Sensitivity analysis for studies transporting prediction models.

Abstract

We consider estimation of measures of model performance in a target population when covariate and outcome data are available from a source population and covariate data, but not outcome data, are available from the target population. In this setting, identification of measures of model performance is possible under an untestable assumption that the outcome and population (source or target) are independent conditional on covariates. In practice, this assumption is uncertain and, in some cases, controversial. Therefore, sensitivity analysis may be useful for examining the impact of assumption violations on inferences about model performance. Here, we propose an exponential tilt sensitivity analysis model and develop statistical methods to determine how measures of model performance are affected by violations of the assumption of conditional independence between outcome and population. We provide identification results and estimators for the risk in the target population under the sensitivity analysis model, examine the large-sample properties of the estimators, and apply them to data on lung cancer screening.