Online supplemental materials
Journal name: PharmacoEconomics
Article title: How consistent is the relationship between improved glucose control and modelled health outcomes for people with type 2 diabetes? A systematic review
Xinyang Hua MSc1, Thomas Wai-Chun Lung PhD1,2, Andrew PalmerMBBS3, Lei SiMSc3, William H. Herman MD4,Philip ClarkePhD1,
1. School of Population and Global Health, University of Melbourne
2. The George Institute for Global Health, University of Sydney
3. Menzies Research Institute, University of Tasmania
4. School of Public Health, University of Michigan
Correspondence to:
Philip Clarke
Address: The University of Melbourne, Level 4, 207 Bouverie Street, Carlton, Victoria 3053, AUSTRALIA
Phone: +61 3 90357897
Fax: +61 3 9348 1174
Email:
Supplemental material 1 –Searching strategy
1. Diabetes Mellitus, Type 2/
2. T2DM.
3. 1 or 2
4. (cost effective* or cost utilit*)
5. (long term outcome* or long term consequence* or health economic* or health evaluation or economic evaluation)
6. 4 or 5
7. (model* or simulat*)
8. (life expectancy or life year* or QALY*)
9. 3 and 6 and 7 and 8
10. limit 9 to English
Supplemental material 2 – Table: Studies included in this research
Study / Year / Model / Intervention / Control / Treatment effectsAfzali, Gray et al. [44] / 2013 / UKPDS OM / High-level involvement of practice nurses in the provision of clinical-based activities / Low level model / HbA1c, blood pressure
Gilmer, O'Connor et al. [45] / 2012 / UKPDS OM / Electronic medical records based clinical decision support / No electronic medical records based clinical decision support / HbA1c
Gao, Zhao et al. [22] / 2012 / UKPDS OM / Liraglutide / Glimepiride / HbA1c, blood pressure, body weight, lipid levels
Gillespie, O'Shea et al. [46] / 2012 / UKPDS OM / Group-based peer support in addition to standardized diabetes care / Standard diabetes care / HbA1c, blood pressure, lipid levels
Coyle, Coyle et al. [47] / 2012 / UKPDS OM / Combined exercise / No program/ aerobic/ resistance exercise / HbA1c
Tunis[48] / 2011 / UKPDS OM / Self-monitoring of blood glucose / No self-monitoring / HbA1c
Cameron, Coyle et al. [49] / 2010 / UKPDS OM / Self-monitoring of blood glucose / No self-monitoring / HbA1c
Waugh, Cummins et al. [50] / 2010 / UKPDS OM / Sitagliptin / Rosiglitazone / HbA1c, BMI/body weight
Brown, Grima[51] / 2014 / CORE / Insulin glargine / Sitagliptin / HbA1c, BMI/body weight, hypoglycaemic events
Tzanetakos, Melidonis et al. [52] / 2014 / CORE / Liraglutide / Sitagliptin/Exenatide / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Shafie, Gupta et al. [53] / 2014 / CORE / Biphasic insulin aspart 30 / Taking glucose-lowering drugs alone / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Giorda, Nicolucci et al. [54] / 2014 / CORE / Associazione Medici Diabetologi-annals initiative / Conventional management / HbA1c, blood pressure, BMI/body weight, lipid levels
Fonseca, Clegg et al. [55] / 2013 / CORE / Exenatide once weekly / Exenatide twice daily/ Insulin Glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Raya, Pérez et al. [56] / 2013 / CORE / Liraglutide 1.2 mg / Sitagliptin 100 mg / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Pollock, Muduma et al. [57] / 2013 / CORE / Laparoscopic adjustable gastric banding / Standard medical management / HbA1c, blood pressure, BMI/body weight, lipid levels
Davies, Chubb et al[58] / 2012 / CORE / Liraglutide / Glimepiride/sitagliptin / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Guillermin, Lloyd et al.[59] / 2012 / CORE / Exenatide once weekly / Sitagliptin/pioglitazone / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Lee, Samyshkin et al. [60] / 2012 / CORE / Liraglutide 1.2 mg/1.8 mg / Sitagliptin / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Pollock, Curtis et al. [61] / 2012 / CORE / Long-acting analog insulin LM75/25 or LM50/50 / Long-acting basal analog insulin / HbA1c
Pollock, Curtis et al. [62] / 2012 / CORE / Long-acting analog insulin LM75/25 or LM50/50 / Long-acting basal analog insulin / HbA1c
Samyshkin, Guillermin et al. [63] / 2012 / CORE / Exenatide once weekly / Insulin Glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Smith-Palmer, Fajardo-Montanana et al. [64] / 2012 / CORE / Detemir / NPH insulin / HbA1c, BMI/body weight, hypoglycaemic events
Valov, Palmer et al. [65] / 2012 / CORE / Biphasic insulin aspart 30 / Human premix insulin / HbA1c, BMI/body weight, hypoglycaemic events
Yang, Christensen et al. [66] / 2012 / CORE / Insulin Glargine / Insulin Detemir / HbA1c, BMI/body weight, hypoglycaemic events
Beaudet, Palmer et al[67] / 2011 / CORE / Exenatide once weekly / Insulin glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Goodall, Costi et al. [68] / 2011 / CORE / Exenatide twice daily / Insulin Glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Lee, Conner et al. [69] / 2011 / CORE / Liraglutide 1.2 mg/1.8 mg + glimepiride / Rosiglitazone + glimepiride / HbA1c, BMI/body weight, hypoglycaemic events
Valentine, Palmer et al. [70] / 2011 / CORE / Liraglutide / Exenatide / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Lee, Conner et al. [71] / 2010 / CORE / Liraglutide / Exenatide / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Palmer, Beaudet al. [72] / 2010 / CORE / Biphasic insulin aspart / Insulin glargine / HbA1c, BMI/body weight, hypoglycaemic events
Palmer, Knudsen et al. [73] / 2010 / CORE / Biphasic insulin aspart / Biphasic human insulin / HbA1c, BMI/body weight, hypoglycaemic events
Pollock, Valentine et al. [74] / 2010 / CORE / Self-monitoring of blood glucose / No self-monitoring of blood glucose / HbA1c
Tunis, Sauriol et al. [75] / 2010 / CORE / Insulin glargine plus oral antidiabetes drugs / Premixed insulin alone / HbA1c, hypoglycaemic events
Tunis, Willis et al. [76] / 2010 / CORE / Self-monitoring of blood glucose / No self-monitoring of blood glucose / HbA1c
Tunis and Minshall[77] / 2010 / CORE / Self-monitoring of blood glucose / No self-monitoring of blood glucose / HbA1c
Tunis, Minshall et al. [78] / 2009 / CORE / Insulin detemir / NPH insuline / HbA1c, hypoglycaemic events
Brandle, Erny-Albrecht et al. [79] / 2009 / CORE / Exenatide / Insulin glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Brandle, Goodall et al. [80] / 2009 / CORE / Pioglitazone / Placebo / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Cameron and Bennett [81] / 2009 / CORE / Insulin analogues / Regular human insulin/ neutral protamine hagedorn insulin / HbA1c, hypoglycaemic events
Goodall, Sarpong et al. [82] / 2009 / CORE / Insulin immediate initiation / Insulin 8-year delay / HbA1c, BMI/body weight, hypoglycaemic events
Lee, Seo et al. [83] / 2009 / CORE / Biphasic insulin aspart 30 / Biphasic human insulin / HbA1c, BMI/body weight, hypoglycaemic events
Mittendorf, Smith‐Palmer et al. [35] / 2009 / CORE / Exenatide / Insulin Glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Scherbaum, Goodall et al. [84] / 2009 / CORE / Pioglitazone / Placebo / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Valentine, Tucker et al. [85] / 2009 / CORE / Pioglitazone / Placebo / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
St Charles, Minshall et al. [86] / 2009 / CORE / Pioglitazone / Rosiglitazone / HbA1c, lipid levels
Ali, White et al. [87] / 2008 / CORE / Biphasic insulin aspart 30 / Human insulin / HbA1c, BMI/body weight, hypoglycaemic events
Chirakup, Chaiyakunapruk et al. [88] / 2008 / CORE / Pioglitazone / Rosiglitazone / HbA1c, lipid levels
Goodall, Jendle et al[89] / 2008 / CORE / Biphasic insulin aspart 70/30 / Glargine / HbA1c, BMI
Minshall, Oglesby et al. [90] / 2008 / CORE / Exentide / Metformin and /or sulfonylurea / HbA1c, blood pressure, BMI/body weight, lipid levels
Palmer, Gibbs et al. [91] / 2008 / CORE / Biphasic insulin aspart 30 / Biphasic human insulin / HbA1c, BMI/body weight, hypoglycaemic events
Palmer, Goodall et al. [92] / 2008 / CORE / Biphasic insulin aspart 30 / Biphasic human insulin / HbA1c, BMI/body weight, hypoglycaemic events
Valentine, Goodall et al. [93] / 2008 / CORE / Insulin detemir / OADs/NPH/insulin glargine / HbA1c, BMI/body weight, hypoglycaemic events
Tunis and Minshall[94] / 2008 / CORE / Self-monitoring of blood glucose / No self-monitoring of blood glucose / HbA1c
Tunis, Minshall et al. [95] / 2008 / CORE / Pioglitazone / Rosiglitazone / HbA1c, lipid levels
Gilmer, Roze et al. [96] / 2007 / CORE / “Project Dulce” / No project / HbA1c, blood pressure, lipid levels
Ray, Boye et al. [97] / 2007 / CORE / Exenatide / Insulin glargine / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Ray, Valentine et al. [98] / 2007 / CORE / Biphasic insulin aspart 70/30 / Glargine / HbA1c, BMI/body weight
Valentine, Erny-Alvrecht et al. [99] / 2007 / CORE / Insulin detemir / OADs/NPH/insulin glargine / HbA1c, BMI/body weight, hypoglycaemic events
Valentine, Bottomley et al. [100] / 2007 / CORE / Pioglitazone / Placebo / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Palmer, Dinneen et al. [101] / 2006 / CORE / Self-monitoring of blood glucose / No self-monitoring of blood glucose / HbA1c
Roze, Valentine et al. [102] / 2006 / CORE / Acarbose / Alacebo / HbA1c, blood pressure, BMI/body weight, lipid levels
Valentine, Palmer et al. [103] / 2006 / CORE / HbA1c reduction
Scenarios / _ / HbA1c
Valentine, Palmer et al. [104] / 2005 / CORE / Biphasic insulin aspart 70/30 / Glargine / HbA1c, BMI
Palmer, Roze et al. [105] / 2004 / CORE / Repaglinide / Nageglinide / HbA1c
Palmer, Roze et al.[39] / 2004 / CORE / HbA1c reduction
scenario / _ / HbA1c
Van, Pompen et al.[29] / 2014 / Cardiff / Dapagliflozin / Placebo / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Elgart, Caporale et al.[26] / 2013 / Cardiff / Saxagliptin / Sulfonylureas / HbA1c, BMI/body weight, hypoglycaemic events
Grzeszczak, Czuprynial et al.[25] / 2012 / Cardiff / Saxagliptin / NPH Insulin / HbA1c, BMI/body weight, hypoglycaemic events
Granstorm, Bergenheim et al.[27] / 2012 / Cardiff / Saxagliptin / Sulfonylureas / HbA1c, BMI/body weight, hypoglycaemic events
Erhardt, Bergenheim et al.[28] / 2012 / Cardiff / Saxagliptin / Sulfonylureas / HbA1c, BMI/body weight, hypoglycaemic events
Neslusan, Teschemaker et al.[106] / 2015 / ECHO / Canagliflozin / Sitagliptin / HbA1c, blood pressure, BMI/body weight, lipid levels, hypoglycaemic events
Gillett, Dallosso et al.[42] / 2010 / Sheffield / Education and self management / No education and self management / HbA1c, blood pressure, BMI/body weight, lipid levels
Ryabov[41] / 2014 / CDC / Intervention involving community health workers / No community health workers / HbA1c
Hoerger, Zhang et al.[107] / 2009 / CDC / With the risk factor changes between 1988-1994 and 2005-2006 in US / Without risk factor changes / HbA1c
Brownson, Hoerger et al.[40] / 2009 / CDC / Self management program / No self management program / HbA1c, lipid levels
The CDC Diabetes Cost-effectiveness Group[18] / 2002 / CDC / Intensive glycemic control / Conventional glycemic control / HbA1c
BMI=body mass index; UKPDS OM=UKPDS outcome model; LM=lispro mix; NPH=neutral protamine hagedorn; OADs=oral antidiabetic dru
Supplemental material 3 --Treatment effects used in cost effectiveness studies in different years and interventions
Figure S1. Percentage of type 2 diabetes cost effectiveness and cost utility studiesusing treatment effects besides of HbA1c, 2002-2015. CE=cost effectiveness; CU=cost utility; BMI=Body Mass Index. Number of studies in each time period: 2002-2006(n=7), 2007-2008(n=14), 2009-2010(n=22), 2011-2012(n=21). 2013-2015(n=12).
Figure S2. Percentage of type 2 diabetes cost effectiveness and cost utility studies using different treatment effects, for different types of interventions. BMI=Body Mass Index. Number of studies in each intervention group: Insulin (n=22); Management intervention (n=18); Oral therapy (n=33). The 3 studies that didn’t specify the intervention were not included in this graph.
Supplemental material 4— Undiscounted outcomes estimation
Methods:
Some studies reported multiple discountingrates outcomes in their sensitivity analysis. For example, a study that reports their ∆QALYs with a base case discounting rate of 3% may also reports their ∆QALYs with a 0% discounting rate and ∆QALYs with a 6% discounting rate in the sensitivity analysis. For each study like this, we calculated the ratio between the undiscounted ∆QALYs and the ∆QALYs in the specific discounting rate ithat the study reported (3% and 6% inthe example):
where irepresents the discounting rate and j represents the study. We then calculated the average ratio for discounting i:
/n
This ratio was used in studies that reported their base case ∆QALYs in discounting rate i and didn’t report undiscounted ∆QALYs in their study. Undiscounted ∆LE was calculated using the same ratio as for ∆QALYs.
Since some studies without undiscounted ∆QALYs reported their base case ∆QALYs in a discounting rate that were not included in any other study’s sensitivity analysis, no ratios were available for them to calculate the undiscounted ∆QALYs. A linear regression of average ratios on different discounting rates was fitted to help calculate those ratios.
Results:
The calculated average ratios for different discounting rates can be found in Table S1. The linear regression can be found in Figure S3.
Table S1Average ratios of the different discounted rate change in QALYs
1.5% / 2.5% / 3.0% / 3.5% / 5.0% / 6.0% / 8.0% / 10.0%Average Ratio / 1.30 / 1.56 / 1.57 / 1.70 / 2.09 / 2.38 / 3.25 / 3.31
N of studies with ratio / 1 / 1 / 25 / 8 / 11 / 24 / 1 / 4
N of studies use the average ratio / 1 / 0 / 20 / 5 / 5 / 0 / 0 / 0
Figure S3Average ratios of the different discounted rate change in QALY