Association between Regulatory Review Times and Adverse Event Reports in New Cardiovascular Devices
Association between Regulatory Review Times and Adverse Event Reports in New Cardiovascular Devices
Monday, June 13, 2016: 8:50 AM
G17 (Claudia Cohen Hall)
Importance: The Food and Drug Administration (FDA) approves new high-risk medical devices after in-depth reviews of safety and effectiveness data gathered in clinical trials. Some have advocated for shorter review times with the hope of encouraging innovation.
Objective: To evaluate whether regulatory review time and product novelty serve as predictors of adverse events involving high-risk cardiovascular devices.
Design: We linked all new 106 high-risk cardiovascular devices approved between 2000-2009, grouped by FDA-assigned product codes, to time spent under regulatory review and all adverse events reported in the FDA’s Manufacturer and User Facility Device Experience (MAUDE) database over the four years following FDA approval. Adverse events were classified as either involving patient injury/death or not. Linear probability models and logistic regression models evaluated whether review times or product novelty predicted adverse event reporting.
Results: Regulatory review times averaged 14.6 months among approved devices (standard deviation: 9.77). 48 devices in 28 product codes (45% of approvals) had at least one reported adverse event (mean: 182, range: 1-2706), yielding a total of 8,721 adverse event reports. Longer review times were associated with a lower probability of any subsequent adverse event report, as well as a lower probability of reports of patient injury/death: a one standard deviation increase in review time was associated with a 13.7 percentage point (30.4%) decrease in the absolute risk of any adverse event reporting (95%CI: 5.9-20.5, p<0.001), a 9.8 percentage point (39.1%) decrease in the absolute risk of events involving patient injury or death (95%CI: 1.0-18.6, p=0.02), and a 5.9 percentage point (23.4%) decrease in the probability of being in the top quartile of the count of total adverse event reports (95%CI: 0.0-11.7, p=0.04). Controlling for review time, novel (first-in-class) products were not more likely to have any adverse events reported, nor were they more likely to have a high count of reports.
Conclusions and Relevance: Among approved high-risk cardiovascular devices, products with longer periods of FDA review were associated with a lower probability of reported adverse events overall, a lower probability of events involving serious injury or death, and a lower probability of a high count of total adverse event reports. These data should inform current proposals aimed at expediting the review of novel high-risk devices.
Objective: To evaluate whether regulatory review time and product novelty serve as predictors of adverse events involving high-risk cardiovascular devices.
Design: We linked all new 106 high-risk cardiovascular devices approved between 2000-2009, grouped by FDA-assigned product codes, to time spent under regulatory review and all adverse events reported in the FDA’s Manufacturer and User Facility Device Experience (MAUDE) database over the four years following FDA approval. Adverse events were classified as either involving patient injury/death or not. Linear probability models and logistic regression models evaluated whether review times or product novelty predicted adverse event reporting.
Results: Regulatory review times averaged 14.6 months among approved devices (standard deviation: 9.77). 48 devices in 28 product codes (45% of approvals) had at least one reported adverse event (mean: 182, range: 1-2706), yielding a total of 8,721 adverse event reports. Longer review times were associated with a lower probability of any subsequent adverse event report, as well as a lower probability of reports of patient injury/death: a one standard deviation increase in review time was associated with a 13.7 percentage point (30.4%) decrease in the absolute risk of any adverse event reporting (95%CI: 5.9-20.5, p<0.001), a 9.8 percentage point (39.1%) decrease in the absolute risk of events involving patient injury or death (95%CI: 1.0-18.6, p=0.02), and a 5.9 percentage point (23.4%) decrease in the probability of being in the top quartile of the count of total adverse event reports (95%CI: 0.0-11.7, p=0.04). Controlling for review time, novel (first-in-class) products were not more likely to have any adverse events reported, nor were they more likely to have a high count of reports.
Conclusions and Relevance: Among approved high-risk cardiovascular devices, products with longer periods of FDA review were associated with a lower probability of reported adverse events overall, a lower probability of events involving serious injury or death, and a lower probability of a high count of total adverse event reports. These data should inform current proposals aimed at expediting the review of novel high-risk devices.