Hidden bias lurks in survival analysis when researchers ignore time patterns

A new simulation study reveals that survival analysis—a statistical method critical for predicting patient outcomes, product reliability, and treatment efficacy—can produce misleading results when researchers fail to account for multiple time-scales in their data. For companies and institutions relying on these analyses to make clinical, regulatory, or business decisions, the finding signals a potential source of systematic error that could affect millions of predictions.

Originaltitel: Potential bias introduced by not including multiple time-scales in survival analysis: a simulation study