Grant: Operating Grant - PA: Drug Safety and Effectiveness (bridge funding)

Meta-analysis of randomized controlled trials (RCTs) is widely accepted by researchers, clinicians and decision-makers as a gold standard for evaluating effectiveness of medical interventions. However, meta-analysis of RCTs when looking at harms outcomes experienced by patients can be problematic due to limitations of RCT data; RCTs often aren't large enough to identify differences in risk of rare harms seen with different treatments, they are often poorly reported, and they often enroll participants who are healthier than patients seen by physicians in practice. As a result, safety data used by decision-makers and clinicians may not be optimal from the perspective of patients. Comparative effectiveness research increasingly involves a need to compare many treatments simultaneously, and traditional methods of combining data across RCTs does not enable this. Thus, the use of an advanced statistical technique called network meta-analysis (NMA) has risen quickly in recent years. So far far NMA has mainly been used to study effectiveness based on RCT data. To compare the safety of three or more treatments, NMA adding observational data is a logical next step that may overcome limitations of RCT data. They often involve active treatment groups, enroll large sample sizes, and patients are more representative of those seen by clinicians each day. Observational research is more common than any other type, and ignoring this information may be wasteful. Combining data from RCTs and observational studies is a new methodology to assess safety, and its benefits require study. This proposal outlines an approach to empirically examine the value of this approach in the context of answering several important drug safety questions chosen in collaboration with physicians. We plan to report clinical findings from our systematic reviews, and we also plan to summarize relevant methodologic findings and recommendations for future work regarding this approach to analyzing safety data.