Association Between Targeted Early EKG Ordering and Emergency Department Length of Stay Among Psychiatric Patients in Southern New Jersey: A Retrospective Study

IntroductionMental health-related visits make up a significant and growing share of emergency department (ED) utilization in the United States.At the Inspira Mullica Hill Emergency Department, a Six Sigma DMAIC analysis of the behavioral health intake process identified EKG ordering as a rate-limiting step in medical clearance, which is required before behavioral health evaluation.This study evaluates whether implementing a targeted early EKG order for psychiatric patients affects length of stay (LOS). Materials and methodsTargeted early EKG ordering was implemented for patients presenting with suspected behavioral health conditions, and outcomes were evaluated over a three-month post-intervention period.A retrospective chart review was conducted to compare the pre-intervention and post-intervention groups.Encounters were compared between a pre-intervention period, January 1, 2025, to March 9, 2025, and a post-intervention period, April 16, 2025, to July 31, 2025.Encounters during the implementation/washout period were excluded.The primary outcome was ED LOS in minutes.Secondary outcomes included LOS >6 hours and LOS >12 hours.Because LOS was non-normally distributed, outcomes were summarized using median (IQR), 90th and 95th percentiles, Mann-Whitney U testing, and chi-square testing.Multivariable log-linear regression was performed using natural log-transformed LOS.Statistical process control and interrupted time-series analyses were performed to evaluate process stability and secular trends. ResultsA total of 412 encounters were included, with 133 pre-intervention and 279 post-intervention encounters.Median LOS decreased from 854 minutes (IQR 389-1301) pre-intervention to 785 minutes (IQR 304-1404) post-intervention, though this difference was not statistically significant (p = 0.245).The proportion of encounters with LOS >6 hours decreased from 78.2% to 71.3% (p = 0.174), and LOS >12 hours decreased from 60.9% to 52.7% (p = 0.144).In the adjusted analysis, the pre-intervention period was associated with longer LOS than the post-intervention period (LOS ratio 1.23, 95% CI 1.02-1.49,p = 0.032), equivalent to approximately 19% lower adjusted LOS in the post-intervention period.Statistical process control analyses demonstrated common-cause variation without special-cause variation, and interrupted time-series analyses did not demonstrate statistically significant level or slope changes. ConclusionsEarly targeted EKG ordering was associated with lower adjusted ED LOS among psychiatric encounters.However, unadjusted comparisons and time-series analyses did not demonstrate statistically significant process-level changes.These findings suggest a potential association between standardized early EKG acquisition and improved psychiatric ED throughput but should be interpreted cautiously given the singlecenter retrospective design and limited operational covariate data.