LO34: System and patient level determinants of EMS offload delay

Abstract

Introduction: Arriving EMS patients often experience offload delay due to a lack of available care spaces. Arrival in an overcrowded ED is the primary cause of offload delay, but patient characteristics may also play a role. Our objective was to describe system and patient level determinants of offload delay. Methods: From July 2013 to June 2016, administrative data was collated from the four Calgary Zone adult EDs. All CTAS level 2 and 3 patients arriving by ambulance were eligible for study. To define patient complexity and illness severity, we captured patient demographic data, living situation (homecare/facility vs. independent), vital signs, complaint category (medical, cardiovascular, mental health/neuro, GI, trauma/MS, other), biochemical parameters (serum Na, K, creatinine, hemoglobin, WBC), patient care needs (IV fluid bolus, IV antibiotics, CT scan, admission) and mortality at 7 and 30 days. Results: 162,002 EMS patients were studied. Of these, 67,785 went to a care space within 15 minutes (minimal offload delay), 53,185 between 15 and 59 minutes (moderate offload delay), and 41,032 at ≥60 minutes (severe offload delay). Vital signs, biochemical and hematologic parameters did not differ between groups. ED site was a strong predictor of offload delay (odds ratio {OR}=1.0, 2.03, 2.14, 3.5 for the 4 EDs), as was arrival on weekday (OR=1.38) or night shift (OR=0.71). After adjusting for site, day and time of arrival, multivariate logistic regression models showed the following associations with offload delays of more than 15 minutes: male sex (OR=0.94), age (OR=1.01 per year of age), dependent living situation (OR=1.15), CTAS 3 acuity (OR=1.27), number of prior ED visits within a year (OR=1.06 per visit), and complaint category: general medical (1.0), cardiovascular (0.90), mental health/neuro (0.90), GI (0.85), trauma/MS (0.61). Odds ratio estimates were precise—all with p<0.001. Offload delay was associated with prolonged time to MD, increased EDLOS and higher LWBS/AMA rates. Delayed patients had similar rates of IV antibiotic use, but lower rates of IV fluid bolus, CT use, admission, and 7-day mortality. Conclusion: The strongest predictor of offload delay is arrival to a crowded ED, but patient factors including female sex, older age, dependent living status and repeat hospital use increase risk. Patients subjected to offload delay also appear to have lesser immediate care needs and lower short-term mortality.