Download Treatment of diabetes mellitus using an external insulin pump in

Available online at www.sciencedirect.com
Diabetes & Metabolism 34 (2008) 425-438
Treatment of diabetes mellitus using an external insulin pump
in clinical practice
N. Jeandidiera, J.-P. Rivelineb, N. Tubiana-Rufic, A. Vambergued, B. Catargie, V. Melkif,
G. Charpentierb, B. Guercig*
a
Hospices Civils, CHU de Strasbourg, ULP Strasbourg, 1, place de l’Hôpital, 67100 Strasbourg, France.
b
Diabétologie, Hôpital de Corbeil Essonnes, 59, bd Henri-Dunant, 91100 Corbeil-Essonnes, France.
c
Hôpital Robert-Debré, AP-HP, 48, bd Sérurier, 75019 Paris, France.
d
Clinique Marc-Linquette, CHRU de Lille, rue du Professeur Laguesse, 59037 Lille Cedex, France.
e
Département d’Endocrinologie, CHU de Bordeaux, place Amélie Raba-Léon, 33076 Bordeaux Cedex, France.
f
Hôpital de Rangueil, CHU de Toulouse, 1, avenue du Professeur Jean-Poulhes, 31403 Toulouse Cedex, France.
g
Hôpital Jeanne-d’Arc, CHU de Nancy, route Nationale 4, BP 90303, 54201 Dommartin-lès-Toul, France.
Abstract
Before the initiation of insulin pump therapy, patients must be aware of the different aspects of this form of intensive insulin therapy. Most
healthcare professionals recommend a sequential approach to inform patients about CSII. Factors that need to be considered in choosing an
insulin pump include its safety features, durability of the device, tolerability and comfort of the catheter, user-friendliness, technical features and appearance. The initial insulin requirements need to be individualized for the given patient, using different methods to determine the
appropriate dosages for the basal rate and prandial boluses. Glycaemic targets and algorithms for insulin dose adaptation need to be learned
by the patients to enable them to avoid and/or correct hypo- and hyperglycaemia/ketosis episodes. Patients are also advised on how to carry
out frequent self-monitoring of blood glucose—and of ketone bodies, if necessary. Insulin pumps are now able to deliver a range of basal
rates and boluses that increase the flexibility of CSII. One specific issue is the approach to meal-planning, based on carbohydrate-counting
or the equivalent: this method of so-called ‘flexible insulin therapy’can improve metabolic control (for instance, by diminishing postprandial
excursions) as well as the quality of life of patients. Evaluation of the knowledge and practices of the patient can be made through a continuous educational programme carried out by experienced nurses and physicians at the start of therapy and during follow-up. In addition, it
may be necessary to identify the reasons for lack of improvement in metabolic control after several months of therapy, which include pump
malfunction, cannula problems, miscalculated insulin dosages and insufficient metabolic control in specific clinical situations with a high
risk of metabolic deterioration (illness, exercise, concomitant drugs). Annual assessment of the patient using an itemized checklist is required to verify the continued efficacy and safety of insulin pump therapy, two main factors of success with CSII treatment.
© 2008 Elsevier Masson SAS. All rights reserved.
Résumé
Traitement du diabète par pompe à insuline externe en pratique clinique
Avant l’initiation du traitement par pompe à insuline, les patients doivent être informés des différents aspects de cette insulinothérapie
intensive. La majorité des professionnels de santé recommandent une approche séquentielle pour l’éducation des patients au traitement par
pompe. Certains facteurs doivent être pris en compte dans le choix de la pompe à insuline, en particulier la sécurité du matériel, sa durabilité, la tolérance et le confort des cathéters, la facilité d’utilisation, les spécificités techniques, et l’attractivité. Les besoins initiaux en doses
d’insuline doivent être individualisés selon différentes méthodes, afin de déterminer respectivement la répartition des doses de base et des
bolus prandiaux. Les objectifs glycémiques et les algorithmes d’adaptation des doses d’insuline sont enseignés aux patients, en particulier
pour prévenir et corriger les épisodes d’hypo- et d’hyperglycémies avec ou sans cétose. Afin de favoriser un autocontrôle glycémique régulier, et la recherche de corps cétoniques si nécessaire, des conseils sont donnés aux patients. Les pompes à insuline d’aujourd’hui permettent
différentes modalités de débits de base et de bolus, afin d’augmenter la flexibilité du traitement par pompe. Une des spécificités repose sur
*Corresponding author.
E-mail Address: [email protected]
© 2008 Elsevier Masson SAS. Tous droits réservés.
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
426
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
un plan alimentaire basé sur le comptage des glucides ou des équivalents glucidiques: cette méthode appelée insulinothérapie flexible peut
améliorer le contrôle métabolique (en particulier les excursions postprandiales), mais aussi la qualité de vie des patients. Une évaluation des
connaissances et des pratiques du patient est réalisée à l’aide d’un programme continu d’éducation sous contrôle d’infirmières et de médecins
expérimentés, au début du traitement et durant tout le suivi des patients. Par ailleurs, il est nécessaire d’identifier au cours du suivi médical
les causes de non amélioration de l’équilibre métabolique après plusieurs mois de traitement par pompe: dysfonction de la pompe, problèmes
de cathéters, erreurs d’adaptation des doses d’insuline, insuffisance de surveillance métabolique dans certaines situations cliniques à risque
(maladie, exercice, traitements intercurrents...). Une réévaluation annuelle à l’aide d’une « check-list » est également requise pour vérifier
l’efficacité et la sécurité du traitement par pompe à insuline, deux points essentiels de succès du traitement par pompe à insuline.
© 2008 Elsevier Masson SAS. Tous droits réservés.
Keywords: External insulin pump; Self-monitoring of blood glucose; Hypoglycaemia; Hyperglycaemia; Ketosis; flexible insulin therapy; Quality of life;
Devices; Review.
Mots clés : Pompe à insuline externe ; Autocontrôle glycémique ; Hypoglycémie ; Hyperglycémie ; Cétose ; Insulinothérapie fonctionnelle ; Qualité de vie ;
Matériel ; Revue générale.
1. Initiation of CSII
As patients should be aware of the various aspects of intensive diabetes management, they need to receive ongoing education and psychological support. In most cases, pump therapy
is initiated when the patient is in hospital for 3 to 5 days [1].
During the initial selection consultation with the patient, the
diabetologist must be clear on several points:
• confirmation of the indication;
• an explanation of what continuous subcutaneous insulin
infusion (CSII) is;
• the fact that the device must be worn 24 hours a day
(except for 1.5 h);
• the need for the patient to be responsible in terms of selfmanagement of blood glucose, which should be measured at
least three times a day.
The patient may then choose a pump and catheter based
on the technical features offered by the devices (possible
infusion rates, dose increments, number of possible different
basal rates per day) and on personal preferences (see General
pump function and comparaison of available insulin pumps
in 2008). The initial treatment is programmed into the device
only when the patient has achieved a complete understanding of the principles behind the workings of the device, and
his technical skills have been evaluated. Rapid-acting insulin analogues such as lispro, aspart or glulisin may be prescribed with the pump [2], except in a few specific medical
situations such as gastroparesis.
1.1. Initial insulin dose determination
It is difficult to specify what comes under bolus action
as opposed to basal rate action, except when the boluses or
basal rates are very low, or at around 6 hours after the insulin
bolus was delivered, suggesting that its action has come to
an end.
1.1.1. Basal rate
Numerous methods can be used to determine the initial
basal rate for a patient currently using multiple daily insulin
injections (MDI). If the patient had been on a ‘basal-bolus’
MDI regimen, then gradually reducing the total daily basal
insulin dose by 10-20% may be considered, while keeping the
same prandial bolus doses using a rapid-acting analogue. As
for other MDI regimens, the total daily insulin dose (including long- and rapid-acting insulin) can be divided into two
parts, with 50-60% becoming the total basal dose, and 4050% becoming the total bolus dose. In both these situations,
the basal rate has to be divided by 24 to arrive at units per
hour (U/h). It is often necessary to progressively increase
the basal rate, notably in the late afternoon and in the second half of the night, according to capillary blood glucose
(Table 1). For example, if a patient has been treated by MDI
with glargine 30 U/d, and lispro 6 U before breakfast, 10 U at
lunch and 10 U before dinner, then the initial basal rate with
CSII could be 20 U/d or 0.8 U/h, with boluses at the same
lispro dosage as before.
The basal rate may also be determined using the patient’s
body weight in cases where the patient’s insulin needs are
unknown. A conservative starting basal dose may be calculated
using 0.22 U/kg body weight/d and dividing the result by
24 hours, to arrive at the number of units per hour (U/h). Up to
four periods of basal rate variations are usually programmed to
mimic physiological insulin secretion as closely as possible.
1.1.2. Prandial bolus
The prandial bolus is administered just before each meal
or extra snack, provided that there is a period of at least 3
hours between boluses to avoid hypoglycaemic episodes due
to insulin overlap. Initially, prandial bolus doses are kept the
same as the rapid-acting analogue doses with MDI. However, when the patient has been treated with regular insulin,
it is recommended to decrease the size of the preprandial
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
Table 1
A method for determining the initial basal rate in a hospitalized patient
1 Initial dose
Reduce basal insulin total daily dose by 10-20%, then divide by 24 to
arrive at units per hour (U/h). Initially, only one basal rate is prescribed
2 Separate the day into two parts if necessary
For fasting glycaemia, adjust the basal rate: 0 h to wake-up time
(around 0700 h)
For glycaemia before dinner (around 1900 h), adjust the basal rate:
0700 h to 0 h
3 If necessary, separate the day in four parts
For glycaemia at 0400 h: 0 h to 0400 h
For fasting glycaemia: 0400 h to wake-up time
For glycaemia before lunch (around 1200 h): 0800 h to lunch time
For glycaemia before dinner (around 2000 h): 1200 h to 0 h
A fifth basal rate may be used but, usually, four basal rates are sufficient.
4 Decrease each basal rate by 10% when the patient leaves hospital
5 At home
The patient can validate these basal rates through complete or glucosefree fasting periods.
boluses of insulin analogues by 10-20%. Bolus calculation
is estimated according to the carbohydrate amount in each
meal, and these dosages may be flexible or not (see Flexibility of insulin therapy). Bolus doses must also take preprandial
glycaemia into consideration. If it is above target, a corrective
dose must be added to the bolus (see below). Preprandial bolus
dosages need to be modified according to the blood glucose
results obtained in the postprandial period (one or two hours
after each meal).
1.2. Glycaemic targets
Insulin-dosing requires that the physician sets a range of
glycaemic targets for each patient; for example, in a young
patient without hypoglycaemic unawareness or unstable
retinopathy, the targets could be as follows [3]:
Fasting and preprandial blood glucose: 70-130 mg/dl
Postprandial blood glucose: < 180 mg/dl
Postprandial glucose measurements should be taken one
to two hours after the beginning of the meal, which is usually
when glucose levels peak in patients with diabetes [3]. Patients
should also be given an individualized target blood glucose
range that may vary over time according to the clinical context
(paediatric, pregnancy, hypoglycaemic unawareness). In cases
of preproliferative retinopathy or with hypoglycaemia-unaware
patients, safety targets should also be laid out, for example:
Fasting and preprandial blood glucose: 100-150 mg/dl
Postprandial blood glucose: 150-250 mg/dl
427
Glycaemic objectives are set independently of how insulin
is administered-in other words, these targets are not specific
to the insulin pump, but to the individual patient.
1.3. Algorithms for insulin dose adaptation
A number of methods can be used to adjust rapid-actinganalogue insulin doses in the pump: retroactive; immediate;
anticipatory; and functional. A blood glucose diary or a computerized equivalent is a necessary tool for the patient and physician alike, as it can be used to analyze blood glucose results
and adjust doses as necessary, and can also suggest modifications according to specific situations (such as sports activities, illness or a dinner party). The rules for dose adaptation
vary according to the treatment centre. In general, however,
there are two main strategies: conventional dose adaptation
uses fixed doses that correspond to a fixed glucose quantity
in each meal; functional or flexible insulin therapy (FIT) is
described in Flexibility of insulin therapy.
1.3.1. Basal rate adaptation (Table 1)
Nocturnal basal rates are adjusted according to glucose
levels at 4am and at wake-up time (hours to be set by the
patient). Basal rates during the day are set according to the
fasting glycaemia before lunch and before dinner, and while
fasting during the day (partial or total). It is advisable not to
use more than these four basal-rate periods. When a significant
increase results in a blood glucose level higher than the target
range for two or three days during the preprandial period, the
basal rate for that period of time needs to be increased by 0.10.2 U/h, depending on the previous basal rate (< or > 1.0 U/h, for
example). On the other hand, on the day after a hypoglycaemic
episode, the insulin basal rate should be decreased by 0.1-0.2 U/h,
depending on the previous basal rate (< or > 1.0 U/h, for
example), and the degree and severity of the hypoglycaemia.
At home, the patient may be able to validate these basal rates
during glucose-free or total fasting periods.
1.3.2. Bolus adaptation according to carbohydrate intake
Boluses should be adapted according to the postprandial
glucose value. It is advisable to take these measurements one
or two hours after the start of a meal, which is usually when
glucose levels peak in patients with diabetes [3].
In patients with a regularized carbohydrate intake at each
meal, bolus doses can be increased by 10% if the postprandial
glycaemia is greater than target levels for more than three days.
However, if the postprandial blood glucose is below target or
in hypoglycaemia, the insulin bolus should be decreased by 14 U the following day. In patients with flexible carbohydrate
intakes, the bolus–or the ratio of carbohydrate units to insulin (see Flexibility of in insulin therapy)–should be adapted
according the postprandial glucose levels.
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
428
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
The patient is considered to have an adequate bolus dose
if the difference between the one-hour postprandial and preprandial blood glucose does not exceed 80 mg/dl (mean
40 mg/dl) [4].
1.3.3. Correction of insulin doses
Preprandial additional or corrective insulin (bolus) doses
are administered when blood glucose is outside of target
range, according to blood glucose measures and total insulin
dosages [5]. It is advisable to determine, for each patient, the
dose required to reduce glycaemia by 100 mg/dl during the
initial hospitalization. For a type 1 diabetic adult patient with
a normal body mass index (BMI), such a required dose is
approximately 2-3 U. When correction doses are required, it
is necessary to adapt the corresponding dose–the basal rate or
bolus–according to the time of the hyperglycaemia to prevent
another similar episode during subsequent days.
If blood glucose is outside of the target range far from mealtimes and/or during the night, the use of corrective boluses
is not recommended due to the high risk of hypoglycaemic
episodes, except in the case of ketosis (see Management and
prevention of hypoglycaemia and ketosis: specific features of
pump treatement).
cases of intense and immediate postprandial physical activity
(such as swimming or intense physical activity), it is also suggested that the preprandial boluses be decreased. An alternative solution is to disconnect the insulin pump from the
catheter, but only for a maximum of two hours.
Glycaemic control is recommended both before and after
exercise. To avoid hypoglycaemic episodes, additional snacks
should be ingested when the physical activity exceeds 30 minutes; an intake of 20 g/h of carbohydrates is usually recommended, whatever the insulin regimen. At the next meal,
patients are then advised to increase their carbohydrate intake
by 50-100%. For intense physical activity or training at the end
of the day (just before or after dinner), a reduction in the basal
rate during the nocturnal period is also recommended.
1.3.5. Other specific situations
In cases of acute illnesses, basal insulin requirements
are usually increased. These higher basal rates should then
be maintained or even further increased if food intake is
reduced or halted. Under such conditions, rapid sugar ingestion is recommended as well as an appropriate reduction in
insulin boluses.
1.4. Other advice
1.3.4. During physical activity
At these times, the most important adaptation is to reduce
the basal rate and apply the temporary basal rate available with
all insulin pumps at least one hour before starting the exercise
until one or two hours after stopping the activity. The basal
rate should be lowered by around 30-50% (and sometimes
more than 70%), depending on the duration and intensity of
the exercise, and the patient’s level of fitness (Table 2). In
Table 2
Dose adaptation during physical activity in patients treated by CSII
1 Use the temporary basal rate
From 1 hour before starting exercise activity and up to 1 or 2 hours
afterwards;
Basal rate should be lowered by about 50% (sometimes > 70%),
depending on the duration and intensity of the exercise, and patient’s
fitness level;
For intensive activity or training scheduled at the end of the day (just
before or after dinner), a reduction in basal rate during the nocturnal
period is also recommended (by approximately 30%).
2 Decrease boluses closest to time of physical activity, before or after
For intense and immediate postprandial physical activity: decrease the
pre- or postprandial boluses by 30–50%, depending on the intensity of
the exercise.
3 Disconnection of insulin pump from catheter site
For a maximum of 2 hours, especially for swimming or other intense
physical activity.
4 If activity is far from mealtime
Additional snacks should be eaten when physical activity exceeds
30 minutes, with an intake of 20 g/h of carbohydrates.
5 Glycaemic control
Required both before and after exercise.
Troubleshooting guidelines for dose adaptation, protocols
in case of severe hyperglycaemic episodes, ketonaemia or
pump failure, and emergency phone numbers should be given
to each patient in writing. If glycaemic deterioration occurs,
the patient should know how to check for ketonaemia or
ketonuria (see Management and prevention of hypoglycaemia and ketosis: specific features of pump treatement). Also,
an ambulatory clinical visit may be scheduled for one month
after starting CSII to check and verify how well the patient
has understood the procedure and advice given.
To avoid insulin absorption problems, a new subcutaneous
infusion site usually needs to be selected every three days,
and the cartridge changed as necessary.
Blood glucose levels and insulin doses should be recorded
on a flow chart and reviewed by the medical team every three
to four months.
Patients should have at hand an at-home ‘survival kit’,
comprising insulin pens (delivering fast- and long-acting
insulin analogues) and needles in case of pump failure (with
the appropriate replacement doses already calculated for the
patient), and a glucagon kit.
1.5. Self-monitoring lood glucose
This may need to be intensified for reasons related to both
treatment efficacy and safety. In addition to preprandial levels,
postprandial blood glucose levels also need to be monitored at
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
429
1.5 to 2 hours after a meal, or earlier, to allow adjustment of
insulin doses if necessary, according to blood glucose, food
intake and activity levels [6]. This is particularly necessary
at the beginning of CSII therapy to determine the patient’s
insulin/carbohydrate sensitivity with flexible insulin therapy
(FIT). During the first month at least, blood glucose levels
should be checked six times a day; subsequently, four or five
capillary glucose tests should be done each day. A bedtime
test of capillary glucose is also useful for checking metabolic
control before the nocturnal period, when the risk of accidental
interruption of insulin delivery is high. In a large number of
studies and in clinical practice, nocturnal blood glucose tests
performed between 2am and 4am were found to provide useful
information for a more precise adjustment of the night-time
basal rate. Patients treated with an external insulin pump are
exposed to the well-known risk of the concomitant occurrence
of unusual hyperglycaemia and symptoms considered by clinicians to suggest ketotic decompensation that may warrant
ketone-body determination. Defining this threshold of hyperglycaemia is, of necessity, arbitrary, and published reports
have not proposed a consensus value for blood glucose (2.23 g.L-1). However, for purposes of simplification, an unusual
blood glucose > 2.50 g.L-1 should prompt testing to detect
ketone bodies in the urine or blood. In fact, it is recommended
to check for ketone bodies once a day to detect any technical
problems related to the device. The most appropriate time to
do this is probably at bedtime [7].
hypoglycaemic episodes especially during physical activity,
which is not always adequately covered in children and adolescents [8].
The guidelines for switching from injections to CSII have
been shown to be different for the paediatric population compared with adults, and prepubertal children in particular. A
reduction (-20%) in total insulin dose is necessary on initiating pump therapy in pubertal, but not prepubertal, children
[9,10]. In addition, although 50% of the daily insulin dose is
given as basal rates, the occurrence of maximum basal rates
is different between adults and children. Several paediatric
studies, involving the use of fast-acting insulin analogues in
pumps in prepubertal children, have shown a maximum basal
rate from after dinner until midnight. In young children, a
lower basal rate is required for the second part of the night.
Nocturnal basal-rate adjustment is an essential instrument for
reducing the risk of hypoglycaemia in diabetic children, especially in high-risk children aged <5 years [9]. The number of
basal rates in this population is frequently reported to be two
or three over 24 hours (early night, late night and daytime).
Paediatricians need to be aware of these specific factors
to optimize the efficacy of CSII therapy. In children, small
changes in bolus and basal rates are recommended to avoid fluctuations in blood glucose. Very small changes, such as 0.05 or
0.1 U/h for basal rates or boluses, can be accurately delivered by a pump-unlike pens and syringes-and are particularly
suitable for very young children.
1.6. Specific situations
1.6.2. During pregnancy
Ideally, insulin pump therapy should be initiated prior to
pregnancy to obtain near-normal glycaemia, thereby reducing
the risk of spontaneous abortion and fetal malformations.
Nevertheless, insulin pump therapy can be started during
pregnancy in carefully selected patients. The criteria include
patients who are highly compliant with capillary glucose
testing (8 to 10 times a day), those who have failed to achieve
acceptable control with MDI therapy and those who have
asked to be treated with an insulin pump. These patients must
also agree to monitor ketone bodies.
Insulin pump therapy should be initiated in specialist centres. The patient has to be well trained in pump use to spot
any mechanical problems such as faulty placement of the
insulin syringe or a kink in the infusion cannula that, if left
unchecked, can interrupt insulin release and lead to hyperglycaemia or ketoacidosis. Hyperglycaemia, which is likely
whenever pump infusion is interrupted, is particularly undesirable during pregnancy.
Although any type of pump may be used, insulin requirements are sometimes very high near the end of pregnancy so,
in general, it is better to use a pump with a large capacity.
Also, catheters should be long even for thin women, and it is
better to use a tangential catheter rather than a straight one.
1.6.1. In the paediatric population
In paediatric cases, patient selection is more a ‘family’
selection, as the family is a key component in the paediatric
patient’s daily management and decision-making process.
Patient and parental motivation is another key factor in the
successful use of the pump. The first step is to provide comprehensive information on the advantages and limitations of
pump therapy, the goals of such intensive therapy, and the
need for at least four blood glucose tests a day, and to assess
the feasibility, degree of support and disease awareness of
the family, their expectations for the CSII treatment and the
commitment to providing suitable follow-up. The parents of a
young patient must be able to quickly respond in the event of
an accidental disconnection. The decision time must be long
enough to form an effective, voluntary partnership.
The success of pump therapy depends to a large extent on
the skills and experience of the multidisciplinary paediatric
team. The members of the team, especially the paediatrician
and diabetes nurses, interact frequently with the patient and
parents. They are also responsible for the initial individual
educational programme about the treatment and for followup. An important issue is prevention of ketoacidosis and
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
430
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
The abdominal wall should be avoided during pregnancy as
insulin absorption becomes unpredictable with the progression
of pregnancy, so alternative infusion sites need to be selected.
Patients should change their pump catheter every 24 to
48 hours, and are instructed to change infusion site if they
have two unexplained blood glucose levels >200 mg/dl.
Approximately 50-60% of the total dose is given as a basal
insulin infusion, with the remaining insulin given as boluses
prior to each meal. The basal rate is usually reduced while
the patient is sleeping and in the early morning hours, but
is increased as morning approaches. The remainder of the
calculated insulin dosage is administered as boluses with
the patient’s meals or snacks. However, protocols may vary
during pregnancy, with a reduction of insulin requirements
up to week 12 of pregnancy and an increase from then up to
week 35 of gestation.
Programmable square-wave bolus doses are useful during
pregnancy as it reduces the risk of glucose excursions and
hyperglycaemia. The square-wave bolus can be administered
at the start of a meal, but extended over several hours to reduce
the hyperglycaemic peak after meals due to delayed gastric
emptying in normal pregnancy as well as gastropathy.
During labour, the pump may be maintained with a single
basal insulin infusion and strict glycaemia monitoring. After
delivery and during the breastfeeding period, the patient’s
requirements return to their prepregnancy levels except that
the basal insulin infusion and boluses are decreased (by about
30%) to avoid hypoglycaemia.
Use of CSII during pregnancy requires skillful professional
care especially by the obstetric staff, as well as careful selection of patients, meticulous patient monitoring and thorough
patient education.
2. Management and prevention of hypoglycaemia
and ketosis: specific features of pump treatment
2.1. Hypoglycaemia
Several studies suggest that the risk of moderate or severe
hypoglycaemia is not increased by pump treatment [11-16].
There is, however, a specific risk of hypoglycaemia, or worsening of preexisting hypoglycaemia, related to the possibility
of excess insulin infusion. However, experience has shown
this risk to be more theoretical than actual, and it can be considerably reduced through proper patient education on the
technical aspects of the pump, and by alarm and other safety
features built into the pumps (Table 3).
Management of a hypoglycaemic episode during pump
therapy is comparable to that used for a multiple-injection
basal-bolus regimen [17,18]. In cases of severe hypoglycaemia and while awaiting specialized care by the medical
team, disconnecting the pump may reduce the time spent in
hypoglycaemia.
Table 3
Specific risks of hypoglycaemia and ketosis related to pump therapy
Hypoglycaemia prolonged by persistent excess insulin infusion
Mechanism
Detection and/or prevention
Pump ‘runaway’ due to too much
pressure in the reservoir
Theoretically impossible
according to material technical
specifications
Incorrect pump programming:
excessively high basal rate;
excessively high or supplemental
bolus; time incorrectly adjusted;
changed insulin concentration
Patient education (i.e. programming, checks); safety features:
automatic pump cut-out; overinfusion alarm; flow interruption
and maximum bolus alerts;
push-button lock; clear programming displays; history recall of
last boluses delivered; ability to
cancel bolus
Ketosis/ketoacidosis due to interruption in insulin infusion
Electronic or mechanical failure
Safety alerts, most frequently
for SMBG and ketone-body
detection
Empty reservoir
Safety alerts
Catheter occlusion
SMBG and ketone-body
detection; delayed occlusion alert
Insulin leakage, air bubbles
SMBG and ketone-body
detection (no alert)
SMBG, self-monitoring of blood glucose
2.1.1. Fine-tuning of insulin doses
The risk of hypoglycaemia is reduced by the exclusive and
continuous administration of fast-acting insulin, and the number of possibilities offered by the pump for adjusting insulin
doses. The basal rate can be modified every 30-60 minutes to
maintain a near-perfect match to the patient’s insulin needs
and variations throughout the day, and to keep blood glucose
levels stable during interprandial periods, thereby allowing
meals to be taken later than usual without major risk of hypoglycaemia. As basal rates can be adjusted to a precision of
0.05-0.10 U/h and boluses by 0.1-0.5 U, when the insulin
is diluted, so even finer modifications can be made, thereby
reducing the risk of hypoglycaemia both in adults with low
daily needs and in children [9].
2.1.2. Temporary basal rates
Temporarily reducing the basal rate can prove useful for
patients engaging in physical activity, especially unplanned
activities. Ideally, the rate should be reduced (by 30-50%,
depending on the type and intensity of the exercise) [18,19]
at least one or two hours before the exercise as well as dur-
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
ing it. For intensive exercise or a day-long activity (such as
skiing or other endurance exercise), basal rates can be lowered by 10-20% to reduce the risk of hypoglycaemia, especially during the following night. For contact or aquatic
sports, the only alternative may be to disconnect the pump
[20], but patients should not remain off the pump for more
than 2 hours to avoid the risk of ketosis. A bolus, along with
carbohydrates, should be administered before or after disconnecting the pump to provide the equivalent insulin dosages
not delivered while the pump is disconnected.
2.1.3. Customizing the bolus
Customizing prandial insulin infusions allows patients to
avoid certain types of hypoglycaemia that can occur at the end
of the meal, such as when the meal lasts longer than expected
or when the carbohydrate content is lower than expected.
Any method can be considered as the number of boluses is
not a problem for the patient. Part of the bolus can be administered at the beginning or halfway through a meal, and the
remainder at the end of the meal or with dessert, or it can
be taken with each course. Bolus infusion over an extended
period (from 30 minutes to 8 hours with some pumps), and
perhaps combined with an immediate bolus, can be a solution
in this particular situation. This approach is especially useful for patients who have documented gastroparesis [21] or
for those whose blood glucose decreases rapidly in the early
postprandial phase to subsequently rise again [22].
431
only attribute glycaemic deterioration to such a banal cause
when defective insulin administration has been dismissed
as a reason.
When hyperglycaemia occurs and the search for ketone
bodies proves positive, then insulin deficiency is confirmed
and needs to be treated immediately. If no ketone bodies are
found, then the diagnosis is uncertain (‘simple’hyperglycaemia,
or the beginnings of insulin deficiency). In either situation,
the infusion kit needs to be checked to eliminate a technical
problem such as pump malfunction, insulin leakage (a dislodged
or damaged catheter) or some element that could lead to
catheter obstruction (for example, blood in the tubing). If
the infusion kit appears to be in proper working order, then
blood glucose should be checked again about two hours later,
together with a test for ketone bodies [27]. As ketonaemia
monitoring allows the earlier detection of ketone bodies and
quicker identification of a deficiency of insulin administration, it should be encouraged during insulin pump therapy
[7, 28-30].
The risks of hyperglycaemia, ketosis and ketoacidosis specifically related to pump therapy are potentially higher than
with MDI treatment [12,23]. If insulin infusion is stopped as a
result of a technical problem with the infusion kit or improper
use of the kit, this may lead to a deficiency of rapid-acting
insulin [24] (Table 3). Thus, the prevention of ketoacidosis
depends on early identification of an insulin deficiency, and
the swift implementation of corrective measures and troubleshooting to determine the cause of the ketoacidosis. It
also requires comprehensive patient education through an
experienced centre.
2.2.2. Symptomatic correction of insulin deficiency
Once the insulin deficiency has been identified, it needs
to be corrected by subcutaneous rapid-acting insulin analogue supplements. The first supplement is delivered using
a prefilled pen. If the pump is to be used again, the cause of
the problem has to be identified and resolved (the catheter
changed, for example) before using it to deliver the supplemental insulin, which may lead to a delay in treatment. Supplements should be renewed every 2-3 hours until no more
ketone bodies are present. The dosage of the insulin supplements will vary according to the levels of blood glucose and
ketone bodies, and the usual practices of the medical team
(Table 4) [30,31]. Should the patient fail to improve (worsening hyperglycaemia, continued presence of ketone bodies)
following the administration of two insulin supplements, or
the clinical signs of worsening appear (presence of confusion,
vomiting), the patient or his family should contact the initiating
centre to decide whether or not to continue with ambulatory
treatment. This is especially necessary with ketonuria > + +
or ketonaemia > 3 mmol/l, when the likelihood of diabetic
ketoacidosis is particularly high [31,32].
2.2.1. Identifying insulin deficiency
A defect of insulin delivery should always be suspected with
each hyperglycaemic episode arising during pump therapy.
Routine testing for ketone bodies should be performed when
hyperglycaemia occurs, whatever the circumstances. Most
cases of ketoacidosis during pump therapy are the result of a
later identified or treated insulin deficiency [25,26], fostered by
patient overconfidence in the insulin infusion and a situation
that, in itself, was likely to cause hyperglycaemia (such as
excess carbohydrate intake). For this reason, patients should
2.2.3. Troubleshooting the cause of and correcting
an insulin administration defect
The attempt to identify and correct the cause of glycaemic
deterioration begins after the first insulin injection by pen,
and involves a thorough check of the infusion kit, including
inspection of the withdrawn catheter/cannula along with the
infusion site. This inspection should first look for an obstruction or folding (by sending a bolus through the kit), and check
that the cannula has not become dislodged. Pump therapy can
be restarted once the catheter, reservoir and, if necessary, the
2.2. Hyperglycaemia and ketosis
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
432
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
Table 4
Management of hyperglycaemia with ketosis
3. General pump function and comparison of available
insulin pumps in 2008
Supplemental doses of rapid-acting insulin as percentage of daily dose
(adapted from Laffel [30])
Blood glucose (mg/dl)
Ketone bodies
< 250
Urine: 0 or trace
No
Blood: < 1 mmol/l
change
Urine: +
Blood: 1–1.4 mmol/l
Urine: ++ or +++
Blood: > 1.5 mmol/l
250-400
> 400
5%
10%
0-5%
10%
15%
0-10%
15-20%
20%
For hyperglycaemia > 250 mg/dl (protocol from Department of
Diabetology, Rangueil Hospital, Toulouse)
Ketone bodies
Urine: 0
Blood: < 0.3 mmol/l
Diagnosis and management
Normal
Re-check blood glucose and set perfusion,
ketone bodies after 2 h
Suspect
Blood: 0.3-0.4 mmol/l
Urine: ⭓ trace
Blood: ⭓ 0.5 mmol/l
Re-check blood glucose and set perfusion,
ketone bodies after 1-2 h
Abnormal
Supplemental doses of insulin every 2-3 h
Urine + or blood 0.8-1 mmol/l g 4 U of insulin (3-5 U)
Urine ++ or blood 1.6-2 mmol/l g 8 U of insulin (6-10 U)
Urine +++ or blood 2.4-3 mmol/l g 12 U of insulin (9-15 U)
insulin have been changed (because of insulin denaturation
and/or old cartridges). Over the next few hours, the patient
may need supplemental insulin doses, which can be administered with the pump as boluses. Should the pump prove (or
appear to be) the cause of the problem, the supplemental doses
should be administered through multiple injections until the
defective item has been replaced by the provider.
2.2.4. Precautionary measures: self-monitoring
and the ‘survival kit’
Patients have to be able to identify and troubleshoot at
an early stage, even before any alarm is triggered, any problems that might interrupt the administration of insulin. For
this reason, they are advised to check their blood glucose at
least four times a day (and sometimes more) and to carry with
them at all times a self-monitoring kit (glucose meter, lancet, strips for testing for capillary blood glucose and ketone
bodies), a rapid-acting-analogue insulin pen and needles. The
recommendations of the paramedical ALFEDIAM, updated
in 2007, insist upon these important aspects of patient monitoring [33].
The goal of insulin pump therapy is to mimic the normal
function of the pancreas as closely as possible. Nowadays,
insulin pump technology is relatively well developed. Modern insulin pumps are about the size of a pager, and are able
to administer rapid-acting insulin analogues to the majority
of patients at precise basal and bolus rates to control blood
glucose levels throughout the day and night. The newer pumps
have even longer-lived batteries, and hold around twice as
much insulin as did the older pumps. More specifically, the
technical sophistication of these devices has made them much
easier to use in young children. This review discusses how
insulin pumps work, and the differences between bolus and
basal rate delivery with the currently available models.
Subcutaneous insulin infusion (CSII) via an insulin pump
helps to overcome the limitations of multiple daily injections
(MDI) by imitating physiological insulin secretion. The two
main components of insulin secretion are basal and prandial
insulin secretion. The pump controls the infusion rates to
create both basal and bolus profiles.
The benefits of CSII may vary according to the patients’
age. For adolescent insulin-pump users, they have the option
to programme a slightly higher insulin infusion rate for the
early morning hours to address a rise in blood glucose at that
time. Elderly patients with longstanding type 1 diabetes who
experience overnight hypoglycaemia can programme the
pump to reduce the amount of basal insulin at night. Lifestyle benefits are a major advantage of insulin pump therapy
in younger patients as they can programme several basal rates
at different times throughout the day. They can also modify
their insulin regimen by using a temporary basal rate during
physical activity, depending upon the intensity and duration of
the exercise. This is possible because of the unique ability of
the latest insulin pumps to allow precise basal rates to be programmed and modified at different times of the day [34].
The pump user can also change the time-action profile of
insulin boluses according to the expected type, size and duration of meals. This confers greater flexibility in meal timing
and quantities. Thus, the modern insulin pump offers several technological features that facilitate fine-tuning of the
daily insulin profile.
3.1. General pump function
3.1.1. Basal insulin rates
Treatment starts with a single basal rate that is monitored
before any additional basal rates are programmed. An extensive glucose profile is important for counteracting the dawn
phenomenon or identifying those patients who are prone to
overnight hypoglycaemia. If readings are > 300 mg/dl, then
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
433
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
checking capillary blood glucose at bedtime, 12am, 3am and
6am, and adjusting the overnight basal rate, is recommended.
Experience shows that adults often need an increase in basal
rate in the dawn hours (from 4-9am) whereas children often
need an earlier increase in basal rate. For basal rate adjustments during the day, it is recommended to check capillary
blood glucose before mealtimes, and it may also be worth
skipping a meal to check it every two hours over the following six hours. Additional adjustment to the daytime basal
rate is necessary if capillary blood glucose is > 300 mg/dl.
Most patients need no more than two or three basal rates per
24 hours or when initiating pump therapy during a stay in a
hospital care unit (Fig. 1).
Bolus
Lunch
Bolus
Dinner
Blood glucose change
from baseline (mg/dl)
Plasma insulin
Breakfast
lin, followed by a sustained administration and then a rapid
decrease in insulin. It can provide adequate insulin coverage
for patients enjoying an extended multiple-course meal or
those who have slow gastric emptying due to gastroparesis.
Pumps can also deliver a combined bolus, which is a standard
bolus followed by an extended bolus. This is helpful for meals
that contain rapidly absorbed nutrients followed by multiple
courses. A dual-bolus profile also appears to be slightly better
at mimicking physiological insulin secretion and reducing
postprandial glucose excursions [22]. In that study, nine type
1 diabetic patients using CSII ate pizza, tiramisu and cake
for four consecutive Saturdays, each time using a different
type of bolus (Fig. 3).
Bolus
Basal infusion
80
60
40
20
0
-20
4:00
8:00
12:00
16:00
20:00
24:00
4:00
8:00
Time
-40
0.5
1.0
1.5
2.0
4.0
5.0
Hours from baseline
Fig. 1. Variable basal rates using a CSII programme.
Single bolus
Double bolus at -10 and 90 mins
Square wave bolus over 2 hours
Dual wave bolus (70% at meal, 30% as 2-h square)
Delta plasma
insulinaemia (mU/l)
3.1.2. Bolus insulin administration
Nowadays, insulin pumps are able to administer different
types of boluses (Fig. 2). The standard insulin bolus has an
immediate delivery followed by a decrease in insulin. However, some patients may be restricted by the time-action profile of fast-acting insulin analogues and may benefit from an
extended bolus. This is characterized by a rapid rise of insu-
80
60
40
20
Time (minutes)
0
60
120
Fig. 3. Comparison of different modes of pump boluses with a highcarbohydrate, high-fat meal (adapted from Chase et al. [22])
A French multicentre randomized study comparing standard and dual-wave boluses is currently looking to determine whether or not a more ‘physiological’ bolus profile
does indeed improve metabolic control in everyday life.
Patients can customize their bolus regimens to their dietary
intake and habits using FIT. The latest pumps have a built-in
bolus calculator that determines the appropriate insulin dose:
patients need only enter their blood glucose target, carbohydrate ratio and insulin-sensitivity factor. This function can
reduce errors in calculation and decrease the number of correction boluses required.
180
3.2. Comparison of currently available insulin pumps
Standard
Square
Dual wave
Fig. 2. Physiological insulin secretion (upper) compared with the different
modes of bolus delivery with insulin pumps (lower).
There are differences between insulin pumps, and it is
important to examine these variations to help patients select
the best pump for their treatment needs and lifestyle (Table 5).
The final choice is important as the outlay for a new pump
is only reimbursed by the French national health insurance
system every four years.
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
434
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
Table 5
Features of currently available pump models
Pump Type
Paradigm 515/715
ACCU-CHEK Spirit
Deltec Cozmo
IR-2020
Manufacturer
Medtronic
Roche Diagnostics
Smiths Medical
Animas Novalab
515 (76 x 50 x 20)
Size (mm) (L x W x H)
715 (94 x 51 x 20)
81 x 20 x 55
88.9 x 19 x 50
74 x 19 x 51
Screen size (mm)
46 x 21
36 x 15
29 x 30
31 x 31
Weight (with battery and
reservoir) (g)
100/108
110
94
98
Reservoir capacity (IU)
180/300
315
300
200
Catheter connection
Specific (Paradigm)
Standard (Luer lock)
Standard (Luer lock)
Standard (Luer lock)
0.1-25.0 U
0.1-25 U
0.05-75 U
0.05-35.0 U
Bolus rate amount
(0.1-U increments)
(0.1-U increments)
(0.05-U increments)
(0.05-U increments)
Bolus type*
N, E, D
N, E, D
N, E, D
N, E, D
Bolus ‘calculator’
Bolus Assistant
Bolus calculator on PDA
Bolus Wizard
Carb Smart
Flexible insulin therapy
Yes
Yes
Yes
Yes
0.0-35 U/h
0.0-20.0 U/h
0.0-35 U/h
0.0-35 U/h
Basal rate range (U/h)
(0.05 U/h increments)
(0.1 U/h increments)
(0.05 U/h increments)
(0.025 U/h increments)
Number of basal rates per day
48
24 of 1 hour each
48
12
Basal delivery intervals
3-min, 0.005
3-min, 0.005
4
Temporary basal patterns
3
15 min to 24 h
Delivery accuracy
± 5%
± 5%
± 5%
± 5%
Last 14 days on display,
Memory: daily total insulin
90 days downloadable
Last 30 days on display, last
6000 events downloadable
4000 events
255 on display
Display format
Menu-driven
Icons and text
Menu-driven
Menu-driven
Display backlight
Yes
Yes
Yes
Yes
Alert types
Beep or vibrate
Beep or vibrate
Beep or vibrate
Beep or vibrate
Occlusion alert
Yes
Yes
Yes
Yes
Low-battery alert
Yes
Yes
Yes
Yes
Low-reservoir alert
Yes for 30, 20, 10 U
remaining
Yes
Yes
Yes
Resistance in the water
Watertight
Watertight (up to 2.50 m
depth and for 60 min)
Waterproof
Waterproof
Yes
Palm OS (ezManager)
Yes
Connection to PC
Yes
(downloading by ACCUCHEK Smart Pix; analysis
by ACCU-CHEK Pocket
Compass)
*N: Normal immediate bolus - E: Extended square-wave bolus - D: Dual-wave bolus
3.2.1. Medtronic MiniMed Paradigm 515/715
The MiniMed Paradigm pump by Medtronic (Northridge,
CA, USA) has a scroll bar on the display face to facilitate
moving through the menus. The pump is able to deliver several types of insulin boluses, including a normal (immediate)
bolus, an extended square-wave bolus and a combination of
normal plus extended dual-wave bolus, depending on the type
of meal and its timing. As part of FIT, patients can enter their
carbohydrate intakes into the pump and the ‘Bolus Assistant’
will calculate the suggested insulin dose.
Paradigm can administer bolus rates from 0.1-25.0 U in
0.1-U increments and also has an audio bolus option. Basal
rates can range from 0.0-35.0 U/h in 0.05-U/h increments.
The pump also has a memory feature that allows the last
24 boluses, and the total insulin dose for the last 14 days, to
be viewed on the pump display, while the last 90 days are
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
downloadable. Patients can use a wireless remote to operate the pump worn discreetly beneath their clothing. Alarms
include infusion-line occlusion, low battery and low insulin
reservoir. The Medtronic MiniMed pump has a watertight
rating (IPX-8) up to a depth of 8 feet (2.44 m) for up to
30 minutes, and uses a standard AAA 1.5-V battery that keeps
maintenance costs down and is convenient to replace. Other
pumps require batteries that are less widely available. On the
other hand, the Paradigm is the only pump that requires a specific catheter. Finally, the choice between the 515 and 715
models depends solely on the patient’s insulin daily needs, as
the two models have different reservoir capacities. The newer
522/722 models can be integrated with Medtronic’s glucose
sensors for real-time continuous glucose readings. However,
these pumps are not yet reimbursable.
3.2.2. Roche Diagnostics/Disetronic ACCU-CHEK Spirit
Disetronic Medical Systems (Burgdorf, Switzerland) makes
the D-TRON and H-TRON lines of pumps, and the latest
ACCU-CHEK Spirit model. These pumps use icons on their
display rather than menus. They have the largest insulin-reservoir capacity, holding 315 U of insulin. The D-TRON is the
first insulin pump to use a prefilled insulin cartridge rather than
a disposable cartridge that requires manual refilling by the
patient. This is a convenience feature that some patients may
well appreciate. The two Disetronic pumps can also deliver an
extended bolus, and have backlit enhanced memory features
and alarms that can be set on vibration mode. The ACCUCHEK Spirit is the first pump able to communicate with a
PDA (personal digital assistant) to help the patient target the
insulin dose. It also has the IPX-8 standard waterproof rating up to a depth of 2.5 m for up to 60 minutes.
3.2.3. Animas Novalab IR-2020
Among the newer insulin pump companies in the American
marketplace is the Animas Corporation (Frazer, PA, USA). The
IR-2020 Pump incorporates some novel features, including
a menu-driven display, a backlight and an audio bolus option
for rapid administration of boluses without having to look
at the display screen. The pump can administer small bolus
amounts of 0.05-35.0 U in 0.05-U increments as well as
extended boluses. One particular feature of the 2020 is insulin delivery in ‘pulses’at 3-minute intervals and in amounts as
tiny as 0.005 U. However, whether this feature, which mimics pancreatic secretion, has any clinical benefit in terms of
metabolic control has yet to be established. The pump also
has an excellent memory, able to store and display the last
600 boluses, and the total 24-hour insulin doses over the last
120 days. Its alarm systems include infusion-line occlusion,
low battery and low insulin reservoir. The Animas pump has
the IPX-8 standard waterproof rating, which may be an important feature for those who indulge in water sports.
435
3.2.4. Deltec Cozmo
Cozmo (Smiths Medical, St. Paul, MN, USA) is the smallest pump on the market, which may be something that many
patients, especially women, will appreciate, and is also waterproof (up to 2.4 m for 330 minutes, or 3 minutes at 3.6 m).
A new insert system is now available. This model also has a
useful option that reminds the user to test for capillary blood
glucose after meals, and can also verify whether or not the
last bolus was delivered (considering the normal time of insulin bolus delivery is from one day to the next).
3.3. Conclusions and future directions
Given the current major progress in insulin pump technology,
it is not surprising that glucose sensors will soon also be built into
these devices [such as the Medtronic MiniMed RT (for real-time)].
At present, patients are still responsible for testing and interpreting their own blood glucose trends, and for modifying their basal
insulin rates and boluses accordingly. Nevertheless, insulin pump
therapy appears to be evolving towards a closed-loop system. As
continuous glucose sensors are developed, they will soon be able
to provide information to the insulin pump. Automatic–or, more
likely, semi-automatic–insulin dose adjustments can then be made
based on these measurements and on other individualized parameters derived from the patient’s previous glucose trends.
4. Flexibility of insulin therapy
Functional or flexible insulin therapy (FIT) represents a form
of intensified insulin therapy that enables basal and prandial
insulin needs to be managed separately through a basal-bolus
or basal-prandial regimen. Such an FIT regimen was carried out initially with two injections of Ultratard (ultralente)
and three injections of regular insulin 30 minutes before each
meal or, even better, as intramuscular injections [35]. These
injections have now been replaced by rapid- and slow-acting
analogues delivered through a pump. A major advantage of this
method is that patients can enjoy total freedom with regard to
meal timings. The preprandial insulin dose is determined according to the amount of carbohydrates to be consumed. Patients
need to learn how to calculate the amount of carbohydrates in
their meals. If a blood glucose level is high, the patient can bring
it back to within the target range using supplemental doses of
fast-acting insulin analogues. All of these adjustments can be
made using algorithms individually tailored to the given patient.
FIT is a highly sophisticated learning process that allows the
patient to become more fully knowledgeable about the system
and to have better control over it through self-adaptation of the
individualized algorithms based on acquired experience. Furthermore, despite the freedom of this method, no glycaemic
deterioration has been reported [36].
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
436
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
Is pump therapy compatible with FIT? Several studies
show that it is [37-40], and one, carried out in children, goes
so far as to suggest that the pump is the most suitable tool
for FIT [41]. Pump therapy has the dual advantage of offering a flexible approach to both prandial needs and basal rates
(Tables 6 and 7).
Table 6
Benefits of insulin pump in flexible insulin therapy (FIT)
1 FIT offers type 1 diabetics freedom in meal timing and food choices
2 Patients have to know their settings and be able to enter them into
the pump
3 Different settings can be entered for morning, midday and evening
meals
4 Pumps can be used for setting:
• meal carbohydrates: the patient enters the number of portions in
the pump (also called ‘equivalent’) or the amount of carbohydrates
in the meal
4 • compensations: the patient enters the preprandial blood glucose
value into the pump
5 The pump automatically proposes a bolus, calculated from the above
two settings.
Table 7
Programming different parameters of flexible insulin therapy (FIT) with
CSII
Parameters
CSII programming
Ratio of
carbohydrate to
insulin units
Enter carbohydrate quantity for 1 U of bolus
or
Number of units for 1 ‘equivalent’ of
carbohydrate, which can be any preestablished
quantity (i.e. 20 g/1 equiv)
It is possible to enter different parameters for
different parts of the day.
Correction of
insulin doses
Enter the decrease of glycaemia by 1 U of bolus
It is possible to enter different parameters for
different parts of the day.
4.1. Boluses
4.1.1. Integrated bolus calculator
The latest insulin pumps are designed for direct FIT operation. The patient’s parameters, such as supplemental insulin
doses and units per portion, can be programmed into most
pumps. Blood glucose and the number of carbohydrate portions for each meal can also be entered, and the total bolus
automatically calculated and proposed to the patient: for example, a patient may need 2 U/20 g of carbohydrate, and his supplemental dose is 3 U to reduce blood glucose by 100 mg/dl.
Before a meal, if his blood glucose level is 200 mg/dl and
he plans to eat three portions (for example, 100 g of potatoes, 40 g of bread and 1 orange), then his bolus will be 2 U
× 3 portions = 6 U, and the supplemental dose of 3 U makes
a total bolus of 9 U (6 U + 3 U). However, the patient could
just enter his blood glucose of 200 mg/dl into the pump along
with the number of carbohydrate portions in his meal (which
has different names according to the pump)–in this case, three
portions. The 9-U dose will then automatically be proposed
to the patient, who can choose to accept it or change it. The
use of this option differs from one pump to another, and is
optional. In real life, however, this method is often not used
by patients as it may appear to be too complicated or the
patient simply lacks the necessary information.
4.1.2. Do proteins need to be counted?
Postprandial glycaemia does not differ whatever the meal’s
protein content, although blood glucose rises considerably
more at night after a patient has had a protein-rich meal compared with a meal containing little protein [42]. The proposed
explanation is that the increase in nocturnal glucagonaemia
has been stimulated by the protein-rich meal [43]. It may be
advisable to administer a dual-wave bolus, delivered over
several hours, if a meal is protein-rich. Algorithms for this
are currently being validated.
4.1.3. Multiple boluses
FIT allows the patient to vary the number of meals he would
like. If the patient is under MDI therapy, this can involve a large
number of injections, which can be uncomfortable. In contrast,
it is easier to administer repeated prandial doses in the form
of boluses.
4.1.4. Bolus before or after the meal?
Patients prefer injecting prandial insulin after a meal as it
is then easier to calculate the portions of ingested carbohydrate. One study, involving 860 type 1 diabetics, compared
the effects on blood glucose control of administering glulisine, a fast-acting insulin analogue, both before and after a
meal [44]. HbA1c was moderately increased by 0.2%, while
postprandial blood glucose levels were less well controlled.
An insulin pump allows patients to get around this difficulty
by splitting the bolus into two parts, delivering a minimal
dose at the start of the meal and the remainder afterwards.
However, this approach needs further investigation before
any definitive conclusions can be made.
4.2. Basal rates
4.2.1. Freedom in meal timing
Basal insulin control is more flexible when insulin is administered continuously over a 24-hour period, and the pump
method has less variability of action than multiple injections
[45]. With a basal-bolus insulin regimen, is it possible to vary
mealtimes without drawbacks? The question seems somewhat
incongruous now, but this was a major issue back in the days
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
when meal timings and content had to be strictly adhered to.
Chantelau et al. carried out detailed studies of the glycaemic profiles of 10 type 1 diabetic patients treated with portable pumps under either a strict regime or a ‘freer’one. Their
results showed that a variation of 90 minutes in meal timings
had no effect on basic blood glucose levels [38].
4.2.2. Validation of basal rates
Apart from the information it provides and greater flexibility in meal timing (especially when factoring in religious considerations), fasting can validate the selected basal
rate(s). Nevertheless, a number of questions remain, such as
the choice between total fasting or carbohydrate fasting (and
then whether or not protein intake needs to be compensated
for by minimal doses of fast-acting insulin), and between a
24- or 36-hour fast [46]. The hyperglycaemia-inducing action
of proteins suggests that total fasting (or partial if necessary)
is preferable. Fasting provides the answers to three questions:
Is the dawn phenomenon properly compensated for (morning
fast)? Is afternoon blood-glucose control appropriate (lunchtime fast)? Is the basal rate during the first part of the night
correct (dinnertime fast)? This means that the insulin pump
is an ideal tool for FIT in diabetic patients (Table 7). Current
pump designs offer features that appear to aid FIT, although
their actual worth remains to be validated.
Conflicts of interest: Nathalie Jeandidier carried out clinical trial as main clinical investigator for Lilly and SanofiAventis.
Jean-Pierre Riveline has no conflict of interest.
Nadia Tubiana-Rufi and Anne Vambergue have not declared
any conflicts of interest.
Bogdan Catargi has no conflict of interest.
Vincent Melki carried out clinical trials as co-investigator
for Medtronic. He attended conferences organized by Abbott
as contributor and by Abbott and Medtronic Inc. as audience
member.
Guillaume Charpentier has no conflict of interest.
Bruno Guerci has no conflict of interest.
References
[1]
[2]
[3]
[4]
Bode B, Sabbah H, Gross T, Fredrickson L, Davidson P. Diabetes management in the new millenium using insulin pump therapy.
Diabetes Metab Res Rev 2002;18(suppl 1):S14-S20.
Guerci B, Hanaire H, Lassmann-Vague V, Leblanc H, Mihaelanu M,
Leutenegger E. Contexte et modalités de mise en route d’un traitement
par pompe à insuline en milieu hospitalier : observatoire de l’insulinothérapie par pompe. Infusystems 2006;23:27-32.
Standards of Medical Care in Diabetes-2008 American Diabetes
Association Diabetes Care 2008 31:S12-S54.
Service FJ, Nelson RL. Characteristics of glycemic stability. Diabetes
Care 1980;3:58-62.
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
437
Guerci B, Jeandidier N, Lassmann-Vague V, Renard E, HanaireBroutin H. Comment utiliser l’analogue de l’insuline lys-pro en perfusion sous-cutanée continue par pompe portable. Diabetes Metab
2000;26:513-9.
Sachon C, Jacqueminet S, Hartemann-Heurtier A, Grimaldi A. Should
diabetic patients be asked to test their blood glucose 90 to 120 minutes
after the beginning of their meals? Diabetes Metab 2006;32:377-81.
Guerci B, Tubiana-Rufi N, Bauduceau B, Bresson R, Cuperlier A,
Delcroix C et al. Advantages to using capillary blood beta-hydroxybutyrate determination for the detection and treatment of diabetic
ketosis. Diabetes Metab 2005;31:401-6.
Admon G, Weinstein Y, Falk B, Weintrob N, Benzaquen H, Ofan R et al.
Exercise with and without an insulin pump among children and
adolescents with type 1 diabetes mellitus. Pediatrics 2005;116:e348e355.
Tubiana-Rufi N, deLonlay P, Bloch J, Czernichow P. Remission of
severe hypoglycaemia incidents in young diabetic children treated
with subcutaneous infusion. Arch Pediatr 1996;3:969-76.
Conrad SC, McGrath MT, Gitelman SE. Transition from multiple
daily injections to continuous subcutaneous insulin infusion in type
1 diabetes mellitus. J Pediatr 2002;140:235-40.
Pickup JC, Kidd J, Burmiston S, Yemane N. Effectiveness of continuous
subcutaneous insulin infusion in hypoglycaemia-prone type 1 diabetes:
implications for NICE guidelines. Pract Diabet Int 2005;22:10-4.
Pickup J, Keen H. Continuous subcutaneous insulin infusion at 25
years: evidence base for the expanding use of insulin pump therapy
in type 1 diabetes. Diabetes Care 2002;25:593-8.
Bode BW, Steed RD, Davidson PC. Reduction in severe hypoglycemia with long-term continuous subcutaneous insulin infusion in type
1 diabetes. Diabetes Care 1996;19:324-7.
Boland EA, Grey M, Oesterle A, Fredrickson L, Tamborlane WV.
Continuous subcutaneous insulin infusion: a new way to lower risk
of severe hypoglycemia, improve metabolic control, and enhance
coping inadolescents with type 1 diabetes. Diabetes Care 1999;22:
1799-84.
Linkeschova R, Raoul M, Bott U, Berger M, Spraul M. Less severe
hypoglycaemia, better metabolic control, and improved quality of life
in type 1 diabetes mellitus with continuous subcutaneous insulin infusion (CSII) therapy: an observational study of 100 consecutive patients
followed for a mean of 2 years. Diabet Med 2002;19:746-51.
Hanaire-Broutin H, Melki V, Bessierre-Lacombe S, Tauber JP. Comparison of continuous subcutaneous insulin infusion and multiple daily
injection regimens using insulin lispro in type 1 diabetic patients on
intensified treatment. Diabetes Care 2000;23:1232-5.
Selam JL. Hypoglycémie et diabète insulino-dépendant. In: André
Grimaldi. Traité de diabétologie. Paris: Éditions Flammarion-médecine-science; 2005. p. 214-26.
Grimaldi A, Slama G, Tubiana-Rufi N, Heurtier A, Selam JL, Scheen
A et al. L’hypoglycémie du patient diabétique. Diabetes Metab
1997;23:100-8.
Zinmann B. Insulin pump therapy with subcutaneous insulin infusion and exercise in patients with type 1 diabetes. In: Ruderman N.
Handbook of exercice in diabetes. Alexandria (Virginia). American
Diabetes Association.
Diabetes Research in Children Network (DirecNet) Study Group.
Prevention of hypoglycemia during exercise in children with type 1
diabetes by suspending basal insulin. Diabetes Care 2006;29:2200-4.
Valensi P. La neuropathie autonome diabétique : quels sont les risques ? Diabetes Metab 1998;24 (Suppl 3):66-72.
Chase HP, Saib SZ, MacKenzie T, Hansen MM, Garg SK. Post-prandial
glucose excursions following four methods of bolus insulin administration in subjects with type 1 diabetes. Diabet Med 2002;19:317-21.
Guerci B. Acute complications of insulin pump therapy. Diabetes
Res Clin Pract 2006;74 (Suppl 2):S104-7.
Guerci B, Meyer L, Sallé A, Charrié A, Dousset B, Ziegler O et al.
Comparison of metabolic deterioration between insulin analog and
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)
438
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
N. Jeandidier, et al / Diabetes & Metabolism 34 (2008) 425-438
regular insulin after a 5-hour interruption of a continuous subcutaneous insulin infusion in type 1 diabetic patients. J Clin Endocrinol
Metab 1999;84:2673-8.
Walter H, Gunther A, Timmler R, Mehnert H. Ketoacidosen unter
langzeittherapie mit insulinpumpen. Häufigkeit, ursachen, umstände
Med Klin 1989;15:565-8.
Peden NR, Braaten JT, McKendry JB. Diabetic ketoacidosis during
long-term treatment with continuous subcutaneous insulin infusion.
Diabetes Care 1984;7:1-5.
Agence Française de Sécurité Sanitaire des Produits de Santé (AFSSAPS): communication sur le risque lié à un d’arrêt de perfusion lors
d’un traitement par pompe à insuline externe. http://agmed.sante.
gouv.fr/htm/10/dm/sdm/glycemie/risque_arret_pompe_insulin.pdf
Matta MP, Melki V, Bessiere-Lacombe S, Hanaire-Broutin H. What are
capillary blood ketone levels in type 1 diabetic patients using CSII in
normal conditions of insulin delivery? Diabetes Metab 2004;30:543-7.
Guerci B, Benichou M, Floriot M, Bohme P, Fougnot S, Franck P
et al. Accuracy of an electrochemical sensor for measuring capillary
blood ketones by fingerstick samples during metabolic deterioration
after continuous subcutaneous insulin infusion interruption in type
1 diabetic patients. Diabetes Care 2003;26:1137-41.
Laffel L. Sick-day management in type 1 diabetes. Endocrinol Metab
Clin North Am 2000;29:707-23.
Taboulet P, Haas L, Porcher R, Manamani J, Fontaine JP, Feugeas
JP et al. Urinary acetoacetate or capillary beta-hydroxybutyrate for
the diagnosis of ketoacidosis in the Emergency Department setting.
Eur J Emerg Med 2004;11:251-8.
Taboulet P, Deconinck N, Thurel A, Haas L, Manamani J, Porcher R
et al. Correlation between urine ketones (acetoacetate) and capillary
blood ketones (3-beta-hydroxybutyrate) in hyperglycaemic patients.
Diabetes Metab 2007;33:135-9.
Recommandations de bonnes pratiques, ALFEDIAM paramedical.
Prise en charge des patients traités par pompes à insuline externe,
portable et programmable (juin 2003 remis à jour mars 2007). http://
www.alfediam.org/membres/recommandations/Pompes.pdf
Catargi B, Meyer L, Melki V, Renard E, Jeandidier N, for the EVADIAC Study Group. Comparison of blood glucose stability and HbA1C
between implantable insulin pumps using U400 HOE 21PH insulin
and external pumps using lispro in type 1 diabetic patients: a pilot
study. Diabetes Metab 2002;28:133-7.
Howorka K, Pumprla J, Schlusche C. Algorithm adaptation under
functional insulin treatment during transfer from regular to insulin
lispro. Diab Nutr Metab 1998;11:266-72.
Franc S, Dardari D, Boucherie B, Hochberg G, Biedzinski M, Lejeune
M et al. Validation des paramètres de l’insulinothérapie fonctionnelle (IF) chez les diabétiques de type 1 dans la vie courante. Diabetes Metab 2007;33(special issue 1):P171.
Bode BW, Tamborlane WV, Davidson PC. Insulin pump therapy in
the 21st century. Strategies for successful use in adults, adolescents,
and children with diabetes. Postgrad Med 2002;111:69-77.
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
Chantelau E, Sonnenberg GE, Stanitzek-Schmidt I, Best F, Altenähr
H, Berger M. Diet liberalization and metabolic control in type 1 diabetic outpatients treated by continuous subcutaneous insulin infusion. Diabetes Care 1982;5:612-6.
Lafrance L, Rabasa-Lhoret R, Poisson D, Ducros F, Chiasson JL.
Effects of different glycaemic index foods and dietary fibre intake
on glycaemic control in type 1 diabetic patients on intensive insulin
therapy. Diabet Med 1998;15:972-8.
Hunger-Dathe W, Braun A, Müller UA, Schiel R, Femerling M, Risse
A. Insulin pump therapy in patients with Type 1 diabetes mellitus:
results of the Nationwide Quality Circle in Germany (ASD) 19992000. Exp Clin Endocrinol Diabetes 2003;111:428-34.
Alemzadeh R, Ellis JN, Holzum MK, Parton EA, Wyatt DT. Beneficial
effects of continuous subcutaneous insulin infusion and flexible multiple daily insulin regimen using insulin glargine in type 1 diabetes.
Pediatrics 2004;114:e91-5.
Laurent C. L’alimentation du diabétique traité par l’insuline. In: Dérot
M, Précis de diabétologie Ed. Masson, Paris 1977:833-47.
Winiger G, Keller U, Laager R, Girard J, Berger W. Protein content
of the evening meal and nocturnal plasma glucose regulation in type1 diabetic subjects. Horm Res 1995;44:101-4.
Garg K, Rosenstock J, Ways K. Optimized basal-bolus insulin regimens in type 1 diabetes: insulin Glulisine versus regular human insulin in combination with basal insulin Glargine. Endocrine Practice
2005;11:11-17.
Pickup JC, Kidd J, Burmiston S, Yemane N. Determinants of glycaemic control in type 1 diabetes during intensified therapy with multiple
daily insulin injections or continuous subcutaneous insulin infusion:
importance of blood glucose variability. Diabetes Metab Res Rev.
2006;22:232-7.
Elgrably F, M’bemba J, Sola Gazagnes A, Rouxel A, Daenen S, Bruzzo
F et al. [Jeûne ambulatoire de 36 h pour la détermination des doses
de base d’insuline avec auto-apprentissage intensifié ambulatoire :
étude chez 27 patients diabétiques de type 1 sélectionnés]. Diabetes
Metab 2004;27(suppl 1):P13.
Appendix
This article is a collaborative work by the following authors,
each of whom has been responsible for writing a specific part
of this report:
Part 1: Nathalie Jeandidier, Jean-Pierre Riveline, Nadia
Tubiana-Rufi, Anne Vambergue and Bruno Guerci
Part 2: Vincent Melki and Bruno Guerci
Part 3: Bogdan Catargi
Part 4: Jean-Pierre Riveline and Guillaume Charpentier.
© 2010 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 02/07/2010 par UNIVERSITE PARIS VI P & M CURIE (15258)