Validation & Evaluation

Validation

For validation, HADES mostly relies on an extensive set of unit tests executed as part of its continuous integration. This assures validity on a continuous basis, at every release. A good reference would be our recent HADES paper:

• Schuemie M, Reps J, Black A, et al. Health-Analytics Data to Evidence Suite (HADES): Open-Source Software for Observational Research Studies in health technology and informatics. 2024 Jan 25:310:966-970.

Evaluation

The population-level estimation methods have been extensively evaluated using the OHDSI Methods Benchmark. The results are published in:

• Schuemie MJ, Cepeda MS, Suchard MA, et al. How Confident Are We About Observational Findings in Health Care: A Benchmark Study Harvard Data Science Review, 2(1).

The results of the evaluation can be reviewed in an interactive web application.