Plenary III-02: Propensity Score-based Sensitivity Analysis Method for Uncontrolled Confounding

Abstract

Background: Uncontrolled confounding remains a major concern for comparative effectiveness and safety results obtained from analyzing observational studies, especially those using existing healthcare databases. Sensitivity analyses are of paramount importance and usefulness in assessing the effect of possible uncontrolled confounding on the estimates of the parameter of interest to adequately evaluate the validity of the corresponding results and inference. Methods: We develop a sensitivity analysis method to address the issue of uncontrolled confounding in observational studies. In this method, we quantify the hidden bias due to uncontrolled confounding using a one-dimensional sensitivity function (SF), which depends on the propensity score only. Propensity score is defined as the conditional probability of being assigned to a selected treatment given the measured confounders. The new method nicely reduces the dimension of the sensitivity function, and makes it much easier to impose reasonable assumptions on the sensitivity functional forms and the values of coefficients. In addition, it offers opportunities for robust inference since one-dimensional continuous functions can be well approximated by low order polynomial structures (e.g., linear, quadratic). Therefore our proposed approach is especially useful when limited information is available on the type and mechanism of unmeasured confounders. We construct SF-corrected inverse probability weighted estimators to draw inference on the causal treatment effect. We demonstrate the use of the new method by implementing it to an asthma study which evaluates the effect of clinician prescription pattern about the use of inhaled corticosteroids (ICs) for children with persistent asthma on selected clinical outcomes. Results: Without considering uncontrolled confounding, daily-use pattern of ICs appears to be harmful than the periodic-use pattern with an increased risk for uncontrolled asthma during the 12-month follow-up period (odds ratio 2.4, 95% CI, 1.7, 3.9). This seemingly harmful effect of daily-use pattern diminishes as the assumed magnitude of hidden bias increases. Nevertheless, within the considered plausible range, the data does not provide evidence suggesting a significant beneficial effect of the daily-use pattern compared to the periodic-use pattern. Conclusions: Our proposed approach provides a practical tool to conduct sensitivity analysis for uncontrolled confounding in observational studies.