Cost effectiveness analysis of six non insulin agents for type 2 diabetes in Iran using a patient level simulation...

Type 2 diabetes mellitus (T2DM) is a global healthcare challenge, requiring cost-effective treatments to improve patient outcomes and resource use. This study assesses the cost-effectiveness of six non-insulin agents (NIAs)-pioglitazone, metformin, dapagliflozin, sitagliptin, empagliflozin, and liraglutide for T2DM management in Iran. Using the UKPDS Outcomes Model 2 (UKPDS-OM2), a 40-year simulation from the healthcare system perspective analyzed data from 440 Iranian patients in the Diabetes Care (DiaCare) study, evaluating direct medical costs and quality-adjusted life years (QALYs). Costs were locally sourced, while treatment effects and complication disutility values came from international evidence. Incremental cost-effectiveness ratios (ICERs) were compared to a threshold of one-time GDP per capita, with uncertainty evaluated via 100 bootstrap iterations. Subgroup analyses were conducted based on sex, age (< 55 vs. ≥55 years), and diabetes duration (< 10 vs. ≥10 years), using the same simulation framework and assumptions. Pioglitazone proved least costly at $1,376.20 PPP with 8.47 QALYs, dominating sitagliptin and empagliflozin, and outperforming metformin, dapagliflozin, and liraglutide, which showed higher costs and limited QALY gains. Subgroup analyses showed that pioglitazone was cost-effective in men, younger patients, and those with longer diabetes duration, while metformin was preferred in women, older patients, and those with shorter disease duration. Bootstrap results showed pioglitazone cost-effective in 52% of iterations and metformin in 46%, with ICERs of $10,554.8-$56,857.13 PPP per QALY. Pioglitazone emerged as the cost-effective option, with metformin as a viable alternative in Iran's resource-constrained setting. Subgroup analyses highlighted the value of individualized treatment strategies, with pioglitazone and metformin demonstrating variable cost-effectiveness across sex, age, and diabetes duration groups.