Use of data mining at the Food and Drug Administration.

OBJECTIVES: This article summarizes past and current data mining activities at the United States Food and Drug Administration (FDA). TARGET AUDIENCE: We address data miners in all sectors, anyone interested in the safety of products regulated by the FDA (predominantly medical products, food, veterinary products and nutrition, and tobacco products), and those interested in FDA activities. SCOPE: Topics include routine and developmental data mining activities, short descriptions of mined FDA data, advantages and challenges of data mining at the FDA, and future directions of data mining at the FDA.

Sex-based differences in outcomes after percutaneous coronary intervention for acute myocardial infarction: a report from TRANSLATE-ACS.

BACKGROUND: Data regarding sex-based outcomes after percutaneous coronary intervention (PCI) for myocardial infarction are mixed. We sought to examine whether sex differences in outcomes exist in contemporary practice. METHODS AND RESULTS: We examined acute myocardial infarction patients undergoing PCI between April 2010 and October 2012 at 210 US hospitals participating in the Treatment with Adenosine Diphosphate Receptor Inhibitors: Longitudinal Assessment of Treatment Patterns and Events after Acute Coronary Syndrome (TRANSLATE-ACS) observational study. Outcomes included 1-year risk of major adverse cardiac events and bleeding according to Global Utilization of Strategies To Open Occluded Arteries (GUSTO) and Bleeding Academic Research Consortium (BARC) definitions. Among 6218 patients, 27.5% (n=1712) were female. Compared with men, women were older, had more comorbidities, and had lower functional status. Use of multivessel PCI and drug-eluting stents was similar between sexes, while women received less prasugrel. Unadjusted cumulative incidence of 1-year major adverse cardiac events was higher for women than for men (15.7% versus 13.6%, P=0.02), but female sex was no longer associated with higher incidence of major adverse cardiac events after multivariable adjustment (hazard ratio 0.98, 95% CI 0.83 to 1.15). Female sex was associated with higher risks of post-PCI GUSTO bleeding (9.1% versus 5.7%, P<0.0001) and postdischarge BARC bleeding (39.6% versus 27.9%, P<0.0001). Differences persisted after adjustment (GUSTO: hazard ratio 1.32, 95% CI 1.06 to 1.64; BARC: incidence rate ratio 1.42, 95% CI 1.27 to 1.56). CONCLUSIONS: Female and male myocardial infarction patients undergoing PCI differ regarding demographic, clinical, and treatment profiles. These differences appear to explain the higher observed major adverse cardiac event rate but not higher adjusted bleeding risk for women versus men.

Disproportionality analysis for signal detection of implantable cardioverter-defibrillator-related adverse events in the Food and Drug Administration Medical Device Reporting System.

BACKGROUND: The Food and Drug Administration (FDA) became aware of lead fracture and inappropriate shock events related to Sprint Fidelis leads in January 2007. The manufacturer announced a voluntary market withdrawal in October 2007. AIM: Our aim was to retrospectively evaluate this safety signal using disproportionality analysis to estimate whether disproportionality analysis could have detected this particular safety signal earlier than actually occurred. MATERIALS AND METHODS: The Manufacturer and User Facility Device Experience (MAUDE) database contains reports on device-related adverse events, of which, FDA receives several hundred thousand every year. For each manufacturer, a list of the top lead brand names was ranked by frequency of reports. We used the Multi-item Gamma Poisson Shrinker (MGPS) method for analysis. We isolated 11 top-reported implantable cardioverter defibrillator (ICD) lead brand names. Using MGPS methodology, we calculated the one-sided 95% lower confidence bound EB05 on the empirical Bayes geometric mean of the reporting ratio. RESULTS: We performed individual MGPS analysis for each of the top reported adverse events in 2006 for ICD leads. Fidelis had the highest EB05 scores for lead fractures and inappropriate shock. DISCUSSION: Through disproportionality analysis of the MAUDE database, we were able to identify known safety signals associated with the Medtronic Sprint Fidelis lead. CONCLUSION: If utilized at the time, this disproportionality analysis would have identified signals earlier for lead fractures, oversensing, high impedance, and inappropriate shock.