Outcomes of Boarding Critically Ill Patients in U.S. EDs: A Systematic Review and Meta-analysis

Outcomes of Boarding Critically Ill Patients in U.S. EDs: A Systematic Review and Meta-analysis


 Authors: Htet NN, Walker JA, Jafari D, Rech MA, Hintze T, Moran M, Bai J, Dinh K, Essaihi A, Wilairat S, Huddleson B, Tran QT


 Journal: American Journal of Emergency Medicine, October 2025

Conclusions:

  • Critically ill ED patients who are boarded after the decision to admit/transfer have no statistically significant increase in all-cause mortality or hospital length of stay compared with non-boarded patients.

  • There is a signal toward harm (OR > 1 for mortality), but the confidence interval crosses 1 and heterogeneity is high—so the true effect remains uncertain.

  • Definitions of “boarding” and practice environments were inconsistent, limiting generalizability and making it hard to set a clean LOS threshold where risk clearly changes.

  • Despite lack of clear statistical harm, the data do not exonerate boarding—publication bias suggests studies showing more harm with boarding may be underrepresented.

Results:

  • 17 studies, 407,178 ED patients total:

    • 194,814 (48%) boarded in the ED.

    • 212,364 (52%) not boarded in the ED.

  • Majority of patients (87.4%) were from urban academic EDs with a resuscitation unit or team—this is not a community-only picture.

  • ED length of stay (LOS):

    • Boarded (EDB): median 6.5 h (IQR 5.18–8.1).

    • Non-boarded (non-EDB): median 4.2 h (IQR 2.38–5.75).

    • Difference significant (p < 0.01), confirming what we feel clinically: boarding = longer ED LOS.

  • Mortality:

    • OR 1.06 (95% CI 0.94–1.19), I² = 69%, p = 0.38 → no statistically significant difference, but moderate–high heterogeneity and a point estimate suggesting a small increase in risk.

  • Hospital length of stay:

    • Mean difference 0.38 days (95% CI −0.75 to 1.50), I² = 61%, p = 0.51 → no significant difference in total LOS.

  • Funnel plot suggested publication bias favoring higher reported mortality in boarded patients, which may mean the pooled effect underestimates harm or is unstable.

Methods:

  • Systematic review and meta-analysis of adult ED patients in the U.S. who were critically ill and had a decision to admit/transfer.

  • Searched major health science databases 2012–December 2024, with medical librarian support.

  • Included observational or randomized trials comparing:

    • ED boarding (held in ED/temporary location after admit/transfer decision) vs.

    • Non-ED boarding (more rapid movement out of ED).

  • Only full-text English-language articles were included.

  • Primary outcomes: all-cause mortality and hospital length of stay.

  • Used random-effects meta-analysis with I² to assess heterogeneity and funnel plots to assess publication bias.

Htet, N.N., Walker, J.A., Jafari, D., Rech, M.A., Hintze, T., Moran, M., Bai, J., Dinh, K., Essaihi, A., Wilairat, S. and Huddleson, B., 2025. Outcomes of boarding critically ill patients in US EDs: A systematic review and meta-analysis. The American Journal of Emergency Medicine.

Back to blog