a b s t r a c t

Purpose: Emergency medical services (EMS) response time is one of prehospital factors associated with survival rate of patients with out-of-hospital cardiac arrest (OHCA). The objective of this study was to determine whether short EMS response time was associated with improved neurologic outcome of patients with OHCA through prospective analysis.

Methods: We performed a prospective observational analysis of collected data from KoCARC registry between October 2015 and December 2016. OHCA patients aged 18 years or older with presumed cardiac etiology by emergency physicians in emergency department were included in this study.

Results: Of 3187 cardiac arrest patients enrolled in the KoCARC registry, 2309 patients were included in the final analysis. Response Time threshold was 11.5 min for prehospital return of spontaneous circula- tion and 7.5 min for survival to discharge and Favorable neurologic outcome. Patients in the <=7.5 min response time group showed increased odds of survival to discharge (OR: 1.54, 95% CI: 1.13-2.10, p = .006) and favorable Neurologic recovery (OR: 2.01, 95% CI: 1.36-2.99, p = .001). When response time was decreased by 1 min, all outcomes were improved (survival to discharge, OR: 1.08; 95% CI: 1.04-1.12, p < .001; favorable neurological outcome, OR: 1.14, 95% CI: 1.07-1.21, p < .001).

Conclusion: We found that shorter EMS response time could lead to favorable neurologic outcome in patients with OHCA of presumed cardiac origin. EMS response time threshold associated with improved favorable outcome was <=7.5 min.

(C) 2019

Introduction

Emergency medical services (EMS) response time, the period of time from when a call is made to the EMS to the point when the EMS arrive at the scene, is one of prehospital factors associated with survival rate of patients with out-of-hospital cardiac arrest (OHCA) [1]. Response time has attracted a lot of attention lately as it is a factor that could be changed sufficiently with policy and support of EMS.

Recently, several reports have shown that shorter EMS response time can improve survival rate of patients with OHCA [1-8]. Ono et al. [1] have reported that response time is closely associated with Favorable neurological outcomes in bystander-witnessed patients with OHCA. However, no study has reported the relation- ship between response time and favorable neurologic outcomes in all patients with OHCA of cardiac origin. Therefore, the objective of this study was to determine whether short EMS response time was associated with improved neurologic outcomes of patients with OHCA through prospective analysis.

* Corresponding author.

E-mail address: raintree@schmc.ac.kr (H.J. Moon).

Methods

Study design and data collection

The Korean Cardiac Arrest Research Consortium (KoCARC) is a Korean collaborative research network developed to conceive var- ious researches conducted in the field of out-of-hospital cardiac arrest [9]. A total of 62 secondary or tertiary hospitals in Republic of Korea participated in this consortium. All OHCA patients who were transported to participated hospitals by emergency medical technician (EMT) were enrolled in the KoCARC registry initiated in October 2015. The KoCARC registry was designed to include patients with OHCA who had a medical etiology identified by emergency physicians in each emergency department. The registry excluded OHCA patients with a terminal illness documented by medical records, patients under hospice care, pregnant patients, and patients with a previously documented ‘Do Not Resuscitate’ card. Patients who suffered cardiac arrest due to non-medical eti- ology, including trauma, drowning, poisoning, burn, asphyxia, or hanging, were also excluded from the registry [9].

Each principal investigator of participating hospital reviewed

hospital records of OHCA patients and filled out their baseline

https://doi.org/10.1016/j.ajem.2019.02.021 0735-6757/(C) 2019

characteristics, prehospital Environmental factors, and emergency medical service (EMS) factors including time intervals, hospital interventions, and clinical outcomes. Characteristics of EMS factors were investigated based on ambulance run sheets entered by EMT. Data Safety and Monitoring Board Committee in KoCARC moni- tored and qualified these data periodically. neurological recovery outcomes were measured at discharge. Cerebral performance cate- gory (CPC) score was used to measure the recovery rate. CPC scores of 1 (good performance) and 2 (moderate disability) were defined as favorable neurological outcome while a score of 3 (severe cere- bral disability), 4 (vegetative state), or 5 (death) was defined as unfavorable neurologic outcome [10,11].

Study setting

This study performed prospective observational analysis of col- lected data from KoCARC registry between October 2015 and December 2016. Additionally, OHCA patients aged 18 years or younger were excluded in this study. Patients with missing or incomplete medical record data (especially time interval) and those whose OHCA episode had been witnessed by EMS personnel were also excluded. This study was approved by the Institutional Review Board of Soonchunhyang University.

Korean EMS system

The Korean EMS system is operated directly by the government. A basic-to-intermediate service level of ambulance is operated by 16 provincial headquarters of the national fire department, cover- ing a population of approximately 50 million [12]. Korean EMS has recently established a centrally based, two-tiered system for OHCA patients in which an ambulance is dispatched with two or three EMTs, including at least one level 1 EMT. Level 1 EMT can execute care that is comparable to EMT-I (intermediate EMT) level in the US, including administering intravenous fluids, endotracheal intu- bation, and laryngeal mask airway insertion under direct medical oversight. However, they can only use medications for advanced cardiac life support (e.g., epinephrine, amiodarone) under direct medical control in limited areas. EMTs dispatched to the field can perform conventional BLS and transfer the patient to the nearest hospital. EMTs cannot declare death or stop CPR on scene or during transport. All patients with OHCA should be transported to the emergency department unless emergency physician instructed EMTs to stop CPR by direct medical instruction [13].

EMS response time was defined by the time interval from the

time of call receipt to EMS dispatcher center to arrival of vehicle at the scene. The time of call receipt was automatically stored in the EMS headquarters, printed in the ambulance run sheet. The time to arrival was electronically stored based on Global Position- ing System (GPS) location and time of the ambulance, whereas if there is a problem with the GPS, EMT could modified it with the correct time.

Statistical analysis

Statistical analysis was performed using SPSS version 21.0 for Windows (SPSS Inc., Chicago, IL, USA). Nominal variables are expressed as counts and percentages of total numbers. Continuous variables are expressed as mean and standard deviation (SD). v²– test (Fisher’s exact test) and Student’s t-test (for continuous vari- ables) at a significance level of p < .05 were used to test differences of observed characteristics and outcomes of patients. Receiver operating characteristic curve analysis with Youden Index were performed to calculate optimal cut-off values for the

response time that predicted prehospital ROSC, survival to dis- charge, and favorable neurological outcomes. We also executed ROC curve analysis for patients with OHCA who received bystander CPR and witnessed by bystander.

Fig. 1. flow charts of the inclusion of patients.

Using logistic regression analysis including seven independent variables (sex, age, witnessed, bystander CPR, place, cardiac rhythm, and drug administration on scene), we assessed the adjusted odds ratios (ORs) and 95% confidence intervals for prehos- pital ROSC, survival to discharge, and favorable neurological outcome.

Results

Patient collection

Of 3187 cardiac arrest patients enrolled in the KoCARC registry, 2309 were included in the final analysis during the study period

Table 1

Characteristics of all patients with OHCA.

Characteristics Total

(N = 2309)

Age, years 67.1 +- 15.3

Male sex 1519 (65.8%)

witnessed arrest 1274 (55.2%)

Occurrence place

Housing facility 1560 (67.6%)

Industrial facility 98 (4.2%)

Sports/leisure facility 123 (5.3%)

Road/motorway 212 (9.2%)

Public facility/restaurants 166 (7.2%)

Hospital/nursing facility 131 (5.7%)

Educational facility 6 (0.3%)

Unknown 13 (0.6%)

Bystander CPR

Chest compression only 1225 (53.1%)

Chest compression and rescue breathing 60 (2.6%)

No CPR 1024 (44.3%)

AED by bystander 58 (2.5%)

initial cardiac rhythm

Ventricular fibrillation or tachycardia 481 (20.8%)

Pulseless electrical activity 423 (18.3%)

Asystole 1405 (60.8%)

EMS response time, minutes 9.28 +- 8.27 Life support by EMT

Defibrillation 623 (27.0%)

Airway management 1934 (83.8%)

Drug administration 337 (14.6%)

Outcome

Prehospital ROSC 301 (13.0%)

Survival to discharge 258 (11.2%)

Favorable neurological outcome 174 (7.5%)

Data was reported as mean +- standard deviation for continuous variables and fre- quency (percentage) for categorical variables.

OHCA, out-of-hospital cardiac arrest; CPR, cardiopulmonary resuscitation; AED, automated external defibrillator; EMS, emergency medical services; EMT, emer- gency medical technician; ROSC, return of spontaneous circulation.

(Fig. 1). A total of 878 patients, including 99 patients whose OHCA episode occurred in ambulance, 67 patients aged <18 years, and 712 patients with missing or incomplete medical record were excluded from final analysis.

Table 1 displays patient characteristics, EMS factor, time inter- val, post resuscitation care, and outcome for all OHCA patients. During the study period, the average response time of EMT was

9.28 +- 8.27 min. Rates of prehospital ROSC, survival to discharge, and favorable neurologic outcome were 13.0%, 11.2%, and 7.5%, respectively.

Cumulative rates for patients’ outcome

Cumulative rates drawn with line graph for prehospital ROSC, survival to discharge, favorable neurologic outcome, and cumula- tive number drawn with bar charts of all OHCA patients are shown in Fig. 2. Cumulative numbers of patients with OHCA receiving bystander CPR and witnessed OHCA patients are also displayed. Cumulative rates for all outcomes in patients with OHCA receiving bystander CPR and witnessed OHCA patients were always higher than those for all OHCA patients. With increasing response time, cumulative rates for all outcomes decreased and plateaued after 10 min.

Response time threshold measured by ROC curve analysis

ROC curve analysis was performed to determine the optimal response time threshold to predict prehospital ROSC, survival dis- charge, and favorable neurological outcome. Response time thresh- old for prehospital ROSC was 11.5 min. It was 7.5 min for survival to discharge and favorable neurologic outcome (Table 2). In patients with OHCA receiving bystander CPR and witnessed OHCA patients, the response time threshold was the same as the thresh- old for all OHCA patients.

Characteristics of patients with response time of 7.5 min and

11.5 min are shown in Tables A and B. There was no significant dif- ference in any variable between the two groups except for the administration of epinephrine.

Adjusted odds ratios for patients’ outcome

Table 3 displays adjusted ORs for prehospital ROSC, survival to discharge, and favorable neurologic outcome in OHCA patients with response time <= 7.5 min or <= 11.5 min per min shorter. Patients in the <=7.5 min response time group showed increased odds of survival to discharge (OR: 1.54, 95% CI: 1.13-2.10, p = .006) and favorable neurologic recovery (OR: 2.01, 95% CI: 1.36-2.99, p = .001) than patients in the >7.5 min response time

group. Patients in the <=11.5 min response time group showed increased odds of all outcomes than patients in the >11.5 min response time group (survival to discharge, OR: 1.91, 95% CI: 1.20-3.04, p = .007; favorable neurological outcome, OR: 3.94, 95% CI: 2.01-7.74, p < .001). When response time was decreased by 1 min, all outcomes were improved (survival to discharge, OR: 1.08, 95% CI: 1.04-1.12, p < .001; favorable neurological outcome,

OR: 1.14, 95% CI: 1.07-1.21, p < .001).

Discussion

This was a prospective observational study that evaluated the relationship between EMS response time and neurologic outcomes among OHCA patients regardless of bystander-witness. We found that shorter EMS response time was associated with favorable neu- rologic outcome in patients with OHCA of cardiac origin. The response time threshold for favorable neurologic outcome was

7.5 min. This study showed that the optimal response time thresh- old to predict outcome of all OHCA patients, OHCA patients receiv- ing bystander CPR, or witnessed OHCA patients was not different. BLS is the corner stone for saving lives after cardiac arrest. Fun- damental aspects of BLS include immediate recognition of sudden cardiac arrest and rapid activation of the EMS, early CPR, and rapid defibrillation with an Automated external defibrillator [14- 16]. If the bystander can recognize cardiac arrest and start CPR and EMT arrives quickly, survival rate and neurologic recovery rate are expected to increase. Eisenberg et al. [17] have assessed the relationship between response time and survival of OHCA patients. They originally reported a response time threshold of 4 min in 1979. Several recent studies have concluded decrease in EMS response time is associated with improved survival of OHCA. How- ever, these reports did not show the optimal threshold for survival of OHCA or neurological outcome. To demonstrate the efficacy of pure response time, the study should be directed to bystander- witnessed cardiac arrest patients. Ono et al. [1] have shown evi- dence to support the hypothesis that a shorter response time is clo- sely associated with favorable neurological outcomes in witnessed patients with OHCA in a retrospective study. They also demon- strated that the response time threshold associated with favorable neurologic outcomes was 6.5 min for all patients with witnessed OHCA and 7.5 min for patients who received bystander CPR. We theorize that a shorter response time is associated with improved neurological outcomes of all OHCA patients of cardiac origin regardless of witness. This study is the first attempt to assess this

response time threshold in a prospective fashion.

The response time threshold associated with favorable neuro- logical outcome in our study was 7.5 min, which was the same as that in a previous study [6]. Nevertheless, whether the threshold

Fig. 2. Cumulative rates of all patients, witnessed patients and patients receiving bystander CPR for (A) prehospital ROSC, (B) survival to discharge and (C) favorable neurological outcome. The cumulative rates (line graphs) for prehospital ROSC, survival to discharge, favorable neurologic outcome, and the cumulative number (bar charts) of all OHCA patients are shown. The cumulative rates of all patients for prognosis decreases over time. CPR, cardiopulmonary resuscitation; ROSC, return of spontaneous circulation.

Table 2

Diagnostic performance of good outcome.

Outcome	THRy	SEN	SPE	ACC	PPV	NPV	AUC (95% CI)
All patients Prehospital ROSC	11.5	263/301 (87.4%)	363/2008 (18.1%)	626/2309 (27.1%)	263/1908 (13.8%)	363/401	0.524
						(90.5%)	(0.490-0.558)
Survival to discharge	7.5	156/258 (60.5%)	1017/2051 (49.6%)	1173/2309 (50.8%)	156/1190 (13.1%)	1017/1119 (90.9%)	0.567
Favorable neurological outcome	7.5	111/174 (63.8%)	1056/2135 (49.5%)	1167/2309 (50.5%)	111/1190	1056/1119 (94.4%)	(0.531-0.604) 0.587
Patients received bystander CPR					(9.3%)		(0.545-0.629)
Prehospital ROSC	11.5	191/221 (86.4%)	174/1064 (16.4%)	365/1285 (28.4%)	191/1081 (17.7%)	174/204	0.512
						(85.3%)	(0.471-0.554)
Survival to discharge	7.5	107/179 (59.8%)	559/1106 (50.5%)	666/1285 (51.8%)	107/654	559/631	0.559
					(16.4%)	(88.6%)	(0.513-0.604)
Favorable neurological outcome	7.5	84/136	579/1149 (50.4%)	663/1285 (51.6%)	84/654	579/631	0.576
Patients witnessed		(61.8%)			(12.8%)	(91.8%)	(0.526-0.626)
Prehospital ROSC	11.5	195/224 (87.1%)	196/1050 (18.7%)	391/1274 (30.7%)	195/1049 (18.6%)	196/225	0.521
						(87.1%)	(0.481-0.562)
Survival to discharge	7.5	128/206 (62.1%)	538/1068 (50.4%)	666/1274 (52.3%)	128/658	538/616	0.585
					(19.5%)	(87.3%)	(0.543-0.627)
Favorable neurological outcome	7.5	91/138 (65.9%)	569/1136 (50.1%)	660/1274 (51.8%)	91/658	569/616	0.607
					(13.8%)	(92.4%)	(0.56-0.654)

y Threshold was computed by Youden’s index.

Sensitivity, specificity, accuracy, PPV and NPV were calculated from the threshold and the 95% CI of AUC was computed by Delong’s method. AUC of the splenic factor was compared to the AUC of splenic volume by Delong’s method and the p-values were corrected by Bonferroni’s method.

THR, threshold; SEN, sensitivity; SPE, specificity; ACC, accuracy; PPV, positive predictive value; NPV, negative predictive value; CI, confidence interval.

Table 3

Adjusted odd ratios for outcomes in all patients.

Outcome response time interval (min)

	Per min shorter			<=7.5			<=11.5
	OR (95% CI)	p-value		OR (95% CI)	p-value		OR (95% CI)	p-value
Prehospital ROSC	0.97 (0.94-1.01)	0.10		1.14 (0.86-1.52)	0.36		1.59 (1.05-2.40)	0.03
Survival to discharge	0.93 (0.89-0.97)	<0.001		1.54 (1.13-2.10)	0.006		1.91 (1.20-3.04)	0.007
Favorable neurological outcome	0.88 (0.83-0.93)	<0.001		2.01 (1.36-2.99)	0.001		3.94 (2.01-7.74)	<0.001

OR, odds ratio; CI, confidence interval; ROSC, return of spontaneous circulation.

for response time that predicts survival to discharge, favorable neurological outcomes in patients with OHCA receiving bystander CPR, and witnessed OHCA patients is the same as that for all OHCA patients remains unclear. This might be due to the fact that the number of patients was not enough to address the relationship between bystander-CPR witnessed and response time. Recent studies of OHCA have reported that bystander CPR with a shorter EMS response time is associated with improved survival and neu- rological outcome [1,5,18]. Bystander-witness and bystander CPR were independently associated with clinical outcomes in our study. Although there was no statistical significance, adjusted ORs were higher in bystander witness and bystander CPR for clin- ical outcomes. Another possibility is that differences might have occurred due to prehospital characteristics of OHCA. lay rescuer AED use is rare in South Korea [19]. The bystander CPR rate was increased by dispatcher assisted bystander CPR using a simplified protocol based on compression only CPR [20]. Although CPR train- ing and education have been implemented every year, the quality of bystander CPR performance and dispatcher CPR instruction remain variable. Further research on the quality of bystander CPR and response time is needed.

This study has several limitations. First, although prospectively

collected data were used, 712 patients (22.3%) with incomplete or missing important variables were excluded in the analysis. This might have resulted in some selection bias. Second, we could not analyze the relationship between bystander CPR witnessed and response time. Further study including more OHCA patients is needed. Third, 1 or 6-month outcome after cardiac arrest was not

included in the analysis. Thus, more accurate clinical outcomes were not evaluated in this investigation.

Conclusions

Shorter EMS response time can lead to favorable neurologic out- come in patients with OHCA of presumed cardiac origin. EMS response time threshold associated with improved favorable out- come was <=7.5 min.

Conflicts of interest

The authors have no conflict of interest to report.

Source(s) of support

This work was supported by a grant funded by the Soonchun- hyang University, College of Medicine (Grant number 20180012).

Acknowledgements

This consortium was supported by the Korean Centers for Dis- ease Control and Prevention during the organizing stage. Currently, KoCARC is partly supported by the Korean Association of Cardiopulmonary Resuscitation.

Appendix A

Table A

Characteristics of all patients according to response time interval 11.5 min

Response time interval	<=11.5	>11.5	p-value
	(N = 1908)	(N = 401)
Age, years	67.2 +- 15.1	66.7 +- 16.0	0.707
Male sex	1252 (65.6%)	267 (66.6%)	0.755
Witnessed arrest	1049 (55.0%)	225 (56.1%)	0.72
Occurrence place			0.125

References

1. Ono Y, Hayakawa M, Iijima H, et al. The response time threshold for predicting favourable neurological outcomes in patients with bystander-witnessed out- of-hospital cardiac arrest. Resuscitation 2016;107:65-70.
2. De Maio VJ, Stiell IG, Wells GA, et al. Optimal defibrillation response intervals for maximum out-of-hospital Cardiac arrest survival rates. Ann Emerg Med 2003;42:242-50.
3. Earnest A, Hock Ong ME, Shahidah N, et al. spatial analysis of ambulance Response times related to prehospital cardiac arrests in the city-state of Singapore. Prehosp Emerg Care 2012;16:256-65.
4. Gold LS, Fahrenbruch CE, Rea TD, et al. The relationship between time to arrival

   Housing facility	1303 (68.3%)	257 (64.1%)	of emergency medical services (EMS) and survival from out-of-hospital
   Industrial facility	80 (4.2%)	18 (4.5%)	ventricular fibrillation cardiac arrest. Resuscitation 2010;81:622-5.
   Sports/leisure facility	92 (4.8%)	31 (7.7%)	[5] Goto Y, Funada A, Goto Y. Relationship between emergency medical services
   Road/motorway	174 (9.1%)	38 (9.5%)	response time and bystander intervention in patients with out-of-hospital

   Public facility/restaurants 143 (7.5%) 23 (5.7%)

   Hospital/nursing facility 100 (5.2%) 31 (7.7%)

   Educational facility 5 (0.3%) 1 (0.2%)

   Unknown 11 (0.6%) 2 (0.5%)

   Bystander CPR 0.073

   Chest compression only	1033 (54.1%)	192 (47.9%)
   Chest compression and rescue	48 (2.5%)	12 (3.0%)
   breathing
   No CPR	827 (43.3%)	197 (49.1%)

   cardiac arrest. J Am Heart Assoc 2018;7.

   O’Keeffe C, Nicholl J, Turner J, et al. Role of Ambulance response times in the survival of patients with out-of-hospital cardiac arrest. Emerg Med J 2011;28:703-6.

5. Pell JP, Sirel JM, Marsden AK, et al. Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study. Bmj 2001;322:1385-8.
6. Valenzuela TD, Roe DJ, Cretin S, et al. Estimating effectiveness of cardiac arrest interventions: a logistic regression survival model. Circulation 1997;96:3308-13.

   AED by bystander	49 (2.6%)	9 (2.2%)	0.841	[9] Kim JY, Hwang SO, Shin SD, et al. Korean cardiac arrest research consortium
   Initial cardiac rhythm			0.16	(KoCARC): rationale, development, and implementation. Clin Exp Emerg Med
   Ventricular fibrillation or	405 (20.2%)	76 (18.9%)		2018;5:165-76.
   tachycardia				[10] Jacobs I, Nadkarni V, Bahr J, et al. Cardiac arrest and cardiopulmonary
   Pulseless electrical activity	360 (18.9%)	63 (15.7%)		resuscitation outcome reports: update and simplification of the Utstein
   Asystole	1143 (59.9%)	262 (65.3%)		templates for resuscitation registries: a statement for healthcare

   Life support by EMT

   Defibrillation 521 (27.3%) 102 (25.4%) 0.481

   Airway management 1598 (83.8%) 336 (83.8%) 1

   Drug administration 256 (13.4%) 81 (20.2%) 0.001

   Outcome

   Prehospital ROSC	263 (13.8%)	38 (9.5%)	0.025
   survival discharge	230 (12.1%)	28 (7.0%)	0.004
   Favorable neurological outcome	162 (8.5%)	12 (3.0%)	<0.001

   Data was reported as mean +- standard deviation for continuous variables and fre- quency (percentage) for categorical variables.

   OHCA, out-of-hospital cardiac arrest; CPR, cardiopulmonary resuscitation; AED, automated external defibrillator; EMS, emergency medical services; EMT, emer- gency medical technician; ROSC, return of spontaneous circulation.

   Table B

   Characteristics of all patients according to response time interval 7.5 min

   Response time interval	<=7.5		>7.5	p-value
   	(N = 1190)		(N = 1119)

   professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Councils of Southern Africa). Circulation 2004;110:3385-97.

   Langhelle A, Nolan J, Herlitz J, et al. Recommended guidelines for reviewing, reporting, and conducting research on post-resuscitation care: the Utstein style. Resuscitation 2005;66:271-83.

7. Kang K, Kim T, Ro YS, et al. prehospital endotracheal intubation and survival after out-of-hospital cardiac arrest: results from the Korean nationwide registry. Am J Emerg Med 2016;34:128-32.
8. Ro YS, Shin SD, Song KJ, et al. A trend in epidemiology and outcomes of out-of- hospital cardiac arrest by urbanization level: a nationwide observational study from 2006 to 2010 in South Korea. Resuscitation 2013;84:547-57.
9. Kleinman ME, Brennan EE, Goldberger ZD, et al. Part 5: adult basic life support and cardiopulmonary Resuscitation quality: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132:S414-35.
10. Perkins GD, Handley AJ, Koster RW, et al. European resuscitation council guidelines for resuscitation 2015: section 2. Adult basic life support and Automated external defibrillation. Resuscitation 2015;95:81-99.

    Age, years	67.4 +- 15.2	66.9 +- 15.4	0.41	[16] Perkins GD, Travers AH, Berg RA, et al. Part 3: Adult basic life support and
    Male sex	782 (65.7%)	737 (65.9%)	0.975	automated external defibrillation: 2015 international consensus on
    Witnessed arrest Occurrence place	658 (55.3%)	616 (55.0%)	0.939 0.006	cardiopulmonary resuscitation and emergency cardiovascular care science with treatment recommendations. Resuscitation 2015;95:e43-69.

    Eisenberg MS, Bergner L, Hallstrom A. Cardiac resuscitation in the community. Importance of rapid provision and implications for program planning. Jama 1979;241:1905-7.

    Housing facility	805 (67.6%)	755 (67.5%)
    Industrial facility	49 (4.1%)	49 (4.4%)
    Sports/leisure facility	48 (4.0%)	75 (6.7%)
    Road/motorway	117 (9.8%)	95 (8.5%)
    Public facility/restaurants	101 (8.5%)	65 (5.8%)
    Hospital/nursing facility	60 (5.0%)	71 (6.3%)
    Educational facility	5 (0.4%)	1 (0.1%)
    Unknown	5 (0.4%)	8 (0.7%)

    Rajan S, Wissenberg M, Folke F, et al. Association of bystander cardiopulmonary resuscitation and survival according to ambulance response times after out-of- hospital cardiac arrest. Circulation 2016;134:2095-104.

11. Bae H. Legal aspects of the application of the lay rescuer automatic external

    Bystander CPR Chest compression only	622 (52.3%)	0.73 603 (53.9%)	defibrillator (AED) program in South Korea. J Emerg Med 2008;34:299-303. [20] Song KJ, Shin SD, Park CB, et al. Dispatcher-assisted bystander
    Chest compression and rescue	32 (2.7%)	28 (2.5%)	cardiopulmonary resuscitation in a metropolitan city: a before-after

    population-based study. Resuscitation 2014;85:34-41.

    breathing
    No CPR	536 (45.0%)	488 (43.6%)
    AED by bystander	30 (2.5%)	28 (2.5%)	1
    Initial cardiac rhythm			0.538
    Ventricular fibrillation or	263 (22.1%)	218 (19.5%)
    tachycardia
    Pulseless electrical activity	217 (18.2%)	206 (18.4%)
    Asystole	710 (59.7%)	695 (62.1%)
    Life support by EMT
    Defibrillation	331 (27.8%)	292 (26.1%)	0.377
    Airway management	1000 (84.0%)	934 (83.5%)	0.755
    Drug administration	149 (12.5%)	188 (16.8%)	0.004
    Outcome
    Prehospital ROSC	165 (13.9%)	136 (12.2%)	0.246
    Survival discharge	156 (13.1%)	102 (9.1%)	0.003
    Favorable neurological outcome	111 (9.3%)	63 (5.6%)	0.001

    Data was reported as mean +- standard deviation for continuous variables and fre- quency (percentage) for categorical variables.

    OHCA, out-of-hospital cardiac arrest; CPR, cardiopulmonary resuscitation; AED, automated external defibrillator; EMS, emergency medical services; EMT, emer- gency medical technician; ROSC, return of spontaneous circulation.