Emergency Department Boarding and Mortality: A Systematic Review and Proportional Meta‐Analysis
Gerardo Altamura, Mario Cesare Nurchis, Silvia Martinelli, Annarita Panebianco, Gabriele Giubbini, Marcello Covino, Francesco Franceschi, Gianfranco DamianiABSTRACT
Background and Aims
Emergency department (ED) crowding, worsened by ED boarding (EDB), is defined as the stay of admitted patients in the emergency department due to the unavailability of inpatient beds. This occurrence poses a significant public health challenge, leading to adverse patient outcomes. Despite efforts to improve patient care pathways, the relationship between EDB and mortality is still contentious. This study aims to synthesize and compare reported in‐hospital mortality (IHM) among ED patients across studies stratified by ED boarding duration thresholds, with particular attention to shorter boarding durations.
Methods
A systematic review and proportional meta‐analysis were conducted following PRISMA guidelines. MEDLINE, EMBASE, Web of Science, and Scopus were searched up to April 2024. Pooled mortality proportions were estimated using random‐effects models with DerSimonian–Laird estimation and Freeman–Tukey double arcsine transformation. Heterogeneity was assessed using I 2 . Leave‐one‐out analyses were performed to assess the robustness of subgroup estimates, with particular attention to longer boarding thresholds.
Results
Twenty studies published between 2007 and 2024 were included. The overall pooled in‐hospital mortality proportion was 8.3% (95% CI: 7.6%–9.0%; I 2 = 99.9%). Subgroup analyses showed pooled mortality estimates of 10.2% for early admission/no boarding, 4.4% for 1‐h boarding, 5.9% for 2‐h boarding, 4.8% for 4‐h boarding, 13.5% for 6‐h boarding, 20.5% for 8‐h boarding, 21.8% for 12‐h boarding, and 11.8% for 24‐h boarding. Shorter boarding thresholds were not associated with higher pooled mortality estimates compared with early admission/no boarding. Although longer boarding thresholds showed higher pooled mortality estimates, the 8‐, 12‐, and 24‐h subgroup estimates were based on few studies and were materially influenced by Verma et al. 2021. After excluding this study, the pooled mortality estimates decreased to 7.9%, 3.9%, and 4.0% for the 8‐, 12‐, and 24‐h subgroups, respectively. Given the extreme heterogeneity, these pooled estimates require cautious interpretation. Because different studies contributed to different threshold‐specific pools, the estimates are descriptive and should not be interpreted as evidence of a threshold or dose‐response relationship.
Conclusion
This study highlights the possible association between prompt admissions, targeted interventions, and outcomes in the context of ED boarding. Improving ED operations and reducing boarding times within these durations remain important organizational objectives. Further research with standardized methodologies is essential to fully understand and address this complex issue. However, the present pooled proportions do not establish a boarding‐duration threshold, dose‐response relationship, or causal association, and their interpretation is limited by extreme heterogeneity (I 2 = 99.9%).