Post-market Review Products Used in the Management of Diabetes Report to Government Stage 2: Insulin Pumps February 2015

4.2 Introduction

In this section, the aim is to address the differences in clinical outcomes between insulin pump therapy, also known as continuous subcutaneous insulin infusion, and multiple daily injections. Different age groups were analysed to determine if there are sub-groups that would benefit more from insulin pump therapy.

In the 2010 Cochrane Collaboration Report, Misso concluded that there may be benefits to using insulin pumps over multiple daily injections for improving glycaemic control and quality of life for people with type 1 diabetes. However, there was insufficient evidence around diabetes late complications, adverse events, mortality and costs, to make clinical recommendations regarding which form of therapy was superior. The systematic review that informed the NHMRC-approved National Evidence-Based Clinical Care Guidelines for Type 1 Diabetes in Children, Adolescents and Adults (Craig 2011), indicated that the slight, but statistically significant, improvement in observed benefits with insulin pump therapy may have been due to the intensive diabetes management plan, as opposed to pump therapy alone.

To clarify some of this uncertainty, the Department engaged the Quality Use of Medicines and Pharmacy Research Centre (QUMPRC) of the University of South Australia to conduct a literature review to evaluate systematically randomised controlled trial (RCT) evidence on the efficacy and safety of insulin pump therapy versus multiple daily insulin injections, the traditional method of administering insulin. Noting the discrepancy between the trial evidence, from the first literature review, which showed only a small benefit of insulin pump therapy in terms of HbA_1c and no effect on hypoglycaemia, and the stakeholder submissions, which indicated benefits to HbA_1c and quality of life, the Reference Group requested an additional literature review of observational studies. Following stakeholder consultation on the draft Insulin Pumps Report, the literature review of observational studies was expanded to consider women with type 1 diabetes who are pregnant or planning a pregnancy.

Most RCTs excluded patients with diabetes complications, other chronic illnesses, pregnancy or hypoglycaemia unawareness. Baseline HbA_1c in the RCTs was frequently ≥ 8.0% (64 mmol/mol) in studies of children. With some exceptions (e.g. Thomas 2007, DeVries 2002), most RCTs were not restricted to patients with severe hypoglycaemia, hypoglycaemia unawareness or persistent poor control on multiple daily injections. The paucity of data from RCTs regarding outcomes in these subgroups has made it difficult to determine if such patients receive greater benefit from insulin pump therapy than people who were previously well controlled on multiple daily injections.

In contrast, observational studies often involved a wider range of patients with indications for insulin pump therapy including high HbA_1c, wide blood glucose variability, frequent or severe hypoglycaemia, dawn phenomenon (see Glossary), development of complications, pregnant or planning a pregnancy, concurrent chronic illness, needle phobia, and irregular lifestyle or need for lifestyle flexibility.

In current Australian clinical practice, initiation to insulin pump therapy often involves consideration of factors such as the capacity to pay for the pump, technological and cognitive competence to manage the pump, and a high level of patient/carer motivation to intensively monitor and manage diabetes (Victorian CSII Working Party 2009, O’Connell 2008). However, the Australian clinical guidelines for type 1 diabetes indicate that patients likely to benefit from insulin pump therapy include pregnant women (ideally preconception), some children and adolescents, and those with hypoglycaemia unawareness or microvascular complications (Craig 2011).

In general, evidence from RCTs should be given greater weight than observational studies. However, it seems likely that the populations included in observational studies may be more reflective of patients using insulin pump therapy in Australia, than patients involved in RCTs, e.g. those with complications, hypoglycaemia unawareness or who are poorly managed on multiple daily injections. An insulin pump user survey in Australia indicated that 88% of patients initiated pump therapy to better control their diabetes (to reduce hypoglycaemic or hyperglycaemic events), and around 60% were recommended to use a pump by a doctor or diabetes educator (AIHW 2012b).

4.3 First literature review: randomised controlled trials (RCTs)

4.3.1 Methodology

The literature review included all systematic reviews and meta-analyses focussing on insulin pump therapy versus multiple daily injections published between 2000 and July 2012, as well as additional RCTs that had not been included in the systematic reviews.

Position statements and guidelines by professional organisations that made recommendations on the use of insulin pump therapy between 2009 and July 2012, were included in the literature review. This will be discussed further in Part 6 of the report addressing ToR 10.

The literature review did not specifically compare different types of insulin used in the insulin pumps.

The literature review aimed to answer three research questions that are relevant to ToR 8:

1. 
   What is the evidence that continuous subcutaneous insulin infusion therapy is more effective than multiple insulin injections in children and adolescents under 18 years old?
   
2. 
   What is the evidence that continuous subcutaneous insulin infusion therapy is more effective than multiple insulin injections in adults?
   
3. 
   What is the evidence that continuous subcutaneous insulin infusion therapy is safer than multiple insulin injections?
   

4.3.2 Results

Five systematic reviews were included in the literature review, all of which examined the efficacy and safety of insulin pump therapy compared to multiple daily injections. One systematic review restricted the analysis to RCTs that had used short-acting insulin analogues in both insulin pump therapy and multiple daily injections groups (Monami 2010), and one review restricted the analysis to RCTs that had used short-acting insulin analogues in the insulin pump therapy groups (Yeh 2012).

Overall, 31 RCTs were evaluated as part of the literature review, including 30 RCTs published between 1982 and 2009, which were included in the five systematic reviews, and one trial published afterwards. Nineteen RCTs included a total of 846 adults, 11 RCTs included a total of 361 children, and one RCT included a total of 23 adults and children. Most trials had a small sample size. There were 15 trials with 1-25 participants, 11 trials with 25-50 participants, four trials with 50-100 participants, and one trial with >100 participants.

Of the 24 RCTs for which the source of funding was reported, 19 were funded at least partly by insulin pump manufacturers who provided material or financial support to patients. The studies were generally assessed as unclear or low risk of bias, except for participant and personnel blinding, which was not possible because of the nature of the intervention.

4.3.2.1 What is the evidence that continuous subcutaneous insulin infusion therapy is more effective than multiple insulin injections in children and adolescents under 18 years old?

HbA_1c

Four meta-analyses assessed HbA_1c in children and adolescents less than 18 years of age. Overall, the observed benefit showed that there was modest improvement in HbA_1c levels for those on insulin pump therapy, but the difference did not reach the 0.5% reduction that is usually considered a clinically significant change in HbA_1c. However, it may be considered important from a public health perspective (Clar 2010). One meta-analysis of seven RCTs published between 1982 and 2008, found a statistically significant decrease in HbA_1c favouring insulin pump therapy of -0.22% (95% CI 0.41% to –0.03%, P = 0.021) (Misso 2010).

Two meta-analyses found a non-significant decrease in HbA_1c favouring insulin pump therapy of -0.20% (95% CI -0.43% to 0.03%) based on eight RCTs published between 2003 and 2007 (Fatourechi 2009), and of -0.17% (95% CI -0.47% to 0.14%) based on seven RCTs published between 2003 and 2008 examining insulin analogues in insulin pump therapy (Yeh 2012).

A fourth meta-analysis found a non-significant decrease in HbA_1c favouring insulin pump therapy in children with an average age of ≤10 years of -0.1% (95% CI -0.5% to 0.3%, P = 0.48) based on 4 RCTs published between 2003 and 2007 (Monami 2010).

Quality of life and treatment satisfaction

Overall, reported quality of life was better with insulin pump therapy compared to multiple daily injections. Five trials assessed quality of life using four different assessment tools including three that were diabetes specific and one designed for measuring health-related quality of life in children and adolescents. Reported superior quality of life in insulin pump therapy over multiple daily injections was consistent; however, the difference only reached statistical significance in two of the scales/scores out of the five trials (the satisfaction scale in Cohen 2003 and the change in worry score in Opipari-Arrigan 2007). All trials had small sample sizes of fewer than 40 participants.

Three RCTs assessed treatment satisfaction using the Diabetes Treatment Satisfaction Questionnaire (Bradley 1994). All found a statistically significant and clinically relevant improvement with insulin pump therapy compared to multiple daily injections, defined as a ≥ fivepoint difference in score. In another RCT, participants assigned to insulin pump therapy reported significantly greater convenience, ease and flexibility of treatment, leading to increased treatment satisfaction with insulin pump therapy compared to multiple daily injections.

Patients’ preferences

Of the assessed patients’ preferences in three trials published between 2003 and 2008, the majority of patients and their families preferred insulin pump therapy to multiple daily injections.

Long-term health outcomes

None of the RCTs assessed long-term health outcomes.

4.3.2.2 What is the evidence that continuous subcutaneous insulin infusion therapy is more effective than multiple insulin injections in adults?

HbA_1c

Overall, similar to the results for children and adolescents, there was a modest observed improvement in HbA_1c levels with insulin pump therapy compared to multiple daily injections. The difference did not reach the 0.5% reduction usually considered a clinically significant change in HbA_1c levels, but may still be considered important from a public health perspective (Clar 2010).

Three meta-analyses reported a statistically significant difference in HbA_1c favouring insulin pump therapy. The Cochrane Review considered the results of 12 RCTs published between 1982 and 2005 and reported a statistically significant difference in HbA_1c of -0.29% (95% CI -0.52% to –0.06%, P=0.0013) (Misso 2010). The results of a meta-analysis based on five RCTs published between 2003 and 2007 also reported a statistically significant difference in HbA_1c of -0.19% (95% CI -0.27% to -0.11%) (Fatourechi 2009).

A meta-analysis of four RCTs involving insulin analogues in the insulin pump therapy group published between 2003 and 2012, reported a statistically significant difference in HbA_1c change from baseline levels favouring insulin pump therapy of -0.30% (95% CI -0.58% to -0.02%, P = 0.038) (Yeh 2012). However, this result was heavily influenced by the results of a single RCT that found a large decrease in HbA_1c, which may be explained by the high baseline HbA_1c level in the cohort of 9.3% (DeVries 2002). High baseline HbA_1c levels may result in a greater reduction with intervention than lower initial HbA_1c levels (Retnakaran 2005).

Quality of life and treatment satisfaction

Six RCTs compared quality of life between insulin pump therapy and multiple daily injections in adults. Overall, quality of life reported was better with insulin pump therapy.

One RCT assessed quality of life with the Medical Outcomes Study 12-Item Short-Form Survey (SF-12) and one RCT used the Medical Outcomes Study 36-Item Short-Form Survey (SF-36). Both studies found significant improvements in the mental health and general health subscales with insulin pump therapy compared to multiple daily injections. One RCT did not provide detailed data. In the second RCT, there were clinically relevant improvements (defined as ≥ a five-point difference) in both the SF-36 general health and mental health subscales with insulin pump therapy compared to multiple daily injections (DeVries 2002).

Three RCTs assessed quality of life with the Diabetes Quality-of-Life questionnaire (The Diabetes Control and Complications Trial Research Group 1988). A significant difference was evident only in the largest trial involving 272 participants (Hoogma 2006). In the two other RCTs, quality of life was often found to be better with insulin pump therapy compared to multiple daily injections, but the differences were not statistically significant.

Three RCTs assessed treatment satisfaction with the Diabetes Treatment Satisfaction Questionnaire (Bradley 1994). Reported treatment satisfaction was significantly greater with insulin pump therapy compared to multiple daily injections in two RCTs and the difference was clinically relevant in at least one RCT (Bruttomesso 2008).

Long-term health outcomes

None of the RCTs assessed long-term health outcomes.

4.3.2.3 What is the evidence that continuous subcutaneous insulin infusion therapy is safer than multiple insulin injections?

The main outcome considered for this question was the occurrence of hypoglycaemia. Four systematic reviews examined the occurrence of hypoglycaemia in insulin pump therapy compared with multiple daily injections. Overall, there was no statistically significant difference in the occurrence of mild, nocturnal and symptomatic hypoglycaemia, or severe hypoglycaemia, between the insulin pump therapy and multiple daily injections groups.

In RCTs, mild hypoglycaemia was often defined as a blood glucose level lower than 60 or 70 mg/dl. Severe hypoglycaemia was most often defined as requiring assistance from another person for recovery or resulting in coma or seizure.

Mild, symptomatic, and nocturnal hypoglycaemia

Three reviews examined mild hypoglycaemia.

One meta-analysis did not find a difference between insulin pump therapy and multiple daily injections in the mean numbers of mild hypoglycaemic episodes in five crossover RCTs published between 2003 and 2008 (Fatourechi 2009). A small significant difference favouring multiple daily injections was found in three parallel RCTs published between 2004 and 2007. This difference may be explained by the differences in study populations and hypoglycaemia data recording between trials.

One systematic review did not perform a meta-analysis because studies used different scales to report hypoglycaemia (Misso 2010). For mild hypoglycaemia, the authors concluded that the data of 17 RCTs published between 1982 and 2008 suggested there was “no relevant benefit” of one intervention over the other.

Another meta-analysis of three RCTs using insulin analogues in the insulin pump therapy groups found a higher risk of symptomatic hypoglycaemia with insulin pump therapy in adults. However, this result was influenced by the results of a single RCT that may not be generalisable to a broader population because it only included participants with at least one episode of severe hypoglycaemia in the past six months (Yeh 2012).

One meta-analysis did not find a difference between insulin pump therapy and multiple daily injections in nocturnal hypoglycaemia (Fatourechi 2009).

Severe hypoglycaemia

Four reviews examined severe hypoglycaemia; however, it should be noted that the event rates for severe hypoglycaemia were low in all RCTs and the RCTs were not powered for this outcome. Therefore, any recorded difference is difficult to interpret.

One meta-analysis did not find a difference between insulin pump therapy and multiple daily injections in the number of people who experienced at least one severe hypoglycaemic episode, based on 12 RCTs between 2003 and 2008 (Fatourechi 2009). The number of people who experienced at least one severe hypoglycaemic episode in each group was low, 5% and 9.4% in the insulin pump therapy and multiple daily injections groups, respectively, which affected the ability to determine the extent of difference between the groups.

Another meta-analysis of 11 RCTs published between 2000 and 2008, did not find a significant difference in severe hypoglycaemia between insulin pump therapy and multiple daily injections (Monami 2010).

The third meta-analysis of RCTs published between 2000 and 2009 with insulin analogues in the insulin pump therapy groups, found no significant difference in the rate of severe hypoglycaemia among adults (three RCTs), children less than 18 years of age (five RCTs), and above and below 12 years of age (three RCTs and two RCTs, respectively) (Yeh 2012).

One systematic review did not perform a meta-analysis because studies used different scales to report hypoglycaemia (Misso 2010). For severe hypoglycaemia, the authors concluded that, based on a summary of 15 RCTs, “…insulin pump therapy may be better than multiple daily injections”.