Statistical analysis

Preoperative and last follow-up values were compared using the paired t-test. Continuous variables between HA and RA groups were compared using Student’s t-test. Discrete variables were compared using the chi squared test and Fisher’s exact test. Preoperative and last follow-up values of flexion contracture and maximum flexion were compared using correlation analysis in both HA and RA groups. Through receiver operating characteristics (ROC) analysis, we aimed to determine the minimum amount of preoperative flexion contracture that increases the risk of residual contracture > 15° after TKA [14] and compare the cutoff value in HA and RA. Residual flexion contracture > 15° was thought to be clinically significant because it affects the patients’ walking ability [29].

The survival rate was evaluated based on life table analysis and the Kaplan–Meier method. The follow-up interval unit was 1 year, and annual success was defined as instances in which the implant remained in place throughout the time period.

All statistical analyses were performed using SPSS software (version 20.0; SPSS, Inc., Chicago, IL). In all analyses, p < 0.05 was considered statistically significant.