Nnadi, Augustine K.

INTRODUCTION: There is a lack of data on the impact of readmission to the same vs a different hospital following an index hospital discharge in cirrhosis patients. METHODS: We sought to describe rates and predictors of different-hospital readmissions (DHRs) among patients with cirrhosis and also determine the impact on cirrhosis outcomes including all-cause inpatient mortality and hospital costs. Using the national readmissions database, we identified cirrhosis hospitalizations in 2013. Regression analysis was used to determine the predictors of DHRs. A time-to-event analysis was performed to assess the impact on subsequent readmissions and all-cause inpatient mortality. RESULTS: In 2013, there were 109,039 cirrhosis readmissions with 67% of these being same-hospital readmissions and 33% being DHRs ( P < 0.001). Two percent of readmitted patients were treated at ≥4 different hospitals. The 30-day readmission rate was 29.1%. Predictors of DHR included Medicaid payer (adjusted odds ratio [OR] 1.07, 95% confidence interval [95% CI] 1.01–1.14), age (OR 0.98, 95% CI 0.978–0.982), elective admission (OR 1.09, 95% CI 1.01–1.17), hepatic encephalopathy (OR 1.20, 95% CI 1.16–1.25), hepatorenal syndrome (OR 1.09, 95% CI 1.03–1.16), and low socioeconomic status (OR 1.15, 95% CI 1.06–1.25). No difference was observed in 30-day readmission risk following a DHR (adjusted hazard ratio 1.044, 95% CI 0.975–1.118). In addition, there was no increased risk of inpatient death observed during a DHR within 30 days (adjusted hazard ratio 1.08, 95% CI 0.94–1.23). However, patients with DHR had significantly higher hospital costs and length of stay. CONCLUSIONS: Majority of cirrhosis readmissions are same-hospital readmissions. Different-hospital readmissions do not increase the risk of 30-day readmissions and inpatient mortality but are associated with higher hospital costs.

Background: Gastrointestinal tract cancers account for a significant proportion of the national cancer burden. Aim: We sought to explore patient- and hospital-level determinants of palliative care utilization among patients hospitalized with metastatic gastrointestinal tract cancers using a national database. Methods: An analysis of the 2012 National Inpatient Sample was performed. International Classification of Diseases, Ninth Revision codes were used to identify hospital discharges associated with metastatic digestive tract cancers and patient/hospital covariates for inclusion in a logistic regression model. Total charges and length of stay were analyzed in a linear regression model. Results: Compared to males, females were more likely to receive inpatient palliative care (adjusted odds ratio [OR] 1.12, P = .002). No difference was seen between white and Asian patients (adjusted OR 1.2, P = .11) or Native Americans patients (adjusted OR 1.4, P = .22). However, relative to white patients, African Americans (adjusted OR 1.13, P = .02) and Hispanics (adjusted OR 1.25, P = .001) had significantly higher odds of inpatient palliative care. Medicare patients were least likely to receive palliative care compared to those with Medicaid or commercial payers. Length of stay during these hospitalizations was longer in African Americans ( P = .0001), Asians ( P = .0001), and Native Americans ( P = .03) compared to white patients. No difference was seen when total charges were compared between white and African American patients ( P = .08). Conversely, total charges were higher in Hispanics ( P = .005) and Asians ( P = .001) relative to white patients. Conclusion: Gender and racial differences exist in utilization of inpatient palliative care among patients hospitalized with metastatic gastrointestinal tract cancers.