The effectiveness of inpatient rehabilitation after uncomplicated total hip arthroplasty: A propensity score matched cohort

Abstract

Background: Inpatient rehabilitation is an expensive option following total hip arthroplasty (THA). We aimed to determine if THA patients who receive inpatient rehabilitation report better hip and quality of life scores post-surgery compared to those discharged directly home. Methods: Prospective, propensity score matched cohort involving 12 private hospitals across five Australian States. Patients undergoing THA secondary to osteoarthritis were included. Those receiving inpatient rehabilitation for reasons other than choice or who experienced significant health events within 90-days post-surgery were excluded. Comparisons were made between those who did and did not receive inpatient rehabilitation for patient-reported hip pain and function (Oxford Hip Score, OHS) and 'today' health rating (EuroQol 0-100 scale). Rehabilitation provider charges were also estimated and compared. Results: Two hundred forty-six patients (123 pairs, mean age 67 (10) yr., 66% female) were matched on 19 covariates for their propensity to receive inpatient rehabilitation. No statistically nor clinically significant between-group differences were observed [OHS median difference (IQR): 0 (- 3, 3), P = 0.60; 0 (- 1 to 1), P = 0.91, at 90 and 365-days, respectively; EuroQol scale median difference 0 (- 10, 12), P = 0.24; 0 (- 10, 10), P = 0.49; 5 (- 10, 15), P = 0.09, at 35-, 90- and 365-days, respectively]. Median rehabilitation provider charges were 10-fold higher for those who received inpatient rehabilitation [median difference $7582 (5649, 10,249), P < 0.001]. Sensitivity analyses corroborated the results of the primary analyses. Conclusion: Utilization of inpatient rehabilitation pathways following THA appears to be low value healthcare. Sustainability of inpatient rehabilitation models may be enhanced if inpatient rehabilitation is reserved for those most impaired or who have limited social supports. Trial registration: ClinicalTrials.gov Identifier: NCT01899443.