TOPLINE

Hospitals were more likely to declare ambulance diversion when neighboring hospitals were already diverting, independent of their own patient load, with an estimated 28% of diversion events attributable to peer diversion, a retrospective study found.

METHODOLOGY

• Researchers conducted a retrospective observational study of all ambulance-receiving hospitals in San Francisco between 2015 and 2020 using city 9-1-1 transport and diversion records, daily walk-in volumes, and hospital inpatient census data.
• The primary outcome was the likelihood that a hospital would declare ambulance diversion according to peer hospital diversion status.
• Researchers estimated the population attributable fraction (PAF) to determine how many diversions would have been prevented if no peer hospitals had ever diverted.
• Covariates included fully interacted hospital fixed effects; time-lagged, hospital-specific ambulance arrivals, walk-in volume, and inpatient census; and hour, weekday, month, and year fixed effects.

TAKEAWAY

• Over 1885 days, nine hospitals averaged 27 daily ambulance arrivals and declared diversion 1.3 times per hospital day, with 1.1 system-wide diversion suspensions per day occurring when a fourth hospital declared diversion.
• The likelihood of declaring ambulance diversion increased with the number of peer hospitals already on diversion — 18% with one peer hospital (adjusted hazard ratio [aHR], 1.18; 95% CI, 1.11-1.25), 29% with two peer hospitals (aHR, 1.29; 95% CI, 1.19-1.39), and 62% with three peer hospitals (aHR, 1.62; 95% CI, 1.41-1.84).
• An estimated 28% of ambulance diversion events were attributable to peer hospital diversion (PAF, 0.28; 95% CI, 0.25-0.30).
• After diversion declarations, adjusted ambulance arrivals reduced at the diverting hospital and increased at undiverted peer hospitals, and system-wide transport intervals increased by less than 1 minute.

IN PRACTICE

"Our findings suggest that greater transparency around ongoing loads and even centralized coordination could be deployed strategically to reduce diversion. Alternatively, abolishing diversion entirely may be a reasonable and much simpler approach," the authors wrote.

SOURCE

The study was led by Jiangheng Lin, PhD, MA, and Gengbin Piao, PhD, MS, Washington University, St. Louis. It was published online on July 6, 2026, in Annals of Emergency Medicine.

LIMITATIONS

The study was limited by its retrospective design, reliance on temporal fixed effects and daily walk-in volumes to approximate more granular patient load data, potential residual confounding, and inclusion of hospitals from a single region, which may have limited the generalizability of the findings.

DISCLOSURES

The study did not receive any funding. The authors reported having no competing interests.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.