Introduction: The ability to predict 30-day readmission for patients following liver resection can improve health outcomes and equity. Methods: A case-control retrospective study was conducted for readmission in a large single center including 743 liver resection patients from 2000-2016. Multivariate propensity score (PS)-adjusted regression was conducted for causal inference to determine the most predictive comorbidities and clinical variables. Regression analysis included age, sex, race, ASA (American Society of Anesthesiologist) score >2, body mass index (BMI), major resection, cirrhosis, hypertension (HTN), diabetes (DM), obesity, tumor size and margin, prior abdominal surgery, and the likelihood of undergoing laparoscopic versus open surgery. Results: The mean age in the study was 57.22 (standard deviation [SD] 13.61), 452 (43.88%) were female, and 100 (13.46%) were readmitted. Patients who were readmitted were significantly more likely to be African American and have American Society of Anesthesiologist scores >2, Medicare or Medicaid instead of commercial insurance, hypertension, and diabetes (all p< 0.01) . In multivariate regression, laparoscopic versus open surgery had no association with readmission. The top readmission predictors were hypertension (OR 8.56, p< 0.001) followed by African American race (OR 2.77, p< 0.001). Conclusions: Our analysis identified significant racial and insurance disparities in readmission following liver resection, in addition to specific comorbidities and clinical variables. This analysis suggests efforts to reduce health disparities and target patients with particular risks factors may produce more effective, equitable care.
Read full abstract