Abstract
Aim
National Joint Replacement Registries, which are important sources for periprosthetic joint infection (PJI) data, report an average PJI incidence ranging from 0.5 to 2.0%. Unfortunately, national registries including the Dutch Arthroplasty Register (LROI), are not specifically designed to register PJI. In the Netherlands, the LROI is a nationwide population-based registry with an overall completeness of more than 95%.3 To ensure usability and reliability of PJI data from the LROI, it is important to evaluate the quality and completeness of these data. From 2013 onwards, eight hospitals in the South-East of the Netherlands, collected their PJI data in a detailed regional infection cohort (RIC), specifically designed for this purpose. This study aimed to determine the accuracy and completeness of PJI registration (hip and knee arthroplasty) in the LROI, by comparing the LROI with the RIC.
Method
All patients registered with an acute PJI in the RIC between 2014–2018 were selected for the study and were matched with the LROI. According to the Workgroup of American Musculoskeletal Infections Society (MSIS), an acute PJI was defined as at least two phenotypically identical pathogens, isolated in cultures from at least two separate tissues, obtained from the affected peri-prosthetic tissue during the DAIR treatment (debridement, antibiotics, irrigation, and retention). Only PJI occurring within 90 days after primary hip or knee arthroplasty were included. The LROI data and completeness was based on the entered procedures and documented reason for revision infection, which was not specially based on the MSIS criteria. After checks on missing and incorrectly data, the completeness of registration in the LROI was calculated by comparing the number of registrations in the LROI with data from the RIC (gold standard).
Results
Of the 639 primary total hip and knee arthroplasty with a suspected PJI registered in the RIC between 2014–2018, 352 cases met the definition of acute PJI. The overall incidence was 1%.
When compared with the LROI, 164 of these cases were also registered in the LROI as PJI revision, resulting in a 53% underestimation of PJI for the LROI.
Conclusions
LROI data on acute PJI shows a significant underestimation, which is comparable to scarce other literature sources. To ensure reliability and usability of national PJI data, a specifically part of the LROI has to be designed for registering PJI similar to the regional cohort.