An Adverse Drug Reaction (ADR) is defined by the World Health Organization as “a noxious response to a medication that is unintended at doses usually administered for diagnosis, prophylaxis, or treatment.” Estimates suggest that such episodes – in which prescription drugs cause negative health consequences – account for more than 2 million hospitalizations and more than 100,000 deaths per year in the United States alone, making ADRs one of the leading causes of death. To put these numbers into perspective: death from treatment with prescription drugs is about 10 times as common as death from suicide. This dissertation aims to understand why these numbers are so high.
Prior work has focused mainly on the politics of drug approval to show that factors such as deadlines, status of pharmaceutical firms, and foreign approval can account for variation in regulatory decision making by the Food and Drug Administration. I take another route and focus on the production of evidence about the safety of prescription drugs. The way in which medical scientists have typically used evidence is by extracting meaning through aggregation or classification of pieces of evidence. The argument that I am making in this dissertation is that rather than aggregating or classifying evidence, one needs to account for the relationships between pieces of evidence. In particular, the dissertation shows how social theories about the structures of evidence production can be used to better understand the harm that drugs can do and, as a result, allow us to identify unsafe drugs more rapidly.
The dissertation presents analyses based on data from the two main sources of evidence that the Food and Drug Administration has at its disposal to identify unsafe drugs. The first is the Adverse Event Reporting System (AERS). AERS is an FDA maintained system through which patients and physicians can voluntarily report ADRs to the FDA. The FDA uses this system by monitoring disproportional increases in the number of ADRs reported for a given drug. The second source of evidence is the scientific literature about prescription drugs. The FDA uses this literature to inform regulatory action.
The first set of findings in this dissertation demonstrate that ADR reports for a specific drug are more likely to be submitted if a drug has been publicly scrutinized or when a drug treats the same health condition as a drug that was publicly scrutinized. Patients and physicians differ in the ways in which their reporting behavior changes in response to increased scrutiny. Preliminary findings suggest that these episodes of changes in reporting behavior are associated with delays in regulatory action compared to drugs in which reporting behavior did not change. These findings are consistent with the hypothesis that the detection of signals in massive yet sparse data benefits from social theories of evidence production.
The second set of findings show that the social structure in which scientific evidence about the safety and efficacy of prescription drugs is not uniformly cumulative. In particular, in some cases the scientific debate about the safety and efficacy of prescription drugs is characterized by a disconnect between the claims made before a drug is approved for marketing and the claims made after approval. Moreover, the results from the study demonstrate that debates characterized by a strong disconnect are more likely to be the target of regulatory action. This suggests that a discontinuity in scientific closure is consistent with the idea that the quality of pre-approval scientific evidence predicts post-approval regulatory action.
In sum, this dissertation identifies salient structures in collective production processes and it demonstrates that the structure of collective production reveals meaning that could reduce ambiguity in interpretation.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8QV3M0R |
| Date | January 2015 |
| Creators | de Vaan, Mathijs |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0112 seconds