| Popis: |
Introduction: Since the 1999 Institute of Medicine report, hospitals have implemented a myriad of measures to protect patients from medical errors. At this point, looking beyond errors may bring additional safety benefits. This study aims to analyze predictors of in-hospital death regardless of underlying diagnoses in an effort to identify additional targets for improvement. Methods: We performed a retrospective study of hospitalizations from the 2016-2019 National Inpatient Sample (NIS) database. Logistic regression analyses were used to calculate adjusted odds ratios (OR) for variables associated with in-hospital death. Results: There were 121,026,484 adult hospital discharges in the database. Multivariable analysis showed the following variables were associated with higher in-hospital death: Age (OR, 1.04), Charlson Comorbidity Index (OR, 1.23), male (OR, 1.16), income Q1 (OR, 1.12), income Q2 (OR, 1.07), west region (OR, 1.07), non-elective admission (OR, 2.01), urban hospital location (OR, 1.17), and weekend admission (OR, 1.14). Percentage of deaths for weekend versus weekday admissions was 2.7% versus 2.1%. Fewer procedures (ICD-10-PCS) were done in first 24 hours of weekend admissions when compared to weekday admissions (34.8% vs 46.8%; p |
