Abstract
Sex-related differences among comorbid conditions associated with surgical site infection (SSI) after total knee arthroplasty (TKA) are unclear. This population-based cohort study used a novel approach with a Phecode system to evaluate preoperative clinical phenotypes (i.e., comorbid conditions) associated with SSI after TKA and delineate sex-related differences in phenotypes. Using the Taiwan National Health Insurance Research Database (2014-2018), 83,870 patients who underwent TKA were identified. Demographic and SSI data during the 90-day postoperative follow-up were obtained. Comorbidities identified by the International Classification of Diseases within 1 year before TKA were recorded and mapped into Phecodes representing phenotypes. The overall rate of 90-day SSI was 1.3%. In total, 1663 phenotypes were identified among 83,870 patients-1585 and 1458 phenotypes for female (n = 62,018) and male (n = 21,852) patients, respectively. According to multivariate logistic regression analysis, the SSI odds ratio significantly increased with the presence of each of the 16 phenotypes. Subgroup analysis revealed that the presence of 10 and 4 phenotypes significantly increased SSI risk in both sexes; only one phenotype was common to both sexes. Therefore, comorbid conditions and sex should be considered in preoperative SSI risk evaluation in patients undergoing primary TKA. These findings provide new perspectives on susceptibility, prevention, and treatment in these patients.