A Physician-led Care Redesign Blunts Predictors of Length of Stay after Total Hip and Knee Arthroplasty

We questioned to what extent traditional predictors of care team burden (via increased length of stay [LOS] after total joint arthroplasty [TJA]) were able to be mitigated through alteration of the care pathway. The impact on LOS of traditional patient risk factors, as well as encounter variables, were analyzed for a consecutive set of patients undergoing surgery before and after a physician-initiated arthroplasty care pathway redesign. We analyzed the impact of these variables on LOS, discharge disposition, and 90-day readmission; separate analyses were performed pre- and post-redesign for LOS. Several patient factors (Risk Assessment and Prediction Tool, body mass index, age, insurance type, smoking) predicted longer LOS in the pre-redesign cohort; post-redesign, only ambulation on the day of surgery and anticoagulation type were predictive. The redesign also lessened the aggregate impact of the patient-specific risk factors, resulting in reduced variation in LOS. Physician leadership of care pathways can reduce the impact of factors that have portended longer LOS, thereby reducing variability in LOS and costs for disparate patient populations while driving improvements in value-based care indices. (Journal of Surgical Orthopaedic Advances 32(2):097-101, 2023).

Arthroplasty Care Redesign Impacts the Predictive Accuracy of the Risk Assessment and Prediction Tool

BACKGROUND

The Risk Assessment and Prediction Tool (RAPT) is used to predict patient discharge disposition after total joint arthroplasty. Following a comprehensive, multidisciplinary redesign, our institution noticed a trend toward home discharge in patients with RAPT scores that historically predicted discharge to acute care facilities, presenting an opportunity to redefine the predictive ranges for RAPT.

METHODS

Retrospectively collected data were analyzed from a single institution in patients undergoing elective primary total joint arthroplasty from January 2016 to April 2017. Predictive accuracy (PA) was calculated for each RAPT score (1-12), RAPT score risk ranges (low, intermediate, and high), as well as overall. Other factors evaluated included patient-reported discharge expectation, body mass index, and American Society of Anesthesiologists scores as related to discharge disposition and the PA of RAPT.

RESULTS

Overall PA of RAPT was 88% (n = 1024 patients). Patients were high risk for acute care facility with a RAPT score of 1 to 3 (PA ≥ 83%), intermediate risk 4 to 7 (PA, 52%-79%), and low risk 8 to 12 (PA ≥ 89%). In multivariable analysis, RAPT score and patient-reported discharge expectation had the strongest correlation with actual discharge disposition.

CONCLUSION

Our multidisciplinary redesign has impacted the PA of RAPT. The original predictive ranges should be modified to reflect the increasing proportion of patients being discharged home following elective arthroplasty procedures. We have identified patient-expected discharge destination as a powerful modulator of the RAPT score and suggest that it be taken into consideration for discharge planning.