Customize your JAMA Network experience by selecting one or more topics from the list below.
Prior reports on complications associated with cochlear implantation have been, for the most part, generated by individual implant centers or manufacturers. Complication rates reported in these studies may reflect specific surgical practices or devices. In this issue of the ARCHIVES, Tambyraja et al1 review cochlear implant complications as reported to the Manufacturer User Facility and Distributor Experience (MAUDE) database. This database is maintained by the Center for Devices and Radiological Health, an arm of the US Food and Drug Administration (FDA). The MAUDE database contains reports of adverse events associated with medical devices. Tambyraja et al assessed the utility of the MAUDE database in reflecting a broader experience of complications related to implantation.
The authors conclude that spontaneous device failure is the most common complication of cochlear implantation. However, it seems likely that the MAUDE database is skewed for adverse sequelae related to device failure. Complications that can be managed medically (ie, cellulitis and otitis media) or those related to surgical technique (ie, tympanic membrane perforation, facial paresis, and loss of taste sensation) are less likely to be reported to the manufacturer or to the FDA. Thus the conclusion that spontaneous device failure comprises the greatest number of adverse events seems inevitable, given present reporting procedures.
Whereas implant centers can report the number of adverse events out of the total number of implants performed, the total number of implanted patients from which the MAUDE reports are derived is unknown. Thus, there is no denominator for determining the frequency of various complications. The authors report a statistically significant decrease in spontaneous device failure and an increase in infections when comparing pre-1998 data to data from the year 2002. The denominator used to make the comparison was the total number of adverse events reported to MAUDE during each period. Presenting the number of device failures as a percentage of adverse sequelae reported is not particularly useful because, as the authors note, increased reporting of other complications could dilute this percentage. This is especially true for the year 2002, when increased awareness of the risk of meningitis would likely have lead to increased reporting of infectious complications. Furthermore, only 2 time points were used in drawing the conclusion that device failures have decreased. The authors argue that the addition of more time points would not enhance the analysis, given the low quality of the data points. However, this begs the question of whether the data points that were used were of sufficient quality to support the conclusions.
Although the conclusions that Tambyraja et al were able to draw from querying the MAUDE database have questionable impact, their efforts are nonetheless significant in that they reveal the weaknesses inherent in the database as it is presently structured. The establishment of an exhaustive and uniform national database, which pools the experience of many centers, would certainly be of value. If MAUDE is to become that entity, it needs to be tailored more specifically to cochlear implants. The authors have initiated a discussion of potential improvements that will enhance MAUDE’s utility. However, the database will have limited impact until information regarding the total number of implants from which these complications derive is more accessible, consensus is established regarding the nature of reportable events, and clear guidelines are established regarding the party responsible for initiating the report.
Correspondence: Dr Raz, Department of Otolaryngology, University of Pittsburgh School of Medicine, 200 Lothrop St, Suite 500, Pittsburgh, PA 15213 (email@example.com).
Financial Disclosure: None.
Raz Y. The Utility of the MAUDE Database in Researching Cochlear Implantation Complications. Arch Otolaryngol Head Neck Surg. 2005;131(3):251. doi:10.1001/archotol.131.3.251
Browse and subscribe to JAMA Network podcasts!
Create a personal account or sign in to: