Chinese Pharmacoeconomic Evaluation Toolkit

The problem

Pharmacoeconomic evaluation often splits into several moving parts: identifying direct medical, direct non-medical, and indirect costs; measuring survival, QALYs, or monetized benefits; selecting decision-tree, Markov, discrete-event simulation, or partitioned survival models; and producing a defensible report. Manual parameter tables, ICER, QALY, and CEAC calculations can miss discounting, sensitivity analysis, and source attribution. This skill organizes the workflow around China's pharmacoeconomic evaluation guidance.

How it works

• Evaluation type: choose CEA, CUA, CBA, or CMA based on the effect measure.
• Parameter management: classify perspective, time horizon, the 4.5% discount rate, costs, utilities, transition probabilities, and simulation settings, with cited sources.
• Modeling: use decision trees for short, single decisions; Markov models for chronic disease progression; DES for individual pathways and resource constraints; partitioned survival models for oncology survival curves.
• Computation: use cost_effectiveness_analysis.py for calculate_icere(), calculate_qaly(), one-way sensitivity, and tornado-plot data; use monte_carlo_simulation.py for PSA, CEAC, and VOI.
• Reporting: compare ICER with China's willingness-to-pay thresholds and report baseline results, uncertainty, limitations, and conflicts of interest per CHEERS.

Boundaries

Best for health intervention cost-effectiveness analysis and decision-model computation, not clinical decision-making or reimbursement policy decisions. Intangible costs are generally not monetized directly but handled through utility analysis; utility values should preferentially use Chinese population data; costs should use actual or standardized charges rather than reimbursement-only prices. If key parameters, distributions, or thresholds are missing, outputs should be treated as structured drafts.