The Food and Drug Administration (FDA) is responsible for communicating prescription drug risks to providers and the public, and the Sentinel Initiative will substantially increase the number of safety signals that the FDA must evaluate. Two of the most important strategies that the FDA uses to communicate these risks are MedWatch Safety Alerts and Public Health Advisories, yet there is limited information regarding their effect. Most research to date has examined the impact of select warnings on the aggregate prescribing rates of drugs such as acetaminophen, droperidol, cisapride, and more recently, selective serotonin reuptake inhibitors. However, many of these studies have been survey based, and even those using administrative data to examine medication use have included a limited assessment of provider, patient, and firm response. This is important because assessing the ultimate impact of an FDA warning is a complex task, since it may: (1) have an impact that varies based on clinical setting and provider characteristics, (2) alter demand among non- targeted patient populations, (3) cause increases in the use of therapeutic substitutes, (4) lead to changes in the use of office-visits and other non-prescription utilization, and (5) lead to alterations in the strategic behavior of pharmaceutical firms, such as in marketing and promotion, which may exert strong independent effects on physician prescribing and patient medication use. Prior work examining FDA warnings leaves numerous other important questions unanswered, including their impact on adherence and therapeutic substitution, the degree to which there is significant physician variability in the response to such communications, and the strategic steps that firms may take when an alert or advisory occurs. We propose to use longitudinal Medicaid and MarketScan private health plan claims as well as audits of office-based prescribing to examine the impact of FDA warnings on the use of antipsychotics, antidepressants, and treatments for attention deficit/hyperactivity disorder (ADHD). In addition to examining changes in prescription and other health care use among non- elderly adults, we will also explore such changes among the children and the elderly, two subpopulations of particular relevance and concern. Finally, we will examine how firms modify their marketing and promotional strategies to providers and consumers in response to these communications. Our analyses, which will be rigorously conducted using advanced econometric techniques, will yield important information for policymakers and other stakeholders committed to ensuring the effective dissemination of safety information regarding these and other widely used classes of prescription drugs.
The FDA implements safety alerts and public health advisories to address emergent safety concerns for prescription drugs. Yet it is poorly understood how these warnings affect provider, patient, and firm behaviors, and hence, how effective these strategies are in minimizing risk. We will analyze a variety of complementary national databases using econometric methods to evaluate the effectiveness of these strategies and to identify factors that can improve risk communication, which will contribute to safeguarding the public health while maximizing clinical innovation.
