Recovery of hypoglycemia awareness in longstanding type 1 diabetes: A multicenter 2×2 factorial randomized controlled trial comparing insulin pump with multiple daily injections and continuous with conventional glucose selfmonitoring (HypoCOMPaSS)
摘要
Diabetes Technology & TherapeuticsVol. 18, No. S1 Original ArticlesFree AccessInsulin PumpsJohn C. PickupJohn C. PickupDiabetes Research Group, King's College London, Faculty of Life Sciences and Medicine, Guy's Hospital, London, UK.Search for more papers by this authorPublished Online:2 Feb 2016https://doi.org/10.1089/dia.2016.2503AboutSectionsPDF/EPUB Permissions & CitationsPermissionsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail IntroductionWhilst the effects of insulin pump therapy (continuous subcutaneous insulin infusion, CSII) on glycemic variables such as HbA1c and severe hypoglycemia frequency are well reported, there has been comparatively little published on the subject of comparative, real-life clinical outcomes during pump treatment and multiple daily insulin injections (MDI). We include in this article one of the first papers to report on cardiovascular and all-cause mortality in type 1 diabetes during these methods of insulin delivery.It was as long as 25 years after the introduction of insulin pump therapy that the first reports of the cost-effectiveness (CE) of this treatment appeared. Nowadays, most manufacturers, health-care professionals, and policy makers within health-care systems and insurance organizations appreciate that CE is a crucial consideration when recommending a diabetes technology for routine usage. It needs early rather than late consideration and study. There have now been several CE analyses of CSII vs. MDI, and we include a systematic review that assesses and summarizes the health economic issues around CSII use in several countries.There is increasing usage of CSII in everyday diabetes management, and technologically sophisticated pump models with enhanced features are being introduced. Regrettably though, surveys report that there are still frequent technical problems—whether through the pump technology itself or because of patient errors in its usage. We have come to realize that there are significant regulatory issues concerning the requirements for safety and efficacy of CSII and other diabetes technologies, both concerning their market introduction and their post-market performance. We therefore include in the Yearbook an important policy statement on safety standards and set of recommendations for action by the European Association for the Study of Diabetes (EASD) and the American Diabetes Association (ADA). Much has been written over the years about predictors of outcome with CSII, both the level of glycemic control achieved and the likelihood of pump discontinuation in some patients, but this has been largely based heretofore on short-term studies. Attention is now turning to the long-term outcomes of CSII, over at least 4 or 5 years, and the factors that may determine success or failure (see last year's Yearbook). We include here two studies that indicate apparently different predictors of long-term outcome in adults and in children and adolescents with type 1 diabetes.Lastly, we discuss a somewhat controversial study on changes in hypoglycemia awareness and severe hypoglycemia frequency during CSII, MDI, continuous glucose monitoring (CGM), and self-monitoring of blood glucose (SMBG). Discussion here turns around whether the package of MDI and SMBG with intensive education and support can lead to as good a clinical result as with CSII and/or CGM.Key Articles Reviewed for this ArticleInsulin pump therapy, multiple daily injections, and cardiovascular mortality in 18,168 people with type 1 diabetes: observational studySteineck I1, Cederholm J2, Eliasson B3, Rawshani A4, Eeg-Olofsson K3, Svensson A-M4, Zethelius B5,6, Avdic T4, Landin-Olsson M7, Jendle J8, Gudbjörnsdóttir S3,4, theSwedish National Diabetes RegisterBMJ 2015;350: h3234Cost-effectiveness of continuous subcutaneous insulin infusion versus multiple daily injections of insulin in type 1 diabetes: a systematic reviewRoze S1, Smith-Palmer J2, Valentine W2, de Portu S3, Nørgaard K4, Pickup JC5Diabetic Medicine 2015. [Epub ahead of print]; DOI 10.1111/dme.12792Insulin pump risks and benefits: a clinical appraisal of pump safety standards, adverse event reporting, and research needs. A joint statement of the European Association for the Study of Diabetes and the American Diabetes Association Diabetes Technology Working GroupHeinemann L1, Fleming A2, Petrie JR3, Holl RW4, Bergenstal RM5, Peters AL6Diabetes Care 2015;38: 716–22Predicting the effectiveness of insulin pump therapy on glycemic control in clinical practice: a retrospective study of patients with type 1 diabetes from 10 outpatient diabetes clinics in Sweden over 5 yearsClements M1,2, Matuleviciene V3,4, Attvall S4,5, Ekelund M6, Pivodic A7, Dahlqvist S3, Fahlén M8, Haraldsson B9, Lind M3,4Diabet Technol Therapeut 2015;17: 21–28The effectiveness and durability of an early insulin pump therapy in children and adolescents with type 1 diabetes mellitusBrancato D, Fleres M, Aiello V, Saura G, Scorsone A, Ferrara L, Provenzano F, Di Noto A, Spano L, Provenzano VDiabet Technol Therapeut 2015;16: 735–41Recovery of hypoglycemia awareness in long-standing type 1 diabetes: a multicenter 2×2 factorial randomized controlled trial comparing insulin pump with multiple daily injections and continuous with conventional glucose self-monitoring (HypoCOMPaSS)Little SA1, Leelarathna L2, Walkinshaw E3, Tan HK4, Chapple O5, Lubina-Solomon A3, Chadwick TJ6, Barendse S7, Stocken DD6, Brennand C6, Marshall SM1, Wood R6, Speight J7–9, Kerr D10, Flanagan D4, Heller SR3, Evans ML2, Shaw JAM1Diabetes Care 2014;37: 2114–22Cardiovascular Mortality On Csii Vs. MdiInsulin pump therapy, multiple daily injections, and cardiovascular mortality in 18,168 people with type 1 diabetes: observational studySteineck I1, Cederholm J2, Eliasson B3, Rawshani A4, Eeg-Olofsson K3, Svensson A-M4, Zethelius B5,6, Avdic T4, Landin-Olsson M7, Jendle J8, Gudbjörnsdóttir S3,4, the Swedish National Diabetes Register1Department of Endocrinology, Aarhus University Hospital, Aarhus, Denmark2Department of Public Health and Caring Sciences/Family and Preventive Medicine, Uppsala University, Uppsala, Sweden3Institute of Medicine, Sahlgrenska University Hospital, University of Gothenburg, Gothenburg, Sweden4National Diabetes Register, Centre of Registers, Gothenburg, Sweden5Department of Public Health and Caring Sciences/Geriatrics, Uppsala University, Uppsala, Sweden6Medical Products Agency, Uppsala, Sweden7Department of Clinical Science, Lund University, Lund, Sweden8Faculty of Health Sciences and Medicine, Örebro University, Örebro, SwedenBMJ 2015;350: h3234BackgroundCardiovascular disease (CVD) is known to be about twice as frequent in diabetes as the general population, with hyperglycemia and hypoglycemia as significant risk factors. Since insulin pump therapy (CSII) improves glycemic control and reduces hypoglycemia, this study aimed to investigate the long-term effect on CVD and mortality of CSII vs. MDI.MethodsDetails from type 1 diabetic patients treated by CSII (n=2,441) or MDI (n=15,727) were entered into the Swedish National Diabetes Register and followed from baseline to death or first CVD event for a mean follow-up of 6.8 years.ResultsFatal coronary heart disease (CHD) was reduced by 45% (adjusted hazard ratio 0.55 [95% confidence interval 0.36–0.83], fatal CVD [stroke or CHD] by 42% [adjusted hazard ratio 0.58, CI 0.40–0.85]) and all cause mortality by 27% (adjusted hazard ratio 0.73 [0.58–0.92]) by CSII vs. MDI. Mean updated HbA1c was similar on CSII vs. MDI (8% [64 mmol/mol] on both treatments) but patients with ≥3 severe hypoglycemic episodes were more likely to be on MDI (p=0.034).ConclusionAmong people with type 1 diabetes, CSII is associated with lower CVD mortality than MDI.CommentThis is the first study to show a reduction in CVD mortality with insulin pump therapy compared to MDI. Nearly all people with type 1 diabetes (about 95%) in Sweden are entered into the National Diabetes Registry, and a reasonable proportion of the adults are receiving CSII (25% of women and 20% of men), so the results are likely to be applicable to the general adult pump population.Whilst it is tempting to attribute the good effect of CSII to the improved diabetes control generally seen with pumps, one should note that HbA1c levels were similar during both treatments. In recent years, the high risk of mortality associated with frequent severe hypoglycemia in type 1 diabetes has been recorded in several studies1 and the reduced hypoglycemia noted with CSII here may have contributed to the beneficial effect on pumps. Improved glycemic variability has also been noted with CSII vs. MDI; it was not measured here but might have been a partial explanation for the improved outcomes on CSII. The frequency of self monitoring of blood glucose and data on patient education were not available in this study and the authors recognize that better education about and practice of diet, exercise, and insulin dosage adjustment in the pump patients might also have been a reason for reduced CVD on CSII.The follow-up in this study was relatively short (mean 6.8 years) and the duration of CSII unknown. As CVD takes a long time to develop, one might expect that outcomes on pump therapy will be especially favorable in patients with long duration of CSII and follow-up. This needs testing.Cost-Effectiveness of CsiiCost-effectiveness of continuous subcutaneous insulin infusion versus multiple daily injections of insulin in type 1 diabetes: a systematic reviewRoze S1, Smith-Palmer J2, Valentine W2, de Portu S3, Nørgaard K4, Pickup JC51HEVA HEOR, Lyon, France2Ossian Health Economics and Communications, Basel, Switzerland3Medtronic International Sarl, Tolochenaz, Switzerland4Department of Endocrinology, Hvidovre University Hospital, Hvidovre, Denmark5Diabetes Research Group, King's College London School of Medicine, Guy's Hospital, London, UKDiabetic Medicine 2015. [Epub ahead of print]; DOI 10.1111/dme.12792BackgroundValue for money is a crucial determinant of usage for diabetes technologies like insulin pump therapy. A number of CE studies comparing CSII and MDI in various settings (different countries and health-care systems) have been published in recent years. The aim of this study was to systematically review these and test whether CSII is cost-effective across different settings.MethodsA literature search for the CE of CSII vs. MDI in type 1 diabetes was performed using Medline, Cochrane Library, and other databases. There were no time or language restrictions. Sensor-augmented pump therapy was excluded. A narrative synthesis was then undertaken and mean CE calculated.ResultsAfter exclusions and duplicates, 11 studies from 8 countries were available for review: United Kingdom (4 studies), United States, Canada, Australia, Spain, Denmark, Italy, and Poland. Nine analyses used the CORE Diabetes Model and two used other methods for calculating CE. CSII was found to be more expensive than MDI, with a mean ratio CSII:MDI cost of 1.4±0.4, but costs in models were partially offset by savings from reduced diabetes complications. CSII was considered cost-effective in all eight settings. For the base case with a mean baseline HbA1c of 8.7% (72 mmol/mol), the mean (95% confidence interval, CI) incremental CE ratio (ICER) was Euros 30,862 (17,997–43,727) per quality-adjusted life year (QALY) gained (USD 40,143 [23,409–56,876], GBP 21,853 [12,744–30,965]). Sensitivity analyses showed the main determinants of CE were either a high baseline HbA1c and consequent large reduction in HbA1c on CSII, or a high baseline frequency and consequent large reduction in severe hypoglycemia on switching to insulin pump therapy.ConclusionsInsulin pump therapy is CE across a number of different countries and health-care systems for patients with type 1 diabetes and poor glycemic control and/or problematic hypoglycemia on MDI.CommentThe UK National Institute for Health and Clinical Excellence (NICE) unofficial “willingness-to-pay” ICER threshold is usually taken to be 1 year, and n=133 selected for analysis. The mean age at diagnosis was 8.9±5.6 years, mean follow-up 4.0±1.8 years, and mean baseline HbA1c 9.5±1.8%.ResultsMean yearly HbA1c was reduced throughout the study compared to baseline. The only predictor of the mean HbA1c achieved throughout the study was the interval between the diagnosis of diabetes and start of CSII, with those commencing pump therapy soon after diagnosis, achieving and maintaining the best control. Age at onset or age at CSII start, diabetes duration, baseline HbA1c, and BMI did not act as predictors of the mean HbA1c on CSII.ConclusionsChildren and adolescents treated by CSII achieve the best reduction in HbA1c when pump therapy is started soon after diagnosis. It is possible that early pump commencement might prolong the honeymoon phase of diabetes, but since C peptide levels were not measured in the study this notion cannot be confirmed or refuted.CommentThese studies appear to reach different conclusions with respect to the predictors of clinical effect on CSII. In adults, Clements et al. found that baseline HbA1c and BMI were significant determinants of pump HbA1c. The fact that the largest reduction in HbA1c is in those patients with the highest HbA1c on MDI has been reported previously by several groups3,4 and is the basis for some national clinical guidelines for pump use, which are largely based on cost-effectiveness and recommend CSII for those with a continued elevated HbA1c on MDI5. Other researchers have also found that CSII “nonresponders” tend to have a higher BMI,6 possibly indicating (as one would expect) a major role for diet, exercise, and insulin resistance in influencing glycemic outcomes on CSII.Clements et al. find a relatively small mean change in HbA1c on CSII vs. MDI (0.25% when the baseline was 9%). This should not be taken as showing that CSII has a small effect on HbA1c reduction in poorly controlled patients—indeed it might be argued that this is a case where the reporting of mean HbA1c levels in unselected patients as a measure of pump effectiveness is misleading. The subjects here were a mixture of those who started on CSII because of elevated HbA1c on MDI and those who had disabling hypoglycemia on CSII—who would presumably have a lower or near-normal HbA1c on MDI yet experience significant clinical benefit from pump therapy. No information on hypoglycemia was available in the study. Equally, factors such as type and intensity of pump education and health-care professional input will influence HbA1c outcome, and this was not recorded here. Indeed, some centers had no improvement in mean HbA1c with pumps, an experience that differs from most established pump clinics.Brancato et al. conclude that in children and adolescents neither baseline HbA1c nor BMI predict the control that can be achieved or maintained on CSII. It is logical that good control (e.g., when CSII is used) might “rest the pancreas;” prolong the honeymoon period; and result in preserved β-cell function, higher C peptide, and lower HbA1c levels. As we have discussed before in the Yearbook (2011), there is evidence for higher C peptide levels in some age groups of children with newly diagnosed type 1 diabetes allocated sensor-augmented pump therapy vs. conventional CSII,7 which hints at better preservation of endogenous insulin production by technology-related strict control. But the issue needs much more study.Hypoglycemia On Csii Vs. Mdi, With and Without Continuous Glucose MonitoringRecovery of hypoglycemia awareness in long-standing type 1 diabetes: a multicenter 2×2 factorial randomized controlled trial comparing insulin pump with multiple daily injections and continuous with conventional glucose self-monitoring (HypoCOMPaSS)Little SA1, Leelarathna L2, Walkinshaw E3, Tan HK4, Chapple O5, Lubina-Solomon A3, Chadwick TJ6, Barendse S7, Stocken DD6, Brennand C6, Marshall SM1, Wood R6, Speight J7,8,9, Kerr D10, Flanagan D4, Heller SR3, Evans ML2, Shaw JAM11Institute of Cellular Medicine, Newcastle University, Newcastle, UK2Wellcome Trust-MRC Institute of Metabolic Science, Metabolic Research Laboratories, University of Cambridge, Cambridge, UK3School of Medicine and Biomedical Sciences, Sheffield University, UK4Peninsula College of Medicine and Dentistry, Plymouth, UK5Bournemouth Diabetes and Endocrine Centre, Royal Bournemouth Hospital, Bournemouth, UK6Newcastle Clinical Trials Unit, Institute of Health and Society, Newcastle University, Newcastle, UK7AHP Research, Hornchurch, UK8The Australian Centre for Behavioural Research in Diabetes, Diabetes Australia–Vic, Melbourne, Australia9Centre for Mental Health and Wellbeing Research, School of Psychology, Deakin University, Burwood, Australia10Centre for Postgraduate Medical Research and Education, Bournemouth University, UKDiabetes Care 2014;37: 2114–22This manuscript is discussed also in the article on Diabetes Technology and the Human Factor, p. S-109.BackgroundCSII is known to reduce severe hypoglycemia vs. MDI but in trials that have not used long-acting insulin analogues as part of the MDI regimen. Previous pilot studies by this group have shown that education and support aimed at rigorous avoidance of biochemical hypoglycemia can reduce severe hypoglycemia and improve hypoglycemia awareness in type 1 diabetes, whether treated by CSII or MDI. The aim of this study therefore was to test whether impaired awareness of hypoglycemia (IAH) can be improved and severe hypoglycemia reduced in patients with type 1 diabetes and IAH when treated by CSII, MDI, real-time CGM, or SMBG using such a hypoglycemia-avoidance program.MethodsAdult patients with type 1 diabetes and known IAH (N=96, mean diabetes duration 29 years) at five centers in the United Kingdom were randomized to one of four treatments over 24 weeks: CSII or MDI using glargine, with CGM or SMBG. All patients received comparable education and support aimed at hypoglycemia-avoidance, including monthly clinic visits and weekly telephone calls.ResultsIn the overall study population, IAH improved and biochemical hypoglycemia (assessed by blinded CGM), fear of hypoglycemia, and severe hypoglycemia frequency decreased without deterioration in HbA1c. At study completion, there was no significant difference in IAH score or reductions in hypoglycemia or fear of hypoglycemia, either between methods of insulin delivery or glucose monitoring. Treatment satisfaction was higher with CSII than MDI, but similar for CGM and SMBG.ConclusionsHypoglycemia awareness and the frequency of severe hypoglycemia can be improved in long-standing patients with type 1 diabetes, with similar outcomes for MDI or CSII and for CGM or SMBG. Treatment satisfaction was, however, higher for CSII.CommentThis study is important in showing that it is possible to reduce severe hypoglycemia in hypoglycemia-prone type 1 diabetes using a special intensive program of education and support aimed at rigorous avoidance of hypoglycemia. Somewhat surprisingly, there was no difference in most outcomes between CSII and MDI, or CGM and MDI. In many previous randomized and observational trials (including Steineck et al. above) and meta-analyses of hypoglycemia-prone patients with type 1 diabetes, severe hypoglycemia has been reported as being significantly reduced by CSII vs. MDI.3In a letter commenting on the Little et al. article, Edelman8 makes the point that the intensive strategies used in the trial, including monthly visits and weekly telephone calls, are not deliverable in routine care. Moreover, the low adherence rate for CGM in the study (57%, with only 17% using CGM >80% of the time) is below the usage rate at which greater CGM effectiveness vs. SMBG is known to occur—at least for HbA1c9 and probably also for hypoglycemia reduction. It is important also that the cohort of patients in the study of Little et al. were not chosen because they had suffered continued disabling hypoglycemia after a period of best attempts with MDI, including structured diabetes education and frequent input from health-care professionals, which is the criterion for a trial of CSII recommended by NICE.5The HypoCOMPaSS trial is valuable in showing what improvement in hypoglycemia is, in theory, achievable with hypoglycemia-prone patients using an intensive education and support program that does not necessarily involve diabetes technology. It does not mean that there is no role for diabetes technologies like CSII and CGM in everyday diabetes management. As we have argued before,9 the pathway of best care in type 1 diabetes is most sensibly and cost-effectively based on a sequential pathway, starting with MDI, then for those that fail to achieve target control on MDI, a trial of CSII, and finally for those who have not achieved target control on CSII, a trial of sensor-augmented pump therapy.Author Disclosure StatementJ.C.P. has received speaker and/or consultancy fees from Medtronic, Roche, CeQur, and Cellnovo—manufacturers of insulin pumps—and from Novo Nordisk and Eli Lilly—manufacturers of insulin formulations.References1 McCoy RG, Van Houten HK, Ziegenfuss JY, Shah ND, Vermers RA, Smith SA. Increased mortality of patients with diabetes reporting severe hypoglycemia. Diabetes Care 2012; 35: 1897–901. Crossref, Medline, Google Scholar2 Pickup JC, Yemane N, Brackenridge A, Pender S. Nonmetabolic complications of continuous subcutaneous insulin infusion: a patient survey. Diabet Technol Therapeut 2014; 16: 1–5. Link, Google Scholar3 Pickup JC, Sutton AJ. Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta-analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabet Med 2008; 25: 765–74. Crossref, Medline, Google Scholar4 Retnakaran R, Hochman J, DeVries JH, Hanaire-Broutin H, Heine RJ, Melki V, Zinman B. Continuous subcutaneous insulin infusion versus multiple daily injections. The impact of baseline A1c. Diabetes Care 2004; 27: 2590–96. Crossref, Medline, Google Scholar5 National Institute for Health and Clinical Excellence. Continuous Subcutaneous Insulin Infusion for the Treatment of Diabetes Mellitus. Technology Appraisal Guidance 151 (Review of Technology Appraisal Guidance 57). London, England: NICE, 2008. Google Scholar6 Nixon R, Folwell R, Pickup JC. Variations in the quality and sustainability of long-term glycaemic control with continuous subcutaneous insulin infusion. Diabetic Med 2014; 31: 1174–77. Crossref, Medline, Google Scholar7 Kordonouri O, Pankowska E, Rami B, Kapellen T, Coutant R, Hartmann R, Lange K, Knip M, Danne T. Sensor-augmented pump therapy from the diagnosis of childhood type 1 diabetes: results of the Paediatric Onset Study (ONSET) after 12 months of treatment. Diabetologia 2010; 53: 2487–95. Crossref, Medline, Google Scholar8 Edelman SV. Comment on Little et al. Recovery of hypoglycemia awareness in long-standing type 1 diabetes: a multicenter 2×2 factorial randomized controlled trial comparing insulin pump with multiple daily injections and continuous with conventional glucose self-monitoring (HypoCOMPaSS). Diabetes Care 2014; 37: e270–71. Crossref, Medline, Google Scholar9 Pickup JC. Banting Memorial Lecture 2014. Technology and diabetes care: appropriate and personalized. Diabetic Med 2015; 32: 3–13. Crossref, Medline, Google ScholarFiguresReferencesRelatedDetails Volume 18Issue S1Feb 2016 InformationCopyright 2016, Mary Ann Liebert, Inc.To cite this article:John C. Pickup.Insulin Pumps.Diabetes Technology & Therapeutics.Feb 2016.S-22-S-28.http://doi.org/10.1089/dia.2016.2503Published in Volume: 18 Issue S1: February 2, 2016PDF download
更多查看译文
关键词
insulin pumps
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要