Abstract 11403: Medical Device Recalls in the U.S

Abstract

Background. Recently the Medical Device Approval Process in the U.S. has been questioned and the expedited 5,10(K) process felt to be a poor testing process being a source of increased recalls.

Methods. To investigate the approval process and number of device recalls, a study was initiated to review all 5,10(K) and PMA (pre market applications) from 2005 to 2011. The number of 5,10(K) approvals and recalls were obtained. Additionally, the time from approval to recall, the commercial period (CP) was determined for 5,10(K) and PMA devices. It was postulated that the CP would be longer for better studied products undergoing PMA testing. Fischer’s exact tests and t tests were used to determine significance.

Results. Between 2005 and 2011 a change in the proportion of recalls was noted. From 2005 to 2009, 0.65% 5,10(K) and 0.47% PMA devices were recalled (p=0.32). From 2010 to 2011, 1.39% 5,10(K) and 0.30% PMA devices were recalled (p<0.0001). The CP 2005-2009 was 30±23 months PMA and 40±46 months 5,10(K) (p=0.45). The CP in 2010-2011 was 70±49 PMA and 34±44 5,10(K) (p=0.113). Contrary to initial expectation, PMA devices were observed to have a shorter time to recall than 5,10(K) devices in 2005 to 2009 while in 2010-2011 CP was shorter for 5,10(K) though the difference was not significant. Among recalled devices 70% were class II devices, 23% class III devices, and 7% class I devices. In 2010 and 2011, cardiovascular (CV) recalled devices were 54% class III and 44% class II. From 2005-2009, 0.52% PMA CV devices were recalled and 1.52% 5,10(K) devices were recalled (p=0.0058) as compared to 2010-2011; 0.53% PMA and 3.13% 5,10(K) (p<0.0006). CP 2005 to 2009 was 30 ±15 months with PMA and 31±34 with 5,10(K) (p=0.96) as compared to 2010-11 PMA 64±58 and 5,10(K) 26±34 (p=0.109).

Conclusion. These data show that recalls were proportionally the same for 5,10(K) and PMA devices and the commercial period was longer for 5,10(K) than PMA devices before recall in 2005-2009. After this period, 5,10(K) percent recalls dramatically increased. CV devices showed an earlier change in number of recalls for 5,10(K) approved devices. Further study is needed to determine the reasons for the increasing number of recalls over time with 5,10(K) devices.