A Novel Online Calculator for Hospital Length of Stay in Brain Tumor Patients. (16th November 2020)

• Record Type: Journal Article
• Authors: Jimenez, Adrian; Khalafallah, Adham M; Huq, Sakibul; Patel, Palak; Sharma, Paarth; Dux, Hayden; White, Taija; Mukherjee, Debraj
• Is Part Of: Neurosurgery. Volume 67(2010)Supplement 1
• Publication Date: 2020-11-16
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1093/neuros/nyaa447_190
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract: INTRODUCTION: Hospital length of stay (LOS) is an important consideration for physicians seeking to provide high-quality, cost-effective care. Accurate preoperative estimation of patient LOS may help contain healthcare costs and optimize scarce resource allocation within hospitals and healthcare systems. METHODS: Patients who underwent brain tumor surgery between 2017–2019 at a single academic institution were included in our study. Bivariate analysis identified significant predictors of LOS, which were then included in multivariate analysis. Logistic regression was used to predict probability of LOS > 7 days and robust MM-type linear regression was used to predict exact LOS in days. Optimal models were selected by minimizing the Akaike information criterion (AIC). P < . 05 was considered to be statistically significant. RESULTS: A total of 1, 061 brain tumor patients were included in our analysis. Our patient cohort was majority male (50.6%) and Caucasian (74.2%). Our logistic regression model predicting risk of LOS > 7 days had an average c-statistic of 0.81 across 5 cross-validation folds. Significant, independent predictors of LOS in both linear and logistic regression included patient race, patient admission source, Karnofsky performance status score, and 5-factor modified frailty index score. Both models were incorporated into an online calculator (https://ajimene8.shinyapps.io/LOS_calc_app1/ ). CONCLUSION: Our study created an online calculator to predict LOS for brain tumor patients. Our work synthesizes prior research to develop a practical tool that can help neurosurgeons provide more cost-effective, resource efficient care while maintaining optimal patient outcomes.