Objective:

Predicting patient needs for extended care after spinal fusion remains challenging. The Risk Assessment and Prediction Tool (RAPT) was externally developed to predict discharge disposition after non-spine orthopedic surgery but remains scarcely used in neurosurgery. The present study is the first to employ coarsened exact matching (CEM) – which incorporated patient characteristics known to independently affect outcomes – for 1:1 matching across a large population of single-level, posterior lumbar fusions, to isolate the predictive value of preoperative RAPT score on postoperative discharge disposition.

Methods:

Preoperative RAPT scores were prospectively calculated for 1066 consecutive single-level, posterior-only lumbar fusion patients within a single, university healthcare system. The primary outcome was discharge disposition. Logistic regression was executed across all patients, evaluating RAPT score as a continuous variable to predict home discharge. Subsequently, patients were retrospectively clustered into predicted risk cohorts – validated within prior orthopedic joint research – based on RAPT score (Lowest, Intermediate, and Highest Risk). CEM was performed among predicted risk cohorts, and outcomes were compared between exact-matched groups.

Results:

Among all patients, single-point increases in RAPT score (i.e., decrease in predicted risk) were associated a 75% increased odds of home discharge (p<0.001). Exact-matched analysis demonstrated increased odds of home discharge by 400% when comparing the Lowest vs. Highest Risk cohorts (p=0.004), by 750% when comparing the Intermediate vs. Highest Risk cohorts (p<0.001), and by 200% when comparing the Lowest vs. Intermediate Risk cohorts (p<0.001).

Conclusion:

RAPT score, captured in preoperative evaluations, can be highly predictive of discharge disposition following single-level, posterior lumbar fusion.