Oncology risk score as predictor of unplanned readmissions.

Abstract

e23166 Background: Almost 75% of our center’s cancer patients with advanced disease are hospitalized at least once in the year following diagnosis. To avoid readmissions, a mark of good quality care, a strategy for risk identification is needed. Our center developed the REDUCE Score (Reducing ED Utilization in the Cancer Experience) based on clinical data to proactively evaluate potential for acute care utilization. REDUCE assigns a patient a numeric score from 0-10, with a score above 2 indicating high risk. We evaluated REDUCE as a tool to stratify patients at risk for unplanned readmissions (UR). Methods: We performed a retrospective analysis of patients with an index admission to a system hospital between January and November 2024. Included were patients with a discharge to home and an encounter with an oncology provider within the past year. We compared the proportion of patients assigned to each REDUCE score who had an unplanned readmission (UR) within the measurement period. We evaluated the relationship between each REDUCE score and risk for readmission as well as comparing previously defined low risk (score 0-1) and high risk (score ≥2+). When comparing risk of UR between low risk and high risk based on the REDUCE score, chi-squared statistical analysis was used with significance defined as p-value less than 0.05. Results: Of the 2,533 patients with a REDUCE score assigned, 542 (21%) had a UR within the measurement periods. 148 (16%) patients with a score of 0 had a UR; 8 (89%) patients with a score of 7 (maximum in cohort) had a UR. The mean REDUCE score among patients with UR is 1.85 (SD = 1.76); the mean score of patients without UR is 1.21 (SD = 1.32) The proportion of patients with UR increases linearly with REDUCE. The risk of admission is significantly higher in the REDUCE score high-risk group as compared to the low-risk group (p-value < 0.00001). Conclusions: These data demonstrate a correlation between REDUCE and UR among cancer patients. REDUCE appears to be a reliable indicator of a patient’s potential for UR and can be useful in determining workflows that intervene with high-risk patients. Our center is exploring effective interventions to prevent unnecessary acute care utilization among cancer patients identified as high risk. Rate of discharges with UR by REDUCE score. REDUCE Score # of discharges with UR Total # of discharges % of discharges with UR 0 148 902 16% 1 129 719 18% 2 104 435 24% 3 73 260 28% 4 31 107 29% 5 28 62 45% 6 21 39 54% 7 8 9 89% Total 542 2533 21%