The FDA launched the Sentinel Initiative with the aim of creating a framework for monitoring the effectiveness and safety of medical products. Monitoring newly marketed products after approval is challenging because the products are not used in randomized controlled settings. Conducting a post-market active surveillance using the exposure and outcome data shortly after approval is particularly needed. However, monitoring and detecting any potential exposure and outcome association is mostly difficult because of (i) fewer expose patients in the early months/years of marketing the product; (ii) rare outcomes (Adverse Events) due to the nature of the outcomes or less exposure; (iii) patients exposed in the early phase of marketing likely differ from those exposed later; (iv) lack of randomization. Confounding may undermine the validity of the emergent treatment comparisons and safety assessments. Therefore, an appropriate adjustment for confounding is critical. Central to the proposed approaches are the use of Propensity Scores (PS) and Disease Risk Scores (DRS). The PS and DRS are used to reduce a large number of covariates into a summary score variable. The summary scores are either used to match users of the newly marketed product with comparators or to stratify users of the new product and comparators into similar groupings. Many such studies evaluate exposure and outcome association in a sequential testing framework. In this framework, two estimation methods, stepwise and cumulative are usually used to build both the PS and DRS models. The stepwise estimation uses the population of current year’s users whereas the cumulative uses the population from the beginning of surveillance to the current year. The objective of this round table is to discuss the validity of these proposed methods.