Article, Emergency Medicine

Scene time interval and good neurological recovery in out-of-hospital cardiac arrest

a b s t r a c t

Objectives: It is unclear whether scene time interval (STI) is associated with better neurological recovery in the emergency medical service (EMS) system with intermediate service level.

Methods: Adult out-of-hospital cardiac arrest (OHCA) patients with presumed cardiac etiology (2012 to 2014) were analyzed, excluding patients not-resuscitated, occurred in ambulance/medical/nursing facility, unknown STI or extremely longer STI (N 60 min), and unknown outcomes. STI was classified into short (0.0-3.9 min), mid- dle (4.0-7.9 min), long (8.0-11.9 min), and very-long (12.0-59.9 min), respectively. The end point was a good Cerebral Performance Category 1 or 2. Multivariable logistic regression by STI group (reference = short)

was performed to calculate adjusted odds ratios (AORs) with 95% confidence intervals (95% CIs) for outcomes with or without interaction term (STI * prehospital return of spontaneous circulation, (PROSC)).

Results: Of 79,832 OHCA patients, 41,054 cases were analyzed; good CPC in the short (3.0%), middle (3.2%), long

(3.0%), and very-long (2.9%) STI groups were similar, respectively (p = 0.55). The AORs (95% CI) for good CPC in the final model without interaction term were 0.74 (0.58-0.95) for the middle, 0.51 (0.39-0.67) for the long, and

0.45 (0.33-0.61) for the very-long STI group (reference = short STI). The AORs in PROSC group were 1.18 (0.97- 1.44) for middle STI group, 0.72 (0.57-0.92) for long group, and 0.56 (0.42-0.77) for very-long group. The AORs in non-PROSC group were 1.22 (1.06-1.40) for middle STI group, 0.82 (0.70-0.96) for long group, and 0.70 (0.57- 0.85) for very-long group.

Conclusion: The middle STI (4-7 min) was associated with the highest odds of neurological recovery for patients who could not be restored in the field. The STI may be a clinically useful predictor of good neurology outcome in victims of cardiac arrest.

(C) 2017

  1. Introduction

Out-of-hospital cardiac arrest (OHCA) is a major public health prob- lem [1]. However, the survival after OHCA remains low b 10% in most communities [2-4]. Rapid dispatch of ambulance and followed by car- diopulmonary resuscitation (CPR) and defibrillation by emergency medical services (EMS) providers has been the most important core

? Funding acknowledgement: This study was supported by the National Emergency Management Agency of Korea and the Korea Centers for Disease Control and Prevention (CDC). The study was funded by the Korea CDC (2012-2014; Grant No.: 2012-E33010- 00; 2013-E33015-00; 2014-E33011-00).

* Corresponding author at: Department of Emergency Medicine, Seoul National University College of Medicine, 101 Daehak-Ro, Jongno-Gu, Seoul 110-744, Seoul, Republic of Korea.

E-mail addresses: [email protected] (K.H. Kim), [email protected] (S.D. Shin).

intervention for improving outcomes [5,6]. The response time interval (RTI), duration from the call for ambulance to arrival at the scene, has known to be the independent risk factor for survival.

However, the association between the scene time interval (STI) from the ambulance’s arrival at the scene to its departure for the hospital and CPR outcomes has been relatively unclear. It is assumed that as the STI increases, the chance of ROSC will increase up to a certain time point and decrease after that point. Longer resuscitation efforts by EMS pro- viders in the field can offset the benefits by management at the ED where could remove the cause of arrest or the cause of sustained arrest status. Shorter STIs will provide chance for much faster access to ED care by emergency physicians but may cost the chance of ROSC with stable, high-quality CPR via a longer CPR duration in the field [7].

If a patient resumes spontaneous circulation very quickly after EMS arrives on the scene, the STI will be very short and will result in good

http://dx.doi.org/10.1016/j.ajem.2017.05.049

0735-6757/(C) 2017

neurological recovery. Therefore, prehospital ROSC will be both the result and the cause of the shorter STI. It is unclear whether longer STIs are associated with outcomes according to the prehospital ROSC (PROSC) status in the field or not. This study aims to determine the association between STI length and good neurological recovery to develop the optimal STI protocol for field CPR. Additionally, we com- pared the effect of the STI across the PROSC on neurological recovery.

  1. Methods

The study was approved by the institutional review board of the study institution, and the Korea Centers for Disease Control and Preven- tion reviewed the study and approved the use of data. Data collection with exemption of personal contents and waiver of individual institu- tional ethic review board were allowed because the data has no private information on the patients and hospital.

Study setting

The Korean EMS, which is a single-tiered and government-based sys- tem, provides a basic-to-intermediate level of ambulance services from the sixteen provincial headquarters of the national fire department and supports a population of approximately 50 million. Ambulance crews and emergency medical technicians can administer CPR at the scene and during transport with automatic external defibrillation and advanced airway management under direct medical control. Advanced cardiac life support including medications is limited in most areas of the country, but it is available in emergency departments (EDs). Most ambu- lances have three crew members, including one level-1 EMT (equivalent to EMT-intermediate in the North American EMS), one level-2 EMT (equivalent to EMT-basic), and one driver. Most level-1 EMTs have grad- uated from an EMT college with three to four years of education lead- ing to EMT certification. Level-2 EMTs are firefighters; they become level-2 EMTs through a shorter training program (approximately 150 h) focusing on basic skills and knowledge. All EMTs in the fire

department are required to participate in 20 h of CME credit per year [8,9]. CPR protocols are described in the national EMS standard for skills and procedures. The national CPR protocol adapted the 2010 American Heart Association Guidelines [10] (Appendix 1). An advanced airway, including supraglottic airway insertion or endotra- cheal intubation and bag valve mask, can be performed by the highest level of EMT. There are a few feedback defibrillators that can provide real-time information about the CPR quality and no mechanical CPR devices available in this EMS setting. The CPR protocol encourages EMTs to stay in the field for at least two cycles of 2 min CPR including 30:2 compressions and ventilations.

Study design and data source

This study is a nation-wide, retrospective, and Cross sectional study. We used the population-based National cardiac arrest registry, which has been compiled since 2006 via collaboration between the Korea Cen- ters for Disease Control and Prevention (CDC), which is under the Min- istry of Health and Welfare, and the Central Fire Services (CFS), the national fire department), which is under the Ministry of Public Safety and Security. Data for all EMS-assessed OHCA patients were collected from the EMS run sheet and the in-depth cardiac arrest EMS registry, and Utstein information was collected from the CFS data server. The medical records for these patients were reviewed to determine their hospital care and outcomes, including survival and neurological recov- ery. Primary EMS records are written by the highest level EMTs in the attending ambulance and are stored on the CFS data server. All cardiac arrest records written by EMTs were obligatory reviewed by a designat- ed EMS medical director who is a part-time physician employed by each local EMS agency. The medical directors reviewed all records on OHCAs and scored the quality of care on the basis of the national standards for medical oversight. This process maintained the quality of EMS CPR data. The medical record reviews were performed by Korea CDC research team members who are professional medical record reviewers trained to use the cardiac arrest registry. Approximately 15 medical record

Fig. 1. Study population. OHCA: out-of-hospital cardiac arrest, EMS: emergency medical services, STI: scene time interval.

Fig. 2. Restricted cubic spline analysis for good neurological recovery by temperature at the scene. Log odds of good neurological recovery according to scene time interval was calculated, using the restricted cubic spline analysis adjusting age group (10-year interval), gender, urbanization (rural and urban versus metropolitan), witnessed, location of arrest (public versus private versus unknown), bystander cardiopulmonary resuscitation, primary ECG (Ventricular fibrillation/ventricular tachycardia versus pulseless electrical activity versus asystole), response time interval, transport time interval, airway management (endotracheal intubation versus supraglottic airway, bag valve mask, passive oxygenation), prehospital defibrillation, level of emergency department (level-1 versus level-2 versus level-3).

reviewers visited between 600 and 700 hospitals to collect the hospital care and outcomes information. This information is stored on the Korea CDC data server, along with the hospital care and outcomes information from the original EMS dataset. The Data quality management committee (QMC) was charged with maintaining data quality via a data Quality management protocol throughout the study project. Monthly QMC meetings provided feedback to all medical record reviewers to ensure the data quality. The QMC was comprised of emergency medicine physicians, EMS physicians, critical care experts, epidemiologists, and bio-statistical experts [9, 11].

Study participants

All adult (18 years and older) OHCA patients with presumed cardiac etiology between January 2012 and December 2014 were included. Pa- tients who had an unknown neurological status at hospital discharge, had not been resuscitated by EMS personnel, collapsed in the ambu- lance or at nursing/medical facilities, were successfully defibrillated by a lay person prior to EMT arrival, had unknown STI information, or had an STI longer than 60 min were excluded.

Data variables

The exposure variable was STI, which was defined from the ambulance’s arrival time at the scene to its departure from the field. The time interval was measured using an automatic Tracking system connected with the dispatcher center. The time interval included the time from ambulance wheel arrival to reaching the patient’s side. The confounding variables were the year of the event; patient age; patient gender; rural versus urban area (population N 1 million); location (pub- lic or private); presence of witnesses; bystander CPR; response time (from call to arrival on the scene); transport time (from departure to ar- rival at ED); prehospital defibrillation; airway management, such as en- dotracheal intubation (ETI), Supraglottic airway , bag valve mask ventilation, and passive ventilation with oxygen mask (PVO2); level of ED (regional ED, local ED, and local emergency room); and prehospital ROSC (PROSC).

Outcome measure

The primary outcome was good neurological recovery, which was defined as a cerebral performance category score of 1 (good performance, no neurological disability) or 2 (moderate disability, can work) on a five-point scale; the remaining categories were 3 (se- vere disability, dependent), 4 (a vegetative state), and 5 (dead) [12]. The CPC score was determined by medical record reviewers employed at the Korea CDC and was based on discharge summary abstracts. Discharge summary abstracts or documentation are drafted by inpatient care doctors, and the summaries are usually reviewed and revised for greater validity by the hospital medical record review team because the summaries are utilized for national health insurance claim data after discharge [13]. The secondary outcome was survival to discharge, which was determined on the basis of medical record and discharge summary by medical record reviewers of the Korea CDC.

Statistical analysis

Descriptive analysis was performed using the chi-square test for cat- egorical variables. The restricted cubic spline (rcspilne) smooth function of good neurological recovery was used as a predictor to obtain the crude relationship between the predicted (smoothed) good neurologi- cal recovery rate and STI interval, and the data were plotted and smoothed using the Stata 11.0 software (StataCorp, College Station, TX, USA). Log odds of good neurological recovery according to scene time interval was calculated, using the rcspline analysis adjusting age group (10-year interval), gender, urbanization (rural and urban versus metropolitan), witnessed, location of arrest (public versus private ver- sus unknown), bystander cardiopulmonary resuscitation, primary ECG (ventricular fibrillation/ventricular tachycardia versus pulseless electri- cal activity versus asystole), response time interval, transport time in- terval, airway management (endotracheal intubation versus supraglottic airway, bag valve mask, passive oxygenation), prehospital defibrillation, level of emergency department (level-1 versus level-2 versus level-3).

From the rcspline function, the STI was categorized into four groups; short (0 b STI b 4 min), middle (4 <= STI b 8 min), long (8

<= STI b 12 min), and very-long (12 <= STI b 60 min). Short STI was

used as the reference group in the multivariable model to determine the association between STI group and neurological recovery status (good CPC), adjusting for the confounders. Adjusted odds ratios (AORs) and 95% confidence intervals (95% CIs) were calculated from the model with adjustment for age group (10-year interval), gender, urbanization, primary ECG, location, witness, bystander CPR, response time and transport time intervals (4 min interval groups), airway management, prehospital defibrillation, prehospital ROSC, and ED level.

We assessed the interaction between main exposure (STI) and PROSC status for main outcomes. To compare the effect size of STI group on the outcomes across the PROSC status, we added the in-

teraction term (STI group * PROSC status) to the above final multi- variable logistic regression mode. AORs with 95% CIs by the STI

groups under negative PROSC and under positive PROSC on out- comes were calculated and compared. Statistical analyses were performed with SAS software version 9.3 (SAS Institute, Inc., Cary, NC, USA).

  1. Results
    1. Demographic findings

From 79,832 eligible OHCA patients, we excluded patients who were younger than 18 years old at the time of arrest (N = 1968), had non-car- diac causes of arrest (N = 21,520), had not been resuscitated (N =

Table 1

Demographic findings of study subjects by scene time interval group.

Factors Total Scene time interval p-Value

Short STI Middle STI Long STI Very-long STI

N

%

0-4

%

4-8

%

8-12

%

12-

%

Total

41,054

100.0

7348

100.0

15,110

100.0

11,415

100.0

7181

100.0

Year

2012

11,952

29.1

2934

39.9

4395

29.1

2835

24.8

1788

24.9

2013

13,815

33.7

2295

31.2

5098

33.7

4020

35.2

2402

33.4

2014

15,287

37.2

2119

28.8

5617

37.2

4560

39.9

2991

41.7

Gender

b0.001

Female

14,688

35.8

2749

37.4

5558

36.8

4075

35.7

2306

32.1

Male

26,366

64.2

4599

62.6

9552

63.2

7340

64.3

4875

67.9

Age, year

b0.001

18-30

546

1.3

109

1.5

179

1.2

164

1.4

94

1.3

31-39

1243

3.0

192

2.6

393

2.6

355

3.1

303

4.2

40-49

3584

8.7

629

8.6

1265

8.4

1000

8.8

690

9.6

50-59

6648

16.2

1195

16.3

2373

15.7

1844

16.2

1236

17.2

60-69

7007

17.1

1225

16.7

2610

17.3

1924

16.9

1248

17.4

70-79

11,855

28.9

2121

28.9

4489

29.7

3255

28.5

1990

27.7

80-

10,171

24.8

1877

25.5

3801

25.2

2873

25.2

1620

22.6

Urbanization

b0.001

Rural and urban

23,017

56.1

4607

62.7

8022

53.1

6085

53.3

4303

59.9

Metropolitan

18,037

43.9

2741

37.3

7088

46.9

5330

46.7

2878

40.1

Witness status

b0.001

No

15,773

38.4

2681

36.5

5955

39.4

4515

39.6

2622

36.5

Yes

17,838

43.5

2990

40.7

6506

43.1

5035

44.1

3307

46.1

Unknown

7443

18.1

1677

22.8

2649

17.5

1865

16.3

1252

17.4

Location of arrest

b0.001

Public

7790

19.0

2074

28.2

2844

18.8

1768

15.5

1104

15.4

Private

33,085

80.6

5232

71.2

12,205

80.8

9604

84.1

6044

84.2

Other

179

0.4

42

0.6

61

0.4

43

0.4

33

0.5

Bystander CPR

b0.001

Yes

21,160

51.5

4508

61.4

7496

49.6

5613

49.2

3543

49.3

No

17,903

43.6

2375

32.3

6901

45.7

5317

46.6

3310

46.1

Unknown

1991

4.8

465

6.3

713

4.7

485

4.2

328

4.6

Primary ECG

b0.001

VF/pulseless VT

3622

8.8

610

8.3

1305

8.6

1039

9.1

668

9.3

PEA

2592

6.3

526

7.2

1026

6.8

661

5.8

379

5.3

Asystole

34,840

84.9

6212

84.5

12,779

84.6

9715

85.1

6134

85.4

EMS response time, minutes

b0.001

b0.001

0-4

2688

6.5

561

7.6

961

6.4

697

6.1

469

6.5

4-8

20,282

49.4

3353

45.6

7644

50.6

5903

51.7

3382

47.1

8-12

9990

24.3

1808

24.6

3693

24.4

2754

24.1

1735

24.2

12-

8094

19.7

1626

22.1

2812

18.6

2061

18.1

1595

22.2

EMS transport time, minutes

b0.001

b0.001

0-4

5478

13.3

880

12.0

1953

12.9

1611

14.1

1034

14.4

4-8

17,861

43.5

2762

37.6

6641

44.0

5288

46.3

3170

44.1

8-12

8722

21.2

1626

22.1

3303

21.9

2360

20.7

1433

20.0

12-

8993

21.9

2080

28.3

3213

21.3

2156

18.9

1544

21.5

Airway management

b0.001

ETI

1136

2.8

76

1.0

212

1.4

383

3.4

465

6.5

SGA

2000

4.9

177

2.4

575

3.8

681

6.0

567

7.9

BVM

30,315

73.8

5405

73.6

11,612

76.8

8365

73.3

4933

68.7

PV

7603

18.5

1690

23.0

2711

17.9

1986

17.4

1216

16.9

Prehospital defibrillation

b0.001

No

31,948

77.8

6021

81.9

12,020

79.5

8734

76.5

5173

72.0

Yes

9106

22.2

1327

18.1

3090

20.5

2681

23.5

2008

28.0

Prehospital ROSC

No

39,361

95.9

7116

96.8

14,486

95.9

10,907

95.5

6852

95.4

b0.001

Yes

1693

4.1

232

3.2

624

4.1

508

4.5

329

4.6

ED level

b0.001

Level 1

4564

11.1

717

9.8

1602

10.6

1344

11.8

901

12.5

Level 2

20,147

49.1

3483

47.4

7653

50.6

5686

49.8

3325

46.3

Level 3

16,343

39.8

3148

42.8

5855

38.7

4385

38.4

2955

41.2

Outcomes

Survival to discharge

2287

5.6

458

6.2

882

5.8

596

5.2

351

4.9

0.001

Good CPC

1256

3.1

217

3.0

484

3.2

348

3.0

207

2.9

0.550

STI: scene time interval, CPR: cardiopulmonary resuscitation, VF: ventricular fibrillation, VT: ventricular tachycardia, PEA: pulseless electrical activity, EMS: emergency medical service, ETI: endotracheal intubation, SGA: supraglottic airway, BVM: bag valve mask, PV: passive ventilation, ED: emergency department, CPC: cerebral performance scale.

7684), had collapsed in the ambulance or at the nursing home/local medical facilities (N = 5387), had incomplete medical records (N

= 2207), and had an unknown neurological status at discharge (N

= 12). The remaining 41,054 patients were included in the final

analysis. Of those, 1693 (3.2%) recovered spontaneous circulation during EMS-provided CPR in the field, and 1256 (3.0%) patients were discharged with a good neurological outcome of CPC 1 or 2 (Fig. 1).

Table 2

Comparison of risk factors for prehospital ROSC and prehospital non-ROSC patients.

Factors Total Prehospital ROSC p-Value

Yes No

N

%

N

%

No

%

Total Year

41,054

100.0

1693

100.0

39,361

100.0

b 0.001

2012

11,952

29.1

393

23.2

11,559

29.4

2013

13,815

33.7

560

33.1

13,255

33.7

2014

Gender

15,287

37.2

740

43.7

14,547

37.0

b 0.001

Female

14,688

35.8

348

20.6

14,340

36.4

Male Age, year

18-30

26,366

546

64.2

1.3

1345

66

79.4

3.9

25,021

480

63.6

1.2

b 0.001

31-39

1243

3.0

105

6.2

1138

2.9

40-49

3584

8.7

327

19.3

3257

8.3

50-59

6648

16.2

507

29.9

6141

15.6

60-69

7007

17.1

317

18.7

6690

17.0

70-79

11,855

28.9

256

15.1

11,599

29.5

80-

Urbanization

10,171

24.8

115

6.8

10,056

25.5

b 0.001

Fig. 2 shows the association between log odds of good neurological recovery according to scene time interval from the adjusted model using the restrict cubic spline analysis. The log odds of good neurological recovery increased according to STI, the highest after about 4 min and decreased after 5 min.

Generally, the four groups divided by the duration of STI had signif- icantly different baseline characteristics. Patients with a short STI were more likely to be younger and male, have a witness, receive bystander CPR, receive prehospital defibrillation, and arrest in a public place. The hospital outcomes did not differ among the STI groups. (Table 1) Ad- vanced airway procedures were more common in the very-long STI group (14.4% for ETI or SGA) compared with the short STI group (3.4% for ETI or SGA; p b 0.001). Compared with the PROSC group, those in the non-PROSC group were more likely to be older, female, and to arrest in a rural and/or public place and were less likely to be witnessed and to receive bystander CPR and prehospital defibrillation. Significant differ- ences in survival rate and neurologic status between the PROSC and the non-PROSC groups were found. Advanced airways were performed more often in the non-PROSC group (14.4% for ETI and SGA) than in the PROSC group (7.4% for ETI and SGA; p b 0.001) (Table 2).

Rural and urban

23,017

56.1

721

42.6

22,296

56.6

3.2. Main analysis

Metropolis

18,037

43.9

972

57.4

17,065

43.4

Witnessed status b 0.001

No

15,773

38.4

264

15.6

15,509

39.4

Yes

17,838

43.5

1277

75.4

16,561

42.1

Unknown

7443

18.1

152

9.0

7291

18.5

Location of arrest b 0.001

Public

7790

19.0

786

46.4

7004

17.8

Private

33,085

80.6

903

53.3

32,182

81.8

Other

179

0.4

4

0.2

175

0.4

Bystander CPR b 0.001

Yes

21,160

51.5

581

34.3

20,579

52.3

No

17,903

43.6

1069

63.1

16,834

42.8

Unknown

1991

4.8

43

2.5

1948

4.9

Primary ECG b 0.001

VF/pulseless VT

3622

8.8

1206

71.2

2416

6.1

PEA

2592

6.3

149

8.8

2443

6.2

Asystole

34,840

84.9

338

20.0

34,502

87.7

In the multivariable logistic model without the interaction term for the whole group of patients, longer scene time intervals were associated with poor outcomes: middle 0.74 (0.58-0.95), long 0.51 (0.39-0.67),

and very-long 0.45 (0.33-0.61) for good CPC, and middle 0.71 (0.61- 0.83), long 0.49 (0.42-0.59), and very-long 0.42 (0.35-0.52) for survival to discharge (Table 3). The PROSC was strongly associated with higher neurological recovery and survival to discharge rates in the model:

31.72 (26.22-38.36 for good CPC and 21.10 (18.14-24.56) for survival to discharge (Table 3). The adjusted odds ratios and 95% confidence in- tervals of the other confounders were calculated in the model (Appendix 2).

3.3. Interaction analysis

EMS response time,

minutes

b 0.001

In the interaction model, the middle STI patients in the non-PROSC

0-4

4-8

2688

20,282

6.5

49.4

182

994

10.8

58.7

2506

19,288

6.4

49.0

group were associated with better neurological recovery (1.22, 1.06-

8-12

9990

24.3

348

20.6

9642

24.5

1.40) and better survival to discharge (1.14, 1.02-1.28) compared

12-

8094

19.7

169

10.0

7925

20.1

with the short STI group. In the PROSC group, a middle STI was not asso-

EMS scene time, minutes 0.820 ciated with better neurological recovery compared with a short STI

EMS transport time, minutes

0-4

7348

17.9

232

13.7

7116

18.1

4-8

15,110

36.8

624

36.9

14,486

36.8

8-12

11,415

27.8

508

30.0

10,907

27.7

12-

7181

17.5

329

19.4

6852

17.4

0.200

(1.18, 0.97-1.44), but it was associated with a better survival to dis- charge rate (1.19, 1.07-1.31).

The longer STI groups showed poor neurological recovery in the both PROSC (0.72, 0.57-0.92) and non-PROSC groups (0.82, 0.70-

0.96) for both 8 <= STI b 12 min and for 12 <= STI b 30 min (PROSC:

0-4

4-8

8-12

5478

17,861

8722

13.3

43.5

21.2

252

742

343

14.9

43.8

20.3

5226

17,119

8379

13.3

43.5

21.3

(0.57, 0.42-0.77); non-PROSC: (0.70, 0.57-0.85)). Compared with the

short STI group, groups with 8 <= STI b 12 min showed lower survival

12-

8993

21.9

356

21.0

8637

21.9

to discharge rates for both PROSC (0.76, 0.68-0.86) and non-PROSC sta-

Airway management b 0.001

ETI

1136

2.8

91

5.4

1045

2.7

SGA

2000

4.9

152

9.0

1848

4.7

BVM

30,315

73.8

1135

67.0

29,180

74.1

PV

7603

18.5

315

18.6

7288

18.5

No 31,948 77.8 326 19.3 31,622 80.3

Yes

9106

22.2

1367

80.7

7739

19.7

Prehospital defibrillation b 0.001

ED level b 0.001

Level 1 4564 11.1 383 22.6 4181 10.6

Level 2 20,147 49.1 1116 65.9 19,031 48.3

Level 3 16,343 39.8 194 11.5 16,149 41.0

Outcomes

Survival to discharge

2287

5.6

1202

71.0

1085

2.8

b 0.001

Good CPC

1256

3.1

977

57.7

279

0.7

b 0.001

STI: scene time interval, CPR: cardiopulmonary resuscitation, VF: ventricular fibrillation, VT: ventricular tachycardia, PEA: pulseless electrical activity, EMS: emergency medical service, ETI: endotracheal intubation, SGA: supraglottic airway, BVM: bag valve mask, PV: passive ventilation, ED: emergency department, CPC: cerebral performance scale.

tus (0.86, 0.76-0.97); for 12 <= STI b 30 min, the survival to discharge rates were much lower for both PROSC (0.64, 0.55-0.74) and non- PROSC status (0.74, 0.64-0.85) (Table 4).

  1. Discussion

The primary purpose of this study was to determine the association between scene time interval and neurological outcomes after OHCA in an EMS system. Longer STI was associated with worse neurological re- covery in the whole group of patients. We found that the middle STI group had more survivors than the short STI group in an interaction model across the PROSC status. Among the patients with non-PROSC, the middle STI group (4-8 min) had a more favorable neurologic out- come compared with the shorter STI group (0-3 min). This time-out- come relationship can make EMS protocol change. For example, EMS directors can encourage the EMS providers to provide the continuous

Table 3

Multivariable logistic regression analysis of hospital outcomes by scene time interval and prehospital ROSC.

Outcome

Exposure group

Total

Outcomes

Unadjusted

Adjusteda

N

n %

OR

95% CI

AOR

95% CI

Good CPC

STI group

Total

Short

7348

217 3.0

1.00

1.00

Middle

15,110

484 3.2

1.09

0.92

1.28

0.74

0.58

0.95

Long

11,415

348 3.0

1.03

0.87

1.23

0.51

0.39

0.67

Very-long

7181

207 2.9

0.98

0.80

1.18

0.45

0.33

0.61

Prehospital ROSC

No 39,361

279 0.7

1.00

1.00

Yes 1693

977 57.7

198.24

170.46

230.55

31.72

26.22

38.36

Survival to discharge

STI group

Short

7348

458 6.2

1.00

1.00

Middle

15,110

882 5.8

0.93

0.83

1.05

0.71

0.61

0.83

Long

11,415

596 5.2

0.83

0.73

0.94

0.49

0.42

0.59

Very-long

7181

351 4.9

0.77

0.67

0.89

0.42

0.35

0.52

Prehospital ROSC

No

39,361

1085 2.8

1.00

1.00

Yes

1693

1202 71.0

107.51

93.84

123.18

21.10

18.14

24.56

AOR: adjusted odds ratio, CI: confidence interval, STI: scene time interval, ROSC: return of spontaneous circulation, CPC: cerebral performance scale.

a Adjusted for age group (10-year interval), gender, urbanization (rural and urban versus metropolitan), witnessed, location of arrest (public versus private versus unknown), bystander cardiopulmonary resuscitation, primary ECG (ventricular fibrillation/ventricular tachycardia versus pulseless electrical activity versus asystole), response time interval, transport time in- terval, airway management (endotracheal intubation versus supraglottic airway, bag valve mask, passive oxygenation), prehospital defibrillation, level of emergency department (level-1 versus level-2 versus level-3).

CPR until achieving ROSC within 8 min. If the patient does not achieve ROSC after 8 min, the EMS provider may decide the immediate trans- port to ED immediately.

From Fig. 2, we found the peak log odds of good neurological recov- ery occurred at 4 to 5 min of STI. This means the good neurological re- covery are expected in the cases restored after two or three times of 2 min CPR done by EMS providers. If the STI increase, the survival chance with good neurological recovery decrease. In this study popula- tion, as predicted, younger age, metropolitan area, witnessed status, public places, bystander CPR, shockable ECG rhythm, prehospital defi- brillation, and higher level of ED were associated with better neurolog- ical recovery and survival to discharge (Appendix 2).

Non-PROSC occurred in 95.9% of the study subjects, which means that most of the patients were transported while receiving CPR on mov- ing ambulance. Only 4.1% of the patients were transported after ROSC, and 50.1% of these PROSC patients had shorter STIs (b 8 min). Therefore, this study’s findings will be interesting to EMS directors and leaders when establishing CPR protocols on STI.

A recent study showed that EMS-provided CPR with an STI from 8 to 16 min was related to improved neurological outcomes after OHCA in

Table 4 Interaction analysis between scene time interval groups across prehospital ROSC and hos- pital outcomes.

PROSC STI Good CPC Survival to discharge

AORa

95% CI

AORa

95% CI

PROSC (+) Short

Ref

Ref

Middle

1.18

0.97 1.44

1.19

1.07 1.31

Long

0.72

0.57 0.92

0.76

0.68 0.86

Very-long

PROSC (-) Short Middle

0.57

Ref 1.22

0.42 0.77

1.06 1.40

0.64

Ref 1.14

0.55 0.74

1.02 1.28

Long

0.82

0.70 0.96

0.86

0.76 0.97

Very-long

0.70

0.57 0.85

0.74

0.64 0.85

AOR: adjusted odds ratio, CI: confidence interval.

a Adjusted for age group (10-year interval), gender, urbanization (rural and urban versus metropolitan), witnessed, location of arrest (public versus private versus un- known), bystander cardiopulmonary resuscitation, primary ECG (ventricular fibrillation/ ventricular tachycardia versus pulseless electrical activity versus asystole), response time interval, transport time interval, airway management (endotracheal intubation versus supraglottic airway, bag valve mask, passive oxygenation), level of emergency department (level-1 versus level-2 versus level-3), prehospital defibrillation, return of spontaneous

circulation status, and interaction term (scene time interval group * prehospital return

of spontaneous circulation).

two Asian metropolitan cities [14]. In the study, the STIs of the two cities were significantly different (86% in Seoul versus 7.2% in Osaka for the proportion of patients with an STI b 8 min). The study was conducted in quite different study settings with different advanced procedures, such as the proportion of advanced airway management (71.0% in Osaka (5377/7517 witnessed non-traumatic OHCAs) versus 8.7% in Korea (2829/32513 EMS-treated cardiac etiology OHCAs) [15,16], the number of higher-level EMTs on the ambulance (95% with one level-1 EMT in Seoul versus 57% with two emergency life-saving technicians (ELSTs; equivalent to level-1 EMTs) or 17% with three ELSTs on the am- bulance in Osaka) [17], and epinephrine injection rates at the field (not permitted in Seoul versus 7.6% in Japan) [18]. These advanced proce- dures and the higher level of the providers were not adjusted for the study.

In this study, advanced airway management was performed more often in the longer STI group and the non-PROSC group, (Tables 1 and 2) meaning that the procedure was selected by EMTs in cases that re- quired longer CPR because of greater resistance to PROSC. Basic airway management was given to much less patients with short STI (1.0% for ETI, 2.4% for SGA) than those with very long STI (6.5% for ETI and 7.9% for SGA). The EMS protocol followed in this study allows EMTs to choose the airway management method according to their preference. We as- sumed that the EMT would select the airway management method ap- propriate for the clinical circumstances: The patients with shockable and witnessed status would be managed with more basic procedures, rather than advanced procedures, and would be treated within an earli- er time window after the EMTs’ arrival on the scene, whereas the pa- tients who were not restored within a few minutes would be managed by advanced airway. Research indicates that 72% of prehospital ROSC patients have a shockable rhythm. They first received defibrillation immediately after EMT arrival on the scene and did not re- quire a long stay at the scene. Therefore, our analysis using the interac- tion model was useful for testing the association between STI across the PROSC status and clinical outcomes. The basic to intermediate EMS ser- vice level showed the best outcomes with EMS-provided CPR with STIs of up to 8 min in the non-PROSC group.

A previous study of the BLS termination rule showed incomplete CPR times at the scene. East Asian EMS models are encouraging EMS pro- viders to stay a short time in the field and to provide very few advanced interventions beyond basic life support [7,19]. Staying longer in the field to provide additional interventions involves trade-offs in terms of ade- quate airway management, circulation, and medication. These ALS

procedures require a high level of performance, more team members and better team quality [20,21]. This study encourages EMS providers to stay in the field for at least 8 min and to move patients to the hospital ED while providing CPR on the ambulance at a basic to intermediate ser- vice level. However, this study did not analyze the effects of advanced procedures performed during field resuscitation. Therefore, the associa- tion between outcomes and time interval should be studied in relation to advanced procedures.

The extracorporeal membrane oxygenator cardiopulmonary resus- citation (E-CPR) has increased for the last several years in Korea. E- CPR can guarantee the higher survival if the patients are not resuscitated and under refractory ventricular fibrillation (VF) [22]. The E-CPR pro- gram can change the EMS CPR protocol including duration of CPR at the field. E-CPR program can encourage EMS providers to transport the patients to ED for E-CPR if the patients has refractory VF. A study re- ported a new protocol “load and go” for selective OHCAs (6%) for poten- tial candidates for E-CPR in ED [23]. The national EMS CPR protocol of Korea, however, has not accepted the selective protocol for E-CPR al- though many hospital EDs are providing E-CPR for in-hospital cardiac arrest patients. The association between STI and E-CPR should be inves- tigated to develop the selective CPR protocol.

Although this study shows the best outcomes in intermediate STI, the clinical implication of the study finding should be considered ac- cording to EMS protocol, resources, competency of EMT performance, accessibility to hospital EDs, and services level. The STI could be consid- ered as a complex and composite measure (surrogate) of the resources used by the pre-hospital care than an independent predictor of outcome or a target for intervention.

Limitation

There are several limitations to this study. First, this study was not randomized or controlled in terms of the effects of STIs on OHCA out- comes. Unmeasured confounders from the non-controlled retrospective study might affect the final results. We cannot conclude the causal asso- ciation. Second, additionally, because the study examined the interme- diate service level of EMS, its findings cannot be generalized to ALS EMS systems. Third, we could not directly measure the treatments and the Quality of CPR provided by EMS providers on the scene. Forth, short STIs could be adequate to provide prehospital defibrillation, but we did not measure the objective time of the procedures or how many defibrillation attempts were made at the scene. Fifth, the STI was considered a proxy time for treatment, and the treatment time itself was not measured. These unmeasured biases will have different effects

on the outcomes. Sixth, we did not test the impact of advanced life sup- port procedures on outcome according to STI group. If we adjusted full range of advanced life support procedures in higher service level, the findings might be changed. Finally, the clinical neurological outcomes obtained by the medical record reviewers from electrical medical re- cords could represent a measurement bias because we did not test the reliability of the measurement.

  1. Conclusion

In a large population-based OHCA cohort study, we found the scene time interval for CPR was significantly associated with the highest odds of neurological recovery for patients who could not be restored in the field. The STI may be a clinically useful predictor of good neurology out- come in victims of cardiac arrest.

Ethic approval

The study was approved by the institutional review board of the study institution, and the Korea Centers for Disease Control and Preven- tion (CDC) reviewed the study and approved the use of data.

Consent for publication

This manuscript is an original article, all of the authors have agreed to the submission and the manuscript is not under consideration by any other journal.

Acknowledgement of funding source

This study was supported by the National Emergency Management Agency of Korea and the Korea Centers for Disease Control and Preven- tion (CDC). The study was funded by the Korea CDC (2012-2014; Grant No.: 2012-E33010-00; 2013-E33015-00; 2014-E33011-00).

Authors’ contribution

All authors contributed to this study; Kim KH analyzed the dataset and drafted the manuscript. Song KJ, and Ro YS have performed data collection. Hong KJ, Kim YJ, and Jeong J contributed to revise the manu- script and provided important discussion. All authors are fully accessible to the data. Shin SD designed the study and has responsible for the study.

Appendix 1. Adjusted odds ratios of exposure and confounders in multivariable logistic regression analysis.

Factors Good neurological recovery Survival to discharge

AOR

95% CI

AOR

95% CI

Total Year

2012

1.00

1.00

2013

1.12

0.89

1.41

1.07

0.93

1.23

2014

Gender Male

1.53

1.00

1.22

1.91

1.06

1.00

0.92

1.23

Female Age, year

20-30

0.36

1.00

0.27

0.49

0.88

1.00

0.77

1.01

31-39

0.74

0.43

1.26

0.89

0.60

1.31

40-49

0.58

0.36

0.93

0.65

0.46

0.93

50-59

0.52

0.33

0.82

0.68

0.48

0.95

60-69

0.26

0.16

0.42

0.47

0.33

0.66

70-79

0.12

0.07

0.20

0.27

0.19

0.38

80-

0.08

0.04

0.15

0.18

0.12

0.26

Urbanization

Appendix 1 (continued)

Factors Good neurological recovery Survival to discharge

AOR

95% CI

AOR

95% CI

Metropolis

1.00

1.00

Rural

0.71

0.59

0.85

0.76

0.67

0.85

Witnessed status

Yes

1.00

1.00

No

1.60

1.27

2.01

1.92

1.67

2.21

Unknown

1.15

0.83

1.60

1.29

1.07

1.57

Location of arrest

Public

1.00

1.00

Private

0.69

0.58

0.83

0.74

0.66

0.84

Other

1.89

0.50

7.10

0.49

0.15

1.57

Primary ECG

VF/pulseless VT

1.00

1.00

PEA

0.36

0.27

0.49

0.63

0.52

0.77

Asystole

0.12

0.09

0.15

0.25

0.22

0.30

Bystander CPR

Yes

1.00

1.00

No

1.05

0.87

1.27

0.85

0.75

0.96

Unknown

0.91

0.56

1.50

0.85

0.62

1.15

EMS response time, minute

0-4

1.00

1.00

4-8

0.55

0.41

0.73

0.73

0.60

0.88

8-12

0.49

0.35

0.68

0.57

0.46

0.71

12-

0.34

0.23

0.50

0.43

0.34

0.56

EMS scene time, minute

0-4

1.00

1.00

4-8

0.51

0.39

0.67

0.71

0.61

0.83

8-12

0.45

0.33

0.61

0.49

0.42

0.59

12-

0.74

0.43

1.26

0.42

0.35

0.52

EMS transport time, minute

0-4

1.00

1.00

4-8

0.91

0.70

1.17

0.92

0.78

1.09

8-12

0.66

0.49

0.90

0.66

0.55

0.81

12-

0.85

0.62

1.15

0.60

0.48

0.74

Airway management

ETI

1.00

1.00

SGA

0.76

0.45

1.28

0.81

0.56

1.17

BVM

0.89

0.57

1.39

0.76

0.56

1.04

PVO2

1.12

0.70

1.81

0.85

0.61

1.19

Pre-hospital defibrillation

No

1.00

1.00

Yes

1.76

1.36

2.28

2.00

1.74

2.31

ED level

Level 1

1.00

1.00

Level 2

0.62

0.50

0.77

0.65

0.57

0.75

Level 3

0.13

0.09

0.18

0.15

0.12

0.18

STI: scene time interval, CPR: cardiopulmonary resuscitation, VF: ventricular fibrillation, VT: ventricular tachycardia, PEA: pulseless electrical activity, EMS: emergency medical service, ETI: endotracheal intubation, SGA: supraglottic airway, BVM: bag valve mask, PV: passive ventilation, ED: emergency department, CPC: cerebral performance scale.

Appendix 2. Cardiopulmonary Resuscitation protocol of Korean emergency medical services

EMS: emergency medical service, BVM: bag valve mask, AED: automatic external defibrillator, ETI: endotracheal intubation, SGA: supraglottic airway, EMT: emergency medical tech- nician, ROSC: return of spontaneous circulation, CPR: cardiopulmonary resuscitation, ED: emergency department.

References

  1. Nichol G, Thomas E, Callaway CW, Hedges J, Powell JL, Aufderheide TP, et al. Regional variation in out-of-hospital cardiac arrest incidence and outcome. JAMA 2008; 300(12):1423-31.
  2. Grasner JT, Lefering R, Koster RW, Masterson S, Bottiger BW, Herlitz J, et al. EuReCa ONE-27 nations, ONE Europe, ONE registry: a prospective one month analysis of out- of-hospital Cardiac arrest outcomes in 27 countries in Europe. Resuscitation 2016; 105:188-95.
  3. McNally B, Robb R, Mehta M, Vellano K, Valderrama AL, Yoon PW, et al. Out-of-hos- pital cardiac arrest surveillance — Cardiac Arrest Registry to Enhance Survival (CARES), United States, October 1, 2005-December 31, 2010. MMWR Surveill Summ 2011;60(8):1-19.
  4. Ong ME, Shin SD, De Souza NN, Tanaka H, Nishiuchi T, Song KJ, et al. Outcomes for out-of-hospital cardiac arrests across 7 countries in Asia: the Pan Asian Resuscitation Outcomes Study (PAROS). Resuscitation 2015;96:100-8.
  5. De Maio VJ, Stiell IG, Wells GA, Spaite DW. Ontario prehospital advanced life support study G. optimal defibrillation response intervals for maximum out-of-hospital car- diac arrest survival rates. Ann Emerg Med 2003;42(2):242-50.
  6. Ono Y, Hayakawa M, Iijima H, Maekawa K, Kodate A, Sadamoto Y, et al. The response Time threshold for predicting favourable neurological outcomes in patients with by- stander-witnessed out-of-hospital cardiac arrest. Resuscitation 2016;107:65-70.
  7. Hock Ong ME, Shin SD, Sung SS, Tanaka H, Huei-Ming M, Song KJ, et al. Recommen- dations on ambulance cardiopulmonary resuscitation in basic life support systems. Prehosp Emerg Care 2013;17(4):491-500.
  8. Shin SD, Ahn KO, Song KJ, Park CB, Lee EJ. Out-of-hospital airway management and cardiac arrest outcomes: a propensity score matched analysis. Resuscitation 2012; 83(3):313-9.
  9. Ro YS, Shin SD, Song KJ, Lee EJ, Kim JY, Ahn KO, et al. A trend in epidemiology and outcomes of out-of-hospital cardiac arrest by urbanization level: a nationwide ob- servational study from 2006 to 2010 in South Korea. Resuscitation 2013;84(5): 547-57.
  10. Neumar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, et al. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association guide- lines for cardiopulmonary resuscitation and emergency cardiovascular care. Circula- tion 2010;122(18 Suppl 3):S729-67.
  11. Song KJ, Shin SD, Park CB, Kim JY, Kim do K, Kim CH, et al. Dispatcher-assisted by- stander cardiopulmonary resuscitation in a metropolitan city: a before-after popula- tion-based study. Resuscitation 2014;85(1):34-41.
  12. Stiell IG, Nesbitt LP, Nichol G, Maloney J, Dreyer J, Beaudoin T, et al. Comparison of the Cerebral performance category score and the health utilities index for survivors of cardiac arrest. Ann Emerg Med 2009;53(2):241-8.
  13. Kim JY, Shin SD, Ro YS, Song KJ, Lee EJ, Park CB, et al. post-resuscitation care and out- comes of out-of-hospital cardiac arrest: a nationwide propensity score-matching analysis. Resuscitation 2013;84(8):1068-77.
  14. Shin SD, Kitamura T, Hwang SS, Kajino K, Song KJ, Ro YS, et al. Association between resuscitation time interval at the scene and neurological outcome after out-of-hospi- tal cardiac arrest in two Asian cities. Resuscitation 2014;85(2):203-10.
  15. Kajino K, Iwami T, Kitamura T, Daya M, Ong ME, Nishiuchi T, et al. Comparison of supraglottic airway versus endotracheal intubation for the pre-hospital treatment of out-of-hospital cardiac arrest. Crit Care 2011;15(5):R236.
  16. Kang K, Kim T, Ro YS, Kim YJ, Song KJ, Shin SD. Prehospital endotracheal intubation and survival after out-of-hospital cardiac arrest: results from the Korean nationwide registry. Am J Emerg Med 2016;34(2):128-32.
  17. Kajino K, Kitamura T, Iwami T, Daya M, Ong ME, Nishiyama C, et al. Impact of the number of on-scene emergency life-saving technicians and outcomes from out-of- hospital cardiac arrest in Osaka City. Resuscitation 2014;85(1):59-64.
  18. Ono Y, Hayakawa M, Wada T, Sawamura A, Gando S. Effects of prehospital epineph- rine administration on neurological outcomes in patients with out-of-hospital cardi- ac arrest. J Intensive Care 2015;3(1):29.
  19. Kim TH, Shin SD, Kim YJ, Kim CH, Kim JE. The scene time interval and basic life sup- port termination of resuscitation rule in adult out-of-hospital cardiac arrest. J Korean Med Sci 2015;30(1):104-9.
  20. McEvoy MD, Field LC, Moore HE, Smalley JC, Nietert PJ, Scarbrough SH. The effect of adherence to ACLS protocols on survival of event in the setting of in-hospital cardiac arrest. Resuscitation 2014;85(1):82-7.
  21. DePriest J, Fee-Mulhearn AL, Teleron A. A novel ACLS team leader checklist imple- mented to improve resuscitation efforts. Resuscitation 2013;84(9):e115.
  22. Johnson NJ, Acker M, Hsu CH, Desai N, Vallabhajosyula P, Lazar S, et al. Extracorpo- real life support as rescue strategy for out-of-hospital and emergency department cardiac arrest. Resuscitation 2014;85(11):1527-32.
  23. Poppe M, Weiser C, Holzer M, Sulzgruber P, Datler P, Keferbock M, et al. The inci- dence of “load&go” out-of-hospital cardiac arrest candidates for emergency depart- ment utilization of emergency Extracorporeal life support: a one-year review. Resuscitation 2015;91:131-6.

Leave a Reply

Your email address will not be published. Required fields are marked *