Comparison Of Benefit-Risk Assessment Methods For Prospective Monitoring Of Newly Marketed Drugs: A Simulation Study

Abstract

A11at market entry of each drug of interest and using a sequential propensity score matched cohort design. We applied four BRA metrics: number needed to treat and number needed to harm (NNT|NNH); incremental net benefit (INB) with maximum acceptable risk [MAR], INB with relative-value adjusted life years [RVALYs], and INB with quality-adjusted life years [QALYs]. We determined whether and when the bootstrapped 99% confidence interval (CI) for each metric excluded zero, indicating net favorability of one drug over the other. Results: For rofecoxib, all four metrics yielded a negative value, suggesting net favorability of ns-NSAIDs over rofecoxib, and the 99% CI for all but the NNT|NNH excluded the null during follow-up. For prasugrel, only the 99% CI for INB-QALY excluded the null, but trends in values over time were similar across the four metrics, suggesting overall net favorability of prasugrel versus clopidogrel. The 99% CI for INB-RVALY and INB-QALY excluded the null in the denosumab example, suggesting net favorability of denosumab over bisphosphonates. ConClusions: Prospective benefit-risk monitoring can be used to determine net favorability of a new drug in electronic healthcare data. In three examples, existing BRA metrics produced qualitatively similar results but differed with respect to alert generation. INB-QALY produced the most conclusive findings across the three examples.

Prospective Benefit-Risk Monitoring of New Drugs for Rapid Assessment of Net Favorability in Electronic Health Care Data

“…With each new data set we also continued to follow patients who remained under observation at the end of the previous period's data set according to the censoring rules described above. Additional details about the prospective monitoring framework can be found elsewhere [10][11][12]14,21].…”

“…We used these cumulative outcome counts to calculate four BRA metrics at each data update. On the basis of previous theoretical and simulation work [21,22], we used the following BRA metrics: 1) number needed to treat and number needed to harm (NNT| NNH); 2) incremental net benefit (INB) with maximum acceptable risk (MAR) weights; 3) INB with relative-value-adjusted life-year (RVALY) weights; and 4) INB with quality-adjusted life-year (QALY) weights. These metrics can accommodate multiple beneficial and harmful outcomes and produce a single numeric value indicating net favorability and are therefore amenable to quantitative prospective BRA monitoring.…”

“…Indeed, each can be formulated using a unified BRA formula in which weights derived from different methods are used [22]. Specific details on these metrics and their performance in simulated data can be found elsewhere [21,22].…”

Sequential surveillance for drug safety in a regulatory environment

“…We consider newly marketed drugs that will be monitored for a limited set of prespecified safety outcomes. 1,2 Our focus is on implementation of sequential surveillance in electronic health data, with consideration of the additional challenges posed by a distributed database environment. Our practical experience is with 2 US programs, the Food and Drug Administration (FDA) Sentinel Initiative and the Centers for Disease Control and Prevention (CDC) Vaccine Safety Datalink (VSD).…”

“…The purpose of this paper is to help regulatory scientists develop and refine recommendations about which drug safety questions are most amenable to sequential surveillance. We consider newly marketed drugs that will be monitored for a limited set of prespecified safety outcomes . Our focus is on implementation of sequential surveillance in electronic health data, with consideration of the additional challenges posed by a distributed database environment.…”