Skip to content
View commentaries on this research

This is a plain English summary of an original research article

People with type 1 diabetes offered insulin pumps did not achieve better blood glucose control compared with those using multiple daily injections. Education remains important.

While both groups saw improvements in blood glucose levels and fewer hypoglycaemic episodes (very low blood sugar) over two years, only one in four participants met NICE blood glucose targets. Insulin pump users showed some modest improvements in satisfaction, dietary freedom and daily hassle.

All participants in this NIHR trial attended a training course on managing their insulin levels before randomisation. This is important because previously observed benefits from pump treatment might actually have been a reflection of the training given to them.

Currently just 10% of adults with type 1 diabetes access these training courses. These results support NICE guidelines around the restricted use of insulin pumps and suggest that training improves self-management of the condition. Efforts should therefore be made to encourage training uptake.


Why was this study needed?

People with type 1 diabetes require lifelong treatment with insulin. Doses are adjusted according to food intake, physical activity, and blood glucose level. High or low blood glucose levels lead to serious short and long term complications, however, many people struggle to maintain blood glucose levels within the target range.

Insulin pumps are small devices, about the size of a mobile phone, that deliver a steady flow of insulin. They offer an alternative to multiple daily injections, and may also help people who are unable to achieve blood glucose control. They are expensive though, costing around £2-3,000, with another £1,500 per year in consumables, and have a life span of four to eight years.

About 6% of UK adults with type I diabetes use pumps, which is lower than in many countries. Proponents of pumps suggest more people should be offered them. The aim of this study was to compare the effectiveness of pumps with multiple daily injections.


What did this study do?

The Relative Effectiveness of Pumps Over MDI and Structured Education (REPOSE) randomised controlled trial included 317 adults with type 1 diabetes.

All participants were first placed on a one week training course (Dose Adjustment For Normal Eating: DAFNE) that taught them how to adjust their insulin dose based on estimated carbohydrate intake. They were then randomly allocated to insulin pump or multiple daily injections for two years if their HbA1c values were 7.5% (59mmol/mol) or higher after the education. Blood glucose HbA1c (glycated haemoglobin) is a measure of blood glucose control over several months and was assessed “blind” at a central laboratory. About a quarter either did not attend classes or had HbA1c values lower than this threshold and this left 235 people for the analyses (119 pump users and 116 in the injections group).

Patients knew which group they were in, as pumps and injections are fundamentally different. The study was designed to be large enough to detect a minimal clinically important difference of 0.5% (5.5mmol/mol) in HbA1c values, if one existed. It was run in eight specialist NHS centres and the findings are relevant to usual practice.


What did it find?

  • Blood glucose control improved in all participants, with no statistically significant difference between groups. The mean reduction in HbA1c was 0.85% with pump treatment and 0.42% with injections (odds ratio 1.22, 95% confidence interval 0.62 to 2.39). Only one in four participants achieved the NICE 2004 target for HbA1c of 7.5% (58mmol/mol) or less by two years (25.0% for the pump group and 23.3% for the injections group).
  • There was no difference between groups in the number of severe low blood glucose (hypoglycaemic) episodes. The average number of episodes per patient per year reduced from 0.17 at the beginning of the study to 0.10 at the end of the study.
  • More people in the pump group suffered from diabetic ketoacidosis, a potentially life-threatening complication caused by lack of insulin, most often due to infection (17 events compared to five events in the injection group). Nearly one in five episodes of ketoacidosis were said to be due to insulin pump failure.
  • Various generic and diabetes specific quality of life measures showed little difference between groups. Pump users reported higher scores in some domains, for example, treatment satisfaction, improved dietary freedom and reduced daily hassle.


What does current guidance say on this issue?

NICE guidance from 2008 recommends that insulin pumps should only be considered for adults with type 1 diabetes who have tried injections but have found that either their blood sugar levels have remained high (HbA1c of 8.5% [69mmol/mol] or above), or they suffer from anxiety about unpredictable episodes of hypoglycaemia (called “disabling hypoglycaemia”).

Pump therapy should only be continued if it results in a sustained improvement in blood sugar control, evidenced by a fall in HbA1c levels, or a sustained decrease in the rate of hypoglycaemic episodes. Appropriate targets should be set by the physician, in discussion with the patient, but are typically close to 6.5% (48mmol/mol) for those at low risk of hypoglycaemia.


What are the implications?

Although participants in both groups showed sustained improvement, blood glucose control remained far short of targets currently recommended by NICE. Only 3% of all participants reached an HbA1c of 6.5% (48mmol/mol) or less. People with type 1 diabetes might be better served by ensuring greater availability of high quality, structured self-management training, which is currently only accessed by around 10% of adults in the UK.

Simply increasing use of insulin pumps may increase resource costs without seeing improvements to diabetes control. If after training individuals still find blood glucose control challenging, they could be offered pump treatment in line with NICE guidance to see if this improves quality of life.


Citation and Funding

REPOSE Study Group. Relative effectiveness of insulin pump treatment over multiple daily injections and structured education during flexible intensive insulin treatment for type 1 diabetes: cluster randomised trial (REPOSE). BMJ. 2017;356:j1285.

This project was funded by the National Institute for Health Research UK Health Technology Assessment Programme (project number 08/107/01).



NICE. Continuous subcutaneous insulin infusion for the treatment of diabetes mellitus. TA151. London: National Institute for Health and Care Excellence; 2008.

NICE. Type 1 diabetes in adults: diagnosis and management. NG17. London: National Institute for Health and Care Excellence; 2015.

The Healthcare Quality Improvement Partnership (HQIP). National Diabetes Audit 2012-2013 Report 1: care processes and treatment targets. London: Health and Social Care Information Centre; 2014.

Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre


  • Share via:
  • Print article
Back to top