Modeling hard clinical end-point data in economic analyses.

Area of Science:

• Health economics
• Clinical trial analysis
• Pharmacoeconomics

Background:

• Increasing availability of hard clinical end-point data (e.g., cardiovascular events in type 2 diabetes) fuels interest in their use for economic analyses.
• Evaluating the suitability of existing modeling approaches for health economic models is crucial.

Purpose of the Study:

• To investigate published methods for modeling hard end-points from clinical trials.
• To assess the applicability of these methods in health economic models with diverse disease characteristics.

Main Methods:

• Systematic review of cost-effectiveness models in cardiovascular diseases, cancer, and chronic lower respiratory diseases.
• Inclusion of studies using hard end-point data from randomized clinical trials.
• Summarization and evaluation of clinical input characteristics and modeling approaches.

Main Results:

• 33 articles analyzed (23 CV, 8 cancer, 2 respiratory) utilized decision trees, Markov models, discrete event simulations, or hybrid approaches.
• Event rates were modeled as constant, time-dependent, or patient-specific.
• Time/patient-dependent risks were employed for major event rates >1%/year in simpler models (<7 health states); constant rates were preferred for infrequent events or complex models.

Conclusions:

• Model complexity and event rate modeling should align with clinical event frequency and characteristics.
• Simplified modeling is suitable for low-risk events.
• Detailed approaches (e.g., individual simulations, time-dependent rates) are better for common, high-risk events in populations like those with cardiovascular disease.