Propensity Scores in Pharmacoepidemiology: Beyond the Horizon

Abstract

Purpose of review Propensity score methods have become commonplace in pharmacoepidemiology over the past decade. Their adoption has confronted formidable obstacles that arise from pharmacoepidemiology's reliance on large healthcare databases of considerable heterogeneity and complexity. These include identifying clinically meaningful samples, defining treatment comparisons, and measuring covariates in ways that respect sound epidemiologic study design. Additional complexities involve correctly modeling treatment decisions in the face of variation in healthcare practice, and dealing with missing information and unmeasured confounding. In this review, we examine the application of propensity score methods in pharmacoepidemiology with particular attention to these and other issues, with an eye towards standards of practice, recent methodological advances, and opportunities for future progress. Recent findings Propensity score methods have matured in ways that can advance comparative effectiveness and safety research in pharmacoepidemiology. These include natural extensions for categorical treatments, matching algorithms that can optimize sample size given design constraints, weighting estimators that asymptotically target matched and overlap samples, and the incorporation of machine learning to aid in covariate selection and model building. Summary These recent and encouraging advances should be further evaluated through simulation and empirical studies, but nonetheless represent a bright path ahead for the observational study of treatment benefits and harms.

Comparative effectiveness of concurrent chemoradiotherapy versus EGFR‐tyrosine kinase inhibitors for the treatment of clinical stage IIIb lung adenocarcinoma patients with mutant EGFR

“…In the first and second AA, we used the propensity score (PS) method, as advocated in the literature, to balance the measured potential confounders to examine the robustness of our findings . We used the covariables in the PS model via various methods (traditional logistic regression models, as well as machine learning, such as neural network [NN] or random forest [RF]) to estimate the possible PS value . We then evaluated the covariate balance via standardized difference, as suggested by several review papers .…”

Adjusted restricted mean survival times in observational studies

“…A limitation of adjusting Kaplan-Meier curves with inverse probability weights or propensity scores is the potential for very large or small weights. 42 Extreme weights typically occur when there is a rare patient profile that is frequent in the adjusting population; a large weight will be assigned to this subgroup. In consequence, the survival curve will have a large variance.…”

Outcomes of Localized Esophageal Squamous Cell Carcinoma Patients Treated With Definitive Concurrent Chemoradiotherapy Using Either Standard or High Radiotherapy Dose: A Retrospective Study Controlling for Organ at Risk Dose

“…The propensity score method was used as advocated in the literature to balance the measured potential confounders (23,24). The above covariables were used in the PS model construction using various methods to estimate the possible PS value as suggested in the literature (25)(26)(27)(28) and then PS matching was performed (PSM, 1:1 paired matching). Finally, the balance of covariate was assessed via standardized difference (SDif) as suggested in several review papers (25,29,30).…”