Metformin is a safe and effective glucose-lowering agent, widely regarded as the first-line treatment option for type 2 diabetes (T2D), and available at very low cost. Recent updates to the American Diabetes Association (ADA) Standards of Care guidance demonstrate a shift towards the earlier use of newer, more expensive, anti-diabetic agents for the treatment of T2D, based on concurrent conditions such as cardiovascular and/or renal complication reduction, and significant weight loss. As of 2026, this has culminated in sodium-glucose cotransporter 2 inhibitors (SGLT2is) being recommended for patients with T2D and established atherosclerosis cardiovascular disease (ASCVD) or indicators of high ASCVD, heart failure or chronic kidney disease, independently of the use of metformin. The main evidence base supporting the use of SGLT2is in these patient populations come from cardiovascular and renal outcome trials, which recruited patients usually with long-standing T2D and high levels of cardiovascular and/or renal comorbidities. Whether the cardiovascular and renal benefits shown during these trials are generalisable to the early T2D or even pre-diabetic population may be debated. Ultimately, this leaves the question as to whether patients with early T2D and no cardiovascular and/or renal complications should receive treatment with SGLT2is, at a higher cost, where clinical evidence of superiority to established therapies like metformin are scarce. As T2D is a chronic, progressive disease, often advancing through pre-diabetes, early T2D and eventually treatment intensification, the cost-effectiveness of treatments at each stage of the disease must also be considered to understand the long-term economic burden. This review explores the available cost-effectiveness evidence, supported by clinical data, for the use of metformin and SGLT2is across the complete T2D continuum.
Campbell et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: