By the end of 1995, the FDA had received 98 405 adverse event reports on silicone gel breast implants (this does not include reports for saline breast implants) . Many of these were reported to FDA after the 1992 Panel meeting described above (Figure 27.2). It is not unusual for media coverage to increase awareness and reporting of adverse event reports. Between 1984 and 1991, there were a total of 3479 reports for silicone gel-filled breast implants. In 1992 alone, there were 27 130 reports entered into the database; this rose to 33 982 reports entered in 1994. Thus, there was a > 2000% increase in manufacturers' reports on silicone gel-filled breast implants from 1991 to 1992. The large increase in reports was attributed to multiple reasons, including: reporting on individual breasts (e.g. if women had two ruptured implants, each rupture was reported as a separate report); multiple reports of the same adverse event when multiple sources reported the same adverse event (e.g. if the same event was reported to multiple manufacturers because the reporter was uncertain of which brand had been implanted); press coverage of breast implants; and growing litigation over breast implants (i.e. reports of alleged injuries). Because of the large number of reports of well-known problems, some manufacturers of implants were allowed to submit adverse event reports as quarterly summaries (ASR) rather than individual reports, starting in 1995 .
Event Reports 15000
Figure 27.2 Adverse event reports for silicone gel-filled breast implants by year, received 1984-1995. Because of a backlog on entering reports, not all reports from 1995 were shown in this report. Additionally, in 1995, manufacturers started reporting through alternative summary reports and these were not available at that time
Because various terms were used to describe implant-related adverse events, the FDA combined terms that appeared to be related to rupture. Silicone gel-filled breast implant rupture or leaking was reported in about 26 661 reports (Table 27.2), accounting for 27% of reports overall. FDA considered this to be the most frequently reported adverse event. The second most commonly reported problem was 'reaction', accounting for 26 106 (26%) reports. The term 'reaction' was used to code the reported adverse event when the description of the event(s) was non-specific. Reaction was defined as 'adverse effect, irritation, or swelling'.
Adverse event reports for medical devices are coded in two different ways, the problem with the device (e.g. implant rupture) and the problem or consequence to the patient
Table 27.2 Implant reports for ruptured or disrupted implants from DEN and MAUDE databases, 1984-1995 
Coded problem report (n)
Burst 16 237
Leak 7 030
Deflate 1 436
Tear, rip, or hole in device 522
Total 26 661
1988 1990 Year
(e.g. additional surgery to remove a ruptured implant). Because historically FDA focused on 'device problems', when illnesses in women were reported to FDA, the list of available codes was inadequate to allow an unambiguous coding of the adverse event. In addition, many of the reports were complex and involved multiple adverse events that had occurred over a period of years, or even decades. Because of this, the term 'reaction' was used by manufacturers to code a wide variety of adverse events, making it very difficult to identify the real problems. New terms were added to the patient problem coding manual, for instance 'capsular contracture', because of the many reports of capsular contracture experienced by women with breast implants. The difficulty in summarizing these reports without actually reading each one makes a poignant argument for adopting and using a standardized system of nomenclature to code adverse events along with detailed training for standardizing their use. Many of these reports were also coded as 'nonspecific' (23 542 reports, 24%) and had no information on the adverse event other than to say that some unspecified problem with silicone gel-filled breast implants was reported.
Was this article helpful?