Clinical trials with high-risk medical devices frequently involve historical controls and other criteria as concurrent control groups in such studies are not feasible or ethical. A survey of about 100 high-risk medical devices approved by the FDA over five years shows some interesting trends in using non-concurrent controls in medical device trials.
Between 2019 and 2023, about two-thirds of the premarket approval (PMA) applications submitted to the FDA for high-risk therapeutic medical devices contained pivotal studies that employed historical controls, objective performance criteria (OPC), or performance goals (PG) instead of traditional concurrent control groups. This trend reflects regulatory flexibility for life-saving technologies where randomized control trials (RCTs) may be impractical, unethical, or infeasible.
The utility of nonconcurrent controls cannot be overstated. In therapeutic areas such as cardiology and neurology, often characterized by high disease burden and limited treatment alternatives, nonconcurrent designs allow regulators and developers to assess device performance using real-world benchmarks or historical comparators. For instance, historical controls leverage legacy data from prior cohorts, OPCs draw from validated thresholds across registries and trials, and PGs define clinically acceptable safety or efficacy endpoints. These practices expedite transformative technologies to meet urgent patient needs.
Among the 52 (out of 88) device approvals relying on nonconcurrent control strategies, the evidence demonstrated exceptional performance: 100% of historical control-based analyses and 98.4% of PG-based analyses met their primary endpoints. While only a minority of these studies explicitly documented justification for omitting concurrent controls, the FDA’s least burdensome provision, mandating the minimum necessary evidence to ensure safety and effectiveness, supports such adaptive approaches. Particularly for breakthrough-designated, implantable, or life-sustaining devices, these alternative methodologies open the door to faster approvals, real-world relevance, and broader patient access.
The authors of the survey raised some concerns regarding the justification of the protocol, demographic matching between historical and current cohorts, and FDA pre-validation of PGs and OPCs, based on the publicly available information. Additionally, the authors contended that up-to-date comparator data may be critical to properly evaluate the nonconcurrent controls used in such studies, as many analyses used in the approved PMAs referenced studies over a decade old. Greater transparency, such as disclosing data provenance and comparator patient characteristics in Summary of Safety and Effectiveness Data (SSEDs), would enhance the scientific rigor behind these approvals.
Importantly, the study does not critique the use of nonconcurrent controls per se, but rather highlights their immense potential when deployed responsibly. As precision medicine and complex device technologies evolve, nonconcurrent control strategies will remain a cornerstone of regulatory science, especially when traditional trial designs lag behind clinical innovation. To ensure these tools fulfill their promise, the FDA, sponsors, and scientific community must collaborate to establish standardized, evidence-based benchmarks that are both dynamic and reflective of contemporary care.
In conclusion, this study affirms the importance of nonconcurrent controls as a strategic asset in clinical trials with high-risk medical devices using innovative new technologies. The FDA agrees and encourages such approaches as they preserve patient safety while providing a clinically meaningful evaluation of high-risk medical technologies.
