Low rate of infected knee replacements in a nationwide series--is it an underestimate?

Abstract

Background and purpose: Specialist hospitals have reported an incidence of early deep infections of < 1% following primary knee replacement. The purpose of this study was to estimate the infection rate in a nationwide series using register-based data.

Methods: The Finnish Arthroplasty Register (FAR) was searched for primary unicompartmental, total, and revision knee arthroplasties performed in 1997 through 2003 and eventual revision arthroplasties. The FAR data on revision arthroplasties was supplemented by a search of the national Hospital Discharge Register (HDR) for debridements, partial and total revision knee replacements, resection arthroplasties, arthrodeses, and amputations.

Results: During the first postoperative year, 0.33% (95% CI: 0.13-0.84), 0.52% (0.45-0.60) and 1.91% (1.40-2.61) of the primary UKAs, primary TKAs, and revision TKAs, respectively, were reoperated due to infection. The 1-year rate of reoperations due to infection remained constant in all arthroplasty groups over the observation period. The overall infection rate calculated using FAR data only was 0.77% (95% CI: 0.69-0.86), which was lower, but was not, however, statistically significantly different from the overall infection rate calculated using endpoint data combined from FAR and HDR records (0.89%; 95% CI: 0.80-0.99). FAR registered revision arthroplasties and patellar resurfacing arthroplasties reliably but missed a considerable proportion of other reoperations.

Interpretation: More reoperations performed due to infection can be expected as the numbers of knee arthroplasties increase, since there has been no improvement in the early infection rate. Finnish Arthroplasty Register data appear to underestimate the incidence of reoperations performed due to infection.

Figure 1.

Flow chart showing the index operations, endpoint events, and the corresponding data sources.

Figure 2.

Flow chart indicating how the data from the Finnish Arthroplasty Register (FAR) and the Hospital Discharge Register (HDR) were combined to create the final dataset. ^aoperation types according to FAR.

Figure 3.

Prosthesis survival after primary unicompartmental arthroplasty (UKA), total knee arthroplasty (TKA), and for aseptic and septic revision TKA with any reoperation (including debridement, resection arthroplasty, partial or total revision arthroplasty, arthrodesis, or amputation) due to infection as endpoint. The data on the operations under follow-up were collected from the Finnish Arthroplasty Register (FAR). Endpoint data were collected from FAR and from the Hospital Discharge Register.