Background: This study aimed to clarify the association between types of knee arthroplasty (KA) (total knee arthroplasty (TKA) or unicompatmental knee arthroplasty (UKA)) and surgical site infection (SSI) with adjustment for various factors, using a Japanese national database.
Methods: Data on 181,608 patients who underwent unilateral primary KA for osteoarthritis from 2010 to 2017 were obtained from the Japanese Diagnosis Procedure Combination database. SSI was identified based on International Classification of Diseases 10th Revision codes. Deep SSI (i.e. periprosthetic joint infection (PJI)) was identified as SSI treated with surgical procedures. Multivariable logistic regression analyses for SSI and PJI were performed, in which dependent variables included types of KA, patient backgrounds (sex, age, body mass index (BMI), smoking status, comorbidities), and seasonality.
Results: Eight percent of analyzed patients underwent UKA, while 92% underwent TKA. The proportions of SSI and PJI after UKA were 0.9% and 0.3%, respectively, both of which were lower than those after TKA (1.9% and 0.6%) (P < 0.001). Multivariable analyses showed lower proportions of SSI for UKA (adjusted odds ratio, 0.47; 95% confidence interval, 0.37-0.60; P < 0.001) and PJI (adjusted odds ratio, 0.47; 95% confidence interval, 0.34-0.65; P < 0.001) than TKA. Other factors associated with both SSI and PJI included male sex, BMI >30 kg/m2, renal dysfunction and summer season.
Conclusion: UKA was associated with lower proportions of SSI and PJI than TKA. Surgeons should carefully consider the indication of UKA before performing TKA, especially in patients with knee unicompartmental osteoarthritis who are at a high risk for SSI or PJI.
Keywords: National database; Periprosthetic joint infection; Surgical site infection; Total knee arthroplasty; Unicompartmental knee arthroplasty.
Copyright © 2020 Elsevier B.V. All rights reserved.