall S, Cousins M, Mitchell
B, Jaros M et al. Randomized, double-blind, placebo-controlled, trial of transforaminal epidural etanercept for the
treatment of symptomatic lumbar disc herniation. Spine
2013;38:1986-94.
51. Ohtori S, Miyagi M, Eguchi Y, Inoue G, OritaS, Ochiai et
al. Epidural administration of spinal nerves with the tumor
necrosis factor-alpha inhibitor, etanercept, compared with
dexamethasone for treatment of sciatica in patients with
lumbar spinal stenosis. Spine 2012;37:439-44.
52. Cohen SP, White RL, Kurihara C, Larkin TM, Chang A,
Griffith SR et al. Epidural steroids, etanercept, or saline in
subacute sciatica: a multicenter, randomized trial. Ann Intern Med 2012;156:551-9.
53. Ohtori S, Miyagi M, Eguchi Y, Inoue G, Orita S, Ochiai
N et al. Efficacy of epidural administration of anti-interleukin-6 receptor antibody onto spinal nerve for treatment of
sciatica. Eur Spine J 2012;21:2079-84.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on May 6, 2014. - Accepted for publication on October 8, 2014. - Epub ahead of print on October 14, 2014.
Corresponding author: H. T. Benzon, Department of Anesthesiology, Northwestern University Feinberg School of Medicine, 251 E.
Huron, Feinberg, 5-704, Chicago, IL, USA 60611. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
909

REVIEW
Preoxygenation and general anesthesia: a review
G. BOUROCHE, J. L. BOURGAIN
Service d’Anesthésie Gustave Roussy, Villejuif, France
ABSTRACT
Because intubation can potentially become a lengthy procedure, the risk of arterial oxygen (O2) desaturation during
intubation must be considered. Preoxygenation should be routine, as oxygen reserves are not always sufficient to
cover the duration of intubation. Three minutes of spontaneous breathing at FiO2=1 allows denitrogenation with
FAO2 close to 95% in patients with normal lung function. Tolerable apnea time, defined as the delay until the SpO2
reaches 90%, can be extended up to almost 10 minutes after 3 minutes of classic preoxygenation. Eight deep breaths
within 60 seconds allow a comparable increase in O2 reserves. For effectiveness, the equipment must be adapted
and tightly fitted. Inadequate preoxygenation (FeO2 <90% after three minutes tidal volume breathing) is frequently
observed. Predictive risk factors for inadequate pre-oxygenation share overlap with criteria predictive of difficult
mask ventilation. In cases of respiratory failure, oxygenation can be improved by positive end expiration pressure or
by pressure support. In morbidly obese patients, preoxygenation is enhanced in a seated position (25°) and by use
of positive pressure ventilation. O2 can also be administered during the intubation procedure; techniques include
pharyngeal O2, special oxygen mask, or even pressure support ventilation for patients with spontaneous ventilation
or positive pressure ventilation to the facial mask for apneic patients. Clinicians (especially anesthesiologists trained
in ENT and traumatology) must be prepared to handle life-threatening emergency situations by alternate methods
including trans-tracheal ventilation. The availability of equipment and training are two essential components of
adequate preparation. (Minerva Anestesiol 2015;81:910-20)
Key words: Intubation, intratracheal - Cell respiration - Laryngeal masks - Ventilation - Anesthesia.
P
reservation of oxygenation during intubation is essential because lack of control of
O2 intake can cause life-threatening complications. Anesthetic induction usually leads to apnea, when tissue oxygenation is maintained by
consumption of the oxygen reserve and by continuous administration of O2. In the majority
of cases, rigorous preoxygenation and face mask
ventilation are provided until muscle relaxation
is sufficient to facilitate intubation in good conditions. In some cases oxygenation cannot be
maintained, either due to pulmonary disease or
to mask ventilation or intubation difficulties.
These critical situations can often be foreseen
and avoided by preparation for alternative methods of oxygenation following a validated algorithm.1
910
Pathophysiology of oxygenation
During anesthesia, oxygenation primarily depends on three parameters: alveolar ventilation
(VA), distribution of ventilation/perfusion ratio,
and consumption of O2 (VO2).
Oxygen reserves
During apnea, tissue oxygenation is maintained at the expense of the body’s O2 reserves,2
which are very low in quantity and are mainly
situated in lungs, plasma, and hemoglobin.
Upon breathing ambient air, the lung O2 reserve
is calculated as follows, for a functional residual
capacity (FRC) of 3000 mL: 0.21x3000=630
mL. After complete pre-oxygenation, the alveo-
MINERVA ANESTESIOLOGICA
August 2015
PREOXYGENATION AND GENERAL ANESTHESIABOUROCHE
lar fraction of O2 (FAO2) is close to 0.95 and the
reserve increases as follows: 0.95x3000=2850
mL. These theoretical figures are maximal values;
the FAO2 is lower in practice because the ventilation/perfusion ratio is heterogeneous.
The plasma O2 reserve of a subject breathing
in ambient air (PaO2=80 mmHg) with a plasma
volume of 3 L is calculated as 0.003x80x3x10=7
mL. At a PaO2 of 500 mmHg, this plasma reserve amounts to 45 mL. The hemoglobin O2
reserve is calculated as follows for a concentration of hemoglobin of 12 g. 100 mL-1 and a total
blood volume of 5 L: 1.34x0.98x12x10x5=788
mL in ambient air (saturation=98%). The value
increases to 804 mL with a FiO2 of 1 (saturation=100%). In cases of anemia, hyperoxic ventilation increases the utilizable O2 by augmenting the dissolved O2.3
Considering the three main physiological O2
reserves, the total O2 reserve is about 1450 mL
when breathing in ambient air and it rises to
approximately 3700 mL when breathing pure
O2. This increase (approximately 2250 mL) is
mainly due to the rise FAO2 in FRC. The theoretical values were confirmed experimentally by
measurement of oxygen uptake breath-by-breath
in healthy volunteers during preoxygenation.
The mean expired fraction of O2 (FEO2) after 3
minutes of breathing O2 was 0.92±0.01, and the
mean additional oxygen taken up was 2.23±0.85
L. This value closely agrees with the physiological model.4
Several factors influence O2 availability: the
initial rise in PaCO2 (Haldane effect), FRC,
FAO2, fraction of shunt, VO2, hemoglobin concentration, and cardiac output. Replacement
of nitrogen by O2 in the lung reservoir during
preoxygenation obeys an exponential law.2 The
change in O2 reserve over time is linear in both
blood and tissue compartments.
O2 consumption
The O2 consumption of an awake subject is
about 300 mL per min and it falls about 15% in
the elderly. After ventilation in ambient air, O2
reserves allow, at maximum, 3 minutes of apnea
without serious impact on O2 transport. This
time can be doubled by correctly performed pre-
Vol. 81 - No. 8
oxygenation. The duration of apnea tolerated is
additionally decreased if O2 reserves are low due
to decreased FRC, low PAO2, and/or high VO2.
Ventilation/perfusion mismatch
Preoxygenation leads to increased shunt and
micro-atelectasis after anesthetic induction.5
High FiO2 is not the only mechanism responsible because atelectasis has also been observed
when a FiO2 0.4 is used.6 The use of a FiO2 of
0.8 does not prevent the appearance of microatelectasis, and it results in a considerably shortened margin of time before unacceptable desaturation compared with the use of 100% oxygen.7
Microatelectasis are reversible by application of
an alveolar recruitment maneuver (tracheal pressure >30 cm H2O for 15 seconds) and they can
be prevented by the addition of a positive end
expiration pressure (PEEP) of 10 cm H2O.8
In morbidly obese patients and in parturients,
shunt can exceed 20% and even increasing FiO2
to 1 does not provide correction of the hypoxemia. Implementation of a microatelectasis prevention strategy of alveolar recruitment maneuvers and PEEP limits the extent in elderly 9 and
obese patients.10
Epidemiology of arterial desaturation
during induction and intubation
Anesthetic induction
Before upper airway control, arterial O2 desaturation occurs when the O2 reserves are insufficient to support the O2 consumption during
the apnea period. There are three mechanisms
responsible (Figure 1): quantitative decrease in
reserves (decrease in FRC, impairment of gas
exchange), increase in VO2 (parturient, fever),
and prolonged apnea. Four high-risk situations
deserve special mention:
—— rapid induction sequence in which mask
ventilation increases the risk of inhalation of gastric fluid (although this has never been demonstrated to occur);
—— predicted difficulty with face mask ventilation;
—— predicted difficulty with intubation due
MINERVA ANESTESIOLOGICA
911
BOUROCHE
PREOXYGENATION AND GENERAL ANESTHESIA
Figure 1.—Mechanism of arterial desaturation in O2 during anesthetic induction.
Figure 2.—Duration of apnea after pre-oxygenation and rapid induction sequence.
The durations in minutes were estimated from the literature for a rapid sequence induction data and have shown in timeline with
pre-oxygenation conditions and duration of apnea up to SpO2<90%. In some cases the apnea time will be less than the duration
of action of anesthetic agents and that an alternative method for oxygenation will become necessary.
to anatomical abnormality or specific technical
considerations (e.g., double lumen tube);
—— obesity or pregnancy.
After rapid sequence induction, the resumption of spontaneous ventilation does not occur
fast enough to allow recovery after a failed intubation procedure, and saturation falls below
90% in 11% of patients (Figure 2).11 After induction by propofol (2 mg.kg-1) and fentanyl (2
912
µg.kg-1), the administration of succinylcholine
(0.56 mg.kg-1 and1 mg.kg-1) increases the risk
of desaturation and apnea duration compared to
placebo.12 In a pharmacodynamic study of succinylcholine (from 0.3 to 1 mg.kg-1), intubation
conditions were found to be excellent at dosages above 0.5 mg.kg-1 (Table I), but the delay
in resumption of spontaneous breathing rose
from 4.0 to 6.16 minutes after administration
MINERVA ANESTESIOLOGICA
August 2015
PREOXYGENATION AND GENERAL ANESTHESIABOUROCHE
Table I.—Period of apnea and onset of desaturation after rapid sequence induction. Naguib used propofol and fentanyl and
Heier thiopental 5mg.kg-1.
% patients with desaturation
Times mean±SD in minutes to spontaneous of
diaphragmatic movements
Naguib 2005 12
Heier 2001 70
Naguib 2005 12
of 0.6 and 1 mg.kg-1, respectively.13 Reversal of
profound high-dose rocuronium-induced neuromuscular block used for rapid sequence induction (1.2 mg/kg) with sugammadex (16 mg/
kg) was significantly faster than spontaneous recovery from succinylcholine (1 mg/kg): 6.2±1.8
minutes versus 10.9±2.4 minutes respectively.14
Rapid sequence induction with rocuronium followed by reversal with sugammadex allowed earlier re-establishment of spontaneous ventilation
than with succinylcholine (216 seconds versus
406 seconds respectively).15 Thus, the choice of
the rocuronium would increase the margin of
safety for a resumption of spontaneous ventilation after a rapid sequence induction.
Anesthesia and spontaneous breathing
Outside of pediatrics, no study has evaluated
the risk factors for desaturation during induction
with spontaneous ventilation. For most hypnotics, alveolar hypoventilation exhibits a dose-dependent effect.16 In the absence of sedation, airway local anesthesia decreases inspiratory flows
for a duration of approximately 45 minutes.17
Succinylcholine
0.65 mg.kg-1
Succinylcholine
1 mg.kg-1
45%
65%
2.7±1.2
4.8±2.5
85%
42%
4.7±1.3
Infection of the upper respiratory tract is noted
to increase the risk of desaturation during induction.18
Resuscitation and prehospital emergency
Variability exists among published incidences
of desaturation in various studies 22, 23 but it is
reported to reach 60% during prehospital intubation.22 In emergency medicine, desaturation
occurs frequently, even in patients who are not
difficult to intubate and for whom the intubation
process is relatively rapid. Decreased FRC related to lung pathology (pulmonary edema, pneumonia, pulmonary contusion) is a determining
factor. Pulmonary aspiration and esophageal intubation are responsible for some cases of severe
desaturation during the intubation process.24 In
a prospective study, preoxygenation was found
to be effective (achieved PaO2>100 mmHg) in
7 of 8 cases in which the indication was the protection of the airway (coma) and in 5 of 34 cases
(15%) in which intubation was indicated due to
respiratory or cardiac failure.23
Preoxygenation
Desaturation in pediatrics
Desaturation episodes occur commonly in
children, with a frequency of 4-10% during induction and 20% during tracheal intubation.18
Desaturation occurs much more rapidly when
the child is young,19, 20 and the duration of apnea
before desaturation is linearly correlated with the
age of the patient. The less the child weighs, the
higher the incidence of severe arterial desaturation after reinstitution of manual ventilation
with 100% oxygen. It has been suggested that
a SpO2 of 95% might be the safe limit for apnea during induction of pediatric anesthesia.21
Vol. 81 - No. 8
Placebo
In cases in which there is a potential risk of
desaturation before securing the airway by endotracheal intubation, pre-oxygenation is highly
recommended during induction of anesthesia. In
its absence, the risk of desaturation is increased.
Preoxygenation techniques
The equipment must be adapted and tightly
fitted to the patient, particularly the face mask.
A morphological mismatch between the mask
and the face of the patient (e.g., inappropriate
mask size, presence of beards or moustaches)
MINERVA ANESTESIOLOGICA
913
BOUROCHE
PREOXYGENATION AND GENERAL ANESTHESIA
Table II.—Comparison of the different techniques of pre-oxygenation in normal subjects.
Preoxygenation technique used
Study
N.
Gambee AM 29
Fleureaux O 33
12 DAWD to 90% min
17 DAWD to 95% min
Pa O2 mmHg
24 DAWD to 95% min
40 DAWD to 90% min
22
Pa O2 mmHg
24
FeO2 %
24
FeO2 %
 5
FeO2 %
20
FeO2 %
20
FeO2 %
Baraka AS 34
Herriger A 30
Gold MI 71
Rooney MJ 32
Nimmagadda U 72
Pandit JJ 4
Gagnon C 73
Tanoubi I 36
Endpoint
TVB
4 DB 30s
8 DB 60s
8.9±10*
3±1*
397±49*
3.73±0.76
9.98±2.25*
350.4±35.8
91.9±3
88±5*
92±1*
89±3*
89±6
6.8±1.8
1.87±0.99
293±86
2.78±0.39
5.21±0.96*
339±33.9
90.8±3
80±5
83±9
76±7
TVB with
PEEP
AI + PEEP
7.83±2.63
87±3*
91±4*
94±4*
TVB: tidal volume breathing; DB: deep breaths; DAWD: duration of apnea without desaturation.
prevents a perfect seal and can cause failure.25
The mask must be applied securely on the face of
the patient; 20% dilution of O2 by ambient air
occurs when the mask is not tightly applied, and
40% dilution occurs when it is held close to the
face.26 The circle system with fresh gas flow (5
L.min– 1) is used as the standard for comparison
in anesthesia studies evaluating the effectiveness
of different circuits because it allows higher inspiratory flow rates.27 Some open circuit (Bain
or Magill) systems have been shown to be much
less effective.27 Before preoxygenation, the circuit and the reservoir should be filled with O2.
Three preoxygenation techniques are used: spontaneous breathing at FiO2 of 1 for 2 to 5 minutes, the “four vital capacities” method, and deep
breaths (Table II).
Spontaneous breathing at FiO2 of 1
The following technique of pre-oxygenation
that was initially proposed by Hamilton in
1955 is still the reference standard: 3 minutes of
spontaneous breathing at FiO2 of 1. In patients
with normal lung function, this provides denitrogenation with an FAO2 of close to 95%. The
denitrogenation is effective from the first minute
of the preoxygenation; nevertheless, circuit leakage cancels these effects by a rapid decrease of
the FiO2.28 Breathing pure O2 for longer than a
minute appears to have little benefit in terms of
SpO2 or denitrogenation alveolar, but positively
influences the duration of apnea before arterial
914
desaturation.29 In experiments performed with
healthy subjects, apnea time (with the exception
of an insufflation to verify tracheal intubation)
that is maintained until the SpO2 reaches over
90%, can be extended to almost 10 minutes after 3 minutes of classic pre-oxygenation. The apnea time can be increased by an additional two
minutes by application of positive pressure during the preoxygenation and by ventilation to the
mask after induction.30
Vital capacity maneuvers
The four vital capacities method is used in
cases in which patient cooperation is lacking.
The duration of apnea without desaturation is
shorter after four capacities maneuvers than with
spontaneous breathing. Technical requirements
are responsible for the limitations of this technic: bag capacity, inspiratory flow and room gas
inspiration. They are partly resolved by the addition of an additional 2 liter bag and a nonrebreathing “Ambu” valve. The vital capacity
maneuver preferably begins with a forced expiration to optimize the elevation of FeO2.31 To be
fully effective, the inspiratory O2 flow should be
greater than the peak inspiratory flow, which is
attained by activating the O2 system “by-pass”
during inspiration; 4 or 5 forced breaths of pure
O2 were found to be as efficient as conventional
pre-oxygenation assessed on the FeO2.32 However, these results were not confirmed when PaO2
was used for comparison; PaO2 was observed to
MINERVA ANESTESIOLOGICA
August 2015
PREOXYGENATION AND GENERAL ANESTHESIABOUROCHE
be lower after the four vital capacity maneuver
(293±86 mmHg) than after spontaneous ventilation in pure O2 (397±48 mmHg) (Table II).33
Deep breathing method
Eight deep breaths within 60 seconds at an
oxygen flow of 10 L per min constitutes a simple method of preoxygenation. This technique
results in a mean arterial oxygen tension of
369±69 mmHg, which is not significantly different from the value achieved by 3 minutes of
tidal volume breathing at an oxygen flow of 5
L per minute.34 The voluntary hyperventilation
technique (1 minute at FiO2 1 followed by 2
minutes of voluntary hyperventilation) has been
proposed to prevent post-apneic hypercapnia.
PaCO2 after intubation was similar, compared
to a control, when either hyperventilation before
induction or 3 min normal breathing was used
as the pre-oxygenation technique.35
Pressure support ventilation
In healthy volunteers, PSV has been shown to
improve the quality of pre-oxygenation by two
mechanisms: acceleration of nitrogen washout
and better contact between the mask and the
face. In a healthy volunteer study,36 FEO2 after 3
minutes of preoxygenation was higher (p<0.001)
with PSV 4 cm H2O/PEEP 4 (94±3%) and PSV
6 cm H2O/PEEP 4 (94±4%) than with the
standard technique (89±6%). One hundred percent and 90% of the participants reached 90%
FEO2 with PSV 4 and 6 cmH20 respectively vs.
65% with spontaneous breathing at FiO2 of 1
(P=0.0013). Clinical tolerance was impaired at
the highest level of pressure tested.
Preoxygenation failure
Inadequate preoxygenation, defined as an
FeO2<90% after three minutes of tidal volume
breathing, is seen frequently in practice (56% in
a sample of 1050 patients).37 The effective FiO2
delivered was observed to be lower in patients
with a FeO2<90%. Risk factors for inadequate
preoxygenation were determined to be bearded
male, beardless male, ASA>1, lack of teeth, and
Vol. 81 - No. 8
age >55 years. These predictive factors overlap
with those previously associated with difficult
mask ventilation.
While SpO2 measurement is not informative regarding the quality of pre-oxygenation
maneuvers, it is essential to identify oxygenation
problems. The FeO2 depends on the tidal volume; small tidal volumes increase the difference
between FeO2 and FAO2, leading to overestimation of FAO2. The CO2 wave shape is informative regarding the quality of the ventilation and
the tightness of the circuit. A FeO2 of <90% indicates incomplete denitrogenation at the FRC
level. In a study of 40 volunteers,25 9 subjects
were unable to attain FeO2>90%. Even if the
mechanisms that cause incomplete denitrogenation are not identified, this monitoring method
has utility in routine practice.2 If the FeO2 cannot be increased above 90%, PSV may be proposed to improve preoxygenation quality.38 In
emergency medicine, the monitoring of preoxygenation is typically based on SpO2 measurement and pre-oxygenation duration, as the FeO2
is not usually available.
Morbidly obese patients
In the obese, the decrease in FRC, increase in
O2 consumption, and heterogeneity of the ventilation/perfusion (V/Q) ratio result in a decrease
in the time required for alveolar denitrogenation 39, 40 and a decrease in O2 stores, which in
turn reduce the duration of apnea tolerance.
After 3 minutes of classic preoxygenation,
obese patients can tolerate apnea of 3 minutes
duration while maintaining SpO2 higher than
90%, and the time required to increase saturation above 96% after desaturation is 37 seconds,
which is longer than the 22 seconds required
in healthy subjects.40 Pulmonary abnormalities are correlated to Body Mass Index and are
responsible for early desaturation, before complete muscle relaxation and intubation.39 The
effectiveness of spontaneous ventilation and
eight deep breaths as preoxygenation methods
are comparable in the obese when regarding Fe
O2 and the duration of apnea before the SpO2
reaches 95%.41 Continuous Positive Airway
Pressure (CPAP) (7.5 cm H2O versus Mapleson
MINERVA ANESTESIOLOGICA
915
BOUROCHE
PREOXYGENATION AND GENERAL ANESTHESIA
circuit) during spontaneous ventilation in pure
O2 were observed not to improve the duration of
apnea (240 and 203 seconds CPAP versus zero
end expiratory pressure, respectively).42 When
CPAP (10 cm H2O) is followed by a PSV with
PEEP, post-intubation PaO2 is significantly improved.43
Because lung volume reduction is the main
factor causing impaired oxygenation in the
obese patient, use of PSV with PEEP has been
proposed. PSV improves the quality of preoxygenation in the obese, probably due to improved
alveolar ventilation and recruitment.44 Compared to five minutes of spontaneous ventilation
with FiO2 of 1, PSV results in increased FeO2
(96.9%±1.3% vs. 94.1%±2.0%) and acceleration of nitrogen elimination (185.3±46.1 vs.
221±41.5 s).45 However, this gain was not associated with an increase in the duration of
tolerable apnea, defined by a 95% SpO2 cutoff
value. This is due to the small increase in O2 reserve, which can be estimated at 58 mL O2 for
a 2000 mL FRC (2000x2, 8%=58 mL). When
associated with a recruitment maneuver, the effectiveness of PSV is statistically significant regarding the arterial oxygenation.46 In morbidly
obese patients, CPAP of 5 cm H2O combined
with PSV of 5 cm H2O during preoxygenation
resulted in better oxygenation compared with
neutral-pressure breathing, and it prevented
desaturation episodes.47 The postintubation
PaO2 was significantly higher in the CPAP/PSV
group (32.2±4.1 kPa) than in the control group
(23.8±8.8 kPa) (P<0.001). In the control group,
the nadir of oxygen saturation was lower (median 98%, range 83-99%) than in the CPAP/PSV
group. The PaO2 is greater and the time of apnea
extended when preoxygenation (either CPAP for
3 minutes 48 or 8 deep breaths) 49 is performed
in a 25° head-up position compared to a supine
position. After preoxygenation in the head-up
position, the time to a SpO2 of lower than 92%
was always measured at longer than 3 minutes.
Preoxygenation in pregnancy
In the parturient, the time required for complete denitrogenation (FeN2=2%) is shorter
than in non-parturient young women as follows:
916
104±30 seconds between 13-26 weeks of pregnancy, 80±20 seconds between 26-42 weeks, and
130±30 seconds in controls, due to the reduction in the FRC during pregnancy.50 Spontaneous ventilation in FiO2 of 1 for 3 minutes and
the four vital capacities method for 30 seconds
give comparable results, whether judged by
PaO2 51 or apnea duration.52 Some women were
observed to have a tolerable apnea duration of
only approximately 60 seconds; this short delay
carries obvious risk.52 The shortening of the time
to FeO2 of 90% (average 107 seconds) is a good
argument supporting recommendation of the 8
deep breaths technique during obstetric emergencies.53
Preoxygenation and chronic obstructive pulmonary
disease
In patients with chronic obstructive pulmonary disease, the time needed to decrease the
alveolar fraction of nitrogen (FAN2) from 78%
to 2% can exceed 30 minutes, as it is inversely
proportional to the peak expiratory flow rate.54
FetO2 monitoring is used to evaluate the time
necessary for preoxygenation in such patients.55
Preoxygenation in pediatrics
In pediatrics, the respiratory physiology of
young children is particularly age-specific. The
inhibition of intercostal tonus by general anesthesia is responsible for a reduction in FRC.
Hypoxemia arises more quickly in children because of a higher VA/FRC ratio, a higher O2
consumption, and lower O2 reserves. Children
exhibit a delay before reaching FeO2 close to
90% of approximately 80 to 90 seconds when
breathing at FiO2 of 1.56 While young children
show a rapid drop in saturation, they also reach
a FeO2 of 0.9 more quickly.56 After a period of
at least 2 minutes breathing at FiO2 of 1 and
after muscle paralysis, the duration of apnea before the SpO2 reaches 90% is found to be 96.5
seconds in children less than 6 months of age,
160.4 seconds in 2- to 5-year-olds, and 382.4
seconds in 11- to 18-year-olds.57 In children
younger than 6 months, even shorter apnea
time limits, on the order of 70-90 seconds, have
MINERVA ANESTESIOLOGICA
August 2015
PREOXYGENATION AND GENERAL ANESTHESIABOUROCHE
been reported.19 The duration of apnea required
to reach a SpO2 of 98%, 95%, or 90%, is significantly increased when the preoxygenation is
extended for 1 to 2 minutes, but no benefit was
found by extension past 3 minutes.21 When the
gas mixture used during pre-oxygenation passes
from an average FiO2 of approximately 93% to
39%, the duration of apnea until a 95% SpO2
decreases from 210 to 71 seconds.58
Intensive care patients and emergency medicine
In emergency medicine, all of the patients
can be considered to be at risk for desaturation
during airway control and thus preoxygenation
should be recommended as part of routine practice. In emergency nonsurgical intubation, preoxygenation is difficult to achieve. The benefit
of the pre-oxygenation is probably greater in
patients who do not have respiratory illness at
the time of intubation.23 Thus, all patients who
are intubated for neurological distress should
benefit from a careful preoxygenation of at least
3 minutes in duration, even if a lack of patient
cooperation limits its effectiveness. During intubation of hypoxemic patients, pre-oxygenation
using PSV is more effective at reducing arterial
desaturation than the usual method.59 At the end
of the preoxygenation period, SpO2 was higher
in the PSV group than in the control group
(98±2 vs. 93±6%, P<0.001). During the intubation procedure, lower SpO2 values were observed in the control group (81±15 vs. 93±8%,
P<0.001). Twelve (46%) of the patients in the
control group and two (7%) in the PSV group
had a SpO2 below 80% (P<0.01).
Apneic oxygenation
It is possible to maintain oxygenation during a long period of apnea by administering 10
to 15 L per min of continuous oxygen into the
pharynx. However, this method is only effective following a complete preoxygenation. Apnea of longer than 30 minutes in duration has
been reported to result in severe hypercapnia
(>150 mmHg) without damage to the patient.
Apneic oxygenation failures are related to failure
of the preoxygenation procedure and to reduced
Vol. 81 - No. 8
FRC.60 An indirect method of apneic oxygenation is the administration of O2 during intubation attempts, and the administration of O2 at a
rate of 3 L per min by a naso- or oro-pharyngeal
catheter can significantly delay the onset of arterial O2 desaturation.61 Similar results have been
reported more recently in ASA 1-2 patients after
preoxygenation. This method is easy to apply
and it confers a definite advantage in patients
without respiratory pathology 62 and probably in
morbidly obese patients as well.63
Management of failures of
preoxygenation and oxygenation
As the risk factors for pre-oxygenation failures
and difficult mask ventilation are similar, such
at-risk patients should be identified and carefully
monitored. If FeO2 is lower than 0.9 after preoxygenation, alternative methods of oxygenation
should be immediately available. Knowing that
none of techniques is 100% reliable, it is essential to be able to provide several methods. The
equipment must be immediately available and
the team must be familiar with its use. The most
popular device is the intubating laryngeal mask
airway (ILMA). The ventilation is of good quality in the vast majority of cases and oxygenation
failures are rare when using this device.64 However, only the No. 3 size exists for use with child
patients, and little data are available about ILMA
use in pediatrics.65. For patients of less than 30
kg, the standard laryngeal mask is used, with the
awareness that implementation is more difficult
and not always successful.
In the event of ventilation failure with the facial or laryngeal mask, rescue trans-tracheal oxygenation is to be considered. Inter-crico-thyroid
membrane puncture is straightforward in 98%
of non-emergency cases.66 Because the use of
transtracheal ventilation in emergency medicine
is very rare, studies on this subject include only
very few patients. The success rate of the emergency puncture procedure is unknown.67
Jet ventilation is administered using a manual injector with operator control or using a jet
ventilator with control of the driving pressure.
The major risk is the possibility of pulmonary
barotrauma by lung overdistension, the impact
MINERVA ANESTESIOLOGICA
917
BOUROCHE
PREOXYGENATION AND GENERAL ANESTHESIA
of which can be serious in this context.68 It is
important to closely monitor the quality of expiration and keep in mind that the outflow of a
14 gauge catheter to a driving pressure of 3 bars
is approximately 600 mL per second.69 With O2
consumption being approximately 300 mL per
minute, the injection duration and respiratory
rate are limited to a minimum.
Conclusions
It is of particular importance to consider the
issues related to oxygenation because O2 reserves
are low and the difficulties of intubating the
patient and providing adequate ventilation are
often associated. The situation becomes critical
when O2 reserves are insufficient. Efficient technique and FeO2 monitoring can improve the effectiveness of the pre-oxygenation and thereby
increase the margin of safety. After pre-oxygenation, supplemental O2 increases the duration
of tolerable apnea in most cases, and this very
simple measure should not be neglected. Failures
of pre-oxygenation must be identified and alternative methods of oxygenation should be available for rapid and facile implementation. To this
end, these methods should be taught and practiced on models or during simulation courses, so
teams are prepared if the need arises.
Key messages
—— Effective preoxygenation (FeO2 >90%)
is essential to avoid hypoxemia during airway
management.
—— Preoxygenation can be improved by
use of a seated position (20° to 30°), PSV,
and/or PEEP, especially in obese patients.
—— Some induction situations present
higher risk: pregnancy, obesity, rapid induction sequence and require special attention.
—— Those situations at risk may be anticipated by identifying risk factors.
References
  1. Bourgain JL, Chastre J, Combes X, Orliaguet G. [Oxygen
arterial desaturation and upholding the oxygenation during
intubation: question 2. Societe Francaise d’Anesthesie et de
Reanimation]. Ann Fr Anesth Reanim 2008;27:15-25.
918
  2. Campbell IT, Beatty PC. Monitoring preoxygenation. Br J
Anaesth 1994;72:3-4.
 3. Lauscher P, Mirakaj V, Koenig K, Meier J. Why hyperoxia matters during acute anemia. Minerva Anestesiol
2013;79:643-51.
  4. Pandit JJ, Duncan T, Robbins PA. Total oxygen uptake with
two maximal breathing techniques and the tidal volume
breathing technique: a physiologic study of preoxygenation.
Anesthesiology 2003;99:841-6.
  5. Reber A, Engberg G, Wegenius G, Hedenstierna G. Lung
aeration. The effect of pre-oxygenation and hyperoxygenation during total intravenous anaesthesia. Anaesthesia
1996;51:733-7.
  6. Serafini G, Cornara G, Cavalloro F, Mori A, Dore R, Marraro G et al. Pulmonary atelectasis during paediatric anaesthesia: CT scan evaluation and effect of positive endexpiratory pressure (PEEP). Paediatr Anaesth 1999;9:225-8.
  7. Edmark L, Auner U, Enlund M, Ostberg E, Hedenstierna
G. Oxygen concentration and characteristics of progressive
atelectasis formation during anaesthesia. Acta Anaesthesiol
Scand 2011;55:75-81.
 8.Neumann P, Rothen HU, Berglund JE, Valtysson J,
Magnusson A, Hedenstierna G. Positive end-expiratory
pressure prevents atelectasis during general anaesthesia even
in the presence of a high inspired oxygen concentration.
Acta Anaesthesiol Scand 1999;43:295-301.
  9. Tusman G, Bohm SH, Vazquez de Anda GF, do Campo JL,
Lachmann B. ‘Alveolar recruitment strategy’ improves arterial oxygenation during general anaesthesia. Br J Anaesth
1999;82:8-13.
10. Pelosi P, Ravagnan I, Giurati G, Panigada M, Bottino N,
Tredici S et al. Positive end-expiratory pressure improves respiratory function in obese but not in normal subjects during anesthesia and paralysis. Anesthesiology 1999;91:122131.
11. Hayes AH, Breslin DS, Mirakhur RK, Reid JE, O’Hare RA.
Frequency of haemoglobin desaturation with the use of succinylcholine during rapid sequence induction of anaesthesia. Acta Anaesthesiol Scand 2001;45:746-9.
12. Naguib M, Samarkandi AH, Abdullah K, Riad W, Alharby
SW. Succinylcholine dosage and apnea-induced hemoglobin desaturation in patients. Anesthesiology 2005;102:3540.
13. El Orbany MI, Joseph NJ, Salem MR, Klowden AJ. The
neuromuscular effects and tracheal intubation conditions after small doses of succinylcholine. Anesth Analg
2004;98:1680-5, table.
14. Lee C, Jahr JS, Candiotti KA, Warriner B, Zornow MH,
Naguib M. Reversal of profound neuromuscular block by
sugammadex administered three minutes after rocuronium:
a comparison with spontaneous recovery from succinylcholine. Anesthesiology 2009;110:1020-5.
15. Sorensen MK, Bretlau C, Gätke MR, Sørensen AM, Rasmussen LS. Rapid sequence induction and intubation with
rocuronium-sugammadex compared with succinylcholine:
a randomized trial. Br J Anaesth 2012;108:682-9.
16. Lee YS. Effects of propofol target-controlled infusion on
haemodynamic and respiratory changes with regard to safety. J Int Med Res 2004;32:19-24.
17. Liistro G, Stănescu DC, Veriter C, Rodenstein DO, D’Odemont JP. Upper airway anesthesia induces airflow limitation
in awake humans. Am Rev Respir Dis 1992;146:581-5.
18. Rolf N, Cote CJ. Frequency and severity of desaturation
events during general anesthesia in children with and without upper respiratory infections. J Clin Anesth 1992;4:2003.
19. Dupeyrat A, Dubreuil M, Ecoffey C. Preoxygenation in
children. Anesth Analg 1994;79:1027.
20. Xue FS, Luo LK, Tong SY, Liao X, Deng XM, An G. Study
of the safe threshold of apneic period in children during
anesthesia induction. J Clin Anesth 1996;8:568-74.
MINERVA ANESTESIOLOGICA
August 2015
PREOXYGENATION AND GENERAL ANESTHESIABOUROCHE
21. Xue F, Luo L, Tong S, Liao X, Tang G, Deng X. Children’s
development effecting blood oxygen desaturation following
apnea. Chin Med J (Engl) 1995;108:434-7.
22. Dunford JV, Davis DP, Ochs M, Doney M, Hoyt DB. Incidence of transient hypoxia and pulse rate reactivity during paramedic rapid sequence intubation. Ann Emerg Med
2003;42:721-8.
23.Mort TC. Preoxygenation in critically ill patients requiring emergency tracheal intubation. Crit Care Med
2005;33:2672-5.
24. Mort TC. Esophageal intubation with indirect clinical tests
during emergency tracheal intubation: a report on patient
morbidity. J Clin Anesth 2005;17:255-62.
25. Berry CB, Myles PS. Preoxygenation in healthy volunteers:
a graph of oxygen «washin» using end-tidal oxygraphy. Br J
Anaesth 1994;72:116-8.
26. McGowan P, Skinner A. Preoxygenation--the importance of
a good face mask seal. Br J Anaesth 1995;75:777-8.
27. Nimmagadda U, Salem MR, Joseph NJ, Lopez G, Megally
M, Lang DJ et al. Efficacy of preoxygenation with tidal volume breathing. Comparison of breathing systems. Anesthesiology 2000;93:693-8.
28. Russell GN, Smith CL, Snowdon SL, Bryson TH. Preoxygenation and the parturient patient. Anaesthesia
1987;42:346-51.
29.Gambee AM, Hertzka RE, Fisher DM. Preoxygenation
techniques: comparison of three minutes and four breaths.
Anesth Analg 1987;66:468-70.
30. Herriger A, Frascarolo P, Spahn DR, Magnusson L. The effect of positive airway pressure during pre-oxygenation and
induction of anaesthesia upon duration of non-hypoxic apnoea. Anaesthesia 2004;59:243-7.
31. McCrory JW, Matthews JN. Comparison of four methods
of preoxygenation. Br J Anaesth 1990;64:571-6.
32. Rooney MJ. Pre-oxygenation: a comparison of two techniques using a Bain system. Anaesthesia 1994;49:629-32.
33. Fleureaux O, Estèbe JP, Bléry C, Douet N, Mallédant Y.
[Effects of preoxygenation methods on the course of PaO2
and PaCO2 in anesthetic post-induction apnea]. Cah Anesthesiol 1995;43:367-70.
34. Baraka AS, Taha SK, Aouad MT, El-Khatib MF, Kawkabani NI. Preoxygenation: comparison of maximal breathing and tidal volume breathing techniques. Anesthesiology
1999;91:612-6.
35. Choinière A1, Girard F, Boudreault D, Ruel M, Girard DC.
Voluntary hyperventilation before a rapid-sequence induction of anesthesia does not decrease postintubation PaCO2.
Anesth Analg 2001;93:1277-80.
36. Tanoubi I, Drolet P, Fortier LP, Donati F. [Inspiratory support versus spontaneous breathing during preoxygenation
in healthy subjects. A randomized, double blind, cross-over
trial]. Ann Fr Anesth Reanim 2010;29:198-203.
37. Baillard C, Depret F, Levy V, Boubaya M3, Beloucif S. et
al. Incidence and prediction of inadequate preoxygenation
before induction of anaesthesia. Ann Fr Anesth Reanim
2014;33:e55-e58.
38. Tanoubi I, Drolet P, Donati F. Optimizing preoxygenation
in adults. Can J Anaesth 2009;56:449-66.
39.Berthoud MC, Peacock JE, Reilly CS. Effectiveness of
preoxygenation in morbidly obese patients. Br J Anaesth
1991;67:464-6.
40. Jense HG, Dubin SA, Silverstein PI, O’Leary-Escolas U.
Effect of obesity on safe duration of apnea in anesthetized
humans. Anesth Analg 1991;72:89-93.
41. Rapaport S, Joannes-Boyau O, Bazin R, Janvier G. Comparaison de la technique de préoxygénation à huit capacités vitales et à volume courant chez les patientes ayant une
obésité morbide. Ann Fr Anesth Reanim 2004;23:1155-9.
42. Cressey DM, Berthoud MC, Reilly CS. Effectiveness of continuous positive airway pressure to enhance pre-oxygenation
in morbidly obese women. Anaesthesia 2001;56:680-4.
Vol. 81 - No. 8
43. Coussa M, Proietti S, Schnyder P, Frascarolo P, Suter M,
Spahn D et al. Prevention of atelectasis formation during
the induction of general anesthesia in morbidly obese patients. Anesth Analg 2004;98:1491-5, table.
44. Solis A, Baillard C. [Effectiveness of preoxygenation using
the head-up position and noninvasive ventilation to reduce
hypoxaemia during intubation]. Ann Fr Anesth Reanim
2008;27:490-4.
45. Delay JM, Sebbane M, Jung B, Nocca D, Verzilli D, Pouzeratte Y et al. The effectiveness of noninvasive positive
pressure ventilation to enhance preoxygenation in morbidly obese patients: a randomized controlled study. Anesth
Analg 2008;107:1707-13.
46. Futier E, Constantin JM, Pelosi P, Chanques G, Massone
A, Petit A et al. Noninvasive ventilation and alveolar recruitment maneuver improve respiratory function during
and after intubation of morbidly obese patients: a randomized controlled study. Anesthesiology 2011;114:135463.
47. Harbut P, Gozdzik W, Stjernfalt E, Marsk R, Hesselvik JF.
Continuous positive airway pressure/pressure support preoxygenation of morbidly obese patients. Acta Anaesthesiol
Scand 2014;58:675-80.
48. Dixon BJ, Dixon JB, Carden JR, Burn AJ, Schachter LM,
Playfair JM et al. Preoxygenation is more effective in the
25 degrees head-up position than in the supine position in
severely obese patients: a randomized controlled study. Anesthesiology 2005;102:1110-5.
49. Altermatt FR, Munoz HR, Delfino AE, Cortinez LI. Preoxygenation in the obese patient: effects of position on tolerance to apnoea. Br J Anaesth 2005;95:706-9.
50. Byrne F, Oduro-Dominah A, Kipling R. The effect of pregnancy on pulmonary nitrogen washout. A study of preoxygenation. Anaesthesia 1987;42:148-50.
51.Norris MC, Dewan DM. Preoxygenation for cesarean
section: a comparison of two techniques. Anesthesiology
1985;62:827-9.
52. Bernard F, Louvard V, Cressy ML. Préoxygénation avant
induction pour césarienne. Ann Fr Anesth Reanim
1994;13:2-5.
53. Chiron B., Laffon M, Ferrandiere M, Pittet JF, Marret H,
Mercier C. Standard preoxygenation technique versus two
rapid techniques in pregnant patients. Int J Obstet Anest
2005;14:79.
54. Selsy DS, Jones JG. Respiratory function and the safety of
anesthaesia. In: Taylor TH, Major E, editors. Hazard and
complications of anaesthesia. Second Edition. London,
UK: Churchill Livingstone; 1995. p. 213-21.
55. Samain E, Biard M, Farah E, Holtzer S, Delefosse D, Marty
J. Monitorage de la fraction expirée d’oxygène lors de la
préoxygénation dans la bronchopathie chronique obstructive. Ann Fr Anesth Reanim 2002;21:14-9.
56.Morrison JE, Jr., Collier E, Friesen RH, Logan L. Preoxygenation before laryngoscopy in children: how long is
enough? Paediatr Anaesth 1998;8:293-8.
57. Patel R, Lenczyk M, Hannallah RS, McGill WA. Age and
the onset of desaturation in apnoeic children. Can J Anaesth 1994;41:771-4.
58. Kinouchi K, Fukumitsu K, Tashiro C, Takauchi Y, Ohashi
Y, Nishida T. Duration of apnoea in anaesthetized children
required for desaturation of haemoglobin to 95%: comparison of three different breathing gases. Paediatr Anaesth
1995;5:115-9.
59. Baillard C, Fosse JP, Sebbane M, Chanques G, Vincent F,
Courouble P et al. Noninvasive ventilation improves preoxygenation before intubation of hypoxic patients. Am J
Respir Crit Care Med 2006;174:171-7.
60. Fraioli RL, Sheffer LA, Steffenson JL. Pulmonary and cardiovascular effects of apneic oxygenation in man. Anesthesiology 1973;39:588-96.
61. Teller LE, Alexander CM, Frumin MJ, Gross JB. Pharyn-
MINERVA ANESTESIOLOGICA
919
BOUROCHE
PREOXYGENATION AND GENERAL ANESTHESIA
geal insufflation of oxygen prevents arterial desaturation
during apnea. Anesthesiology 1988;69:980-2.
62. Taha SK, Siddik-Sayyid SM, El-Khatib MF, Dagher CM,
Hakki MA, Baraka AS. Nasopharyngeal oxygen insufflation
following pre-oxygenation using the four deep breath technique. Anaesthesia 2006;61:427-30.
63. Ramachandran SK, Cosnowski A, Shanks A, Turner CR.
Apneic oxygenation during prolonged laryngoscopy in
obese patients: a randomized, controlled trial of nasal oxygen administration. J Clin Anesth 2010;22:164-8.
64.Ferson DZ, Rosenblatt WH, Johansen MJ, Osborn I,
Ovassapian A. Use of the intubating LMA-Fastrach in 254
patients with difficult-to- manage airways. Anesthesiology
2001;95:1175-81.
65. Weiss M, Schwarz U, Dillier C, Fischer J, Gerber AC. Use
of the intubating laryngeal mask in children: an evaluation
using video-endoscopic monitoring. Eur J Anaesthesiol
2001;18:739-44.
66. Bourgain JL, Desruennes E, Fischler M, Ravussin P. Transtracheal high frequency jet ventilation for endoscopic airway
surgery: a multicentre study. Br J Anaesth 2001;87:870-5.
67. Combes X, Jabre P, Margenet A, Merle JC, Leroux B, Dru
M et al. Unanticipated difficult airway management in the
prehospital emergency setting: prospective validation of an
algorithm. Anesthesiology 2011;114:105-10.
68. Parmet JL, Colonna-Romano P, Horrow JC, Miller F, Gonzales J, Rosenberg H. The laryngeal mask airway reliably
provides rescue ventilation in cases of unanticipated difficult tracheal intubation along with difficult mask ventilation. Anesth Analg 1998;87:661-5.
69. Bourgain JL, Chollet M, Fischler M, Gueret G, Mayne A;
membres du conseil du club en anesthésie en ORL. [Guide
for the use of jet-ventilation during ENT and oral surgery].
Ann Fr Anesth Reanim 2010;29:720-7.
70. Heier T, Feiner JR, Lin J, Brown R, Caldwell JE. Hemoglobin desaturation after succinylcholine-induced apnea: a
study of the recovery of spontaneous ventilation in healthy
volunteers. Anesthesiology 2001;94:754-9.
71. Gold MI, Duarte I, Muravchick S. Arterial oxygenation in
conscious patients after 5 minutes and after 30 seconds of
oxygen breathing. Anesth Analg 1981;60:313-5.
72.Nimmagadda U, Chiravuri SD, Salem MR, Joseph NJ,
Wafai Y, Crystal GJ et al. Preoxygenation with tidal volume
and deep breathing techniques: the impact of duration of
breathing and fresh gas flow. Anesth Analg 2001;92:133741.
73. Gagnon C, Fortier LP, Donati F. When a leak is unavoidable, preoxygenation is equally ineffective with vital capacity or tidal volume breathing. Can J Anaesth 2006;53:8691.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript.
Received on October 28, 2014. - Accepted for publication on June 3, 2015. - Epub ahead of print on June 5, 2015.
Corresponding author: J. Bouroche, Département d’Anesthésie, Gustave Roussy Cancer Campus Grand-Paris, 114 rue Edouard Vaillant,
94800 Villejuif. E-mail: [email protected]
920
MINERVA ANESTESIOLOGICA
August 2015

REVIEW
Statin therapy in critically-ill patients
with severe sepsis: a review and metaanalysis of randomized clinical trials
G. THOMAS 1, 2, S. HRAIECH 1, 2, A. LOUNDOU 3, J. TRUWIT 4, P. KRUGER 5
D. F. MCAULEY 6, 7, L. PAPAZIAN 1, 2, A. ROCH 1, 2
1Assistance Publique–Hôpitaux de Marseille, Hôpital Nord, Réanimation des Détresses Respiratoires et des Infections
Sévères; 2UMR-CNRS 7278, Aix-Marseile Université, Marseille, France; 3APHM, Faculté de Médecine Timone,
Laboratoire de Santé publique, Marseille, France; 4Pulmonary and Critical Care Medicine, Froedtert and Medical
College of Wisconsin, Milwaukee, WI, USA; 5Intensive Care Unit, Princess Alexandra Hospital, Wooloongabba,
Brisbane, Australia; 6Centre for Infection and Immunity, Queen’s University of Belfast, Belfast, UK; 7Regional Intensive
Care Unit, Royal Victoria Hospital, Belfast, Northern Ireland, UK
ABSTRACT
While statins are indicated to reduce blood cholesterol levels, they also have anti-inflammatory and immunomodulatory effects. Several observational cohort studies suggested that statins may improve survival and reduce complications in patients with sepsis. Recent randomized controlled studies in critically ill patients have been conducted
and published. In this paper we present a meta-analysis of these randomized trials. Methods: An electronic article
search through PubMed was performed. Only randomized controlled trials including critically ill adult patients with
severe sepsis were retained. A meta-analysis was performed as detailed in text below. Overall analysis including 1818
patients total from 4 studies showed that there was no difference in 60-day mortality between statins (223/903) and
placebo (233/899) [risk ratio, 0.930; 95% CI, 0.722 to 1.198]. Similarly, no difference in 28-day mortality was
observed between groups (statins 191/907, placebo 199/911; risk ratio 0.953; 95% CI, 0.715 to 1.271). The results
of this meta-analysis confirm that the use of statin therapy should not be recommended in the management of severe
sepsis in critically ill patients. Statins should be continued with caution and only if necessary, as one study reported
that the statin group had a higher rate of hepatic and renal failure. (Minerva Anestesiol 2015;81:921-30)
Key words: Sepsis - Intensive care - Meta-analysis.
M
ortality rate related to severe sepsis and
septic shock has decreased significantly as
a result of aggressive intravenous fluid management, appropriate and early broad spectrum antibiotics and early removal of the source of infection.1 Nonetheless, mortality of sepsis remains
high despite these improvements in clinical
management and new therapies are still needed.
Despite a better understanding of the mechanisms involved in septic response, targeted immunotherapies have not improved survival rates.
Statins have antioxidant and antiapoptotic ef-
Vol. 81 - No. 8
fects and have demonstrated the ability to reduce
the production of pro-inflammatory cytokines
known to be detrimental in the development
and progression of sepsis.2
To date, numerous observational studies on
the interest of statins in sepsis have been published. Meta-analyses 3-7 suggest potential beneficial effect of statins on mortality. More recent meta-analyses have also pooled data from
randomized studies.6, 7 However, those studies
were conducted in patients with variable severity of sepsis. Several, randomized controlled tri-
MINERVA ANESTESIOLOGICA
921
THOMAS
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSIS
als of statins versus placebo, have been recently
conducted in patients with severe sepsis,8 with
ventilator-associated pneumonia 9 and with
acute respiratory distress syndrome.10, 11 The aim
of this review was to perform a meta-analysis of
randomized controlled studies conducted in this
population and summarize the literature on the
role of statins in improving outcome of sepsis in
critically-ill adult patients.
Rationale for statin therapy in sepsis
Statins inhibit the conversion of hydroxy-methyl-glutaryl-coenzyme A to mevalonate, an early rate-limiting step in cholesterol biosynthesis,
thereby reducing total cholesterol.2 Beyond their
lipid-lowering effect, statins also have pleiotropic
properties including anti-inflammatory, antioxidant, immunomodulatory and antithrombotic
effects 12-14 which could prevent or curtail sepsis.
Traditionally, the host immune response to
sepsis has been described as an overly exuberant inflammation responsible for endothelial
dysfunction and injury leading to organ failures.15 The concept of sepsis as a cytokine storm
emerged, as several trials described increased
concentration of various cytokines in patients
with sepsis.16, 17 Later, Bone et al. advanced
the idea that the initial inflammatory response
gave way to a subsequent “compensatory antiinflammatory response syndrome”.18 However,
recent studies have shown that infection triggers
a much more complex and prolonged host response where both pro-inflammatory and antiinflammatory responses occur early and simultaneously.19, 20 Although a large part of septic
patients die from multiorgan failure, a persistent
deficiency of both innate and adaptive immunity leads a marked immunosuppressive state,21
resulting in deaths from inability to clear primary infections and/or development of secondary
infections. A persistent activation of innate immunity which results in prolonged hyperinflammation is responsible for organ injury and late
deaths.22 Therapies with immunomodulatory
properties such as statins may favorably influence the evolution of sepsis. First, an improvement in endothelial dysfunction and apoptosis,
which play a crucial role in the pathogenesis of
sepsis,23 has been shown with statins (Table I).
Second, statins have various immunomodulatory effects. Although these properties could
reduce ability to clear infection, they have been
associated with an improvement of prognosis in
animal models of sepsis.24, 25 Statins reduce geranylpyrophosphate and farnesylpyrophosphate
availability, by interacting in cholesterol biosynthesis pathway, which are major components of
subcellular binding sites for small GTP-binding
protein. GTP-binding proteins have crucial
roles in intracellular inflammatory signaling.2
Moreover, statins reduce the expression of proinflammatory cytokines 26, 27 and also reduce Creactive protein level (CRP), which is associated
with organ dysfunction and death in critically
ill patients.28, 29 Third, statins inhibit leucocyte
movement by reducing adhesions molecules expression.30 Statins could finally modulate adaptive immunity through the direct inhibition of
MHC-II expression by monocytes and macrophages.31
Whether some statins have more anti-inflammatory power than others is unclear since they
have rarely been compared. However, simvastatin has been used for a majority of experimental studies and has exhibited the largest range of
effects and in vivo properties.2 Different statins
may have sometimes opposing effects. For exam-
Table I.—Main immunomodulatory and therapeutic actions of statins.
Improvement in endothelial dysfunction and apoptosis 23
Reduction of components of subcellular binding sites for small GTP-binding protein, affecting intracellular inflammatory signaling 2
Reduction of the expression of proinflammatory cytokines (IL-1, IL-6 and TNF alpha) 24, 25
Inhibition of leucocyte movement by reducing adhesions molecules expression 28
Inhibition of MHC-II expression by monocytes and macrophages 29
Modulation of coagulation by blunting monocyte tissue factor expression and reducing plasminogen activator inhibitor (PAI)-1
(cross-talk with inflammation)
Promotion of a favorable balance between constitutive and inducible NOS leading to improved hemodynamic stability 23
922
MINERVA ANESTESIOLOGICA
August 2015
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSISTHOMAS
ple, simvastatin reduced, whereas atorvastatin
enhanced superantigen-mediated T-cell activation in healthy volunteers.30
During sepsis, significant alteration of both
coagulation system and cells that regulate this
system also occur.33, 34 Statins have been shown
to modulate coagulation by blunting monocyte
tissue factor expression and reducing plasminogen activator inhibitor (PAI)-1 which interact
with the fibrinolytic system.2 The main mediators of inflammation-induced activation of coagulation are pro-inflammatory cytokines.35, 37
There is increasing evidence that extensive crosstalk between inflammation and coagulation exists, whereby inflammation leads to activation
of coagulation and components of the coagulation system may modulate the inflammatory
response.34 Therefore, the ubiquitary effects of
statins both on inflammation and coagulation
may be synergistic in sepsis.
Finally, during sepsis, endothelial constitutive
nitric oxide synthase (NOS) activity decreases,
while a delayed increase in inducible NOS leads
to an overproduction of nitric oxide, which is
responsible for vasodilatation, loss of vascular
resistance and vascular leak.38 Statin could contribute to a favorable balance between constitutive and inducible NOS leading to improved
hemodynamic stability.23
Clinical applications
Statins for prevention of severe sepsis
Observational studies of acute ward patients
suggest a preventive role for statins in sepsis.
The first published work was a prospective observational cohort study, in which all consecutive patients admitted to a medicine ward with a
known or presumed bacterial infection were included and divided into two groups on the basis
of whether they were taking statins for at least
one month before admission or not.39 Of the
361 patients enrolled, 82 (22.7%) were treated
with statins. Severity scores were similar between
groups. The primary outcomes were the development of severe sepsis, which was significantly
lower in the statin group than in the non-statin
group (2.4% vs. 19% respectively, P<0.001),
Vol. 81 - No. 8
and admission to an intensive care unit (ICU),
which was 3.7% in the statin group versus 12.2%
in the group without statins (P=0.025).39 The
same authors conducted a prospective observational population based-study to evaluate the
infection-related mortality in 11,362 patients
with atherosclerotic disease.40 Patients were divided into two groups on the basis of whether
they were taking statins or not. Infection-related
mortality, defined as death due to an acute infection occurring within 30 days of admission,
was significantly lower in the statin compared
with the non-statin group (0.9% vs. 4.1%) with
a relative risk of 0.22 (95% confidence interval,
0.17-0.28).36
Several studies have also been performed
in critically-ill patients. In a two center, randomized, open-label trial, Makris et al.41 compared pravastatin (40 mg/d) with a placebo in
ICU patients without infection but receiving
mechanical ventilation. Six patients (8.4%) in
the pravastatin group and 16 (19.8%) in the
control group died during the 30-day treatment
period (P=0.06). In a cross-sectional analysis of
a prospective cohort, O’Neal et al.42 evaluated
the impact of statins received before ICU admission in critically-ill patients. Of 575 patients,
149 (26%) were on statin therapy prior to hospitalization. In a multivariable logistic regression
model including age, gender, race, current tobacco use, prehospital aspirin use, and APACHE
II score, prehospital statin use was significantly
associated with a lower rate of diagnosis of severe
sepsis (OR 0.62, 95% CI 0.40-0.96, P=0.03).42
Harbi et al.43 conducted a nested cohort study
within two randomized controlled trials unrelated to statins. Of the 763 patients enrolled
in the study, 107 (14%) received statins during
their ICU stay, and 656 (86%) did not. In the
statin group, patients were older (69±11 years
vs. 49±22 years, P<0.0001) and presented higher
APACHE II score than in the non-statin group
(27±7 vs. 23±8 P<0.0001). There was no significant association between statin use and the development of sepsis or severe sepsis.43 Fernandez
et al.44 retrospectively reviewed the ICU charts
of patients receiving mechanical ventilation for
more than 96 hours. Patients were classified into
statin group, which consisted of patients taking
MINERVA ANESTESIOLOGICA
923
THOMAS
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSIS
statins before ICU admission and continuing on
statin therapy throughout the course of hospitalization, or the non-statin group. Of the 438
patients included, the 38 statin-treated patients
were older (71±8 years vs. 61±18 years P=0.001)
and tended to have a higher median APACHE
II Score (21 vs. 17, P=0.07). The ICU-acquired
infection rate in statin-treated patients was not
significantly lower (29% vs. 38%, P=0.3) nor
delayed (median 12 vs. 10 days, P=0.6).44 No
differences were found regarding the source of
infections.44 Finally, a recent meta-analysis made
from 11 randomized controlled trials totaling
30,947 patients showed no effect of statins on
the risk of infections (RR=1.00, 95% CI 0.961.05) or on infection related deaths (0.97, 0.83
to 1.13).5 Thus, the effect of statin in preventing severe sepsis remains uncertain especially in
critically-ill patients.
Continuation of statin therapy in sepsis
In ambulatory settings, elevated hepatic
transaminases generally occur in 0.5% to 2.0%
of cases and are dose-dependent.45 Another untoward effect is myopathy but little is known
about the fundamental mechanisms of statinassociated myopathy. Severe myopathy is rare
and its incidence is reported to be 0.08% with
lovastatin and simvastatin.46 Current prescribing
guidelines suggest caution in the continued use
of statins in critically-ill patients because of concern regarding serious side effects and toxicity
that might be greater than in the general population. However, should statins influence the
inflammatory response to sepsis, cessation may
cause an inflammatory rebound leading to worse
outcomes. Very few studies have explored this
topic and particularly in critically-ill patients.
Mekontso-Dessap et al.47 conducted a retrospective cohort study among patients admitted for
severe sepsis and septic shock in the ICU and
with ongoing statin therapy (initiated at least
one month before ICU admission and continued with no interruption until ICU admission).
Of 76 patients included in the final analysis, 44
had statin therapy continued and 32 had not.
After propensity-matching or multivariable adjustment, there was no association of statin con-
924
tinuation with organ failure-free days (beta coefficients with 95% CI of 2.37 [-0.96 to 5.70],
P=0.20 and 2.24 [-0.43 to 4.91], P=0.11 respectively).47 Another observational study suggested
a better outcome in prior statin users presenting
with ventilator-associated pneumonia who continued statin therapy in the ICU compared with
those who stopped statin therapy.48
Kruger et al. conducted the only prospective
randomized double-blind placebo-controlled
trial on this subject.29 They randomized atorvastatin (20 mg) or matched placebo in 150 patients
on preexisting statin therapy requiring hospital
admission for infection. The primary end point
was the progression to severe sepsis. Severe sepsis was present at baseline in 32% (24 of 75) of
patients in both groups. Presence of severe sepsis significantly decreased over time (P<0.01) in
each group with no significant difference between treatment groups at each follow-up time
point (day 3, 5, 10, 14) (overall P=0.6). Twentyfour patients (16%) of the cohort required ICU
admission (atorvastatin group 13/75, placebo
group 11/75). Investigators also explored inflammatory markers (IL-6 and CRP) for both groups
at baseline and follow-up time-points. Median
IL-6 for the cohort at baseline was 43.8 pg/mL
(interquartile range 16.9-100.5), with no significant difference between the groups. IL-6 level
decreased in both groups over time (P<0.01)
with no significant difference between groups
at any follow-up time-point (overall P=0.7).29
Finally, in a trial which randomized patients to
receive atorvastatin 20mg per day or placebo, the
subgroup of prior statin users with continued
atorvastatin therapy had a significantly lower 28day mortality (5% vs. 28%; P=0.01). However,
the difference was not statistically significant at
day 90 (11% vs. 28%; P=0.06).8
Statins to improve outcome of sepsis
Most studies that evaluated the role of statins
in patients with sepsis are retrospective cohort
trials and few focused on critically-ill patients.
Wan et al.7 performed a meta-analysis of 5 randomized and controlled trials and of 27 observational studies. Among randomized controlled
trials, statins did not improve 28-day mortality
MINERVA ANESTESIOLOGICA
August 2015
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSISTHOMAS
(RR, 0.93; 95% CI, 0.46 to 1.89) while observational studies indicated a significant decrease
in mortality with adjusted data (RR, 0.65; 95%
CI, 0.57 to 0.75). Janda et al.4 conducted another meta-analysis of 20 studies whose 18 were
cohort studies. Pooled odds ratios were all in favor of statin versus non-statin use: 0.61 (95% CI,
0.48 to 0.73) for 30-day mortality (N.=7) and
0.40 (95% CI, 0.23 to 0.57) for sepsis-related
mortality. These results suggested a protective effect of statins in patients with sepsis. However,
these meta-analyses are limited by the cohort
design of the selected studies and the high heterogeneity among them regarding the type and
severity of patients, dosage and duration of statin administration, and type of infection.49 Since
randomized controlled trials have been recently
conducted to evaluate the potential interest of
statins in treating patients with severe infections
requiring ICU admission, we performed a metaanalysis limited to this type of trial.
Methods for meta-analysis
We performed an electronic article search
through Pubmed. We used combinations of keywords related to statins (“hydroxymethylglutaryl-CoA Reductase Inhibitor” or “statin” or “simvastatin” or “rosuvastatin” or “pravastatin” or
“atorvastatin” or “fluvastatin” or “cerivastatin”or
“pitavastatin” or “lovastatin”) AND the associated disease (“infection” or “sepsis” or “severe
sepsis” or “septic shock”) AND the type of patients (“Intensive care” or “critical care” or “critically ill”). All searches were limited to “English
language” and “humans”. Only randomized
controlled trials were included and must have
met the following criteria: adult patients admitted to ICU, experienced severe sepsis at study
enrollment, statins compared with a control and
data available on the mortality. Abstracted data
included: characteristics of the studies, characteristics of the included patients and outcomes
of the studies. The endpoint was mortality (60day mortality or in-hospital mortality at 60-day
and 28-day mortality). Authors were contacted
to obtain data and results that were not in the
manuscript. The methodological quality of studies was evaluated using the Jadad Scale,50 which
Vol. 81 - No. 8
is a 0 to 5 point scale used to independently assess the methodological quality of a clinical trial.
In examining the associations between statins and infection/ sepsis mortality, results were
expressed as risk ratios with 95% confidence
intervals (CIs). Heterogeneity across trials was
assessed by means of the I2 statistic, with significance being set at I2>50%. Random effects
models were used for statistical analysis.
Results
The flowchart of included studies and selection progress is presented in Figure 1. Eight randomized studies were identified. Among them
3 were not retained for analysis because they
included only or mainly ward patients 29, 51, 52
and one since it did not include patients with
sepsis.41 Finally, 4 studies 8-11 were retained for
analysis (Table II).
Figure 2 shows the pooled results from random effects models combining the risk ratios for
mortality. Overall analysis including 1818 patients total from 4 studies 8-11 showed that there
was no significant difference between statins
(223/903) and placebo (233/899) in terms of 60day mortality (risk ratio, 0.930; 95% CI, 0.722
to 1.198), with moderate heterogeneity among
the studies (I2=52%, P=0.09). Similarly, no 28day mortality difference was observed between
groups (statins 191/907, placebo 199/911; risk
ratio 0.953; 95% CI, 0.715 to 1.271) with moderate heterogeneity among the studies (I2=54%,
P=0.09).
The first prospective randomized double
blind, placebo controlled was a phase II trial 8
designed to assess the biological and clinical effect of atorvastatin therapy in critically-ill patients with severe sepsis (Table II). A total of
250 patients admitted in 21 ICUs underwent
randomization. Seventy seven patients (30.8%
of the cohort) were on prior statin therapy (37
randomized to treatment, 40 to placebo). Mean
APACHE II Score was 22.1 and 23.5 and SOFA
score was 8.3 and 8.0 in statin and placebo group
respectively. For the entire cohort, as for de novo
therapy, there was no statistically significant difference in ICU, hospital, 28-day or 90-day mortality and in length of stay between groups.
MINERVA ANESTESIOLOGICA
925
THOMAS
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSIS
Figure 1.—Flowchart of included studies and selection progress.
Table II.—Outcome data of randomized controlled trials included in the meta-analysis of statin for sepsis in critically ill patients.
Study
design
Date
Jadad Score
Inclusion criteria
Kruger et al. (2013) 8
Prospective, randomized,
double blind placebocontrolled, phase II,
multicenter trial (21 ICU)
July 2007-August 2010
4
–– Suspected or proven
infection and severe sepsis
–– Prior statin users (N.=77)
Papazian et al. (2013) 9
Prospective, randomized,
placebo-controlled, double
blind, parallel-group,
multicenter trial (26 ICU)
January 2010 -March 2013
5
–– Mechanical ventilation
>48 hours and suspected
ventilator-associated
pneumonia (CPIS >5)
Prior statin users (N.=26)
NHLBI ARDS clinical trial
network Network (2014) 10
Prospective, randomized,
double blind placebocontrolled, multicenter trial
(44 hospitals)
March 2010 - September 2013
5
–– Mechanical ventilation
and PaO2 to FiO2 ratio
<300 mmHg for 24 hours
and known or suspected
infection
–– Prior statin users (N.=109)
McAuley et al. (2014) 11
Prospective, randomized,
double blind placebocontrolled, multicenter trial
(40 hospitals)
December 2010 – March 2014
5
–– Mechanical ventilation and
ARDS <48 hours (PaO2 to
FiO2 ratio <300 mmHg)
–– Prior statin users (N.=0,
exclusion criteria)
Papazian et al.9 conducted a multicenter
prospective randomized double-blind placebocontrolled trial evaluating simvastatin (60 mg/
926
day) versus placebo in critically-ill patients who
underwent ventilator-associated pneumonia.
Patients were free of statin therapy at the mo-
MINERVA ANESTESIOLOGICA
August 2015
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSISTHOMAS
Figure 2.—Forest plot for the risk ratio of 60-day mortality (A) and 28-day mortality (B) from randomized controlled trials.
Intervention
Atorvastatin 20 mg/d until
discharge from ICU or day 14
Outcome
Results
–– IL-6 concentration (primary
endpoint)
–– ICU, hospital, 28-day and
90-day mortality (secondary
endpoints)
–– N.=250 (123 atorvastatin/127
placebo)
–– No difference in IL6 concentration
(P=0.76)
–– No difference for ICU/hospital/28
and 90-day mortality
Simvastatin 60 mg/day until
28-day mortality
–– Trial stopped for futility
discharge from ICU or death or
(primary endpoint)
–– N.=284 (146 simvastatin/138
day 28
placebo)
–– Day 28 mortality: 21.2% in
simvastatin group vs. 15.5% in
placebo group (P=0.10)
Rosuvastatin 40 mg at day 1, then 60-day mortality
–– Trial stopped for futility
20 mg/day
(primary endpoint)
–– N.=745 (379 rosuvastatin/366
until 3 days after discharge from
placebo)
ICU or discharge or death on
–– Day-60 mortality
day 28
28.5% in rosuvastatin group, vs. 24.9%
in placebo group
(P=0.21)
Simvastatin 80 mg/day until
–– Number of ventilator-free days –– N.=540 (259 simvastatin/281
discharge from ICU or death or
(VFD) to day 28
placebo)
day 28
(primary endpoint)
–– 28-day mortality 22% in simvastatin
–– 28-day mortality, death before
group vs. 26.8% in placebo group
discharge from ICU or hospital
(P=0.23)
(secondary endpoints)
–– Death before discharge from hospital
25.9% in simvastatin group vs.
32.1% in placebo group (P=0.13)
ment of intubation. Mean overall SOFA Score
on day 1, was 7.0. The trial was stopped for futility at the first scheduled interim analysis. Three
Vol. 81 - No. 8
Adverse effects
None
None
Less hepatic and renal failure
free days in rosuvastatin
group (P=0.003 and P=0.01
respectively)
Elevated creatinine kinase and
transaminases more common
in simvastatin group than in
placebo group (17 vs. 8, OR
2.5 [0.9-7], P=0.05, and 26
vs. 16, OR [0.9-4.3], P=0.08)
hundred patients were enrolled, and 284 were
analyzed, 146 in the simvastatin group and 138
in placebo group. 28-day mortality was not dif-
MINERVA ANESTESIOLOGICA
927
THOMAS
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSIS
ferent between groups although it was higher in
the statin group (21.2% vs. 15.2% in the placebo group, P=0.10). ICU and hospital mortality were not different. Results were similar in the
subgroup naive to statins at admission.
The Acute Respiratory Distress Syndrome
(ARDS) Network recently conducted a multicenter trial in patients with sepsis-associated
ARDS.10 They randomly assigned patients to
receive rosuvastatin or placebo in a double-blind
way. The study was stopped because of futility
after enrollment of 745 patients (379 in rosuvastatin group and 366 in placebo group). Mortality at day 60, ventilator-free days and ICU
free days to day 28 did not differ significantly
between the two groups. One hundred and nine
patients had used statins previously (54 in rosuvastatin group and 55 in placebo group). Rosuvastatin therapy had no effect on mortality in this
subgroup (17/54 deaths in rosuvastatin group,
11/55 deaths in placebo group, P=0.14).
Finally, McAuley et al.11 published the results
of a multicenter trial which randomized 540
ARDS patients to receive either simvastatin (80
mg/day) or placebo. Patients having received a
statin within 2 weeks of meeting ARDS criteria
were excluded. There was no significant difference between the study groups in the number
of ventilator-free days (12.6±9.9 with simvastatin and 11.5±10.4 with placebo, P=0.21), in
28-day mortality (22% and 26.8% respectively, p=0.23) or in mortality before discharge
from hospital (25.9% and 32.1% respectively,
P=0.13). Although 404 of included patients
presented with sepsis (189/259 in simvastatin
group and 218/280 in placebo group), there was
no interaction between outcome and presence
or absence of sepsis. Therefore, the results of the
whole population have been included in the present meta-analysis.
Regarding adverse effects of statins (alanine
aminotransferase or creatine kinase elevation
during the study period), there were no significant differences between statin and placebo
groups in three trials.8-10 However, one trial reported fewer days free of hepatic failure or renal
failure in the statin group.51 The rate of adverse
events in the most recent study 11 was higher in
the simvastatin group than in the placebo group
928
(OR 2.2 [1.1-4.2], P=0.02). The majority of the
adverse events were related to elevated creatine
kinase and hepatic aminotransferase levels.
Implications for clinical practice
In numerous cohort trials, statins have been
suggested to decrease mortality or the risk of infection in septic patients.39, 42 Four prospective
randomized double-blind placebo-controlled
trials in the ICU have failed to demonstrate benefit with statins in patients with severe sepsis.8-11
Two of them were stopped for futility.9, 10 A trend
in higher 28-day mortality in the statin group in
one study 9 and in 60-day mortality in the statin
group in another one 10 prompted our desire to
perform a meta-analysis of available randomized
trials. Our results confirm the lack of beneficial
effects but do not exhibit any significant detrimental effect of statins on mortality.
A limitation of this meta-analysis is that trials
had different inclusion criteria, respectively severe
sepsis,8 ventilator-associated pneumonia 9 and sepsis-associated ARDS.10, 11 However, populations
had similar severity scores and mortalities, ranging from 14% to 24%. Another limitation is that
these trials tested 3 different statins, respectively
atorvastatin,8 simvastatin 9, 11 and rosuvastatin.10
These molecules had been chosen by investigators since immunomodulatory effects had been
described with each molecule and since they had
a safety profile compatible with administration in
critically-ill patients. Nevertheless, the results of
the three studies were similar with 3 different molecules, reinforcing the idea that statin lack of effect
on the course of sepsis is not molecule dependent.
Like previous observational studies, a very recent large cohort study suggested that patients
treated with high regimens of statins at least one
month before the occurrence of sepsis had lower
hospital and one-year mortality.53 This result
supports the concept that statin premedication
may confer additional protection in community-acquired sepsis that is not related to atherosclerotic diseases. This preventive potential has
not been evaluated in recent trials. Moreover, the
potential of statins as the part of a multimodal
approach to improve prognosis of sepsis remains
to be evaluated.
MINERVA ANESTESIOLOGICA
August 2015
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSISTHOMAS
Conclusions
In conclusion, the results of this meta-analysis
confirm that the use of statin therapy should not
be recommended in the management of severe
sepsis in critically ill patients. Since one study
reported less hepatic and renal failure free days
in the statin group, statins should be continued
with cautious at ICU admission only if considered as necessary.
Key messages
—— The use of statin therapy should not
be recommended in the management of severe sepsis in critically ill patients.
—— Statins should be continued with cautious at ICU admission only if considered as
necessary.
—— The potential of statins given for a long
time before sepsis remains to be clarified.
References
  1. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach
H, Opal SM et al. Surviving Sepsis Campaign Guidelines
Committee including the Pediatric Subgroup. Surviving
sepsis campaign: international guidelines for management
of severe sepsis and septic shock: 2012. Crit Care Med
2013;41:580-637.
  2. Terblanche M, Almog Y, Rosenson RS, Smith TS, Hackam
DG. Statins and sepsis: multiple modifications at multiple
levels. Lancet Infect Dis 2007;7:358-68.
  3. Tleyjeh IM, Kashour T, Hakim FA, Zimmerman VA, Erwin PJ, Sutton AJ et al. Statins for the prevention and treatment of infections: a systematic review and meta-analysis.
Arch Intern Med 2009;169:1658-67.
  4. Janda S, Young A, Fitzgerald JM, Etminan M, Swiston J.
The effect of statins on mortality from severe infections and
sepsis: a systematic review and meta-analysis. J Crit Care
2010;25:656.e7-22.
  5. Van den Hoek HL, Bos WJ, de Boer A, van de Garde EM.
Statins and prevention of infections: systematic review and
meta-analysis of data from large randomised placebo controlled trials. BMJ 2011;343:d7281.
  6. Ma Y, Wen X, Peng J, Lu Y, Guo Z, Lu J. Systematic review
and meta-analysis on the association between outpatient
statins use and infectious disease-related mortality. PLoS
One 2012;7:e51548.
  7. Wan YD, Sun TW, Kan QC, Guan FX, Zhang SG. Effect
of statin therapy on mortality from infection and sepsis: a
meta-analysis of randomized and observational studies. Crit
Care 2014;18:R71.
  8. Kruger P, Bailey M, Bellomo R, Cooper DJ, Harward M,
Higgins A et al. ANZ-STATInS Investigators–ANZICS
Clinical Trials Group. A multicenter randomized trial of
atorvastatin therapy in intensive care patients with severe
sepsis. Am J Respir Crit Care Med 2013;187:743-50.
  9. Papazian L, Roch A, Charles PE, Penot-Ragon C, Perrin
Vol. 81 - No. 8
G, Roulier P et al. STATIN-VAP Study Group. Effect of
statin therapy on mortality in patients with ventilatorassociated pneumonia: a randomized clinical trial. JAMA
2013;310:1692-700.
10. Truwit JD, Bernard GR, Steingrub J, Matthay MA, Liu
KD, Albertson TE et al. National Heart, Lung, and Blood
Institute ARDS Clinical Trials Network. Rosuvastatin for
sepsis-associated acute respiratory distress syndrome. N
Engl J Med 2014;370:2191-200.
11. McAuley DF, Laffey JG, O’Kane CM, Perkins GD, Mullan B, Trinder TJ et al. Simvastatin in the acute respiratory
distress syndrome. N Engl J Med 2014;371:1695-703.
12.Beri A, Sural N, Mahajan SB. Non-atheroprotective effects of statins: a systematic review. Am J Cardiovasc Drugs
2009;9:361-70.
13. Jerwood S, Cohen J. Unexpected antimicrobial effect of
statins.J Antimicrob Chemother 2008;61:362-4.
14. Blanco-Colio LM, Tuñón J, Martín-Ventura JL, Egido J.
Anti-inflammatory and immunomodulatory effects of statins. Kidney Int 2003;63:12-23.
15. Hotchkiss RS, Karl IE. The pathophysiology and treatment
of sepsis. N Engl J Med 2003;348:138-50.
16. Girardin E, Grau GE, Dayer JM, Roux-Lombard P, Lambert PH. Tumor necrosis factor and interleukin-1 in the
serum of children with severe infectious purpura. N Engl J
Med 1988;319:397-400.
17. Hatherill M, Tibby SM, Turner C, Ratnavel N, Murdoch
IA. Procalcitonin and cytokine levels: relationship to organ
failure and mortality in pediatric septic shock. Crit Care
Med 2000;28:2591-4.
18. Bone RC, Grodzin CJ, Balk RA. Sepsis: a new hypothesis for
pathogenesis of the disease process. Chest 1997;112:23543.
19. Munford RS, Pugin J. Normal responses to injury prevent
systemic inflammation and can be immunosuppressive. Am
J Respir Crit Care Med 2001;163:316-21.
20. Stearns-Kurosawa DJ, Osuchowski MF, Valentine C, Kurosawa S, Remick DG. The pathogenesis of sepsis. Annu Rev
Pathol 2011;6:19-48.
21. Hotchkiss RS, Schmieg RE Jr, Swanson PE, Freeman BD,
Tinsley KW, Cobb JP et al. Rapid onset of intestinal epithelial and lymphocyte apoptotic cell death in patients with
trauma and shock. Crit Care Med 2000;28:3207-17.
22. Hotchkiss RS, Monneret G, Payen D. Sepsis-induced immunosuppression: from cellular dysfunctions to immunotherapy. Nat Rev Immunol 2013;13:862-74.
23.Almog Y. Statins, inflammation, and sepsis: hypothesis.
Chest 2003;124:740-3.
24. Merx MW, Liehn EA, Janssens U, Lütticken R, Schrader J,
Hanrath P, Weber C. HMGCoA reductase inhibitor simvastatin profoundly improves survival in a murine model of
sepsis. Circulation 2004;109:2560-5.
25. Merx MW, Liehn EA, Graf J, van de Sandt A, Schaltenbrand M, Schrader J et al. Statin treatment after onset of
sepsis in a murine model improves survival. Circulation
2005;112:117-24.
26. Wang HR, Li JJ, Huang CX, Jiang H. Fluvastatin inhibits
the expression of tumor necrosis factor-alpha and activation
of nuclear factor-kappaB in human endothelial cells stimulated by C-reactive protein. Clin Chim Acta 2005;353:5360.
27. Novack V, Eisinger M, Frenkel A, Terblanche M, Adhikari
NK, Douvdevani A et al. The effects of statin therapy on inflammatory cytokines in patients with bacterial infections:
a randomized double-blind placebo controlled clinical trial.
Intensive Care Med. 2009;35:1255-60.
28. Lobo SM, Lobo FR, Bota DP, Lopes-Ferreira F, Soliman
HM, Mélot C et al. C-reactive protein levels correlate with
mortality and organ failure in critically ill patients. Chest
2003;123:2043-9.
29. Kruger PS, Harward ML, Jones MA, Joyce CJ, Kostner
MINERVA ANESTESIOLOGICA
929
THOMAS
STATIN THERAPY IN CRITICALLY-ILL PATIENTS WITH SEVERE SEPSIS
KM, Roberts MS et al. Continuation of statin therapy in
patients with presumed infection: a randomized controlled
trial. Am J Respir Crit Care Med 2011;183:774-81.
30. Pruefer D, Scalia R, Lefer AM. Simvastatin inhibits leukocyte-endothelial cell interactions and protects against
inflammatory processes in normocholesterolemic rats. Arterioscler Thromb Vasc Biol 1999;19:2894-900.
31. Kwak B, Mulhaupt F, Myit S, Mach F. Statins as a newly recognized type of immunomodulator. Nat Med
2000;6:1399-402.
32. Fehr T, Kahlert C, Fierz W, Joller-Jemelka HI, Riesen WF,
Rickli H et al. Statin-induced immunomodulatory effects
on human T cells in vivo. Atherosclerosis 2004;175:83-90.
33. Levi M, ten Cate H, van der Poll T, van Deventer SJ. Pathogenesis of disseminated intravascular coagulation in sepsis.
JAMA 1993;270:975-9.
34. Levi M, van der Poll T. Inflammation and coagulation. Crit
Care Med 2010;38:S26-34.
35. Van der Poll T, Levi M, Hack CE, ten Cate H, van Deventer SJ, Eerenberg AJ et al. Elimination of interleukin 6
attenuates coagulation activation in experimental endotoxemia in chimpanzees. J Exp Med 1994;179:1253-9.
36. Van Deventer SJ, Büller HR, ten Cate JW, Aarden LA,
Hack CE, Sturk A. Experimental endotoxemia in humans:
analysis of cytokine release and coagulation, fibrinolytic,
and complement pathways. Blood 1990;76:2520-6.
37. Boermeester MA, van Leeuwen PA, Coyle SM, Wolbink
GJ, Hack CE, Lowry SF. Interleukin-1 blockade attenuates
mediator release and dysregulation of the hemostatic mechanism during human sepsis. Arch Surg 1995;130:739-48.
38. Bateman RM, Sharpe MD, Ellis CG. Bench-to-bedside review: microvascular dysfunction in sepsis-hemodynamics,
oxygen transport, and nitric oxide. Crit Care 2003;7:35973.
39. Almog Y, Shefer A, Novack V, Maimon N, Barski L, Eizinger M et al. Prior statin therapy is associated with a decreased rate of severe sepsis. Circulation 2004;110:880-5.
40. Almog Y, Novack V, Eisinger M, Porath A, Novack L, Gilutz H. The effect of statin therapy on infection-related
mortality in patients with atherosclerotic diseases. Crit Care
Med 2007;35:372-8.
41. Makris D, Manoulakas E, Komnos A, Papakrivou E, Tzovaras N, Hovas A et al. Effect of pravastatin on the frequency of ventilator-associated pneumonia and on intensive care
unit mortality: open-label, randomized study. Crit Care
Med 2011;39:2440-6.
42. O’Neal HR Jr1, Koyama T, Koehler EA, Siew E, Curtis BR,
Fremont RD et al. Prehospital statin and aspirin use and the
prevalence of severe sepsis and acute lung injury/acute respiratory distress syndrome. Crit Care Med 2011;39:134350.
43. Al Harbi SA, Tamim HM, Arabi YM. Association between
statin therapy and outcomes in critically ill patients: a nested cohort study. BMC Clin Pharmacol 2011;11:12.
44. Fernandez R, De Pedro VJ, Artigas A. Statin therapy prior
to ICU admission: protection against infection or a severity
marker? Intensive Care Med 2006;32:160-4.
45. Pasternak RC, Smith SC Jr, Bairey-Merz CN, Grundy SM,
Cleeman JI, Lenfant C, American College of Cardiology;
American Heart Association; National Heart, Lung and
Blood Institute. ACC/AHA/NHLBI Clinical Advisory on
the Use and Safety of Statins. Circulation 2002;106:1024-8.
46. Hunninghake DB. Drug treatment of dyslipoproteinemia.
Endocrino lMetab Clin North Am 1990;19:345-60.
47. MekontsoDessap A, Ouanes I, Rana N, Borghi B, Bazin
C, Katsahian S et al. Effects of discontinuing or continuing
ongoing statin therapy in severe sepsis and septic shock: a
retrospective cohort study. Crit Care 2011;15:R171.
48. Bruyere R, Vigneron C, Prin S, Pechinot A, Quenot JP, Aho
S et al. Impact of prior statin therapy on the outcome of
patients with suspected ventilator-associated pneumonia:
an observational study. Crit Care 2014;18:R83.
49. Kouroumichakis I, Papanas N, Proikaki S, Zarogoulidis P,
Maltezos E. Statins in prevention and treatment of severe
sepsis and septic shock. Eur J Intern Med 2011;22:125-33.
50. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds
DJ, Gavaghan DJ et al. Assessing the quality of reports of
randomized clinical trials: is blinding necessary? Control
Clin Trials 1996,17:1-12.
51. Novack V, Eisinger M, Frenkel A, Terblanche M, Adhikari
NK, Douvdevani A et al. The effects of statin therapy on inflammatory cytokines in patients with bacterial infections:
a randomized double-blind placebo controlled clinical trial.
Intensive Care Med 2009;35:1255-60.
52. Patel JM, Snaith C, Thickett DR, Linhartova L, Melody
T, Hawkey P et al. Randomized double-blind placebocontrolled trial of 40 mg/day of atorvastatin in reducing
the severity of sepsis in ward patients (ASEPSIS Trial). Crit
Care 2012;16:R231.
53. Ou SY, Chu H, Chao PW, Ou SM, Lee YJ, Kuo SC et al.
Effect of the use of low and high potency statins and sepsis
outcomes. Intensive Care Med 2014;40:1509-17.
The views expressed in this article are those of the authors and not necessarily those of the Medical Research Council (MRC), National
Health Service, National Institute for Health Research (NIHR), or Department of Health.
Funding.—The HARP-2 study was funded by the Efficacy and Mechanism Evaluation (EME) programme (08/99/08), which is funded
by the MRC and NIHR with contributions from the CSO in Scotland, NICCHR in Wales and the HSC R&D, Public Health Agency
in Northern Ireland and is managed by the NIHR, based at the University of Southampton. This study was also funded in Ireland by a
Health Research Award from the Health Research Board, Dublin, Ireland (HRA_POR-2010-131). The HSC R&D, Public Health Agency
in Northern Ireland and the Intensive Care Society of Ireland provided additional funding.
Conflicts of interest.—J. Thomas, S. Hraiech, A. Loundou, L. Papazian, A. Roch and P. Kruger have no conflicts of interest. D. F. Mcauley
reports receiving fees from GlaxoSmithKline for serving on advisory boards, consulting fees from GlaxoSmithKline and Peptinnovate, and
clinical trial support from GlaxoSmithKline paid to his institution; he is also a named inventor on a pending unlicensed patents for the use
of a pharmacotherapy (not a statin) for the treatment f acute repiratory distress syndrome. J. Truwit has received fundings from NHBLI in
ARDS and from AstraZeneca for community-acquired pneumonia.
Received on July 18, 2014. - Accepted for publication on February 12, 2015. - Epub ahead of print on February 18, 2015.
Corresponding author: A. Roch, Assistance Publique, Hôpitaux de Marseille, Hôpital Nord, Réanimation des Détresses Respiratoires et
des Infections Sévères, Chemin des Bourrely, 13015 Marseille, France. E-mail: [email protected]
930
MINERVA ANESTESIOLOGICA
August 2015

LETTER TO THE EDITOR
Association between steroid particle sizes and
serious complications during epidural injections
M. A. GHOBADIFAR 1, 2
1Zoonoses Research Center, Jahrom University of Medical Sciences, Jahrom, Iran; 2Research Center for Social
Determinants of Health, Jahrom University of Medical Sciences, Jahrom, Iran
Dear Editor,
I
have read with curiosity the intriguing article entitled “Epidural steroid injections: update on efficacy, safety, and newer
medications for injection” published in this issue of Minerva
Anestesiologica.1 Kozlov et al. skillfully reviewed the literature
on efficacy, safety, and newer medications for epidural steroid
injections (ESIs). This review is a helpful article for resolving
the interpretations about safety and efficacy of using ESIs during the management of radicular pain. However, I would like
to add some additional points about association between steroids particle sizes and serious complications, which are evident
in the results discussed below.
It was showed that larger particle sizes of steroids might
have a higher risk for small vessels occlusion of arterial tree as
a potential sequel after incidental steroid injection in radiculomedullary artery. Notably, infarction of the spinal cord, cerebellum, and the brainstem may occur. One might contemplate
that particles of steroid smaller than an erythrocyte diameter
are safer.2
Injectable preparations including betamethasone sodium
phosphate, triamcinolone acetonide, betamethasone acetate,
methylprednisolone acetate, and dexamethasone sodium phosphate were analyzed by light microscopy in vitro. Derby et
al.2 have revealed that triamcinolone acetonide particles were
greater than median-sized erythrocytes. Betamethasone and
triamcinolone particles were packed with an extensive particle
aggregation. Methylprednisolone in a mixture with lidocaine
and iodinated contrast media and alone showed that particles
with their aggregations were smaller than red blood cells. Nevertheless, methylprednisolone particles were packed, signifying
the potency to form emboli with following occlusion of small
arterioles. Sodium phosphate dexamethasone particles without
the tendency of particle aggregation were smaller than erythrocytes. A mixture of dexamethasone, iodinated contrast media,
lidocaine showed the same properties 2 and did not precipitate.
On the other hand, MacMahon et al.3 evaluated dexamethasone sodium phosphate, triamcinolone acetonide, and methylprednisolone acetate nondilute injectable particle sizes and
after mixing with human plasma and local anesthetic. Again,
corticosteroids in a mixture with local anesthetics did not
change their properties. Additional plasma showed significant
reduction in the aggregate sizes eventually related to repulsion
effects and coating effects of albumin; however, the insoluble
corticosteroid particle sizes was unchanged.3 Consequently, the
steroid aggregates might play the role of potential embolization
agents within spinal cord arterioles.4
As is obvious from the above discussion, after incidental
intra-arterial injection, soluble dexamethasone is less likely to
cause capillary or arterial occlusion. Interestingly, more studies
are required for a deeper understanding of the association between steroids particle sizes and serious complications.
References
 1. Kozlov N, Benzon HT, Malik K. Epidural steroid injections: update on efficacy, safety, and newer medications for
injection. Minerva Anestesiol 2015;81:901-9.
  2. Derby R, Lee SH, Date ES, Lee JH, Lee CH. Size and aggregation of corticosteroids used for epidural injections.
Pain Med 2008;9:227-34.
  3. MacMahon PJ, Shelly MJ, Scholz D, Eustace SJ, Kavanagh EC. Injectable corticosteroid preparations: an embolic
risk assessment by static and dynamic microscopic analysis.
AJNR Am J Neuroradiol 2011;32:1830-5.
  4. Gazelka HM, Burgher AH, Huntoon MA, Mantilla CB,
Hoelzer BC. Determination of the particulate size and aggregation of clonidine and corticosteroids for epidural steroid injection. Pain Physician 2012;15:87-93.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on December 11, 2014. - Accepted for publication on December 15, 2014. - Epub ahead of print on December 17, 2014.
Corresponding author: M. A. Ghobadifar, Zoonoses Research Center, Medicine School, Jahrom University of Medical Sciences, Motahari
Avenue, Jahrom, Iran. E-mail: [email protected]
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
931

LETTER TO THE EDITOR
Missed citations
D. CALDIROLI, E. F. ORENA
Department of Neuroanesthesia and Intensive Care, Fondazione IRCCS Istituto Neurologico Carlo Besta, Milan, Italy
Dear Editor,
T
he interesting study from Pieters et al.1 about the forces
applied by the laryngoscope on patients’ teeth opens the
discussion on some issues that, in our opinion, need to be
specified.
The authors suggest in the discussion that the Mac blades
create more room for intubation compared to other videolaryngoscopes (VLS), making intubation easier and faster.
While time is a continuous variable that was measured, the
ease of laryngoscopy is a subjective element which does not
find any reference in the results and is therefore an inference.
We have recently published a study on the activation of the
upper limb muscles recorded through surface electromyography (SEMG) during laryngoscopy in manikins, which included
also the assessment of workload with the NASA Task Load Index to compare direct laryngoscopy and Glidescope® (GLS).1
Our results showed that the force used by the muscles of the
left upper limb during videolaryngoscopy is much lower than
during direct laryngoscopy and the workload showed the same
trend. Unfortunately, this paper is not mentioned by Pieters
et al.
Furthermore, the differences in the results are the consequence of the decision not to use the stylet with both McGrath®
and GLS and of the lack of expertise with the two VLS.
While with direct laryngoscopy it is crucial to create a wide
room by only pushing the soft tissue for passing the tube, this is
not necessary with the angulated blades as it is the shape of the
malleable stylet that adapts the tube to the pharyngeal anatomy
and allows its progression with respect of the soft palate and the
tonsillar pillars.
The required bimanual task to accomplish videolaringoscopic intubation needs a long training to reach expertise and
the abandonment of the stylet for fear of soft tissue injury is the
proof that the expertise was not yet reached with either GLS or
with McGrath® before the study began, and this bias affected
the results and the conclusions of the study.
In further support of this argument, we would like to point
out that in the methods the three investigators were defined as
“experts” because they have done at least 100 intubations with
each of the three VLS.
Therefore, before the study started each operator would
have had at least 300 intubations done for a total of 900 consecutive intubations.
However, only 150 consecutive patients were enrolled in a
5 months period (May-September 2012).
This means that each operator would have taken at least 10
months to reach the expertise, for a total of 30 months for the
three operators, and the training would have started in 2009.
Even considering this likely, the authors should have provided at least the total number of intubations and the training
period for supporting their expertise with the three VLS, as
suggested by Behringher in 2012 2 and by Cortellazzi in the
2014.3
In our opinion, the authors should have also taken into
consideration the evidences from the recent literature that are
missing.
References
  1. Pieters B, Maassen R, van Eig E, Maathuis B, van den Dobbelsteen J, van Zundert A. Indirect videolaryngoscopy using
Macintosh blades in patients with non –anticipated difficult
airway results in significantly lower forces exerted on teeth
relative to classic direct laryngoscopy; a randomized cross
over trial. Minerva Anestesiol 2015;81:846-54.
  2. Caldiroli D, Molteni F, Sommariva A, Frittoli S, Guanziroli E, Cortellazzi P et al. Upper limb muscular activity and
perceived workload during laryngoscopy: comparison of
Glidescope® and Macintosh laryngoscopy in manikin: an
observational study. Br J Anaesth 2014;112:563-9.
  3. Behringer EC, Cooper RM, Luney S, Osborn IP. The comparative study of video laryngoscopes to the Macintosh laryngoscope: defining proficiency is critical. Eur J Anaesthesiol 2012;29:158-9.
  4. Cortellazzi P, Caldiroli D, Byrne A, Sommariva A, Orena
EF, Tramacere. Defining and developing expertise in tracheal intubation using a GlideScope® for anaesthetists with
expertise in Macintosh direct laryngoscopy: an in-vivo longitudinal study. Anaesthesia 2015;70:290-5.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on January 15, 2015. - Accepted for publication on January 19, 2015. - Epub ahead of print on January 19, 2015.
Corresponding author: E. F. Orena, Via Celoria 11, 20133, Milan, Italy. E-mail [email protected]
932
MINERVA ANESTESIOLOGICA
August 2015

LETTER TO THE EDITOR
Videolaryngoscopy offers us more
than classic direct laryngoscopy
A. A. J. VAN ZUNDERT 1, B. M. A. PIETERS 2
1Department of Anesthesia and Perioperative Medicine, Royal Brisbane and Women’s Hospital, Brisbane, Queensland,
Australia; 2Department of Anaesthesiology, University Hospital of Leuven, Leuven, Belgium
Dear Editor,
T
hank you for allowing me the privilege to respond to the
questions posed by Drs. Caldiroli and Orena 1 in relation
to our recent publication 2 in this journal.
While authors should do their utmost to provide the latest literature, this is not always possible due to the delay between submission and publication dates. Hence, at the time
of acceptance of our paper, the cited publication 3 was not yet
available, and we considered another cited publication 4 of no
additional value to our paper.
We thank Caldiroli et al.5 for demonstrating less muscular
activity and perceived workload with the use of the GlideScope® compared to classic direct laryngoscope. However, we
regret that only one videolaryngoscope was used. The results
of their current study do not surprise us, since the oral, pharyngeal and laryngeal axes have to be brought in one line when
performing direct laryngoscopy using a Macintosh blade, requiring more force. Having to bring these axes in one line has
a bigger implication than “only pushing the soft tissue for passing the tube”.1 With videolaryngoscopy these axes do not need
to be brought into one line, and so significantly less force is
applied to the maxillary incisors.2
We fully agree with Drs. Caldiroli and Orena,1 that only
experienced clinicians with equal expertise should execute
studies. The three clinicians involved in our study 2 have been
using the three videolaryngoscopes (C-MAC®, GlideScope®,
McGrath®) extensively on a daily basis. The authors’ experience with videolaryngoscopy ranges from 1000 to >4000 clinical cases in the operating theatre plus regular practice and studies on manikins.
We disagree with Drs. Caldiroli and Orena 1 that the abandonment of the stylet for fear of soft tissue injury is the proof
that the expertise was not yet reached with either the GlideScope® or McGrath® videolaryngoscopes. We based our decision precisely on our long-time experience with both videola-
Vol. 81 - No. 8
ryngoscopes and, more importantly, proved that even without
stylet, endotracheal intubation with both GlideScope® (39%)
and McGrath® (40%) is possible.2
The reasons why it took several months to complete our
study 2 were notdue to a lack of expertise but merely to its particular setting (special equipment to measure pressure exerted),
the availability of the correct team (experienced staff member,
research nurse and specialised technician), and the availability
of the videolaryngoscopes for research.
We agree with Cortellazzi et al.3 that frequent practice is essential to maintain expertise. We therefore plea for routine use
of videolaryngoscopy for endotracheal intubation in all cases.
It is time to discourage the use of less successful ‑ potentially blind ‑ direct laryngoscopy. The anesthesia community
now has to prove which blade design is most optimal. Indeed,
all videolaryngoscopes provide a better laryngeal view, but this
does not necessarily result in an easy and successful intubation. In our opinion, anesthetists would benefit from blades
designed so that they provide sufficient illumination for good
visualization at the distal end of the blade (e.g. Macintosh design) and which create more room, necessary for intubation
and use of adjuncts (e.g. Magill forceps, bougie).
References
  1. Caldiroli D, Orena EF. Missed citations. Minerva Anestesiol
2015;81:932.
  2. Pieters B, Maassen R, van Eig E, Maathuis B, van den Dobbelsteen J, van Zundert A. Indirect videolaryngoscopy using
Macintosh blades in patients with non-anticipated difficult
airways results in significantly lower forces exerted on teeth
relative to classic direct laryngoscopy: a randomized crossover trial. Minerva Anestesiol 2015;81:846-54.
  3. Cortellazzi P, Caldiroli D, Byrne A, Sommariva A, Orena
EF, Tramacera I. Defining and developing expertise in tracheal intubation using a GlideScope® for anaesthetists with
expertise in Macintosh direct laryngoscopy: an in vivo longitudinal study. Anaesthesia 2014 [Epub ahead of print].
MINERVA ANESTESIOLOGICA
933
LETTER TO THE EDITOR
  4. Behringer EC, Cooper RM, Luney S, Osborn IP. The comparative study of video laryngoscopes to the Macintosh
laryngoscope: defining proficiency is critical. Eur J Anaesthesiol 2012;29:158-9.
  5. Caldiroli D, Molteni F, Sommariva A, Frittoli S, Guan-
ziroli E, Cortallazzi P, Orena EF. Upper limb muscular
activity and perceived workload during laryngoscopy:
comparison of Glidescope® and Macintosh laryngoscopy
in manikin: an observational study. Br J Anaesth 2014;
112:563-9.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material discussed
in the manuscript.
Received on January 25, 2015. - Accepted for publication on February 6, 2015. - Epub ahead of print on February 13, 2015.
Corresponding author: A. van Zundert, The University of Queensland – Royal Brisbane and Women’s Hospital, Herston-Campus, Brisbane, Queensland 4029, Australia. E-mail: [email protected]
934
MINERVA ANESTESIOLOGICA
August 2015

T O P 5 0 M I N E RVA A N E S T E S I O L O G I C A R E V I E W E R S
Most active reviewers between January 2015-June 2015
Surname
Name
Verheijde
Peris
Brogly
Faraoni
Cattano
Alston
Della Rocca
Del Sorbo
Lirk
Deflandre
Caruselli
Torgersen
Chelazzi
Gómez-Ríos
Barbas
Bertini
Langer
Leone
Turnbull
Brazzi
Cata
Agrò
Dalfino
Loop
Vincent
Sbaraglia
Pabelick
Aceto
Dauri
Zanella
Savoia
Caldiroli
Gristina
Tritapepe
Masclans
Jellish
Schwartz
Cavallone
Li Bassi
Abad Gurumeta
Vaida
Ranucci
Diaz-Gomez
Kim
Vannucci
Page
Max
Bruder
Boer
Staals
Joseph L.
Adriano
Nicolas
David
Davide
Theodore
Giorgio
Lorenzo
Philipp
Eric P.
Marco
Christian
Cosimo
Manuel Ángel
Silvia Valente Carmen
Pietro
Thomas
Marc
David
Luca
Juan P.
Eugenio Felice
Lidia
Torsten
Jean Louis
Fabio
Christina Maria
Paola
Mario
Alberto
Gennaro
Dario
Giuseppe R.
Luigi
Joan R
W. Scott
Arthur
Laura Francesca
Gianluigi
Alfredo
Sonia
Marco
Jose L.
Tae W.
Andrea
Valerie
Martin
Nicolas
Christa
Lonneke
Vol. 81 - No. 8
MINERVA ANESTESIOLOGICA
Number
of revisions
8
8
8
7
7
6
6
6
6
6
6
5
5
5
5
5
5
5
4
4
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
935