Strength of evidence used to approve CV devices questioned

The FDA gave the green light to nearly two-thirds of cardiovascular devices approved between January 2000 and December 2007 on the basis of a single study, results of a new analysis reveal.

Researchers from the University of California, San Francisco reviewed 123 publicly-available summaries of safety and effectiveness data for 78 high-risk CV devices that were granted premarket approval during this time period. The researchers found that only 27% of the 123 studies that were used to support approval were randomized, and just 14% were blinded.

Primary endpoints were identified in 14% of these studies, data indicated, and 52% of the 213 documented primary endpoints were compared with controls. Furthermore, 31% of the 111 controls were retrospective.

“Premarket approval of cardiovascular devicesby the FDA is often based on studies that lack adequate strength and may be prone to bias,” the researchers wrote.

The researchers also took issue with the use of surrogate endpoints (88%), stating that they may not be reliable predictors of patient benefits.

“Although surrogate outcomes are attractive because they decrease the time and costs required to do a study they must be linked to a meaningful endpoint to be valid,” according to the analysis.

Overall, the researchers stated that 15% of primary endpoints were noninterpretable, mainly due to a lack of a target goal for device performance (75%). “In some instances, end points were interpreted to meet their targets when they may have met only part of them,” the researchers wrote.

“Although there has been recent scrutiny of evidence used in the U.S. FDA drug approval process, less attention has been paid to the approval process for medical devices,” they wrote.

The researchers suggested several possibilities that may account for less rigorous device-approval standards. These include less experience approving devices, increases in the number and complexity of devices up for approval, and a device’s unique position on “the FDA continuum” between drugs that require “strict criteria for approval” and surgical operations that do not require approval.

Because insurance companies often extend coverage to include a given product once the FDA approves it, products are often rapidly disseminated after approval and off label use is common, the researchers called for more rigorous research to support CV device approval. “[T]he bar for evidence of benefit should be higher for devices because they are implanted and cannot simply be discontinued, as drugs can.”

Although the researchers noted some information in the available data may be incomplete, which may limit the study findings, they emphasized that improved access to complete FDA reviews is necessary. “[Summaries of safety and effectiveness data] are the sole basis […] used for systemic review and guideline development.”

Dhruva SS. JAMA.2009;24:2679-2685.