Abstract
Introduction
With the removal of total hip arthroplasty (THA) from the Centers for Medicare & Medicaid Services (CMS) inpatient-only list, understanding predictors of length of stay (LOS) after THA is critical. Thus, we aimed to determine the influence of patient- and procedure-related risk factors as predictors of >1-day LOS after THA.
Methods
A prospective cohort of 5,281 patients underwent primary THA between January 2016 and April 2019. Risk factors increased LOS were categorized as patient-related (demographics, smoking status, baseline Veterans RAND 12 Item Health Survey Mental Component Summary score [VR-12 MCS], Charlson Comorbidity Index [CCI], surgical indication, baseline Hip Injury and Osteoarthritis Outcome Score [HOOS] pain subscore and baseline HOOS physical function shortform (HOOS-PS), range of motion, and predicted discharge disposition) or procedure-related (hospital site, surgeon, approach, day of surgery, and surgery start time). By using the Akaike information criterion (AIC) and internally-validated concordance probabilities (C-index) for discriminating a 1-day LOS from a >1-day LOS, we compared performance between a patient-related risk factors only model and a model containing both patient- and procedure-related risk factors.
Results
A >1-day LOS was statistically significantly associated with older age (p<0.001), female sex (p<0.001), higher body mass index (BMI) (p<0.001), higher CCI (p<0.001), Medicare status (p=0.012), higher baseline HOOS-PS (P<0.001) and lower baseline VR-12 MCS scores (p<0.001). The C-index after 1,000 bootstrap iterations were 0.693 and 0.883 for patient-related and patient plus procedure-related factors, respectively. Upon addition of procedure-related risk factors to the model, the AIC decreased by approximately 1,100 units, indicating that procedure-related risk factors (especially hospital site and surgical approach) explain LOS more effectively than patient-related risk factors alone.
Conclusions
Although patient-related risk factors provide substantial predictive value for LOS following THA, procedure-related risk factors (mainly hospital site and surgical approach) remain the main drivers of predicting LOS. These results can help clinicians select appropriate candidates for short-stay and outpatient THA.