Systematic review update and economic evaluation for the New Zealand setting

Transcription

1 Systematic review update and economic evaluation for the New Zealand setting February 2008 Subcutaneous insulin pump therapy Suzanne Campbell Arsupol Suebwongpat Lachlan Standfield Adele Weston

2 This report should be referenced as follows: Campbell S, Suebwongpat A, Standfield L, Weston A. Systematic review update and economic evaluation for the New Zealand setting: Subcutaneous insulin pump therapy. HSAC Report 2008; 1(3). Health Services Assessment Collaboration (HSAC), University of Canterbury ISBN (Online) ISSN (Online)

3 i Review Team This review was undertaken by the Health Services Assessment Collaboration (HSAC). HSAC is a collaboration of the Health Sciences Centre of the University of Canterbury, New Zealand and Health Technology Analysts, Sydney, Australia. This report was authored by Dr Suzanne Campbell, Health Outcomes Manager, who developed and undertook the literature search, extracted the data, conducted the critical appraisals, and prepared the clinical component of the report. A preliminary economic evaluation was undertaken by Arsupol Suebwongpat, Health Economist and Lachlan Standfield, Senior Health Economist. Adele Weston assisted with the interpretation of the clinical and economic components of the report. Acknowledgements Dr Ray Kirk and Dr Adele Weston (as HSAC Directors) reviewed the final draft. Cecilia Tolan (Administrator) provided administrative support. The current review was conducted under the auspices of a contract funded by the New Zealand Ministry of Health. This report was requested by the Long Term Conditions Policy and Strategy team of the Clinical Services Directorate of New Zealand s Ministry of Health. We thank Sandy Dawson and Christine Andrews for their assistance in developing the scope of the review and providing clinical practice and cost information for the review. The systematic review of the evidence will ultimately be used by the Long Term Conditions Policy and Strategy team to inform policy decision making in conjunction with other information. The content of the review alone does not constitute clinical advice or policy recommendations. Copyright Statement & Disclaimer This report is copyright. Apart from any use as permitted under the Copyright Act 1994, no part may be reproduced by any process without written permission from HSAC. Requests and inquiries concerning reproduction and rights should be directed to the Director, Health Services Assessment Collaboration, Health Sciences Centre, University of Canterbury, Private Bag 4800, Christchurch, New Zealand HSAC take great care to ensure the accuracy of the information in this report, but neither HSAC, the University of Canterbury, Health Technology Analysts Pty Ltd nor the Ministry of Health make any representations or warranties in respect of the accuracy or quality of the information, or accept responsibility for the accuracy, correctness, completeness or use of this report. The reader should always consult the original database from which each abstract is derived along with the original articles before making decisions based on a document or abstract. All responsibility for action based on any information in this report rests with the reader.

4 ii This report is not intended to be used as personal health advice. People seeking individual medical advice should contact their physician or health professional. The views expressed in this report are those of HSAC and do not necessarily represent those of the University of Canterbury New Zealand, Health Technology Analysts Pty Ltd, Australia or the Ministry of Health. Contact Details Health Services Assessment Collaboration (HSAC) Health Sciences Centre University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand Tel: Fax: hsac@canterbury.ac.nz Website: healthsac.net (under development)

5 iii Executive summary Introduction The systematic review update was requested by the Long Term Conditions Policy and Strategy team of the Clinical Services Directorate of New Zealand s Ministry of Health. The update was to be based on one of two systematic reviews provided by the requestor: a Canadian review from Agence d'evaluation des technologies et des modes d'intervention en santé (AETMIS) (Côté & St-Hilaire, 2005), or the NHS technology assessment report (Colquitt et al, 2004) which informed the National Institute for Health and Clinical Excellence (NICE) recommendations for the use of continuous subcutaneous insulin infusion (CSII) in diabetes. Although the NHS report is not as current as the AETMIS report, it was chosen as the basis of the systematic review update because it considers the use of CSII in people with type 1 and type 2 diabetes. Methods The aim of this review was to evaluate the relative effectiveness, safety, and costeffectiveness of CSII in patients with insulin-treated diabetes, when compared with optimised multiple daily injections (MDI). The primary research question to be addressed by this review was: Is subcutaneous insulin pump therapy effective, safe, and cost-effective compared with multiple daily injections? A systematic method of literature searching, study selection, data extraction and appraisal was employed in the preparation of this report. The literature was searched using the following bibliographic databases: Medline and Embase. In addition, the bibliographies of included papers were examined for relevant studies. The Cochrane Library was also searched to help identify existing systematic reviews and original studies. Searches were limited to English-language material published from January 2002 to August 2007 inclusive. Studies were included if they were randomised trials of crossover or parallel-group design, which compared CSII therapy with optimised MDI therapy (at least three injections per day) for at least 10 weeks in insulin-treated type 1 or type 2 diabetic patients (including adults, children, adolescents, and pregnant women with preexisting diabetes). The key outcomes examined were: HbA 1c levels, lipid levels, quality of life, severe hypoglycaemic episodes, and ketoacidotic episodes. NHMRC dimensions of evidence, levels of evidence and quality assessment criteria were used to evaluate each of the included studies. Data was extracted onto standardised data extraction forms by one reviewer. A systematic search of the published literature was also conducted to identify any relevant economic evaluations of CSII versus MDI. The literature was searched using the following bibliographic databases: Medline and Embase. Searches were limited to English-language material published through January Relevant publications were briefly reviewed to inform a qualitative discussion of the incremental costs and outcomes likely to be associated with CSII relative to MDI.

6 iv An economic analysis was also undertaken to examine the cost-effectiveness of CSII compared with MDI for a patient with Type 1 diabetes. The economic model was based mainly on the method and approach presented in the NHS assessment report (Colquitt et al, 2004), with modification to inputs in order to populate the model with NZ data wherever available. Cost-effectiveness is presented as the incremental cost per severe hypoglycaemic attack avoided over six years (ie, the pump life-time). The total additional costs of using CSII can also be considered relative to the improvements seen in glycaemic control, and any reduction in diabetic complications that this may confer. Key results The search strategy for clinical evidence identified a total of 726 potentially relevant publications. After consideration of titles and abstracts using the pre-defined study selection criteria, 32 full papers were retrieved and scrutinised in detail for possible inclusion in the review. For excluded studies, the reasons for exclusion were documented. As a result 12 publications, referring to 11 randomised trials, were ultimately included in the review. Three publications (two original reports of clinical trials and one systematic review) compared CSII with MDI in adults with type 1 diabetes, four publications (all original reports referring to three clinical trials) compared CSII with MDI in children and/or adolescents with type 1 diabetes, four publications (all original reports of clinical trials) compared CSII with MDI in adults with type 2 diabetes, and one publication (a systematic review) compared CSII with MDI in pregnant women with diabetes. Efficacy In trials of adults with type 1 diabetes, therapy with CSII was associated with a greater reduction in HbA 1c levels compared with MDI, at all timepoints examined (4, 6, and 8 months); however, although the magnitude of the difference was small. Although the publications reported a statistically significant difference between CSII and MDI in terms of endpoint HbA 1c levels, meta-analysis of baseline adjusted data from the two trials at a common timepoint (four months) favoured CSII but failed to reach statistical significance. Quality of life endpoints were poorly reported but indicate that treatment with CSII may be associated with significantly improved quality of life compared with MDI. The effect of treatment on lipid levels was not reported in either trial. In children and/or adolescents with type 1 diabetes, HbA 1c levels were lower at endpoint after CSII than MDI, although the magnitude of the difference was small. Meta-analysis of baseline adjusted data at a common timepoint (3-4 months) confirmed the trend in favour of CSII but did not reveal a significant difference between treatment modalities in terms of HbA 1c levels at endpoint. Quality of life was assessed in all three trials using a youth-specific questionnaire. Whilst two studies found no statistically significant difference between CSII and MDI, the third reported a significant difference in favour of CSII in the satisfaction subscale (P< 0.05), but not other subscales. The effect of treatment on lipid levels was not reported. In adults with type 2 diabetes, the change from baseline in HbA 1c levels was greater with CSII than MDI. Meta-analysis of data from the four trials (using the longest

7 v timepoint in each trial) confirmed the trend in favour of CSII but the difference was not statistically significant. One of the four trials reported quality of life outcomes and showed no difference between treatment modalities. Of the two studies that reported the effect of treatment on lipid levels, there was no significant difference between CSII and MDI. In pregnant women with diabetes, HbA 1c levels were lower in women treated with MDI compared with CSII, in both second and third trimester, but the difference was not significant. No other outcomes of relevance were reported in the review and therefore no firm conclusions can be drawn with respect to best practice in pregnant women with diabetes. Safety In adults with type 1 diabetes, episodes of severe hypoglycaemia were infrequent in both trials but were consistently lower during treatment with CSII compared with MDI. Ketoacidotic events were also uncommon but occurred more frequently during treatment with CSII than MDI. In children and/or adolescents with type 1 diabetes, all three trials reported that episodes of severe hypoglycaemia were less frequent during treatment with CSII compared with MDI. Ketoacidotic events were uncommon with both treatment modalities and therefore no firm conclusions can be drawn. Overall, in adults with type 2 diabetes, the total number of severe hypoglycaemic episodes across the four trials was higher in patients treated with MDI; however, this was driven by only one trial. Two of the trials reported no cases of severe hypoglycaemia and one trial reported more episodes in during treatment with CSII. There was insufficient evidence to draw any conclusions regarding the effect of treatment on the incidence of ketoacidosis. Economic implications The search strategy for economic evaluations identified a total of four relevant publications published in English. All four studies reported results favouring insulin pump therapy over MDI and concluded that CSII could be a worthwhile investment, however these conclusion were heavily dependent upon the magnitude of clinical benefit assumed. The cost-effectiveness analysis undertaken for this report was conducted as a whatif scenario. It is postulated, however unproven, that CSII may reduce the number of severe hypoglycaemic attacks a patient experiences compared with MDI. If CSII were to reduce the number of hypoglycaemic attacks by 0.5 per patient per year, the incremental cost per severe hypoglycaemic event avoided would be ~NZ$6,000 per hypoglycaemic attack avoided over a 6-year time horizon. In addition, based of the clinical evidence the CSII pump appears to improve a patient s HbA 1c by around 0.37% compared to MDI after approximately four months of therapy. It is unknown, but not inconceivable that this benefit will be sustained in the long-term, bringing potential benefits of reduced disease sequelae in the long-term. These benefits come at an additional cost of approximately NZ$16,000 per patient over six years. Although not quantified in the current analyses, when a realistic magnitude of benefit (eg, a

8 vi 0.37 improvement in HbA 1c ) is applied within the models of others, the cost/qaly is likely to well exceed NZ$100,000. Conclusions The review conclusions are based on the current evidence available from this report s critical appraisal of literature published since the NHS assessment report on the efficacy, safety, and cost-effectiveness of CSII compared with MDI. In general, the findings of the current review support the findings of the original NHS assessment report, that is, an improvement in glycaemic control that is of small magnitude and borderline statistical significance. When compared with optimised MDI, CSII results in a modest but potentially worthwhile improvement in glycosylated haemoglobin levels in all patient groups assessed (ie, adults with type 1 diabetes, children and adolescents with type 1 diabetes, and adults with type 2 diabetes). Due to the short duration of the clinical trials is not possible to evaluate the longer term benefits of such a difference in HbA 1c levels; however, there is an expectation that it would be reflected in a reduction in long-term complications. Although more immediate primary benefits from CSII may be associated with an impact on the incidence of severe hypoglycaemic events and improved quality of life (through greater flexibility of lifestyle), there is limited evidence to support this from the studies identified in this update. However, despite the limited evidence it is postulated that CSII may reduce the number of severe hypoglycaemic attacks a patient experiences compared with MDI. According to the results of the cost-effectiveness analysis conducted herein, it is estimated that if every patient who changed from MDI to CSII therapy was able to avoid one severe hypoglycaemic attack every two years (ie, an improvement of 0.5 events per annum), the incremental cost per severe hypoglycaemic event avoided would be approximately $6,000. The total incremental cost associated with the introduction of CSII compared to MDI for a patient with type 1 diabetes is approximately $16,000 over six years (the approximated life of the pump).

9 vii Table of Contents Review Team...i Acknowledgements...i Copyright Statement & Disclaimer...i Contact Details...ii Executive summary... iii Introduction... iii Methods... iii Key results...iv Efficacy...iv Safety...v Economic implications...v Conclusions...vi Table of Contents...vii List of Tables...ix List of Figures...xi List of Figures...xi List of Abbreviations and Acronyms...xii Introduction...1 Objective...1 Insulin-dependent diabetes...1 Continuous subcutaneous insulin infusion...2 Structure of report...2 Methods...3 Research questions...4 Literature search...6 Bibliographic databases...6 Review databases...6 HTA Groups...6 Clinical Practice Guidelines...6 Assessment of study eligibility...7 Appraisal of included studies...8 Dimensions of evidence...8 Data extraction...11 Data synthesis...11 Limitations of the review methodology...12 Evaluation of economic implications...14

10 viii Results...17 Overview...17 Clinical evidence: study characteristics and outcomes...17 Adults with type 1 diabetes...17 Children and/or adolescents with type 1 diabetes...20 Adults with type 2 diabetes...22 Pregnant women with diabetes...25 Clinical evidence: results and meta-analyses...27 Glycosylated haemoglobin...27 Lipid levels...36 Quality of life...36 Severe hypoglycaemia...37 Ketoacidosis...37 Economic considerations...39 Results of the literature search...39 Background to the economic evaluation...40 Costs incorporated in the model...40 Health benefits incorporated in the economic model...43 Cost-effectiveness...44 Sensitivity analyses...44 Limitations of economic evaluation...46 Summary and conclusions...47 Summary of evidence for evidence review...47 Adults with type 1 diabetes...47 Children and/or adolescents with type 1 diabetes...47 Adults with type 2 diabetes...47 Pregnant women with diabetes...48 Summary of results from economic evaluation...48 Limitations of evidence base...48 Conclusions...49 References...50 Glossary...53 Appendix A: Included Studies...55 Appendix B: Excluded Studies Annotated by Reason for Exclusion...57 Appendix C: Data extraction tables...97 Studies of adults with type 1 diabetes...97 Studies of children and/or adolescents with type 1 diabetes Studies of adults with type 2 diabetes Studies of pregnant women with pre-existing diabetes Appendix D: Forest plots...127

11 ix List of Tables Table 1: Recent systematic reviews, candidates for updating...3 Table 2: Criteria for determining study eligibility...5 Table 3: Nature of the evidence...5 Table 4: Search strategy for relevant clinical trials...7 Table 5: Application of selection criteria to citations...8 Table 6: NHMRC Dimensions of evidence...9 Table 7: Table 8: NHMRC Interim Levels of Evidence (NHMRC 2005) for Evaluating Intervention Studies...10 Quality criteria for different levels of evidence...11 Table 9: Reporting biases in systematic reviews a...13 Table 10: Search strategy for economic evaluations...15 Table 11: Summary of evidence for adults with type 1 diabetes...19 Table 12: Summary of evidence for children and/or adolescents with type 1 diabetes...21 Table 13: Summary of evidence for adults with type 2 diabetes...23 Table 14: Table 15: Table 16: Table 17: Table 18: Table 19: Table 20: Table 21: Table 22: Table 23: Table 24: Summary of evidence for pregnant women with diabetes...26 Detailed data extraction from Level II evidence glycosylated haemoglobin at endpoint...29 Detailed data extraction from Level II evidence change from baseline to endpoint in glycosylated haemoglobin...31 Glycosylated haemoglobin at endpoint: Studies comparing the same insulin type with CSII and MDI a...33 Glycosylated haemoglobin at endpoint: All included studies, regardless of type of insulin a...34 Change in glycosylated haemoglobin: Studies comparing the same insulin type with CSII and MDI a...35 Change in glycosylated haemoglobin: All included studies, regardless of type of insulin a...36 Data extraction from Level II evidence episodes of severe hypoglycaemia and ketoacidosis...38 Existing economic evaluations: methods...39 Existing economic evaluations: results...40 Costs of insulin pump and consumables...41 Table 25: Cost of a patient education programme for CSII...41 Table 26: Cost of additional outpatient management...42

12 x Table 27: Reduction of insulin use with continuous subcutaneous insulin infusion...42 Table 28: Cost of managing a severe hypoglycaemic event...42 Table 29: Table 30: Table 31: Incremental costs associated with the introduction of CSII over year Incremental costs associated with the introduction of CSII over 6-years...43 Incremental cost per severe hypoglycaemic event avoided per patient per annum...44

13 xi List of Figures Figure 1: Figure 2: Figure 3: Figure 4: Figure 5: Figure 6: Figure 7: Figure 8: Figure 9: Figure 10: Figure 11: Figure 12: Figure 13: Incremental cost per severe hypoglycaemic attack avoided over 6 years...44 ICER following changes in the pump price and cost of consumables...45 ICER following removal of the requirement for a hospital admission for pump training...45 Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 1 diabetes: Base case unadjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in children and adolescents with type 1 diabetes: Base case unadjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 2 diabetes: Base case unadjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in children and adolescents with type 1 diabetes: All studies, unadjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 2 diabetes: All studies, unadjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 1 diabetes: Base case adjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in children and adolescents with type 1 diabetes: Base case adjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 2 diabetes: Base case adjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in children and adolescents with type 1 diabetes: All studies, adjusted data - WMD (random effects model) Meta-analysis of the effect of CSII vs MDI on glycosylated haemoglobin in adults with type 2 diabetes: All studies, adjusted data - WMD (random effects model)...131

14 xii List of Abbreviations and Acronyms AE BMI CI CIT CSII DQoL QOLCTQ EOT HbA 1c HDL ICER ITT LDL MDI NHMRC NHS NICE NPH nr ns QOL SD SF-12 SF-36 WMD adverse event body mass index confidence interval conventional insulin therapy continuous subcutaneous insulin infusion Diabetes Quality of Life Diabetes Quality of Life Clinical Trial Questionnaire end of treatment glycosylated haemoglobin A 1c high density lipoprotein incremental cost-effectiveness ratio intention-to-treat low density lipoprotein multiple daily injections National Health and Medical Research Council National Health Service National Institute for Health and Clinical Excellence neutral protamine Hagedorn not reported not significant quality of life standard deviation 12-item Short Form Health Survey 36-item Short Form Health Survey weighted mean difference

15 1 Introduction Objective The purpose of this systematic review update is to provide a summary of the recent evidence pertaining to the relative effectiveness, safety, and cost-effectiveness of CSII in patients with insulin-treated diabetes, when compared to optimised MDI. The systematic review update is inclusive of evidence relating to type 1 and type 2 diabetes. The cost-effectiveness analysis considers individuals with type 1 diabetes only. Insulin-dependent diabetes Diabetes is a chronic metabolic disorder characterised by hyperglycaemia due to impaired insulin secretion and/or impaired insulin action. There are two major types of diabetes: type 1 and type 2. People with type 1 diabetes generally require daily insulin injections to survive because the islet b cells in the pancreas (which produce insulin) have been destroyed through an autoimmune mechanism. Type 2 diabetes results from failure of insulin production to overcome reduced tissue sensitivity to insulin (known as insulin resistance). In the early stages, type 2 diabetes can be managed through diet and exercise alone. However, it is a progressive disease and nearly all individuals with type 2 diabetes will require oral glucose lowering drugs over time and most will eventually need insulin to maintain satisfactory blood glucose levels. An impaired insulin effect results in increased levels of glucose in the blood (hyperglycaemia), which can cause microvascular and macrovascular damage if prolonged. Common complications of diabetes include visual impairment, kidney failure, angina, myocardial infarction, stroke, foot ulceration, and erectile dysfunction. The principal goals of treatment for diabetes are therefore to prevent the acute and late complications of diabetes and thus improve quality of life and avoid premature diabetes-associated death. These goals may be achieved through better glycaemic control of blood glucose levels and through reductions in other macrovascular risk factors. There are four main types of insulin preparation, which are classified on the basis of their duration of action, onset of action, and peak action time: rapid-acting insulin analogues (eg, lispro, insulin aspart): onset min, peak min, duration 3-5 hr; used premeal short-acting insulins, also referred to as human soluble insulin or Regular insulin: onset min, peak 2-3 hr, duration 4-6 hr; used premeal intermediate-acting insulins (eg, neutral protamine Hagedorn (NPH) insulin): onset 2-4 hr, peak 4-8 hr, duration hr; used to control blood glucose between meals and during the night long-acting insulin analogues, also referred to as basal insulins (eg, glargine): onset 1-2 hr, peak none, duration 24 hr; given in the morning or at bedtime to control blood glucose all day

16 2 Conventional insulin therapy involved twice-daily injections of a mixture of slowacting and short-acting insulins (ie, meal-time and a basal insulin together). However, intensive therapy by multiple daily injection (MDI) of slow-acting and short-acting insulins (ie, 4-5 injections daily of meal-time insulin before the main meals and 1-2 injections of basal insulin) is now considered superior. The timing and frequency of insulin injections depend upon a number of factors, including the type of insulin, the amount and type of food eaten, the person's level of physical activity, experience of hyperglycaemia and hypoglycaemia, and the person s preference and appropriateness to his/her lifestyle. Living with diabetes requires the diligent monitoring of blood glucose levels, constant attention to diet and, in individuals receiving insulin therapy, multiple daily injections. Insulin and glucose intake have to be carefully balanced to avoid hyperglycaemia or episodes of hypoglycaemia, which can cause weakness, confusion, dizziness, unconsciousness, and seizures. The ongoing threat of short-term and long-term complications of diabetes can cause anxiety, fear and depression. Furthermore, the inflexibility of lifestyle often has a negative impact on quality of life. In adolescence, lifestyle and attitudes can make maintenance of tight blood glucose control more difficult. In younger children, good control of blood glucose is difficult to attain, partly because intermediate-acting insulins are absorbed too rapidly. Continuous subcutaneous insulin infusion Insulin pump therapy is technically referred to as continuous subcutaneous insulin infusion (CSII). CSII devices are external pumps comprising a programmable pump and insulin storage reservoir to which the patient is continuously connected. Insulin is administered to the patient via a needle or cannula inserted under the skin. The pump delivers insulin continuously at a constant or variable basal rate with an additional boost dose delivered at meal times. The objectives of CSII therapy include improved glycaemic control through lower mean blood glucose levels, less variation in blood glucose levels, and fewer episodes of hypoglycaemia. CSII also allows a greater degree of flexibility in lifestyle (activity and meal-planning). Structure of report This report is divided into sections. The next section describes the review s Methods and includes the search strategy, inclusion and exclusion criteria, the data extraction, appraisal and synthesis methods, and the methodological limitations of the evidence review. The results section considers the included appraised studies, and includes a meta-analysis of relevant outcomes where appropriate. Study characteristics and findings are reported in separate tables and synthesised in the text. This is followed by a section describing the methods and results of the economics evaluations. The final section summarises results, briefly discusses the limitations of the evidence base, and presents key conclusions. A Glossary and detailed appendices follow, including a list of included studies, all excluded papers annotated by reason for exclusion, the completed data extraction tables for included papers, and forest plots.

17 3 Methods The systematic review update was to be based on one of two systematic reviews provided by the requestor (Table 1). Although the literature search span of the Canadian review (Côté & St-Hilaire, 2005) is considerably more recent, it is not suitable for updating as the requestor has advised because it does not consider the use of CSII in people with type 2 diabetes. Thus, the basis of this update is the NHS (National Health Service) technology assessment report (Colquitt et al, 2004), which informed the National Institute for Health and Clinical Excellence (NICE) recommendations for the use of CSII in diabetes. Table 1: Recent systematic reviews, candidates for updating HTA organisation Country Title Literature search end date Agence d'evaluation des technologies et des modes d'intervention en santé ((AETMIS) Côté & St-Hilaire, 2005) Southampton Health Technology Assessments Centre under auspices of the NHS R&D HTA Programme (Colquitt et al, 2004) Québec, Canada UK Comparison of the insulin pump and multiple daily insulin Clinical and costeffectiveness of continuous subcutaneous insulin infusion for diabetes July 2004 May 2002 NICE recommended the use of insulin pump therapy as one option for people with type 1 diabetes provided that: multiple-dose insulin therapy (including using insulin glargine when it is appropriate) has failed, and the patient is willing and able to use insulin pump therapy effectively NICE considers that multiple-dose insulin therapy has failed when someone has been carefully trying to manage their diabetes but hasn t been able to keep their blood glucose levels within recommended levels without having disabling hypoglycaemia. This means that they have repeated and unpredictable hypoglycaemic episodes for which they need help from other people, and which make them anxious about the episodes occurring again and significantly spoil their way of life. The NICE recommendations regarding insulin pump therapy for type 1 diabetes are also valid for children, adolescents, pregnant women and women who are intending to become pregnant. However, pregnant women and women who are intending to become pregnant should only change to insulin pump therapy when under the care of the specialist team. NICE does not recommend insulin pump therapy for people who have type 2 diabetes and need to take insulin. It is noteworthy that the evidence upon which NICE based their recommendation for CSII was far from convincing in terms of reduction in levels of glycosylated haemoglobin (HbA 1c ), which is a standard indicator of the quality of glycaemic control over the past two to three months. In adults with type 1 diabetes, eight studies reported glycosylated haemoglobin after 10 weeks to 4 months of follow-up, and five provided sufficient data for inclusion in a meta-analysis. Using the random effects

18 4 model to account for heterogeneity between the trials, a non-significant reduction in glycosylated haemoglobin after up to four months of treatment with CSII compared with MDI was observed (P= 0.05). After six and 12 months of treatment with either CSII or MDI, a similar non-significant trend in favour of CSII was seen (P= 0.10 from four studies at six months, and P= 0.08 from three studies at 12 months). No relevant studies were available for children with type 1 diabetes, or adults with type 2 diabetes. Of two studies in adolescents with type 1 diabetes, one showed significantly lower glycosylated haemoglobin with CSII (9.3%) than with MDI (9.6%) at four months of follow-up (P< 0.05), whereas the other study found no difference between the treatments after six months of follow-up. The evidence for superiority of CSII over MDI was not overwhelming for other outcomes of relevance either. In adults with type 1 diabetes, only two of the included studies reported cholesterol levels after treatment with CSII or MDI and therefore the results in terms of the relative effect of treatment on lipid profile are inconclusive. The two studies in adolescents did not report lipid outcomes. Quality of life was reported in only one of the studies included in the NHS assessment. In this study (adult with type I diabetes), no significant differences between CSII and MDI were found overall or on any of the subscales of the Diabetes Quality of Life (DQOL) tool. Severe hypoglycaemic episodes differed little between treatments in nine of the twelve included studies of adults with type 1 diabetes. Likewise, the two studies of adolescents with type 1 diabetes reported no difference in the number of hypoglycaemic events between treatments. Cases of diabetic ketoacidosis were rare across all studies and therefore it was not possible to draw firm conclusions with regard to the relative incidence of ketoacidosis with CSII and MDI. The Appraisal Committee for NICE considered that the evidence from clinical trials comparing CSII therapy with MDI therapy showed that either CSII therapy is no more effective than MDI therapy, or if CSII therapy is more effective, then at best the difference between the two therapies is small. However, it was accepted that the comparisons combined evidence from studies using modern pump technology with studies using CSII technology that is now obsolete. In addition, many of the people enrolled in the trials were not from the group most likely to benefit from CSII therapy (ie, people with the most erratic control of their blood glucose levels). In making their recommendation, the Committee that for a small subgroup of people with type 1 diabetes who could not achieve adequate glycaemic control with MDI therapy, a much more satisfactory quality of life could be achieved using CSII therapy. Research questions The clinical question to be answered by this review was defined by staff from the Long Term Conditions Policy and Strategy team of the Clinical Services Directorate of New Zealand s Ministry of Health in conjunction with the reviewers. In general, the aim of this review was to evaluate the relative effectiveness, safety, and costeffectiveness of CSII in patients with insulin-treated diabetes, when compared to optimised MDI.

19 5 The primary research question to be addressed by this review was: Is subcutaneous insulin pump therapy effective, safe, and cost-effective compared with multiple daily injections? The review questions are defined according to the PICO (or PICOT) criteria: patient population intervention comparator outcomes time consideration For inclusion in the current review, the evidence had to fulfil the criteria outlined in Table 2 and Table 3. These criteria were developed a priori and described in the scoping protocol prepared prior to commencement of the review proper. Although the NHS technology assessment report (Colquitt et al, 2004) also included nonrandomised crossover studies, the current review includes randomised trials only to minimise bias. Table 2: Patient population Intervention Comparator Outcomes Table 3: Publication type Study design Study duration Sample size Criteria for determining study eligibility Insulin-treated diabetic patients (type 1 or type 2). Includes adults, children, adolescents, and pregnant women with pre-existing diabetes. Newly diagnosed patients are excluded. Gestational diabetes was excluded. Continuous subcutaneous insulin infusion (CSII), any manufacturer. Comparisons of conventional therapy, implantable pumps, and hospital inpatient pumps are excluded. Optimised multiple insulin injections (MDI), at least 3 injections per day. Efficacy outcomes: HbA 1c, lipid levels, quality of life (QOL) Safety outcomes: severe hypoglycaemic episodes, ketoacidotic episodes Nature of the evidence Studies published in the English language, including primary (original) research published as full original reports and secondary research (systematic reviews and meta-analyses) appearing in the published literature. Papers for which an abstract is not available for review via the bibliographic database are excluded. Those that provide at least Level II evidence according to the National Health and Medical Research Council (NHMRC) interim levels of evidence for intervention research questions (2005). This includes randomised controlled trials (Level II evidence) of crossover or parallel-group design, and systematic reviews of Level II evidence. At least 10 weeks on each treatment. At least 10 evaluable patients per study arm (or exposed to both treatments). It is noteworthy that the NHS technology assessment claims that comparison of CSII using analogues with MDI using soluble would introduce a confounding factor and thus comparisons of intensive therapy should only be made with the same type of insulin. Although there is a clear difference in the insulin profile of rapid-acting analogues and soluble insulin, the most appropriate comparator in practice is MDI using the insulin type most commonly used, which may well be short-acting rather then a rapid-acting analogue. In this review, studies of rapid-acting analogues in CSII with short-acting soluble insulin using MDI are included in additional all-inclusive analyses.

20 6 Literature search A systematic method of literature searching and selection was employed in the preparation of this review update. The NHS technology assessment report (Colquitt et al, 2004) upon which this update was based had a search date end of May/June For this update, searches were limited to English language material published from January 2002 onwards. The searches were completed on 27 August, Therefore, studies published after this date were not eligible for inclusion in the systematic review. The following databases were searched: Bibliographic databases Embase Medline Review databases Cochrane Database of Systematic Reviews Cochrane Central Register of Controlled Trials Database of Abstracts of Reviews of Effectiveness Health Technology Assessment database NHS Economic Evaluation database HTA Groups INAHTA website database: MSAC: ANZHSN: NZHTA: NICE: AHRQ/USPSTF: CADTH: Clinical Practice Guidelines National Guideline Clearing House database: The reference lists of key papers were searched to identify any peer-reviewed evidence that may have been missed in the literature search. In accordance with the Scoping Protocol, hand-searching was not undertaken and the grey literature and unpublished material such as conference abstracts were not searched due to the short time frame for this review. Search terms were searched for as keywords, exploded where possible, and as free text within the title and/or abstract, in the Embase and Medline databases. Variations on these terms were used for the Cochrane library and HTA website databases, modified to suit their keywords and descriptors. The search terms, search strategy and citations identified are presented in Table 4.