학술논문

Hospital Readmission and Mortality in a Brazilian Tertiary Public Hospital.
Document Type
Article
Source
Quality Management in Health Care. Apr-Jun2020, Vol. 29 Issue 2, p76-80. 5p.
Subject
*CHI-squared test
*CONFIDENCE intervals
*FISHER exact test
*INTERNAL medicine
*LONGITUDINAL method
*MEDICAL records
*MULTIVARIATE analysis
*POISSON distribution
*PUBLIC hospitals
*QUESTIONNAIRES
*STATISTICS
*T-test (Statistics)
*RELATIVE medical risk
*RETROSPECTIVE studies
*PATIENT readmissions
*DATA analysis software
*DESCRIPTIVE statistics
*HOSPITAL mortality
*ACQUISITION of data methodology
*TERTIARY care
*MANN Whitney U Test
Language
ISSN
1063-8628
Abstract
Background and Objectives: Data on mortality associated with hospital readmission are imprecise and highly variable. This study aimed to describe the rate of nonelective 30-day readmission and associated hospital mortality of patients discharged from the Internal Medicine Unit of a Brazilian tertiary public hospital. Methods: This retrospective cohort study included all patients discharged from the Internal Medicine Unit of our institution between September and November 2017 who were nonelectively readmitted within 30 days. Results: A total of 1047 hospital discharges were analyzed. The rate of nonelective 30-day readmission was 13.7%. Of these, 41 (28.5%) were early readmissions (0-7 days) and 103 (71.5%) were late readmissions (8-30 days). The hospital mortality rate during readmission was 27.8%, being significantly higher during early readmissions (41.5% vs 22.3%; P = .035). Early (as compared with late) readmission was associated with mortality during readmission (relative risk [RR] 1.95; 95% confidence interval, 1.18-3.22; P = .002), regardless of age and Charlson comorbidity index. Conclusion: The Readmission rate was 13.7%, with an associated mortality of 27.8%. Early readmission was an independent predictor of mortality (RR 1.95) in relation to late readmission. Larger studies are needed to better identify this group of patients with an aim to adopt preventive measures. [ABSTRACT FROM AUTHOR]