Improving Drug Safety

Abstract

Spontaneous reporting systems (SRS) as a form of post-market drug-safety passive surveillance were created several decades ago. SRS do have several limitations. Somewhat common but serious events will not necessarily prompt the filing of a report. Reporting is voluntary, leading to under-reporting of adverse events, and submitted reports are often lacking in detail. Post-market active drug-safety surveillance is now emerging as a realistic and powerful complement to SRS. Recent examples of active surveillance systems are presented in this paper; these have some, but not all, the elements of an ideal active surveillance programme.

An ideal programme should allow rigorous confirmation of suspected risks and have adequate scope for timely and validated signal detection of emerging adverse events. It should allow for both hypothesisgenerating and hypothesis-confirming investigations and be based on a well defined population including not only those with drug exposures but also those who are unexposed. In terms of analysis, epidemiological methods would be utilized that accurately control for confounding-by-indication, and statistical techniques would be used that account for multiple testing among several potential outcomes assessed at different points in time. All of the necessary information should be contained in readily accessible electronic medical record databases. There also must be patient data privacy protections that conform to law and ethical guidelines. However, there is much work to be done before active surveillance becomes the paramount tool in post-market pharmacovigilance.

Among recent recommendations from US drug-safety experts were the establishment of the following: a distributed data network from the public and private sectors; working definitions of what constitutes a drug-safety signal; screening algorithms and criteria and strategies to confirm or refute signals; guidelines regarding communications to physicians and patients about safety signals; improved methods to identify and evaluate potential safety signals; and a sustainable and knowledgeable workforce to conduct the studies and to understand how to interpret the results. Some of these recommendations were incorporated in the recently announced US FDA Sentinel Initiative. However, stable and sufficient funding and demand from healthcare consumers must be strong to support long-term efforts in this regard. Public confidence must be high to support widespread drug-safety active surveillance. Patient data privacy protections must be maintained by government and private research staff. There should also be strong incentives for health insurance companies, hospitals, pharmacies and the pharmaceutical industry to fully support widespread and effective active drug-safety surveillance.