- Study design
This study was a single-center, observational, registry-based cohort study conducted at the ED of a university-affiliated teaching hospital in Seoul, Korea, which receives approximately 110,000 patients per year. We retrospectively analyzed prospectively collected registry data of all consecutive adult (age ≥ 18 years), non-traumatic IHCA and OHCA patients treated in the ED between January 2014 and June 2017. Patients were excluded if they were younger than 18 years, had do-not-resuscitate status, were dead upon arrival at the ED, or if return of spontaneous circulation (ROSC) occurred before arrival at the ED. Our institutional review board approved the study and waived the requirement for informed consent (study number S2017-1004).
Our ED had 7 zones with a total of 65 beds, including 2 beds in the CPR zone, 3 beds in the isolation rooms for airborne-transmitted disease, 37 beds in the intensive care zone, and 23 in the zone for boarding or ambulatory care. For simplification of analysis, we classified the ED zone into two zones: the critical zone for patients in need of CPR, isolation, or intensive care and the urgent zone for patients intended to be admitted to the hospital, discharged to home or another facility, or to receive ambulatory care. During the study period, the ED had an average of 16 board-certified emergency physicians, 24 residents, and 90 nurses. The study facility had 24-hour consultants including cardiologists, vascular surgeons, neurologists, neurosurgeons, interventional radiologists, and orthopedic surgeons.
- Data collection
All patients presenting with OHCA and IHCA were registered in the CPR registry of our hospital. For each patient, an emergency physician on duty recorded a CPR report using the Utstein style, and the data were verified and entered into the web-based registry by the principal investigator.
Data included demographic characteristics, medical history, and characteristics of CPR, such as do-not-resuscitate (DNR) during or after CPR, if death occurred in the ED, whether the arrest was witnessed, the initially reported rhythm, time of first defibrillation, time of first epinephrine injection, time of intubation, total CPR duration, and whether ROSC occurred. Also, the registry contained information about hospital treatment and clinical outcomes, such as disposition of the hospital discharge, length of stay, and neurologic status at discharge. Achievement of sustained ROSC was declared when patients had a noticeable pulse for longer than 20 minutes. Neurologic status was quantified based on cerebral performance category (CPC) scales at the time of hospital discharge.[16]
We used ED occupancy rate as an overcrowding index. ED occupancy rate was defined as the ratio of the total number of ED patients to the number of beds in the ED.[17] Although there is no universal consensus on overcrowding measurement, ED occupancy rate is one of the most promising quantifiable methods.[18,19] It is essential that the ED occupancy rate is automatically updated periodically since this data has a time-series nature; our electronic medical record system automatically collects variables necessary for ED occupancy rate calculation at one-hour intervals. Previous studies using ED occupancy rate only measured occupancy at ED presentation.[20–22] To evaluate if other times during an ED cardiac arrest were associated with crowding levels, we measured ED occupancy rates of both IHCA and OHCA cases at presentation, at time of arrest, at time of maximum occupancy, and the average during ED stay. We then calculated the accumulated ED stay time of all patients. Finally, cardiac arrest occurrence at each specific ED occupancy rate time point was calculated based on the number of cardiac arrest patients at a certain range of ED occupancy rate (e.g., 0.9-1.1) divided by the accumulated time of each ED occupancy rate during the study period.
The primary outcome was the correlation between the four ED occupancy rate measurements and the occurrence rate of unexpected cardiac arrest during ED stay. The secondary outcome was comparing clinical outcomes of IHCA and OHCA.
- Statistical analysis
All continuous variables were expressed as median with interquartile range (IQR). The Mann-Whitney U test was used to compare the values of continuous variables. Categorical variables were analyzed with the chi-square test or with the Fisher exact test. Normality of distribution was examined by the Kolmogorov–Smirnov test. Differences between means were calculated by one-way analysis of variance or Kruskal-Wallis test. We assessed the correlation between ED occupancy rate and cardiac arrest occurrence by calculating the Spearman rank correlation coefficient. For all analyses, a two-sided P value of < .05 was considered to indicate a statistically significant difference. Statistical analyses were performed by using R version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria).