Increased risk of readmission but not mortality for patients discharged from the emergency deaprtment with symptom- or suspicion-diagnoses: A single center cohort study.

Increased risk of readmission but not mortality for patients discharged from the Emergency Department with symptom- or suspicion-diagnoses: A single center historic cohort study. Background: Most patients are discharged from an emergency department (ED) after brief examination, observation, and treatment. Some patients have unknown causes of their symptoms after workup and are discharged with symptom- or suspicion-diagnoses. These patients might have underlying, unrecognized diseases causing increased risk of mortality and readmission, a hypothesis which has not yet been investigated. This study was aimed at examining the usage of symptom- and suspicion-discharge-diagnoses along with mortality and readmission within 3 days from discharge. Methods: In this historic, register-based study, we identified all patients discharged from the ED at Bispebjerg and Frederiksberg Hospital (Copenhagen, Denmark) in years 2016-2018 from the National Patient Register. Generally, workups, initial treatment, and assessment for need of further admission are intended completed within 6 hours at the Emergency Department. For discharged patients, we assessed mortality and in-hospital readmission within 3 days from discharge as outcomes, calculated as risk (95% CI) using the cumulative incidence function, accounting for death as a competing risk for readmission. The exposure was the primary ICD-10 discharge diagnoses, grouped by R-diagnoses (symptoms), Z-diagnoses (suspicion of disease), or other. Some of the most common symptom-diagnoses were also grouped for sub-group analyses. Results: We identified 251,424 primary contacts at the ED, 42,324 (16.3%) were admitted, 85 died during stay in the ED, and 209,015 (52.5% female) with a median age of 36.3 years (Q1-Q3: 26.1-54.1) were discharged. A R-diagnosis was the primary diagnosis in 28,858 (13.8%) contacts and a Z-diagnosis in 21,398 (10.2%) contacts. Risk of death within 3 days was 0.04% (0.04-0.05%) for all discharged patients and stratified 0.06% (0.04-0.08) for R-diagnoses, 0.06% (0.04-0.09%) for Z-diagnoses, and 0.04% (0.03-0.05%) for other. Risk of in-hospital readmission within 3 days was 2.9% (2.8-2.9%) for all discharged patients and stratified 5.1% (4.9-5.3%) for R-diagnoses, 3.6% (3.5-3.8%) for Z-diagnoses, and 2.4% (2.3-2.4%) for other. R-diagnoses at readmission were more common for patients with R-diagnoses at index admission at 43.6%, compared to 12.6% for Z-diagnoses and 8.9% for other diagnoses at index (p<0.001). The highest risk of readmission in the sub-group symptom-diagnoses was seen in “Fever” at 12.7% (10.5-15.5%), followed by “Abdominal pain” at 9.2% (8.6-9.8%), “Headache” at 4.9% (4.0-5.9%), “Cough and abnormal breathing” at 4.2% (3.6-4.8%), “Vertigo and dizziness” at 4.1% (3.4-4.9), “Malaise” at 3.1% (2.4-4.0%), “Fainting” at 3.0% (2.2-3.9%), “Chest pain” at 2.8% (2.3-3.3%), and “Abnormal heart rhythm” at 2.2% (1.6-2.9%). Risk of death in the symptom-groups were imprecise due to a small number of fatal outcomes. Discussion & conclusion: Risk of death was low for all patients discharged from the emergency department at 4 per 10.000 contacts within 3 days, not significantly different for R or Z-diagnoses. Readmissions was significantly more frequent for patients discharged with R- or Z-diagnoses. This suggest, a higher proportion of patients with underlying disease which should be of further focus although we cannot conclude on the reasons for readmissions. In conclusion, patients discharged from the ED after brief examination with a R- or Z-diagnosis had an increased risk of readmission but not mortality within 3 days.