To examine whether hospital occupancy was associated with increased testing and treatment during emergency department (ED) evaluations, resulting in reduced admissions.

We analyzed the electronic health records of an urban academic ED. We linked data from all ED visits from October 1, 2010, to May 29, 2015, with daily hospital occupancy (inpatients/total staffed beds). Outcome measures included the frequency of laboratory testing, advanced imaging, medication administration, and hospitalizations. We modeled each outcome using multivariable negative binomial or logistic regression, as appropriate, and examined their association with daily hospital occupancy quartiles, controlling for patient and visit characteristics. We calculated the adjusted outcome rates and relative changes at each daily hospital occupancy quartile using marginal estimating methods.

We included 270,434 ED visits with a mean patient age of 48.1 (standard deviation 19.8) years; 40.1% were female, 22.8% were non-Hispanic Black, and 51.5% were commercially insured. Hospital occupancy was not associated with differences in laboratory testing, advanced imaging, or medication administration. Compared with the first quartile, the third and fourth quartiles of daily hospital occupancy were associated with decreases of 1.5% (95% confidence interval [CI] −2.9 to −0.2; absolute change −0.6 percentage points [95% CI −1.2 to −0.1]) and 4.6% (95% CI −6.0 to −3.2; absolute change −1.9 percentage points [95% CI −2.5 to −1.3]) in hospitalizations, respectively.

The lack of association between hospital occupancy and laboratory testing, advanced imaging, and medication administration suggest that changes in ED testing or treatment did not facilitate the decrease in admissions during periods of high hospital occupancy.