Early outpatient follow-up and 30-day outcomes in patients hospitalized with cirrhosis

Abstract

Preventing readmission has been the focus of numerous quality improvement efforts across many conditions. Early outpatient follow-up has been proposed as the best mechanism for reducing readmissions. The extent to which early outpatient follow-up averts readmission or improves outcomes in cirrhosis is not known. We evaluated the relationship between early outpatient follow-up and short-term readmission and mortality in patients with cirrhosis. We conducted a retrospective cohort study of patients with cirrhosis who were hospitalized with a liver-related diagnosis and discharged to home from 122 Veterans Administration hospitals between 2010 and 2013. We defined early follow-up as an outpatient visit with a clinician within 7 days after discharge. We propensity matched patients who received early visit with those who did not have any visit and examined the associations between early follow-up and all-cause readmission and mortality within 8-30 days after discharge. Of 25,217 patients hospitalized with cirrhosis, 8,123 (32.2%) had an early follow-up visit within 7 days of discharge. A total of 3,492 (13.8%) patients were readmitted and 1,185 (4.6%) died between 8 and 30 days after discharge. In the propensity-matched sample (N = 16,238), patients with early outpatient follow-up visit had a slightly higher risk of readmission (15.3% vs. 13.8%; hazard ratio [HR] =1.10; 95% confidence interval [CI] = 1.02-1.19), but significantly lower risk of mortality (3.2% vs. 5.2%; HR = 0.60; 95% CI = 0.51-0.70) than those without early visit. The findings persisted in several subgroup and sensitivity analyses. Conclusions: Early outpatient follow-up after discharge was associated with a small increase in readmissions but lower overall mortality in patients with cirrhosis. Transitional care may be effective in improving short-term outcomes in patients with cirrhosis, but readmission performance measures would miss this effect. (Hepatology 2016;64:569-581).