|
Back to 2016 Annual Meeting
A Novel Prediction Score Identifies Patients at High Risk for Readmission after Hepatic Resection
Mustafa Raoof*, Aaron G. Lewis, Leanne Goldstein, Sinziana Dumitra, Susanne Warner, Gagandeep Singh, Yuman Fong, Laleh Melstrom
Surgery, City of Hope National Medical Center, Duarte, CA
Background: To cut healthcare costs there is an increasing push to reduce readmissions after complex operations. Data on factors that predict or reduce readmission after hepatectomy are limited.
Methods: This is a retrospective study of a prospectively maintained institutional database. All patients undergoing liver resection with or without ablation from Jan 2009- Jun 2015 were included. Hepatectomy was defined as major if 3 or more segments were resected. Complications are graded on Clavien-Dindo Classification Scheme and scored using Comprehensive Complication Index (CCI). Univariate analyses are performed using Chi square and Student t-test. Multvariate analysis is performed using logistic regression.
Results: Of the total 257 patients that met the inclusion criteria, 51% were female, 51% were white and median age was 62 years (IQR 53-69). Most patients had a minor hepatectomy (73.5%). Most common indication was metastatic colorectal cancer (53.3%). Overall 30-day complication rate was 40.8%. The incidence of grade 4/5 complications was 6.2% and there were no mortalities. Overall 30-day readmission rate was 10.5%, most common reason was intra-abdominal sepsis (44%) and median time to readmission after discharge was 8.5 days (IQR 3-15) (Figure). On stepwise multivariate logistic regression CCI, Age>60 years, Maximum temperature 48 hours prior to discharge, White blood cell count and Total bilirubin 48 hours prior to discharge were significant predictors of 30-day readmission. These variables were used to develop a readmission prediction score (RPS) (R 0.81; 95% CI 0.70-0.92, p<0.001). An RPS of >0.063 had a sensitivity of 80% and specificity of 71% in predicting 30-day readmission.
Conclusion: We have developed a unique predictive tool that identifies patients at high risk of readmission. Patients with RPS >0.063 should be followed closely after discharge.
 Figure. Timing of complications in readmitted patients after hepatic resections.
Back to 2016 Annual Meeting
|
