Cost Effectiveness of Sodium-Glucose Cotransporter-2 (SGLT2) Inhibitors, Glucagon-Like Peptide-1 (GLP-1) Receptor Agonists, and Dipeptidyl Peptidase-4 (DPP-4) Inhibitors: A Systematic Review

Abstract

Objective

This study aimed to systematically review cost-effectiveness studies of newer antidiabetic medications.

Methods

The PubMed/MEDLINE, EMBASE, CINAHL Plus, Cochrane Library–NHS Economic Evaluation Database (Wiley), Cochrane Library–Health Technology Assessment Database (Wiley), Cochrane Library–Database of Abstracts of Reviews of Effects (Wiley), and the Cost-Effectiveness Analysis Registry databases (from 1 January 2000 to 1 June 2018) were searched. The search strategies included the Medical Subject Heading (MeSH) term ‘economics’, and the MeSH entry terms ‘cost’, ‘cost effectiveness’, ‘value’, and ‘cost utility’, as well as all names for GLP-1 receptor agonists, DPP-4 inhibitors, and SGLT2 inhibitors. Inclusion criteria included (1) cost-effectiveness studies of the newer antidiabetic medications, including sodium-glucose cotransporter-2 (SGLT2) inhibitors, glucagon-like peptide-1 (GLP-1) receptor agonists, and dipeptidyl peptidase-4 (DPP-4) inhibitors; and (2) full-text publications in English. Two reviewers independently screened the titles, abstracts, and full-text articles to select studies for data extraction. Discrepancies were resolved by discussion and consensus. The quality of reporting cost-effectiveness analyses was assessed using the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) guideline.

Results

Among 85 studies selected, 82 clearly stated the types of diabetes model used (e.g. CORE model), and 70 studied used validated diabetes models. Seventy-four (87%) studies were funded by pharmaceutical companies, and 72 (85%) studies were conducted from a payer’s perspective. Seventy-six (89%) studies presented were of good quality (20–24 CHEERS items), and nine were of moderate quality (14–19 items). Thirty studies compared newer antidiabetic medications with insulin, 3 studies compared newer antidiabetic medications with thiazolidinediones (TZDs), 15 studies compared newer antidiabetic medications with sulfonylureas, 40 studies compared new antidiabetic medications with alternative newer antidiabetic medication, and 9 studies compared other antidiabetic agents that were not included above. Newer antidiabetic medications were reported to be cost-effective in 26 of 30 (87%) studies compared with insulin, and 13 of 15 (87%) studies compared with sulfonylureas.

Conclusions

Most economic evaluations of antidiabetic medications have good reporting quality and use validated diabetes models. The newer antidiabetic medications in most of the reviewed studies were found to be cost effective, compared with insulin, TZDs, and sulfonylureas.

Appendices

Appendix 1: Details of the search strategies for the identification of studies

(1) (GLP-1) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

(2) (exenatide OR Byetta OR Bydureon OR liraglutide OR Victoza OR Saxenda OR lixisenatide OR Lyxumia OR albiglutide OR Tanzeum OR dulaglutide OR trulicity OR emaglutide OR Ozempic) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

(3) (DPP-4) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

(4) (Sitagliptin OR Januvia OR Vildagliptin OR Galvus OR Saxagliptin OR Onglyza OR Linagliptin OR Tradjenta OR Gemigliptin OR Zemiglo OR Anagliptin OR Suiny OR Teneligliptin OR Tenelia OR Alogliptin OR Nesina OR Vipidia OR Trelagliptin OR Zafatek OR Omarigliptin OR MK-3102 OR Evogliptin OR Suganon OR Gosogliptin OR SatRx) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

(5) (SGLT2) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

(6) (Dapagliflozin OR Farxiga OR Empagliflozin OR Jardiance OR Canagliflozin OR Invokana OR Ipragliflozin OR ASP-1941 OR Suglat OR Tofogliflozin OR Apleway OR Deberza OR Ertugliflozin OR Steglatro) AND (cost effectiveness OR cost utility OR economics OR value OR cost).

Appendix 2: Summary of sensitivity analyses in the included studies

References	Comparison	Probabilistic sensitivity analysis (Y/N)	Cost-effectiveness acceptability curve (Y/N)	Sensitive factors 1	Sensitive factors 2	Sensitive factors 3	Sensitive factors 4	Sensitive factors 5
Ray et al., 2007 [20]	Exenatide vs. insulin	N	Y	Exenatide price	HbA1c change	Insulin dose	Time horizon
Shaya et al., 2007 [21]	Exenatide vs. placebo; exenatide vs. no additional treatment	N	N	HbA1c change	Discount rate
Minshall et al., 2008 [22]	Exenatide vs. no additional treatment beyond metformin and/or a sulfonylurea	Y	Y	Time horizon	HbA1c change	BMI change	Lipid change
Schwarz et al., 2008 [88]	Sitagliptin vs. rosiglitazone; sitagliptin vs. sulfonylurea	N	N	Lipid change	SBP change	Congestive heart failure risk	Sitagliptin efficacy	SBP change
Woehl et al., 2008 [75]	Exenatide vs. insulin glargine in three scenarios	N	N	Hypoglycemic events	Coefficient of durability of sitagliptin
Brandle et al., 2009 [23]	Exenatide vs. insulin glargine	Y	Y	Time horizon	Discount rate	Cost of exenatide	Complication costs
Mittendorf et al., 2009 [24]	Exenatide vs. insulin glargine	Y	Y	Baseline BMI	Baseline HbA1c	Time horizon	HbA1c effect sustainability
Sullivan et al., 2009 [25]	Liraglutide vs. rosiglitazone	N	N	HbA1c change
Lee et al., 2010 [26]	Liraglutide vs. exenatide	N	N	Discount rate	Time horizon	HbA1c reduction	Disutility for BMI
Sinha et al., 2010 [89]	Exenatide vs. sitagliptin; Sitagliptin vs. glyburide	N	N	Disutility for injectable medication	Disutility for injectable medication, and cost of medication	Utility for weight change	Third-line therapy change
Beaudet et al., 2011 [27]	Exenatide vs. insulin glargine	Y	Y	Time horizon	Cost of complications	Discount rate	Exenatide price
Goodall et al., 2011 [28]	Exenatide vs. insulin glargine	Y	Y	Time horizon	HbA1c effect sustainability	Discount rate	HbA1c reduction
Klarenbach et al., 2011 [76]	DPP-4 vs. no additional treatment	N	Y	Clinical effects of HbA1c, weight, hypoglycemia	Treatment intensification	Price of DPP-4 inhibitors	Cost of complications
Lee et al., 2011 [29]	liraglutide vs. rosiglitazone	N	N	Discount rate	Time horizon	HbA1c reduction	(Dis)utility for BMI change
Valentine et al., 2011 [30]	Liraglutide vs. exenatide	N	Y	Time horizon	Discount rate	HbA1c and minor hypoglycemia benefits only	HbA1c reduction
Bergenheim et al., 2012 [56]	Saxagliptin vs. sulfonylurea	Y	N	Cost and disutility for weight gain and hypoglycemic event	Cost and disutility for hypoglycemic event	HbA1c effect sustainability, cost and disutility for hypoglycemic event	HbA1c effect sustainability, cost and disutility for weight gain and hypoglycemic event
Davies et al., 2012 [31]	Liraglutide vs. glimepiride; liraglutide vs. sitagliptin	N	Y	Treatment duration	Discount rate	Weight gain	BMI disutility
Erhardt et al., 2012 [57]	Saxagliptin vs. sulfonylurea	Y	Y	Model entry on second-line treatment	Weight change	BMI disutility	Discount rate
Gaebler et al., 2012 [87]	Exenatide vs. insulin	N	N	Time horizon
Gao et al., 2012 [77]	Liraglutide vs. glimepiride	N	N	Time horizon	Discount rate	Complication costs	Initial utility score: use UKPDS utilities
Granstrōm et al., 2012 [58]	Saxagliptin vs. sulfonylurea;	Y	Y	Hypoglycemia rates	Cost of hypoglycemia	BMI disutility	Sustainability of weight change
Grzeszczak et al., 2012 [59]	Saxagliptin vs. insulin	Y	Y	Model entry on second-line treatment	Baseline HbA1c	Weight change	Discount rate
Lee et al., 2012 [32]	Liraglutide vs. sitagliptin	Y	Y	Time horizon
Samyshkin et al., 2012 [33]	Exenatide vs. insulin glargine	Y	N	Time horizon	HbA1c reduction	BMI change	Convenience of less frequent administration of exenatide
Elgart et al., 2013 [60]	Saxagliptin vs. sulfonylureas	Y	Y	HbA1c values	age	Cost values	Utility values
Fonseca et al., 2013 [34]	Exenatide once weekly vs. exenatide twice daily; exenatide once weekly vs. insulin glargine	Y	Y	Time horizon	95% upper CI for HbA1c	(Dis)utility for BMI/nausea
Langer et al., 2013 [93]	Liraglutide vs. sitagliptin	Y	N	Discount rate
Raya et al., 2013 [35]	Liraglutide vs. sitagliptin	N	Y	Time horizon	Discount rates	Over- or underestimating the unit costs of diabetes complications
Teramachi et al., 2013 [102]	Sitagliptin 100 mg vs. sitagliptin 50 mg; vildagliptin vs. sitagliptin; alogliptin vs. sitagliptin	N	N	The standard deviation of HbA1c	Price of each drug
Brown et al., 2014 [36]	Sitagliptin vs. insulin glargine	Y	Y	Changes to the number of testing strips
DeKoven et al., 2014 [103]	Liraglutide vs. exenatide	N	N
Ektare et al., 2014 [94]	Canagliflozin vs. sitagliptin	Y	N	Drug costs	Treatment discontinuations	BMI	LDL-c	Discount rates
Kiadaliri et al., 2014 [80]	GLP-1 vs. DPP-4; GLP-1 vs. NPH insulin; DPP-4 vs. NPH insulin	Y	N	Drug costs	Treatment discontinuations	BMI	LDL-c	HbA1c reduction
Steen Carlsson et al., 2014 [81]	Liraglutide vs. sulfonylurea; liraglutide vs. sitagliptin	Y	Y	Frequency of hypoglycemia	Baseline BMI	HbA1c at which the second-line treatment was initiated		Cost of glucose monitoring strips
Van Haalen et al., 2014 [61]	Dapagliflozin vs. placebo	Y	Y	Uncertainty around the QALY point estimate	Price of test strips
Viriato et al., 2014 [78]	Vildagliptin vs. sulfonylurea	Y	Y	Coefficient of treatment failure	Drug costs	HbA1c	Discount rates	Risk of congestive heart failure
Charokopou et al., 2015 [62]	Dapagliflozin vs. DPP-4	Y	Y	HbA1c change from baseline	Weight change from baseline	BMI utility values	Total non-drug costs	Risk of congestive heart failure
Charokopou et al., 2015 [63]	Dapagliflozin vs. sulfonylurea	Y	Y	Change in HbA1c from baseline	Change in weight from baseline	Disutility for minor hypoglycemia
Deng et al., 2015 [64]	Exenatide vs. insulin glargine	Y	Y	HbA1c level at baseline	Health utilities decrement	BMI at baseline	Disutility for minor hypoglycemia
Gu et al., 2015 [65]	Saxagliptin vs. glimepiride	Y	Y	Utility	HbA1c	BMI	Age	Disutility for minor hypoglycemia
Huetson et al., 2015 [37]	Lixisenatide vs. basal-bolus insulin	Y	N	Discount rates	Reductions in HbA1c	Application of code-2 disutility for excess BMI		Disutility for minor hypoglycemia
Neslusan et al., 2015 [84]	Canagliflozin vs. sitagliptin	N	Y	QALY gains for canagliflozin	Time horizon	Disutility for overweight/obesity	Disutility for minor hypoglycemia
Pérez et al., 2015 [38]	Liraglutide vs. sitagliptin	N	N	Time horizon	Disutility for minor hypoglycemia
Sabale et al., 2015 [101]	Dapagliflozin vs. sulfonylurea	Y	Y	Assumptions on body weight progression over time	Utility weights associated with this change
Wang et al., 2015 [90]	Liraglutide vs. exenatide	Y	Y	the cost of liraglutide and exenatide	the liraglutide and exenatide AE rates	the cost of treating gastrointestinal AEs	the effectiveness of liraglutide and exenatide in patients who completed the full treatment duration
Bruhn et al., 2016 [54]	Albiglutide vs. insulin lispro; albiglutide vs. sitagliptin	Y	Y	Time horizon	Discount rate	Complication costs	Us ethnic breakdown	hypoglycemia disutility
Chuang et al., 2016 [66]	Exenatide vs. dulaglutide; exenatide vs. liraglutide; exenatide vs. lixisenatide	Y	Y	HbA1c reduction	Weight loss	Treatment discontinuation due to AE	Nausea as an AE
Davies et al., 2016 [55]	IDegLira vs. basal insulin + liraglutide; IDegLira vs. insulin glargine + IAsp; IDegLira vs. uptitrated insulin glargine	Y	Y	Alternative time horizon	Alternative discount rate	Alternative treatment switching	Alternative BMI progression	Hypoglycemia disutility
Gordon et al., 2016 [40]	Metformin + alogliptin vs. metformin + sulfonylurea	Y	Y	Time horizon	Costs	Utility	Discount rate	Age at model entry
Gordon et al., 2016 [67]	Exenatide vs. lispro	Y	Y	Time horizon	Discount rate	Complication utility	Complication costs
Gu et al., 2016 [68]	Dapagliflozin vs. acarbose	Y	Y	Utility	Age	HbA1c	Ethnicity
Gu et al., 2016 [69]	Saxagliptin vs. acarbose	Y	Y	Utility	Age	HbA1c	Ethnicity
Permsuwan et al., 2016 [17]	Saxagliptin vs. metformin; sitagliptin vs. metformin; vildagliptin vs. metformin; saxagliptin vs. sulfonylurea; sitagliptin vs. sulfonylurea; vildagliptin vssulfonylurea	Y	Y	HbA1c change of DPP-4 inhibitors	Discount rate	Risk of severe hypoglycemia of DPP-4 inhibitors	Cost of sexagliptin	HbA1c reduction
Roussel et al., 2016 [41]	Liraglutide vs. sitagliptin; liraglutide vs. glimepiride	Y	Y	Time horizon	Discount rate	Alternative costs of complications	HbA1c difference	Rule of treatment intensification
Sabapathy et al., 2016 [85]	Canagliflozin vs. sitagliptin	Y	Y	Discount rate	Time horizon	eGFR drift	Alternative treatment thresholds	Insulin dose
Tzanetakos et al., 2016 [70]	Dapagliflozin vs. sulfonylurea; dapagliflozin vs. DPP-4	Y	Y	the assumptions around the HbA1c switching threshold	the utility weights applied to BMI progression	utilities t2 dm complications	Discount rate	Insulin dose
Zhang et al., 2016 [71]	Liraglutide vs. exenatide	Y	Y	Discount rates	Simulated treatment period
Cazarim et al., 2017 [91]	Linagliptin vs. none; linagliptin + metformin vs. none; sitagliptin vs. none; sitagliptin + metformin vs. none; vildagliptin vs. none; vildagliptin + metformin vs. none; saxagliptin vs. none	N	N	NA
Dilla et al., 2017 [39]	dulaglutide vs. liraglutide	Y	Y	Treatment effects	Treatment duration	Economics	Disutility exclusions	Discount rate
Ericsson et al., 2017 [82]	IDegLira vs. insulin glargine; IDegLira vs. NPH insulin; IDegLira vs. insulin aspart + insulin glargine; IDegLira vs. insulin aspart + NPH insulin; IDegLira vs. liraglutide + insulin glargine; IDegLira vs. liraglutide + NPH insulin	Y	Y	Absolute treatment effect on the HbA1c level	Annual absolute drift of the HbA1c level, HbA1c	Initial absolute treatment effects on other biomarker levels	Absolute drift of the other biomarker levels
Gu et al., 2017 [72]	Exenatide vs. insulin glargine	Y	N	HbA1c threshold value for therapy	BMI-associated disutility	BMI-related prescription costs	Incidence, costs and disutility of the AEs
Hunt et al., 2017 [42]	Exenatide vs. insulin glargine	Y	Y	Time horizon	Upper and lower limits of HbA1c change	HbA1c progression	BMI progression
Hunt et al., 2017 [43]	Liraglutide vs. lixisenatide	Y	Y	Time horizon	Discount rates	HbA1c difference	Blood pressure difference	Lipid difference
Hunt et al., 2017 [95]	Liraglutide vs. lixisenatide	N	N	NA
Hunt et al., 2017 [44]	IDegLira vs. liraglutide	Y	N	No influential parameters identified could influence base-case results
Hunt et al., 2017 [45]	IDegLira vs. insulin glargine	Y	N	Time horizon	HbA1c difference abolished	Cost of neutral protamine Hagedorn
Hunt et al., 2017 [96]	IDegLira vs. insulin glargine	N	N	NA
Hunt et al., 2017 [46]	Liraglutide vs. lixisenatide	Y	N	Time horizon	HbA1c difference abolished
Hunt et al., 2017 [47]	Liraglutide vs. exenatide; liraglutide vs. lixisenatide	N	N	No influential parameters identified could influence the base-case results
Kousoulakou et al., 2017 [79]	Vildagliptin vs. glimepiride	N	N	No influential parameters identified could influence the base-case results
Kvapil et al., 2017 [48]	IDegLira vs. basal insulin	N	N	No influential parameters identified could influence the base-case results
Lasalvia et al., 2017 [97]	Dulaglutide vs. liraglutide; dulaglutide vs. glargine	Y	N	Non-daily injection utility	Monthly dulaglutide cost	Percentage of patients achieving < 7% HbA1c with glargine	Glucometries per month	Blood pressure
Mezquita-Raya et al., 2017 [49]	Liraglutide vs. lixisenatide	Y	Y	HbA1c difference abolished	Blood pressure
Psota et al., 2017 [50]	IDegLira vs. basal bolus	Y	Y	No influential parameters identified could influence the base-case results	Blood pressure
Shao et al., 2017 [73]	Dapagliflozin vs. glimepiride	Y	Y	No influential parameters identified could influence the base-case results
Vega-Hernandez et al., 2017 [51]	Liraglutide (dual therapy) vs. dapagliflozin (dual therapy); liraglutide (triple therapy) vs. dapagliflozin (triple therapy)	Y	Y	Treatment switch year	Time horizon
Basson et al., 2018 [52]	Dulaglutide vs. exenatide	Y	Y	No influential parameters identified could influence the base-case results
Chakravarty et al., 2018 [98]	Dapagliflozin vs. DPP-4; dapagliflozin vs. GLP-1	Y	Y	No influential parameters identified could influence the base-case results
Drummond et al., 2018 [92]	IDegLira vs. basal-bolus	N	N	NA
Ericsson et al., 2018 [83]	Liraglutide vs. lixisenatide	Y	Y	Between-treatment difference in HbA1c abolished
Gourzoulidis et al., 2018 [99]	Empagliflozin vs. standard of care (background glucose-lowering therapy)	Y	Y	Time horizon
Ishii et al., 2018 [53]	Dulaglutide vs. insulin glargine	Y	Y	No influential parameters identified could influence the base-case results
Neslusan et al., 2018 [86]	Canagliflozin vs. dapagliflozin	Y	Y	No influential parameters identified could influence the base-case results
Nguyen et al., 2018 [100]	Empagliflozin vs. standard treatment	Y	N	Cost of empagliflozin	Death rate of standard treatment
Tzanetakos et al., 2018 [74]	Exenatide vs. insulin glargine; exenatide vs. liraglutide	Y	Y	No influential parameters identified could influence the base-case results

1. HbA1c hemoglobin A1c, BMI body mass index, eGFR estimated glomerular filtration rate, SBP systolic blood pressure, CI confidence interval, Y yes, N no, DPP-4 dipeptidyl peptidase-4, GLP-1 glucagon-like peptide-1, UKPDS UK Prospective Diabetes Study, QALY quality-adjusted life-year, LDL-c low-density lipoprotein-cholesterol, AE adverse event, NA not available