The Research on Adverse Drug Events and Reports (RADAR) evaluation of the quality of reports to the FDA’s MedWatch system for oncology related adverse drug reactions (ADRs): Extensive improvements are needed

Abstract

6028 Background: MedWatch, the Adverse Event Reporting System for the FDA has been the main source of drug safety information since its introduction in 1993. It is especially important for cancer drugs, which have a low therapeutic index. Herein, we evaluated the quality of oncology-associated ADR reports contained in MedWatch and two other data sources. Methods: For new cancer drugs, we evaluated: 1) the completeness of ADR reports from MedWatch, clinical trials, and the published literature, 2) frequency with which ADRs submitted to MedWatch were subsequently retrievable from external queries. Mean frequency of reporting of > 8 ADR-related clinical variables for each ADR such as drug dose & schedule, clinical & laboratory findings, days to ADR onset, and outcomes were calculated. Results: For three oncology-related ADRs, mean completeness scores were best for ADRs described in published case reports (Lit), intermediate for ADRs which occurred in clinical trials (RCT), and poorest for ADRs which occurred outside of the clinical trial setting and were submitted to MedWatch (Table). Several ADR reports which described bisphosphonate-associated osteonecrosis and gemtuzumab-associated veno-occlusive disease were submitted by physicians to the FDA, but were not retrievable when RADAR investigators queried the MedWatch database. Conclusions: ADR reports for cancer drugs contained in the MedWatch database are frequently of poor quality, and in some instances, are not retrievable. Policy-makers should consider establishing an independent pharmaceutical safety monitoring board. Moreover, policy-makers should also consider establishing an on-line source for MedWatch reports, similar to the on-line database established by the FDA for devices. No significant financial relationships to disclose.