Panel Paper:
How Does FDA Utilize Drug Information Acquired through Postmarketing Studies?
*Names in bold indicate Presenter
This is an important topic because PMR/PMC is becoming norm and very costly. A postmarketing study is a study of a drug after FDA approved it. FDA can require a firm to conduct such studies in certain situations (PMR) or studies might be commissioned because a sponsor and FDA agreed (PMC). Since the 2007 FDA Modernization Act, the practice of PMR/PMCs became norm rather than exception. The number of PMR/PMCs has been increasing and they became larger and more complex thus more costly. A deeper understanding of the regulatory effects of PMR/PMCs could provide useful information to regulators how to tame but not destroy such practice or how to enforce firms to comply their PMR/PMCs. Currently, the FDA lacks much information and analysis about their practice of PMR/PMCs and how to improve its effectiveness and efficiency.
In this paper, I will assess how well FDA disseminates the newly acquired information. FDA disseminates and communicates safety and efficacy information by changing labels. By looking at label changes and languages it uses when PMR/PMC study results are provided, I can test to what degree FDA utilizes additional information acquired from PMR/PMCs. I will use the following datasets: (1) random sample drugs (300) in oncology and cardiovascular agents approved by FDA since 1990; (2) PMR/PMC data associated with those drugs; (3) publications data through PubMed (NIH research database); (4) FDA regulatory action approval database; and (5) FDA Advisory Committee meeting log data.
Hypothesis1: When comparing safety-related PMR/PMC, a negative study result (more risk) is likely to increase the possibility of label changes(warnings).
Hypothesis2: Strength of a study result is associated with the strength of a regulatory action. Studies showing larger risk are more likely to result in stronger warning in their labels compared with studies showing smaller risk.
Hypothesis3: Strength of a study result is associated with the speed of a regulatory action. The effect size of a study is positively associated with the speed of regulatory action. Studies showing larger risk are more likely to result in faster label changes.
My preliminary analysis employing negative binomial models (from 261 sample drugs in oncology, cardiovascular, and central nervous system agents) shows that, on average, a standard deviation increase in PMR/PMC (1.6 studies) increases the expected rate of warnings by 0.184 (p-value <0.001), controlling for the quality/length of studies, drug class, firm characteristics, and safety signals. Since this shows only that the number of PMR/PMCs are likely to have more warnings, I will test strength of warnings, label change contents, and time to warning with the PMR/PMC study result (how surprising it is compared to expectation), the difficulty of approval decisions in the first place, the quality and length of PMR/PMC study and other variables to control.
