A recent audit of the Food and Drug Administration’s 510(k) program by the Center for Devices and Radiological Health (CDRH) has found that inadequate device performance data submissions from sponsors make up a substantial majority--80%--of Not Substantially Equivalent (NSE) determinations made by the agency between 2005 and 2010.

The analysis covered more than 700 NSE determinations made during the course of that six-year period. NSE determinations were based either on failure to provide appropriate performance data to support substantial equivalence (SE) determinations, or to demonstrate equivalent performance compared to predicate devices.

The remaining 20% of NSE determinations between 2005 and 2010 were associated with FDA requests for additional data that had not been made for predicate devices. These requests stemmed from different indications, new technologies or new safety issues associated with devices under review.

A chart comparing percentages of SE, NSE and undefined “other” decisions from 2001 to 2011, however, shows a relatively flat (and low) trend line in terms of actual numbers of NSE determinations over that time period. Only 3.5% of 510(k) submissions earned NSE determinations between 2001 and 2010; that percentage increased to 8% in 2010, but has so far fallen to 5% for 2011. Agency report authors argue that although NSE determinations remain small in number compared to the overall volume of 510(k) applications reviewed, these determinations take up an inordinate amount of resources and should be minimized.

Conducting a more granular analysis of 109 510(k) submissions in 2009, investigators found that NSE determinations stemmed from the following issues:

• For 88% of cases determined to be NSE, the FDA had to conduct additional review cycles because sponsors did not adequately address the agency’s initial questions, or because sponsors’ responses raised additional safety concerns
• For 18% of submissions, the agency requested data that had not been required for those devices’ predicates, due to issues such as safety concerns arising from pre-clinical testing results
• In 16% of cases, data from sponsors indicated their devices performed more poorly than their predicates