Operative Times Can Predict and Are Correlated With Lengths-of-Stay in Primary Total Knee Arthroplasty: A Nationwide Database Study

The specific influence of operative times on lengths-of-stay (LOS) has not been well assessed and is therefore an important topic of evaluation for multiple reasons, including potential economic implication. The purpose of this study is to (1) identify predictors of longer operative times; (2) identify predictors of longer LOS; and (3) evaluate the effects of operative times on LOS in primary total knee arthroplasty (TKA).

The National Surgical Quality Improvement Program database was queried for primary TKAs performed between 2008 and 2016, yielding 225,344 cases. Multivariate analysis was performed to determine the effect of operative times on LOS after adjusting for patient factors. Univariate analyses were performed with 1-way analysis of variance (ANOVA), linear regression, and independent sample t-tests where appropriate. Multivariate analysis was performed with a multiple linear regression model adjusted for patient age, sex, and body mass index (BMI).

The mean operative time was 93 ± 35 minutes and the mean LOS was 3 ± 3 days. Young age, male sex, and obesity were predictors of longer operative times ( P < .001). Similarly, there were significant associations of age, sex, and BMI with LOS ( P< .001). Increasing 30-minute operative time intervals were associated with increasing LOS ( P < .001). Longer operative times had significant associations with longer LOS even after adjusting for patient factors (β = 0.076, SE < 0.001, P value < .001). Out of all study covariates, operative times had the greatest effect on LOS.

This study provides unique insight by directly correlating operative times to LOS in over 200,000 TKAs. The results from this study indicate that there is a strong correlation between operative times and LOS, so that operative time had a greater association than age, sex, and BMI with postoperative TKA LOS.