• John T Schousboe, Adrienne Landsteiner, Tyler Drake, Shahnaz Sultan, Lisa Langsetmo, Anjum Kaka, Maylen Anthony, Charles J Billington, Caleb Kalinowski, Kristen Ullman, and Timothy J Wilt.
• Park Nicollet Clinic and HealthPartners Institute, Bloomington, and Division of Health Policy and Management, School of Public Health, University of Minnesota, Minneapolis, Minnesota (J.T.S.).
• Ann. Intern. Med. 2024 May 1; 177 (5): 633642633-642.

BackgroundIn the United States, costs of antidiabetes medications exceed $327 billion.PurposeTo systematically review cost-effectiveness analyses (CEAs) of newer antidiabetes medications for type 2 diabetes.Data SourcesBibliographic databases from 1 January 2010 through 13 July 2023, limited to English.Study SelectionNonindustry-funded CEAs, done from a U.S. perspective that estimated cost per quality-adjusted life-year (QALY) gained for newer antidiabetic medications. Two reviewers screened the literature; disagreements were resolved with a third reviewer.Data ExtractionCost-effectiveness analyses were reviewed for treatment comparisons, model inputs, and outcomes. Risk of bias (RoB) of the CEAs was assessed using Drummond criteria and certainty of evidence (CoE) was assessed using GRADE (Grading of Recommendations Assessment, Development, and Evaluations). Certainty of evidence was determined using cost per QALY thresholds predetermined by the American College of Physicians Clinical Guidelines Committee; low (>$150 000), intermediate ($50 to $150 000), or high (<$50 000) value per QALY compared with the alternative.Data SynthesisNine CEAs were eligible (2 low, 1 high, and 6 some concerns RoB), evaluating glucagon-like peptide-1 agonists (GLP1a), dipeptidyl peptidase-4 inhibitors (DPP4i), sodium-glucose cotransporter-2 inhibitors (SGLT2i), glucose-dependent insulinotropic peptide agonist (GIP/GLP1a), and insulin. Comparators were metformin, sulfonylureas, neutral protamine Hagedorn (NPH) insulin, and others. Compared with metformin, GLP1a and SGLT2i are low value as first-line therapy (high CoE) but may be of intermediate value when added to metformin or background therapy compared with adding nothing (low CoE). Insulin analogues may be similarly effective but more expensive than NPH insulin (low CoE). The GIP/GLP1a value is uncertain (insufficient CoE).LimitationsCost-effectiveness analyses varied in methodological approach, assumptions, and drug comparisons. Risk of bias and GRADE method for CEAs are not well established.ConclusionGlucagon-like peptide-1 agonists and SGLT2i are of low value as first-line therapy but may be of intermediate value when added to metformin or other background therapy compared with adding nothing. Other drugs and comparisons are of low or uncertain value. Results are sensitive to drug effectiveness and cost assumptions.Primary Funding SourceAmerican College of Physicians. (PROSPERO: CRD42022382315).

25 keywords...

hide…