Logotyp: till Uppsala universitets webbplats

Background and importance

Crowded emergency departments (EDs) and long waiting times may lead to delayed care for acute conditions and increased mortality in patients leaving without being seen by a physician (LWBS).

Objectives

The aim of this study was to characterize patients LWBS from the ED at a Swedish university hospital and report their mortality.Design, setting, and participantsA cohort study was conducted. Data from electronic medical records of patients LWBS (exposure) were compared with those who stayed through completion of treatment. Data were gathered from 151 228 patients making 363 308 visits to the Uppsala University Hospital ED from 1 January 2010 to 30 June 2017.

Outcome measures and analysis

Short- and long-term mortality of patients LWBS were studied using Cox regression analysis. Sex, age, triage level, arrival mode, length of stay, presenting complaint, crowding, and time of day were assessed as regards effect on the probability of LWBS and as covariates in the mortality analysis.

Results

In our dataset, 9058 patient visits to the ED (2.5% of all ED visits) ended in LWBS. These patients were younger than average and had lower triage levels. No significant difference was found between males and females. The highest rates were seen in patients presenting with a psychiatric disorder (12.8%), victims of physical assault (11.7%), or alcohol/drug withdrawal (9.9%). The likelihood of LWBS was highest in the evening and lowest in the early morning. Short-term mortality at 28 days was not significantly affected by LWBS, but a 34% lower long-term mortality risk was seen for patients LWBS (hazard ratio 0.66, 95% confidence interval 0.45-0.95, P = 0.028) in median follow-up of 3.5 years.

Conclusion

Patients LWBS were younger, had lower triage levels, and more frequently presented with psychiatric or substance use-related complaints. They had lower short- and long-term mortality compared with those who completed treatment. This observation does not imply a causal relationship but highlights differences in patient characteristics.