Design and validation of a data simulation model for longitudinal healthcare data

Area of Science:

• Health Informatics
• Biostatistics
• Real-World Data Analysis

Background:

• Evaluating analytical methods for treatment effects in longitudinal healthcare data lacks objective benchmarks.
• Real-world data has complex relationships not precisely quantified, limiting method development.
• Existing simulated data often fails to capture real-world complexities.

Purpose of the Study:

• To develop and evaluate a model for simulating longitudinal healthcare data that accurately reflects real-world complexities.
• To provide a benchmark for evaluating analytical methods in healthcare research.
• To improve the development of methods for identifying treatment effects.

Main Methods:

• An empiric design was employed, using characteristics from a real healthcare database as input for simulation.
• The model was developed to capture complex relationships among diseases and treatments.
• Performance of the simulation model was evaluated against real-world data characteristics.

Main Results:

• The developed model successfully simulates longitudinal healthcare data with known characteristics.
• The simulated data adequately reflects complex relationships among diseases and treatments.
• The model demonstrates the potential for creating realistic benchmarks for method evaluation.

Conclusions:

• Simulated longitudinal healthcare data, when based on real-world characteristics, can effectively mirror complex patient data.
• This approach offers a viable solution for developing and validating analytical methods for treatment effect identification.
• The model advances the capability to create objective benchmarks for real-world evidence studies.