(BSPOT08) Enhancing Early Detection of Medical Device Safety Signals: A Case Study of Implantable Cardioverter Defibrillator Leads

Background: Real-world evidence (RWE) offers a vital avenue to expedite the fulfillment of the Food and Drug Administration (FDA) post-market requirements and informs medical device coverage decisions, benefiting patients and fostering innovation. The case study of a defibrillator lead failure highlights RWE's role in early safety signal detection.

Objectives: This study aims to quantify the time saved in detecting safety signals for defibrillation leads by utilizing data from the National Cardiovascular Disease Registry (NCDR).

Methods: In this retrospective cohort analysis, we compared the time taken to detect a safety signal under two scenarios: one utilizing real-world data from the NCDR (counterfactual) and the other reflecting the actual time taken for signal detection, leading to the voluntary recall of the device 36 months post-approval. The counterfactual scenario involved estimating the time required to identify the same signal if the NCDR Implantable Cardioverter Defibrillator (ICD) registry had been prospectively monitored for this safety signal using active surveillance. Three simulation scenarios ("conservative," "likely," and "optimistic") were developed to model outcomes based on varying proportions of timely reported ICD lead implants and different lead failure rates. The Data Extraction and Longitudinal Analysis (DELTA) active surveillance system monitored accruing data from the simulated ICD Registry, calculating risk-adjusted adverse event rates reported by each center.

Results: Detection of a significantly increased failure rate for Fidelis leads, compared to other high-energy ICD leads, would have occurred within six months of monitoring. Considering the delay between clinical event occurrence and data submission, DELTA would have alerted the FDA and manufacturer to the higher failure rate within 11 months post-launch, saving over 790 days as compared to what actually occurred. Approximately 200,000 patients could have avoided exposure to the defective device if active surveillance had been employed. Even in the most conservative scenario, DELTA would have alerted regulators and manufacturers approximately 700 days post-launch, significantly earlier than the FDA's eventual awareness of the issue.

Conclusions: The relationship between the time-to-signal metric and time-to-recall is nuanced. While the shorter time required to signal detection with full NCDR utilization aligns with documented cases leading to recalls, not all signals lead to regulatory action. RWE enhances early safety detection, crucial for patient safety and regulatory decision-making.