Integrated AHP-Delphi and Mini-HTA for selection of metabolic drugs in China's national medical insurance catalog: a framework towards quantitative dynamic adjustment.

BACKGROUND: China's high burden of metabolic disease requires more systematic approaches to selecting medicines for inclusion in the National Medical Insurance Catalog. Existing processes face difficulties in quantitatively balancing clinical value, target population needs, pharmacoeconomic evidence, and resource constraints, and may therefore rely heavily on subjective judgment. METHODS: We developed a decision framework integrating the analytic hierarchy process (AHP), the Delphi method, and mini-health technology assessment (mini-HTA) principles. Candidate criteria were identified from the WHO Essential Medicines List, India's National List of Essential Medicines, the Kenya Essential Medicines List, China's National Medical Insurance Catalog, and relevant policy-oriented literature on drug selection in China. These criteria were used to construct a systematic and locally adapted framework. Three rounds of consultation were conducted with a multidisciplinary expert panel comprising one clinician, three pharmacists, and three health insurance administrators. Indicator weights were calculated using the geometric mean method. Expert agreement and judgment consistency were assessed using Kendall's coefficient of concordance (W) and consistency ratios (CR), respectively. RESULTS: The framework included 14 primary indicators and 42 secondary indicators. Expert consensus was statistically significant (Kendall's W = 0.57, p < 0.001), and all consistency ratios were below 0.1, indicating acceptable consistency and robustness in weight allocation. CONCLUSIONS: The proposed AHP-Delphi-mini-HTA framework provides a quantitative and transparent approach to selecting metabolic drugs for insurance coverage in China. Although prospective pilot testing is still required, the framework supports standardized assessment, reduces reliance on subjective judgment, and may facilitate future integration of real-world evidence and AI-assisted decision-making.