Evidence Strength in FDA Premarket Approval of Cardiovascular Devices

To the Editor: In their systematic review, Dr Dhruva and colleagues¹ examined studies supporting premarket approval (PMA) of cardiovascular devices by the US Food and Drug Administration (FDA). They concluded that the FDA often uses studies that lack adequate strength as the basis for device approvals.

We believe that the premise of the review—that the normative standard for ascertaining device safety and effectiveness should be multiple, randomized, double-blinded, prospectively controlled clinical trials that are compelling in themselves—greatly oversimplifies the realities of medical device development and regulation. That premise emulates the FDA paradigm for new drugs,² concluding that “criteria on which FDA device approval is based appear to be less rigorous than those for drug approvals.”¹

Unlike drugs, devices have a broad range of human exposure, from momentary contact to permanent implantation. Regulatory mechanisms reflect this difference. Preclinical data and experience for mature device technologies (such as prosthetic heart valves) can form robust standards and objective performance criteria for similar new devices seeking the same indication.³ By insisting on prospective, contemporaneous controls, the authors ignore these important relationships.

The evidence used by regulators should represent best available science. The systematic review spanned 2000 to 2007, a period when FDA was formalizing its approach to existing evidence. A guidance for Bayesian PMA designs was not released until 2006.⁴ Bound by Congress to pursue least burdensome approaches, regulators must endorse rigorous yet efficient methods to demonstrate reasonable assurance of safety and effectiveness. Inefficient approaches can unduly delay US access to devices otherwise available worldwide, which itself presents a safety concern. With many devices, shrinking adverse event rates (eg, 0.5% for drug-eluting stent thrombosis⁵ ) render the conventional pharmaceutical paradigm impractical.

The review isolated the statistical dimension from the more complete and complex decision making that engenders reasonable assurance of safety and effectiveness and protects public health. How much risk, benefit, and uncertainty are reasonable for approvable devices is a challenging calculus that defies such constructs.