1. Herbold, Janet A. PT, MPH
  2. Bonistall, Kristen MS
  3. Walsh, Mary Beth MD


Purpose: To determine whether clinical outcomes and reimbursement for care differed between patients with hip fracture, total knee replacement (TKR), and total hip replacement (THR) undergoing an inpatient rehabilitation facility (IRF) versus skilled nursing facility (SNF).


Method: A total of 541 patients (IRF = 409, SNF = 131) with unilateral hip fracture, TKR, and THR were recruited. The IRF and SNF patients were matched on age, sex, diagnosis, severity index, and ambulation Functional Independence Measure (FIM) score on admission. Comparisons of discharge motor FIM scores, length of stay, discharge ambulation devices, discharge disposition, use of home health services, transfer to acute care, and total reimbursement for the inpatient stay were carried out between matched pair groups.


Results: From a sample of 541 patients, 102 matched IRF-SNF pairs were created. The mean length of stay for those receiving care in IRF was 10.7 (4.2) days, compared to 25.5 (16.5) days for those receiving care in SNF (P < .001). Costs of care in the IRF setting were $11,984 ($5254) compared to that in the SNF setting, that is, $10,001 ($7141) (P = .008). As compared to patients receiving care in the SNF setting, those in the IRF were more likely to ambulate independently (87.5% vs 74.0%; P = .019), manage stairs independently (68.4% vs 34.7%; P < .001), require less home care (33.7% vs 76.4%; P < .001), and were less likely to use a walker at discharge (41.7% vs 67.7%; P < .001). There were no differences between settings in terms of transfers to acute care, ability to dress the lower body, toilet transfers, and discharge to home.


Conclusion: When patients were matched for age, gender, operative diagnosis, severity index, and admission ambulation FIM score, those who received rehabilitation in the IRF had shorter length of stay and superior functional outcomes than those in the SNF setting. Cost of stay in an IRF was, however, significantly greater.