@article {Bagepallye001020, author = {Bhavani Shankara Bagepally and Usa Chaikledkaew and Yogesh Krishnarao Gurav and Thunyarat Anothaisintawee and Sitaporn Youngkong and Nathorn Chaiyakunapruk and Mark McEvoy and John Attia and Ammarin Thakkinstian}, title = {Glucagon-like peptide 1 agonists for treatment of patients with type 2 diabetes who fail metformin monotherapy: systematic review and meta-analysis of economic evaluation studies}, volume = {8}, number = {1}, elocation-id = {e001020}, year = {2020}, doi = {10.1136/bmjdrc-2019-001020}, publisher = {BMJ Specialist Journals}, abstract = {Objectives To conduct a systematic review and meta-analysis and to pool the incremental net benefits (INBs) of glucagon-like peptide 1 (GLP1) compared with other therapies in type 2 diabetes mellitus (T2DM) after metformin monotherapy failure.Research design and methods The study design is a systematic review and meta-analysis. We searched MEDLINE (via PubMed), Scopus and Tufts Registry for eligible cost{\textendash}utility studies up to June 2018, adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guideline. We conducted a systematic review and pooled the INBs of GLP1s compared with other therapies in T2DM after metformin monotherapy failure. Various monetary units were converted to purchasing power parity, adjusted to 2017 US$. The INBs were calculated and then pooled across studies, stratified by level of country income; a random-effects model was used if heterogeneity was present, and a fixed-effects model if it was absent. Heterogeneity was assessed using Q test and I2 statistic.Results A total of 56 studies were eligible, mainly from high-income countries (HICs). The pooled INBs of GLP1s compared with dipeptidyl peptidase-4 inhibitor (DPP4i) (n=10), sulfonylureas (n=6), thiazolidinedione (TZD) (n=3), and insulin (n=23) from HICs were US$4012.21 (95\% CI US$-571.43 to US$8595.84, I2=0\%), US$3857.34 (95\% CI US$-7293.93 to US$15 008.61, I2=45.9\%), US$37 577.74 (95\% CI US$-649.02 to US$75 804.50, I2=92.4\%) and US$14 062.42 (95\% CI US$8168.69 to US$19 956.15, I2=86.4\%), respectively. GLP1s were statistically significantly cost-effective compared with insulins, but not compared with DPP4i, sulfonylureas, and TZDs. Among GLP1s, liraglutide was more cost-effective compared with lixisenatide, but not compared with exenatide, with corresponding pooled INBs of US$4555.09 (95\% CI US$3992.60 to US$5117.59, I2=0) and US$728.46 (95\% CI US$-1436.14 to US$2893.07, I2=0), respectively.Conclusion GLP1 agonists are a cost-effective choice compared with insulins, but not compared with DPP4i, sulfonylureas and TZDs.PROSPERO registration number CRD42018105193.}, URL = {https://drc.bmj.com/content/8/1/e001020}, eprint = {https://drc.bmj.com/content/8/1/e001020.full.pdf}, journal = {BMJ Open Diabetes Research and Care} }