Insulin pumps


  • Disclosures: JP has received consulting or speaker fees from Medtronic, Animas, CeQur and Nilimedix, manufacturers of insulin pumps.

  • Endorsed by the International Conference on ATTD organized by Kenes International.

John Pickup,
King’s College London School of Medicine, Guy’s Hospital, London, UK
Tel.: +44(20)78486024


The last year has seen a continued uptake of insulin pump therapy in most countries. The USA is still a leader in pump use, with probably some 40% of type 1 diabetic patients on continuous subcutaneous insulin infusion (CSII), but the large variation in usage within Europe remains, with relatively high use (> 15%) in, for example, Norway, Austria, Germany and Sweden and low use (< 5%) in Spain, the UK, Finland and Portugal. There is much speculation on the factors responsible for this variation, and the possibilities include physician attitudes to CSII and knowledge about its benefits and indications for its use (and inappropriate beliefs about dangers), the availability of reimbursement from insurance companies or funding from national health services, the availability of sufficient diabetes nurse educators and dietitians trained in pump procedures, and clear referral pathways for the pump candidate from general practitioner or general hospital to specialist pump centre. There are now several comprehensive national guidelines on CSII use (see ATTD Yearbook 2009) but more work needs to be done in unifying uptake and ensuring all those who can benefit do so.

Technology developments recently include increasing use of pumps with continuous glucose monitoring (CGM) connectivity (see elsewhere in this volume) and the emergence of numerous manufacturers developing so-called ‘patch pumps’, often for the type 2 diabetes market. Interestingly, the evidence base for CSII in this group is not well established, and for this reason the selected papers on CSII in this section include several in this area. The use of CSII in diabetic pregnancy is a long-established practice, in spite of the lack of evidence that it is superior to multiple daily injections (MDI), and few randomised controlled trials have been done in recent years. Several papers in this field this year continue the debate about the usefulness of CSII in diabetic pregnancy and are reviewed here.

It is pleasing to see more research on the psychosocial aspects of CSII during the year, both from the point of view of how psychological beliefs influence outcomes on CSII (is there a type of patient who does particularly well or poorly on CSII?) and how CSII affects psychological factors like mood, behaviour and quality of life. Quality of life is a difficult topic with doubts that the instruments always capture the aspects of quality of life important to the patient, and there have been conflicting results over the years about whether CSII alters quality of life. Patients in the clinic usually say that it does, and more evidence for quality of life improvement in pump therapy is reviewed here.

CSII and type 2 diabetes

Insulin pump therapy in patients with type 2 diabetes. Safely improved glycaemic control using a simple insulin dosing regimen

S. V. Edelman,1B. W. Bode,2T. S. Bailey,3M. S. Kipnes,4R. Brunelle,5X. Chen,6J. P. Frias6

1University of California, San Diego, CA, USA,2Atlanta Diabetes Associates, Atlanta, GA, USA,3North County Endocrine, Escondido, CA, USA,4Cetero Research, San Antonio, TX, USA,5Bowsher Brunelle Smith LLC, Carmel, IN, USA, and6Lifescan Inc., Milpitas, CA, USA

Diabet Technol Ther 2010; 12: 627–33

Background: The purpose of this study was to investigate the insulin dosing patterns that are necessary to achieve good control in type 2 diabetic patients treated by both insulin injections and oral antidiabetic agents (OADs) who are switched to CSII.

Methods: Type 2 diabetic subjects (= 56) with suboptimal glycaemic control and treated by OADs, basal insulin ± OADs or MDI ± OADs were switched to CSII for 16 weeks and control was optimised to achieve the best glycaemic levels.

Results: The insulin dose increased throughout the period of pump treatment in all three groups, with the mean basal:bolus ratio being 55:45 at completion. Most subjects (88%) were controlled on two or fewer basal rates. HbA1c declined in all three groups (mean −1.2%, p < 0.001), with those with an HbA1c > 8.5% having the greatest improvement (mean –2.1%, p < 0.001). There was no severe hypoglycaemia, and the mean weight increase at week 16 was 1.9 kg (p < 0.001). Treatment satisfaction improved on pump therapy.

Conclusions: CSII using a simple one or two basal rate regimen significantly improves glycaemic control in poorly controlled type 2 diabetic patients whether treated by OADs, MDI or basal insulin only.

Using insulin pump therapy in poorly controlled type 2 diabetes

P. Wolff-McDonagh, J. Kauffman, S. Foreman, S. Wisotsky, J. A. Wisotsky, C. Wexler C

Graduate School of Nursing, Robert Morris University, Moon Township, PA, USA

Diabetes Educator 2010; 36: 657–65

Background: Many type 2 diabetic patients remain poorly controlled on injection therapy but reimbursement for CSII usage is restricted in this group by many healthcare providers. This study was a retrospective analysis of the effect of CSII in poorly controlled type 2 diabetes in the setting of a suburban endocrine practice.

Methods: Data were examined from 15 type 2 diabetic subjects with HbA1c >8% on MDI with or without OADs, who were switched to CSII and followed up over 1 year.

Results: Although the HbA1c increased in one subject on pump therapy, for the remaining 14 people the HbA1c decreased significantly (from a mean of about 9.3% to 8.3%). The body mass index increased significantly on changing to CSII, reaching a plateau at 6 months. The basal insulin dose decreased over the course of pump treatment. Calculating cost of therapy based on pump or injection supplies and insulin use alone over 4 years (i.e. the lifetime of a pump, and not taking into account the effect of lowered HbA1c on outcomes), CSII was more expensive than MDI by about $5000 when daily insulin usage was < 100 units, but it was less expensive by about $600 when insulin usage was 100–150 units/day and less expensive by about $12,000 when usage was > 150 units/day.

Conclusions: CSII achieves improved glycaemic control in most type 2 diabetic patients poorly controlled on MDI. It is likely to be cost effective in those with high insulin requirements (> 100 units/day).

Insulin pump use in type 2 diabetes

B. W. Bode

Atlanta Diabetes Associates, Atlanta, GA, USA

Diab Technol Ther 2010; 12 (Suppl 1): S17–21

This review discusses a number of topics relating to pump use in type 2 diabetes including the limited number of randomised controlled trials comparing CSII and MDI in type 2 diabetes, the evidence that simpler insulin dosing regimens can be used with type 2 diabetes during CSII, such as a fixed basal rate, the use of U-500 insulin and the development of simpler patch pumps that might be more suitable in this group. The review concludes that the evidence base for the use of CSII in type 2 diabetes is not consistent, and suggests randomised controlled trials comparing CSII vs. MDI after the latter therapy has failed, and more studies on the cost effectiveness of CSII in type 2 diabetes.

  • Comment: The use of CSII in type 2 diabetes is likely to be of increasing interest in the coming years. Although the evidence for the superiority of pumps over MDI is inconsistent as judged by randomised controlled trials and therefore funding by healthcare providers is very limited, there are an accumulating number of uncontrolled studies that show that HbA1c is often reduced when poorly controlled, usually insulin-resistant, patients on MDI are switched to CSII (as shown by Wolff-McDonagh et al.). As with type 1 diabetes, CSII seems to work best in those worst controlled on MDI, as indicated by Edelman et al.

points out that pump manufacturers are developing smaller and potentially simpler patch pumps that may be more usable in type 2 diabetes. It seems that one fixed basal rate may be sufficient in many patients. As with type 1 diabetes, the issue of cost effectiveness will be important in ensuring the routine use of CSII in type 2 diabetes, and little work has been done in this area. Wolff-McDonagh et al. show that, because the basal insulin dosage is less on pumps than MDI, the costs savings are high for those who are initially insulin resistant on injection therapy. Given the large number of very poorly controlled type 2 diabetic people, an improvement in control in only a small proportion would result in a huge increase in CSII uptake.

CSII in pregnancy

Continuous subcutaneous insulin infusion versus multiple daily injections in pregnant women with type 1 diabetes

S. González-Romero,1,2I. González-Molero,1M. Fernández-Abellán,3M. E. Domínguez-López,1S. Ruis-de-Adana,1,2G. Olveira,1,2F. Soriguer1,2

1Endocrinology and Nutrition Department, Hospital Regional Universitario Carlos Haya, Málaga, Spain,2CIBER de Diabetes y Enfermedades Metabólicas Asociadas, Barcelona, Spain, and3Obstetrics and Gynecology Department, Universitario Carlos Haya, Málaga, Spain

Diab Technol Ther 2010; 12: 263–9

Background: The aim of this study was to compare glycaemic control and obstetric and perinatal outcomes in pregnant women with type 1 diabetes who were treated by CSII compared to a control group managed by MDI.

Methods: A case–control study of 35 pregnancies treated by CSII and 64 pregnancies treated by MDI was performed; women had similar age, diabetes duration and microvascular status.

Results: Pregnancy planning was commoner in the CSII group who had a lower HbA1c at the start of pregnancy (6.6% vs. 7.6%, p < 0.05), but control during pregnancy was similar. The frequency of hypoglycaemia and ketoacidosis was not significantly different during MDI- and CSII-managed pregnancy. No differences in diabetes, obstetric or perinatal outcomes were seen, including retinopathy progression, pre-eclampsia, spontaneous miscarriage, congenital abnormalities, birth weight, neonatal hypoglycaemia and stillbirth. Although gross macrosomia was uncommon, many of the newborns were large for gestational age in both groups.

Conclusions: The authors conclude that CSII is safe and effective during pregnancy in type 1 diabetes but there is no evidence for improved outcomes.

Glycaemic control and selected pregnancy outcomes in type 1 diabetes women on continuous subcutaneous insulin infusion and multiple daily injections: the significance of pregnancy planning

K. Cyganek,1,2A. Hebda-Szydlo,1,2B. Katra,1,2J. Skupien,3T. Klupa,1,2I. Janas,1I. Kaim,2,4J. Sieradzki,1,2A. Reron,2,4M. T. Malecki1,2

1Department of Metabolic Diseases, Jagiellonian University Medical College, Krakow, Poland,2University Hospital, Krakow, Poland,3Section on Genetics and Epidemiology, Joslin Diabetes Center, Boston, MA, USA, and4Department of Obstetrics and Perinatology, Jagiellonian University Medical College, Krakow, Poland

Diab Technol Ther 2010; 12: 41–7

Background: The aim of this study was to assess the safety and efficacy of CSII compared to MDI in an observational study of women with type 1 diabetes.

Methods: Data were analysed from 116 planned and 153 unplanned pregnancies – 157 women were treated by MDI, 42 by CSII and 70 switched from MDI to CSII during the first trimester. Groups were of similar age, but CSII subjects had longer diabetes duration.

Results: Those who planned pregnancy had a lower HbA1c in the first trimester, whether or not they were receiving CSII or MDI. Treatment by CSII or MDI resulted in similar HbA1c during pregnancy, but CSII-treated women gained more weight. There were no differences in birth weight, malformations, stillbirths or other outcomes.

Conclusions: Both CSII and MDI can provide excellent control during pregnancy. Planning is beneficial independent of therapy.

Insulin pumps and their use in pregnancy

A. D. Wollitzer, H. Zisser, L. Jovanovič

Sansum Diabetes Research Institute, Santa Barbara, CA, USA

Diab Technol Ther 2010; 12 (Suppl 1): S33–6

This review again points out the lack of trial evidence for improved outcomes when CSII is used in diabetic pregnancy, but then moves to a useful discussion of the potential advantages and disadvantages of CSII during pregnancy in diabetic women. Much of this is extrapolated from non-pregnant people with type 1 diabetes and may or may not apply in pregnancy, but there is the possibility of less hypoglycaemia, less glycaemic variability, better adherence and treatment satisfaction. And on the downside, the learning curve required for the patient to become ‘pump-educated’ may be long compared to the short 20–25 weeks of pump use in many pregnancies. Another potential disadvantage that is discussed is the risk of ketoacidosis if pump insulin delivery is interrupted – especially dangerous in pregnancy.

  • Comment: Insulin pump therapy in diabetic pregnancy continues to be a poorly investigated subject. There are still few randomised controlled trials of CSII vs. MDI during pregnancy. The above studies are useful in confirming what has been known for several decades, that average glycaemic control and pregnancy and fetal outcomes are similar during CSII and MDI, and CSII may be safely used in pregnancy. What we do not know (for formal trial evidence) is whether there are subsets of patients who would gain particularly by being switched to CSII during pregnancy. It would be useful, for example, to have a randomised controlled trial of those who have failed to achieve satisfactory control on MDI during the first trimester because of elevated HbA1c or frequent hypoglycaemia on MDI who are then allocated to either CSII or continued MDI. Many healthcare professionals have anecdotal observations from their own clinic that there are women who have been very troubled by, for example, hypoglycaemia during pregnancy who have improved substantially by a switch to CSII. It is curious that the trial evidence to confirm or refute this indication for CSII continues to be weak.

Psychological aspects of CSII

Psychological aspects in continuous subcutaneous insulin infusion: a retrospective study

I. Aberle,1U. Scholz,2B. Back-Kliegel,3C. Fischer,4M. Gorny,5K. Langer,5M. Kliegel1

1Department of Psychology, Technische Universität, Dresden, Germany,2Department of Psychology, University of Zurich, Switzerland,3Department of Metabolism Center, Hirslanden Clinic, Zurich, Switzerland,4School Counseling Center, Schwäbisch Hall, Germany, and5Department of Internal Medicine, City Hospital Darmstadt, Germany

J Psychol 2009; 143: 147–60

Background: The purpose of this study was to examine in type 1 diabetic subjects treated by CSII the influence of psychological factors such as coping strategy and locus of control on both metabolic control and psychological outcomes such as treatment satisfaction, depressive symptoms and quality of life.

Methods: Psychological outcomes and HbA1c were examined in type 1 diabetic patients (= 51) on CSII at a diabetes centre in Germany.

Results: HbA1c was significantly lower on CSII than intensified conventional therapy (7.1% vs. 8.2%, p < 0.001). High HbA1c on CSII correlated with high external locus of control (i.e. the individuals had the general belief that they had little control over their lives and life events such as disease were dependent on external factors, luck and coincidence). High HbA1c was also correlated with more depressive symptoms. Treatment satisfaction and quality of life were associated with high self-efficacy (belief that one is competent to tackle difficult or novel tasks and cope with adversity in demanding situations).

Conclusions: Psychological factors are important in the outcomes of CSII. Patients who feel more responsible for the process of CSII have better control and those with more self-efficacy have a better quality of life and treatment satisfaction.

Investigation of quality of life and family burden issues during insulin pump therapy in children with type 1 diabetes mellitus – a large-scale multicentre pilot study

E. Müller-Godeffroy, S. Treichel, V. M. Wagner on behalf of the German Working Group for Paediatric Pump Therapy

Department of Paediatrics, Division of Paediatric Endocrinology and Diabetology, University of Luebeck, Luebeck, Germany

Diab Med 2009; 26: 493–501

Background: Although it is assumed that CSII provides psychological and psychosocial as well as metabolic benefits, quality of life effects, for example, have given conflicting results to date. The aim of this study was to investigate changes in psychological factors and the family burden of diabetes in children switched from MDI to CSII.

Methods: In a multicentre study, metabolic and psychosocial data were collected from 117 children before and after transfer from MDI to CSII.

Results: Although general quality of life did not improve, parent-reported or self-reported diabetes-specific quality of life increased significantly, particularly in school-aged and younger children. Parents reported less frequent parenting stress, hypoglycaemia worry, and overall diabetes burden with regard to themselves and the child. Family conflict was not altered. Parents of younger children reported fewer problems with nutrition management.

Conclusions: CSII has significant psychosocial benefits for both the child with diabetes and the family.

Improvements in cognition, mood and behaviour following commencement of continuous subcutaneous insulin infusion therapy in children with type 1 diabetes mellitus: a pilot survey

S. Knight,1,2E. Northam,1–3S. Donath,1A. Gardner,4N. Harkin,4C. Taplin,4P. Joy,4F. J. Cameron,1,3G. R. Ambler4,5

1Murdoch Children’s Research Institute, Melbourne, Vic, Australia,2School of Behavioural Science, University of Melbourne, Vic, Australia,3Royal Children’s Hospital, Melbourne, Vic, Australia,4Institute of Endocrinology and Diabetes, Children’s Hospital at Westmead, Westmead, NSW, Australia, and5Discipline of Paediatrics and Child Health, University of Sydney, NSW, Australia

Diabetologia 2009; 52: 193–9

Background: Children with type 1 diabetes are at risk of cognitive and behavioural difficulties that may be influenced by glycaemic fluctuations and long-term metabolic control. The aim of this study was to investigate whether improved glycaemic control with CSII was associated with improvements in cognition, mood and behaviour in children.

Methods: Children (= 32) with type 1 diabetes at two centres in Australia underwent behavioural and cognitive assessments before and 18 weeks after starting CSII. A 72-h period of CGM was used to assess glycaemic variability and hyperglycaemia and hypoglycaemia.

Results: HbA1c improved from 8.2% to 7.5% after commencing CSII. Glycaemic variability was also reduced. There were significant improvements in a number of measures of cognition such as perceptual reasoning, selective attention and working memory. From parents, teachers and self-reports, there were fewer mood-related symptoms and parents reported fewer behavioural problems. None of the changes in mood, behaviour or cognition correlated with improved HbA1c.

Conclusions: The improvement in HbA1c afforded by CSII in children with diabetes is accompanied by improvements in mood, behaviour and cognition.

  • Comment: These interesting papers show that psychosocial factors determine (to some extent) the quality of control on CSII, and that the effects of CSII extend beyond reduced HbA1c, hypoglycaemia and microvascular risk to various interrelated psychological and social outcomes such as quality of life and mood.

It is obvious that not all patients do well on CSII and there are several possible reasons why some have an unsatisfactorily high HbA1c. Aberle et al. show that such patients tend to have a high external locus of control, i.e. they usually feel less responsible for diabetes and CSII, that the diabetes is beyond their control and that they do not need to take action. Such studies emphasise that psychological care must be seen as an important part of diabetes education before and during CSII. We may well be able to improve metabolic outcome on CSII by attention to psychological health.

It is striking how often patients in the clinic report improvements in thinking, mood and well-being after the start of CSII. This has been comparatively poorly studied to date, and the results from Knight et al. are important in confirming it. Particularly interesting is the apparent lack of correlation with HbA1c improvements or other glycaemic variables, suggesting that the effect might be either related to other CSII-related changes (insulin, other metabolites) or more probably the result of changes in hypoglycaemia exposure or blood glucose variability not assessed by the short period of CGM.