RCS-E is Predictive of Outcomes and Resource Allocation in a Large Neuromotor Rehabilitation Setting

To assess the reliability of the Rehabilitation Complexity Scale-Extended (RCS-E) as a measure of rehabilitation complexity. Secondary outcomes include identifying which baseline data, including the RCS-E, are most predictive of rehabilitation stay effectiveness and exploring its role as a predictor of reimbursement for hospital admissions.

A retrospective observational study. Demographic and clinical variables were collected, including length of stay, modified Rankin Scale, modified Barthel Index (mBI), RCS-E, and reimbursement of stay.

Study on inpatients admitted to sixteen neuromotor rehabilitation units.

Of 5870 hospitalizations, 4091 cases (mean age, 71.8 ± 13y; 42.2% men) met the inclusion criteria. The pre-post analysis was performed on 3792 patients (mean age, 71.8 ± 12.8y; 41.6% men).

Rehabilitation program for neurologic or orthopedic condition (median [1st-3rd quartile], duration: 27 days [20-41]).

RCS-E scores on admission were categorized as follows: <8 (low complexity), 8-10 (medium complexity), and >10 (high complexity). For each patient, we calculated the mBI-derived parameters: mBI gain, rehabilitation efficiency, and rehabilitation effectiveness.

A significant moderate Spearman correlation was found between RCS-E score and mBI (rho=-0.53, P<.0001), length of stay (rho=0.41, P<.0001), and premorbid mRS (rho=0.23, P<.0001). Rehabilitation efficiency and effectiveness were significantly lower in the high complexity group. Both RCS-E and mBI at admission significantly correlated with total reimbursement (both P<.0001, rho=0.48, and rho=-0.43, respectively). Regression models indicated a predictive effect of RCS-E at admission on mBI at discharge.

This study demonstrates the reliability and utility of RCS-E in assessing neurologic and orthopedic rehabilitation complexity, predicting outcomes, and informing funding models.