- Email: [email protected]
Association between coronary angiography with or without percutaneous coronary intervention and outcomes after out-of-hospital cardiac arrest
Resuscitation 127 (2018) 21–25
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Association between coronary angiography with or without percutaneous coronary intervention and outcomes after out-of-hospital cardiac arrest☆
T
⁎
Tyler F. Vadeboncoeura, , Vatsal Chikanib,1, Chengcheng Huc,2, Danial W. Spaitec,3, Bentley J. Bobrowb,c,d,e,1,4 a
Department of Emergency Medicine, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL, 32224, United States The Arizona Department of Health Services Bureau of Emergency Medical Services and Trauma System, Phoenix, AZ, United States c Arizona Emergency Medicine Research Center, The University of Arizona College of Medicine, Tucson, AZ, United States d Maricopa Medical Center, Phoenix, AZ, United States e The University of Arizona Sarver Heart Center, Tucson, AZ, United States b
A R T I C LE I N FO
A B S T R A C T
Keywords: Cardiac arrest Post-resuscitation care Coronary angiography
Aim: The aim of our study was to assess the impact of coronary angiography (CAG) after out-of-hospital cardiac arrest (OHCA) without ST-elevation (STE). Methods: Prospective observational study of adult (age ≥ 18) OHCA of presumed cardiac etiology from 1/01/ 2010–12/31/2014 admitted to one of 40 recognized cardiac receiving centers within a statewide resuscitation network. Results: Among 11,976 cases, 1881 remained for analysis after exclusions. Of the 1230 non-STE cases, 524 (43%) underwent CAG with resultant PCI in 157 (30%). Survival in non-STE cases was: 56% in cases without CAG; 82% in cases with CAG but without PCI; and 78% in those with PCI (p < 0.0001). In cases without STE the aOR for survival with CAG alone was 2.34 (95% CI 1.69–3.24) and for CAG plus PCI was 1.98 (95% CI 1.26–3.09). The aOR for CPC 1/2 with CAG alone was 6.89 (95% CI 3.99–11.91) and for CAG plus PCI was 2.95 (95% CI 1.59–5.47). After propensity matching, CAG was associated with an aOR for survival of 2.10 (95% CI 1.30–3.55) and for CPC 1/2 it was 5.06 (95% CI 2.29–11.19). Conclusion: In OHCA without STE, CAG was strongly and independently associated with survival regardless of whether PCI was performed. The association between CAG and positive outcomes remained after propensity matching.
Introduction Out-of-hospital cardiac arrest (OHCA) remains a major public health problem in the United States [1]. Successful resuscitation requires a synchronized set of interdependent actions (the “chain of survival”) which in the latest 2015 American Heart Association (AHA) Guidelines includes targeted temperature management (TTM) and coronary angiography (CAG) after return of circulation [2]. Early revascularization is thought to be the primary benefit of emergent CAG after cardiac arrest [3]. Numerous (> 15) studies have reported improved rates of survival
to hospital discharge with emergent CAG in patients with ST-segment elevation (STE) [2]. As such, the AHA endorses emergent CAG for OHCA patients with suspected cardiac etiology and STE regardless of whether they are comatose or awake [2]. Several studies have reported lesions amenable to percutaneous coronary intervention (PCI) in a significant minority of patients without STE on their initial post-cardiac arrest electrocardiogram (EKG) [4–7]. While studies have associated emergent CAG in patients without STE after cardiac arrest with improved survival, PCI has not always been shown to provide added benefit [6,8]. The purpose of this study was to assess the association between CAG
☆
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.03.023. Corresponding author. E-mail addresses: [email protected] (T.F. Vadeboncoeur), [email protected] (V. Chikani), [email protected] (C. Hu), [email protected] (D.W. Spaite), [email protected] (B.J. Bobrow). 1 Bureau of Emergency Medical Services, Arizona Department of Health Services, 150 N. 18th Avenue, #540, Phoenix, AZ 85007, United States. 2 University of Arizona, PO Box 245057, 1501 N. Campbell, Tucson, AZ 85724-5057, United States. 3 Department of Emergency Medicine, University of Arizona, Box 245057, 1501 N. Campbell, Tucson, AZ 85724-5057, United States. 4 Address for reprints: Arizona Department of Health Services, 150 N 18th Avenue, Suite 540, Phoenix, AZ, United States. ⁎
https://doi.org/10.1016/j.resuscitation.2018.03.023 Received 17 October 2017; Received in revised form 12 January 2018; Accepted 12 March 2018 0300-9572/ © 2018 Elsevier B.V. All rights reserved.
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
alone and CAG plus PCI on outcomes after OHCA without initial STE.
Table 1 Demographics and event characteristics.
Methods
N = 1881
Setting
n (%) Male Age, yrs-median (Q1, Q3) Arrest witnessed Bystander CPR performed Initial rhythm VF/VT No-STEMI Intervention CAGa with PCIb CAG without PCI No CAG GCS < =8 >8 Targeted temperature management Survival Response Interval, minutes- median (Q1, Q3) CPC scale All 1 2 3 4 5
Arizona had 6.7 million residents in 2014 [9]. The Arizona Department of Health Services (ADHS) establishes the EMS scope of practice and provider certification. EMS protocols are implemented on regional level. Crew configuration, vehicle deployment, dispatch, and response intervals vary widely across the state and some local OHCA bypass protocols are in place. About 100 EMS agencies and 40 Cardiac Receiving Centers (CRCs) responding to approximately 80% of the Arizona’s population voluntarily participated in the state-sponsored Save Hearts in Arizona Registry and Education (SHARE) Program during the study period. SHARE has been described previously [10–14]. In order to be recognized as a CRC, a hospital must have: 1) primary 24/7 PCI capability with a protocol including calling cardiology for OHCA, 2) a dedicated TTM protocol for OHCAs that remain comatose, 3) an evidence-based termination of resuscitation protocol which includes a 72-h moratorium on termination of care for patients receiving TTM, and 4) commitment to on-going data submission for all OHCA patients (www.azshare.gov). There are no specific mandates or protocols regarding the timing or selection of patients for coronary angiography. The decision is left to the treating providers which includes required cardiology involvement.
a b
Missing, n (%)
1289 (68.5) 63.0 (52.0, 72.0) 837 (63.3) 677 (53.5) 930 (52.6) 1230 (71.1)
0 0 559 (29.7) 615 (32.7) 113 (6.0) 152 (8.1)
482 (25.6) 525 (27.9) 874 (46.5)
0
1427 (86.6) 221 (13.4) 655 (34.8) 1192 (63.4) 5.0 (4.0, 6.0)
233 (12.4)
1822 (96.9) 740 (40.6) 182 (10.0) 62 (3.4) 149 (8.2) 689 (37.8)
59 (3.1)
0 0 568 (30.2)
CAG – coronary angiography. PCI – percutaneous coronary intervention.
Study design and population Department of Health Services Human Subjects Review Board have determined that, by virtue of being a public health initiative, neither the interventions nor their evaluation constitute human subjects research and have approved the publication of de-identified data.
This is a prospective, multicenter, observational cohort study of consecutive adult (aged ≥18 years) patients with OHCA between January 1, 2010, and December 31, 2014, who were transported initially or transferred to a CRC. Cases were excluded if prehospital resuscitation was not initiated, the cause of the arrest was presumed to be non-cardiac (e.g., known respiratory arrest, suicide, trauma, drowning, or drug overdose), the patient had a Do-Not-Resuscitate order, or the patient died in the emergency department (ED).
Statistical analysis To maximize the available subjects for analysis, multiple imputation was carried out in SAS (SAS, Version 9.3, SAS Institute, Cary, North Carolina) to impute missing data. Multiple imputation (MI) has been shown to generate less biased estimates with more statistical efficiency when compared with alternative methods of handling incomplete data (e.g., complete-case analysis, single imputation, missing indicator regression) [16,17]. MI involves three distinct phases: 1) the missing data are filled in m times to generate m complete data sets, 2) the m datasets are analyzed by using standard procedures, and 3) the results from the m complete datasets are combined for the inference. MI procedure replaces each missing value with a set of plausible values that represent the uncertainty about the right value to impute, instead of filling in a single value for each missing value. We used all the variables in Table 1 for MI. Twenty imputed data sets were generated and model fit and diagnostics were evaluated. Missing data fit an arbitrary missing pattern and we used Fully Conditional Specification method to impute data. Linear regression was used to impute all time intervals. Logistic regression was used to impute categorical variables, witnessed arrest, bystander cardiopulmonary resuscitation (CPR), shockable rhythm and STE. The study population was divided into three groups, patients without CAG (reference group), patients with CAG and without PCI, and patients with CAG and PCI. Descriptive statistics were used to describe the study population and are reported as proportion for categorical and binary data and as median and interquartile range for continuous data. All analyses used the imputed data, accounting for variance across imputed data sets using Rubin’s rules [18] (using Proc MI in SAS). To control for indication bias, propensity score was calculated to predict the probability of receiving CAG using a multivariable logistic regression model, including age, gender, bystander CPR
Data processing and study approval EMS data are obtained from the patient care reports and outcomes are obtained either from the hospitals or from the State Office of Vital Records. SHARE includes an Utstein-style OHCA EMS database linked with in-hospital post-arrest care and outcome data. Consistent with Utstein methodology, each OHCA in which EMS attempts resuscitation is included. EMS data are cross-referenced between first responding EMS agencies, private ambulance transport companies, and the CRC database. In order to perform continuous quality improvement for inhospital post-arrest care a data tool was developed to collect in-hospital patient information for all OHCA patients brought to a recognized CRC. The CRC data include details on CAG including the initial EKG findings, the timing of CAG, and whether or not PCI is performed. It does not include angiographic data. It also includes the final patient outcome including the Cerebral Performance Category (CPC) score [15]. The data forms were completed by CRC clinical personnel using a secure web-based data entry system. Each hospital individual completing data forms was trained in person and each form was secondarily reviewed by a SHARE data coordinator for completeness and accuracy before entry into the database. Any inconsistencies were addressed in follow-up by examining the hospital medical record. SHARE is an Arizona Department of Health Services–sponsored public health initiative. As such the program is exempt from the requirements of the Health Insurance Portability and Accountability Act, which allows linkage of EMS and hospital data, tracking of OHCA events, and evaluation of efforts to improve resuscitation care. The University of Arizona Institutional Review Board and the Arizona 22
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
performed, arrest witnessed, initial rhythm, TTM, GCS and response time as covariates. The multivariate logistic regression analysis was then carried out to investigate the association between CAG with or without PCI, and survival with propensity score included as a covariate in the model along with the other covariates (adjusted unpaired analysis). Next, patients were categorized into two groups based on their CAG status (CAG vs no CAG) and were matched by propensity scores using a 1:1 matching procedure without replacement. Among the matched patients, the standardized difference for each of the covariates between the two groups was < 0.1 suggesting a negligible difference in the prevalence of the covariates between the two groups and, hence, successful matching [19]. Finally, we compared the survival and favorable neurologic outcomes between the CAG and no CAG groups for the matched pairs using the multivariable logistic regression model (propensity match adjusted analysis). Data were analyzed using Statistical Analysis System (SAS version 9.3, SAS Institute, Inc. Cary, NC). P value of < 0.05 was considered statistically significant.
initial rhythm VF/VT and to receive TTM. They were more likely to be comatose upon arrival to the ED (Table 3). To account for indication bias as the three groups differ significantly on most of the covariates, we estimated the propensity of receiving CAG for each of the non-STE patients using a multivariate logistic regression model. The adjusted unpaired logistic regression analysis with propensity score as a covariate showed that CAG was associated with improved survival and neurologic outcomes whether or not PCI was performed. The aOR for survival with CAG alone was 2.34 (95% CI 1.69–3.24) and for CAG with PCI was 1.98 (95% CI 1.26–3.09). The aOR for CPC 1/2 with CAG alone was 6.89 (95% CI 3.99–11.91) and for CAG with PCI was 2.95 (95% CI 1.59–5.47) (Table 4). After matching on propensity score, the survival benefit of CAG persisted. CAG conferred an aOR for survival of 2.10 (95% CI 1.30–3.55) and for CPC 1/2 of 5.06 (95% CI 2.29–11.19). Results were similar when including only comatose (GCS ≤ 8) patients and when analyzing the subgroup without initial STE but with an initial shockable rhythm.
Results
Discussion
There were 11,976 OHCA cases during the study period. After exclusions, the study population was 1881 (Fig. 1), 1007 (54%) of which underwent CAG. Of those undergoing CAG 482 (48%) had PCI (Table 1). Overall, 63% of cases survived to hospital discharge and 81% of survivors had a CPC 1/2. Ninety percent of (450/499) STE cases underwent CAG, 70% of which had PCI. Of the 1230 cases without initial STE 43% underwent CAG, 157 (30%) of which had PCI performed (Table 2). Fig. 2 presents a histogram depicting the timing of CAG. Overall survival to hospital discharge was 53% for those without CAG and 72% for those who underwent CAG whether or not PCI was performed (p < 0.0001) (Table 2). In the 1230 cases without initial STE survival was again better in those who underwent CAG whether or not PCI was performed (78.3% and 81.5%) compared to the group without CAG (56.0%, p < 0.0001). Table 2 also presents the results by CPC score which demonstrates favorable neurologic outcomes in those undergoing CAG, but no difference between CAG with PCI or without PCI. After restricting the analyses to only patients who did not have STE (1230), those without CAG were less likely to be witnessed, to have
The AHA Guidelines recommend that OHCA with initial STE undergo emergent CAG whether or not they are comatose [2]. It remains less clear whether OHCA patients without initial STE should be taken for early CAG. While observational data regularly show associations between CAG and improved outcomes even without STE, authors have suggested that indication bias has impacted these findings [2,5,6,20]. The fact that CAG with PCI has not consistently been associated with better outcomes when compared to CAG alone lends further support to the argument that there is bias [6,8]. In this statewide analysis encompassing 40 CRCs and over 100 EMS agencies we used logistic regression and propensity matching in an effort to control for confounding. CAG was widely utilized across the state with 90% of OHCA with STE and 43% of OHCA without initial STE undergoing CAG. These numbers are strikingly similar to previous observational studies reporting rates of CAG for OHCA (93% with STE and 42% without STE) [4]. We did not assess the impact of CAG in STE because only a small minority did not undergo CAG. In cases without STE, CAG was associated with better survival to discharge and neurologic outcomes, but PCI did not confer additional benefit. (Table 2) As shown in Table 3, subjects who underwent CAG we more likely to have a witnessed arrest, to have an initial shockable rhythm, and to receive TTM. They were also less likely to be comatose upon arrival to the ED. Because, all of these factors have been associated with better outcomes, a propensity analysis was performed to assess whether indication bias played a role in our findings. After matching, CAG was still associated with a doubling of survival to hospital discharge and an even stronger association with a positive neurologic outcome (Table 4). Previous studies have been conflicting regarding whether OHCA without STE should undergo early CAG. In a recent meta-analysis of 11 articles Millin et al reported that 32.2% of OHCA without STE taken to CAG had an acute culprit lesion requiring intervention [4]. Consistently, in the PROCAT II study, Dumas et al brought all OHCA without an obvious non-cardiac cause to immediate CAG and found lesions amenable to PCI in 29% (199/695) [7]. There was a favorable outcome in 43% of those undergoing PCI and 33% without PCI. Conversely, Hollenbeck et al reviewed 269 OHCA due to ventricular arrhythmia without STE who were treated with TTM and found that early CAG was strongly associated with survival (65.6% vs. 48.6%; p = 0.017), but that PCI did not confer an additional benefit over CAG alone [6]. Finally, Casella et al reviewed 278 comatose OHCA treated with TTM and found that CAG, but not PCI, was associated with improved outcomes [8]. The propensity analysis combined with the large size of this study, including multiple EMS agencies and hospitals, makes these findings as strong and generalizable as any previous study evaluating CAG and PCI
Fig. 1. Study population profile. 23
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
Table 2 Survival and neurologic outcomes by coronary angiography (CAG) and percutaneous coronary intervention (PCI). Overall (N = 1881)
No-CAG CAG without PCI CAG with PCI P-value
No STEMI (n = 1230)
No STEMI VF/VT (n = 557)
n
Survival
CPC Scale (1/2)
n
Survival
CPC Scale (1/2)
n
Survival
CPC Scale (1/2)
874 525 482
% (N) 53.1% (464) 72.2% (379) 72.4% (349) < 0.0001
30.6% (268) 64.0% (336) 66.0% (318) < 0.0001
706 367 157
% (N) 56.0% (395) 81.5% (299) 78.3% (123) < 0.0001
34.1% (241) 75.2% (276) 68.2% (107) < 0.0001
209 233 115
% (N) 72.7% (152) 87.1% (203) 82.6% (95) < 0.001
49.3% (103) 82.4% (192) 73.0% (84) < 0.0001
findings. For example, we chose to include all cases whether or not they remained comatose and we controlled for GCS. The vast majority of cases, however, were comatose and, as such, when we ran the analysis on only the comatose cases the findings were nearly identical. We also chose to include all cases with CAG during the index hospitalization rather than only those with CAG in the first 2, 12 or 24 h. The histogram in Fig. 2 shows that the vast majority of CAG was in the first 12 h. Clearly late CAG requires survival at least until that point, however it is also true that late CAG is not the same as no CAG. Finally, multiple imputation was required because we were not able to match all CRC cases with their out-of-hospital data. As such, there was a significant amount of “missing” prehospital data.
after OHCA without STE. Our findings add to the body of literature suggesting that CAG is a beneficial component of post-cardiac arrest care even without initial STE. Our results, however, do not make clear that revascularization is the primary driver of the association between CAG and improved outcomes. As such, it is possible that the association with improved outcomes is a result of other factors such as a higher intensity of care with early CAG or even that CAG facilitates making alternative diagnoses (such as pulmonary embolism) or informs other treatment decisions. In our study, CAG patients were more likely to be treated with TTM suggesting a higher intensity of care. Hollenbeck et al found that patients treated with CAG were more likely to receive mechanical support with intraaortic balloon counterpulsation and were also treated with more aggressive anticoagulation [6]. It has also been suggested that CAG might be associated with timely venous access, invasive hemodynamic monitoring and rapid titration of vasoactive medications [6]. The resuscitation community is anxiously awaiting the results of the randomized DISCO and PEARL trials to help determine which and/or whether OHCA patients without initial STE should go to early CAG. Until that time, authors such as Rab et al have attempted to create risk stratification tools for comatose cardiac arrest patient to determine which non-STE patients are most likely to benefit early CAG [21].
Conclusions In conclusion, we found that coronary angiography was widely utilized after out-of-hospital cardiac arrest in this statewide cardiac receiving system. In cases without initial ST-elevation, a significant minority of cases had a lesion amenable to percutaneous coronary intervention and coronary angiography was associated with improved outcomes with or without PCI. After propensity matching, coronary angiography remained significantly associated with higher survival to hospital discharge and positive neurologic outcomes.
Limitations As with any observational study, this analysis has several limitations. Despite the large size of this study and the use of propensity matching, unaccounted for bias still remains a possibility. We also selected specific inclusion criteria which could potentially impact the
Role of funding None.
Fig. 2. Emergency department arrival to coronary angiography time. 24
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
Table 3 Demographics and event characteristics by type of intervention: No-STEMI patients. No-STEMI patients (N = 1230)
No CAG (n = 706)
Male Age, yrs-median (Q1, Q3) Arrest witnessed Bystander CPR performed Initial rhythm VF/VT GCS < = 8 Targeted temperature management Survival Response Interval, minutes- median (Q1, Q3)
CAG with PCI (n = 157)
N (%)
Missing (%)
N (%)
429 (60.8) 62 (49–74) 296 (58.2) 258 (51.7) 209 (31.3) 562 (79.6) 221 (31.3) 395 (56.0) 5 (4–6)
0 0 197 (27.9) 207 (29.3) 38 (5.4) 76 (10.8) 0 0 196 (27.8)
123 (78.3) 65 (55–72) 83 (70.3) 62 (57.4) 115 (75.2) 119 (75.8) 59 (37.6) 123 (78.3) 5 (3–6)
Table 4 Adjusted and propensity match adjusted association of coronary angiography (CAG) with survival and good neurologic outcome. No-STEMI patients
Adjusted unpaired analysis No CAG (Ref) CAG without PCI CAG with PCI Propensity match adjusted No CAG (Ref) CAG
Survival 95% CI
OR
95% CI
1.00 2.34 1.98
– 1.69–3.24 1.26–3.09
1.00 6.89 2.95
– 3.99–11.91 1.59–5.47
1.00 2.10
– 1.30–3.39
1.00 5.06
– 2.29–11.19
39 49 4 18
42
0 0 (24.8) (31.2) (2.5) (11.5) 0 0 (26.8)
N (%)
Missing (%)
P value
259 (70.6) 63 (52–71) 181 (67.3) 130 (50.6) 233 (66.8) 254 (69.2) 166 (45.2) 299 (81.5) 5 (4–6)
0 0 98 (26.7) 110 (30.0) 18 (4.9) 52 (14.2) 0 0 98 (26.7)
< 0.0001 0.2978 < 0.05 0.4754 < 0.0001 < 0.001 < 0.0001 < 0.0001 0.6082
[7] Dumas F, Bougouin W, Geri G, Lamhaut L, Rosencher J, Pene F, et al. Emergency percutaneous Coronary intervention in post-cardiac arrest patients without STsegment elevation pattern: insights from the PROCAT II registry. JACC Cardiovasc Interv 2016;9:1011–8. [8] Casella G, Carinci V, Cavallo P, Guastaroba P, Pavesi PC, Pallotti MG, et al. Combining therapeutic hypothermia and emergent coronary angiography in out-ofhospital cardiac arrest survivors: optimal post-arrest care for the best patient. Eur Heart J Acute Cardiovasc Care 2015;4:579–88. [9] http://quickfacts.Census.Gov/qfd/states/04000.html, Accessed 10-14-15. [10] Bobrow BJ, Vadeboncoeur TF, Clark L, Chikani V. Establishing Arizona's statewide cardiac arrest reporting and educational network. Prehosp Emerg Care 2008;12:381–7. [11] Bobrow BJ, Clark LL, Ewy GA, Chikani V, Sanders AB, Berg RA, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA 2008;299:1158–65. [12] Bobrow BJ, Spaite DW, Berg RA, Stolz U, Sanders AB, Kern KB, et al. Chest compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA 2010;304:1447–54. [13] Spaite DW, Stiell IG, Bobrow BJ, de Boer M, Maloney J, Denninghoff K, et al. Effect of transport interval on out-of-hospital cardiac arrest survival in the OPALS study: implications for triaging patients to specialized cardiac arrest centers. Ann Emerg Med 2009;54:248–55. [14] Spaite DW, Bobrow BJ, Stolz U, Berg RA, Sanders AB, Kern KB, et al. Statewide regionalization of postarrest care for out-of-hospital cardiac arrest: association with survival and neurologic outcome. Ann Emerg Med 2014;64:496–506. [15] Jacobs I, Nadkarni V, Bahr J, Berg RA, Billi JE, Bossaert L, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare 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. [16] Haukoos JS, Newgard CD. Advanced statistics: missing data in clinical research–part 1: an introduction and conceptual framework. Acad Emerg Med 2007;14:662–8. [17] Newgard CD, Haukoos JS. Advanced statistics: missing data in clinical research–part 2: multiple imputation. Acad Emerg Med 2007;14:669–78. [18] Rubin DB. Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons; 1987. [19] Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar Behav Res 2011;46:399–424. [20] Patel N, Patel NJ, Macon CJ, Thakkar B, Desai M, Rengifo-Moreno P, et al. Trends and outcomes of coronary angiography and percutaneous coronary intervention after out-of-hospital cardiac arrest associated with ventricular fibrillation or pulseless ventricular tachycardia. JAMA Cardiol 2016;1:890–9. [21] Rab T, Kern KB, Tamis-Holland JE, Henry TD, McDaniel M, Dickert NW, et al. Cardiac arrest: a treatment algorithm for emergent invasive cardiac procedures in the resuscitated comatose patient. J Am Coll Cardiol 2015;66:62–73.
CPC 1 or 2
OR
Missing (%)
CAG without PCI (n = 367)
Model is adjusted for age, gender, bystander CPR performed, arrest witnessed, initial rhythm, targeted temperature management, GCS and response time.
Conflicts of interest None. References [1] Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. American Heart Association statistics C and stroke statistics S. heart disease and stroke statistics–2015 update: a report from the American Heart Association. Circulation 2015;131:e29–322. [2] Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132:S465–82. [3] Lotun K, Kern KB. How much is enough… what more is needed? Circ Cardiovasc Interv 2015:8. [4] Millin MG, Comer AC, Nable JV, Johnston PV, Lawner BJ, Woltman N, et al. Patients without ST elevation after return of spontaneous circulation may benefit from emergent percutaneous intervention: a systematic review and meta-analysis. Resuscitation 2016;108:54–60. [5] Kern KB, Lotun K, Patel N, Mooney MR, Hollenbeck RD, McPherson JA, et al. Outcomes of comatose cardiac arrest survivors with and without ST-segment elevation myocardial infarction: importance of coronary angiography. JACC Cardiovasc Interv 2015;8:1031–40. [6] Hollenbeck RD, McPherson JA, Mooney MR, Unger BT, Patel NC, McMullan Jr. PW, et al. Early cardiac catheterization is associated with improved survival in comatose survivors of cardiac arrest without STEMI. Resuscitation 2014;85:88–95.
25
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Association between coronary angiography with or without percutaneous coronary intervention and outcomes after out-of-hospital cardiac arrest☆
T
⁎
Tyler F. Vadeboncoeura, , Vatsal Chikanib,1, Chengcheng Huc,2, Danial W. Spaitec,3, Bentley J. Bobrowb,c,d,e,1,4 a
Department of Emergency Medicine, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL, 32224, United States The Arizona Department of Health Services Bureau of Emergency Medical Services and Trauma System, Phoenix, AZ, United States c Arizona Emergency Medicine Research Center, The University of Arizona College of Medicine, Tucson, AZ, United States d Maricopa Medical Center, Phoenix, AZ, United States e The University of Arizona Sarver Heart Center, Tucson, AZ, United States b
A R T I C LE I N FO
A B S T R A C T
Keywords: Cardiac arrest Post-resuscitation care Coronary angiography
Aim: The aim of our study was to assess the impact of coronary angiography (CAG) after out-of-hospital cardiac arrest (OHCA) without ST-elevation (STE). Methods: Prospective observational study of adult (age ≥ 18) OHCA of presumed cardiac etiology from 1/01/ 2010–12/31/2014 admitted to one of 40 recognized cardiac receiving centers within a statewide resuscitation network. Results: Among 11,976 cases, 1881 remained for analysis after exclusions. Of the 1230 non-STE cases, 524 (43%) underwent CAG with resultant PCI in 157 (30%). Survival in non-STE cases was: 56% in cases without CAG; 82% in cases with CAG but without PCI; and 78% in those with PCI (p < 0.0001). In cases without STE the aOR for survival with CAG alone was 2.34 (95% CI 1.69–3.24) and for CAG plus PCI was 1.98 (95% CI 1.26–3.09). The aOR for CPC 1/2 with CAG alone was 6.89 (95% CI 3.99–11.91) and for CAG plus PCI was 2.95 (95% CI 1.59–5.47). After propensity matching, CAG was associated with an aOR for survival of 2.10 (95% CI 1.30–3.55) and for CPC 1/2 it was 5.06 (95% CI 2.29–11.19). Conclusion: In OHCA without STE, CAG was strongly and independently associated with survival regardless of whether PCI was performed. The association between CAG and positive outcomes remained after propensity matching.
Introduction Out-of-hospital cardiac arrest (OHCA) remains a major public health problem in the United States [1]. Successful resuscitation requires a synchronized set of interdependent actions (the “chain of survival”) which in the latest 2015 American Heart Association (AHA) Guidelines includes targeted temperature management (TTM) and coronary angiography (CAG) after return of circulation [2]. Early revascularization is thought to be the primary benefit of emergent CAG after cardiac arrest [3]. Numerous (> 15) studies have reported improved rates of survival
to hospital discharge with emergent CAG in patients with ST-segment elevation (STE) [2]. As such, the AHA endorses emergent CAG for OHCA patients with suspected cardiac etiology and STE regardless of whether they are comatose or awake [2]. Several studies have reported lesions amenable to percutaneous coronary intervention (PCI) in a significant minority of patients without STE on their initial post-cardiac arrest electrocardiogram (EKG) [4–7]. While studies have associated emergent CAG in patients without STE after cardiac arrest with improved survival, PCI has not always been shown to provide added benefit [6,8]. The purpose of this study was to assess the association between CAG
☆
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.03.023. Corresponding author. E-mail addresses: [email protected] (T.F. Vadeboncoeur), [email protected] (V. Chikani), [email protected] (C. Hu), [email protected] (D.W. Spaite), [email protected] (B.J. Bobrow). 1 Bureau of Emergency Medical Services, Arizona Department of Health Services, 150 N. 18th Avenue, #540, Phoenix, AZ 85007, United States. 2 University of Arizona, PO Box 245057, 1501 N. Campbell, Tucson, AZ 85724-5057, United States. 3 Department of Emergency Medicine, University of Arizona, Box 245057, 1501 N. Campbell, Tucson, AZ 85724-5057, United States. 4 Address for reprints: Arizona Department of Health Services, 150 N 18th Avenue, Suite 540, Phoenix, AZ, United States. ⁎
https://doi.org/10.1016/j.resuscitation.2018.03.023 Received 17 October 2017; Received in revised form 12 January 2018; Accepted 12 March 2018 0300-9572/ © 2018 Elsevier B.V. All rights reserved.
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
alone and CAG plus PCI on outcomes after OHCA without initial STE.
Table 1 Demographics and event characteristics.
Methods
N = 1881
Setting
n (%) Male Age, yrs-median (Q1, Q3) Arrest witnessed Bystander CPR performed Initial rhythm VF/VT No-STEMI Intervention CAGa with PCIb CAG without PCI No CAG GCS < =8 >8 Targeted temperature management Survival Response Interval, minutes- median (Q1, Q3) CPC scale All 1 2 3 4 5
Arizona had 6.7 million residents in 2014 [9]. The Arizona Department of Health Services (ADHS) establishes the EMS scope of practice and provider certification. EMS protocols are implemented on regional level. Crew configuration, vehicle deployment, dispatch, and response intervals vary widely across the state and some local OHCA bypass protocols are in place. About 100 EMS agencies and 40 Cardiac Receiving Centers (CRCs) responding to approximately 80% of the Arizona’s population voluntarily participated in the state-sponsored Save Hearts in Arizona Registry and Education (SHARE) Program during the study period. SHARE has been described previously [10–14]. In order to be recognized as a CRC, a hospital must have: 1) primary 24/7 PCI capability with a protocol including calling cardiology for OHCA, 2) a dedicated TTM protocol for OHCAs that remain comatose, 3) an evidence-based termination of resuscitation protocol which includes a 72-h moratorium on termination of care for patients receiving TTM, and 4) commitment to on-going data submission for all OHCA patients (www.azshare.gov). There are no specific mandates or protocols regarding the timing or selection of patients for coronary angiography. The decision is left to the treating providers which includes required cardiology involvement.
a b
Missing, n (%)
1289 (68.5) 63.0 (52.0, 72.0) 837 (63.3) 677 (53.5) 930 (52.6) 1230 (71.1)
0 0 559 (29.7) 615 (32.7) 113 (6.0) 152 (8.1)
482 (25.6) 525 (27.9) 874 (46.5)
0
1427 (86.6) 221 (13.4) 655 (34.8) 1192 (63.4) 5.0 (4.0, 6.0)
233 (12.4)
1822 (96.9) 740 (40.6) 182 (10.0) 62 (3.4) 149 (8.2) 689 (37.8)
59 (3.1)
0 0 568 (30.2)
CAG – coronary angiography. PCI – percutaneous coronary intervention.
Study design and population Department of Health Services Human Subjects Review Board have determined that, by virtue of being a public health initiative, neither the interventions nor their evaluation constitute human subjects research and have approved the publication of de-identified data.
This is a prospective, multicenter, observational cohort study of consecutive adult (aged ≥18 years) patients with OHCA between January 1, 2010, and December 31, 2014, who were transported initially or transferred to a CRC. Cases were excluded if prehospital resuscitation was not initiated, the cause of the arrest was presumed to be non-cardiac (e.g., known respiratory arrest, suicide, trauma, drowning, or drug overdose), the patient had a Do-Not-Resuscitate order, or the patient died in the emergency department (ED).
Statistical analysis To maximize the available subjects for analysis, multiple imputation was carried out in SAS (SAS, Version 9.3, SAS Institute, Cary, North Carolina) to impute missing data. Multiple imputation (MI) has been shown to generate less biased estimates with more statistical efficiency when compared with alternative methods of handling incomplete data (e.g., complete-case analysis, single imputation, missing indicator regression) [16,17]. MI involves three distinct phases: 1) the missing data are filled in m times to generate m complete data sets, 2) the m datasets are analyzed by using standard procedures, and 3) the results from the m complete datasets are combined for the inference. MI procedure replaces each missing value with a set of plausible values that represent the uncertainty about the right value to impute, instead of filling in a single value for each missing value. We used all the variables in Table 1 for MI. Twenty imputed data sets were generated and model fit and diagnostics were evaluated. Missing data fit an arbitrary missing pattern and we used Fully Conditional Specification method to impute data. Linear regression was used to impute all time intervals. Logistic regression was used to impute categorical variables, witnessed arrest, bystander cardiopulmonary resuscitation (CPR), shockable rhythm and STE. The study population was divided into three groups, patients without CAG (reference group), patients with CAG and without PCI, and patients with CAG and PCI. Descriptive statistics were used to describe the study population and are reported as proportion for categorical and binary data and as median and interquartile range for continuous data. All analyses used the imputed data, accounting for variance across imputed data sets using Rubin’s rules [18] (using Proc MI in SAS). To control for indication bias, propensity score was calculated to predict the probability of receiving CAG using a multivariable logistic regression model, including age, gender, bystander CPR
Data processing and study approval EMS data are obtained from the patient care reports and outcomes are obtained either from the hospitals or from the State Office of Vital Records. SHARE includes an Utstein-style OHCA EMS database linked with in-hospital post-arrest care and outcome data. Consistent with Utstein methodology, each OHCA in which EMS attempts resuscitation is included. EMS data are cross-referenced between first responding EMS agencies, private ambulance transport companies, and the CRC database. In order to perform continuous quality improvement for inhospital post-arrest care a data tool was developed to collect in-hospital patient information for all OHCA patients brought to a recognized CRC. The CRC data include details on CAG including the initial EKG findings, the timing of CAG, and whether or not PCI is performed. It does not include angiographic data. It also includes the final patient outcome including the Cerebral Performance Category (CPC) score [15]. The data forms were completed by CRC clinical personnel using a secure web-based data entry system. Each hospital individual completing data forms was trained in person and each form was secondarily reviewed by a SHARE data coordinator for completeness and accuracy before entry into the database. Any inconsistencies were addressed in follow-up by examining the hospital medical record. SHARE is an Arizona Department of Health Services–sponsored public health initiative. As such the program is exempt from the requirements of the Health Insurance Portability and Accountability Act, which allows linkage of EMS and hospital data, tracking of OHCA events, and evaluation of efforts to improve resuscitation care. The University of Arizona Institutional Review Board and the Arizona 22
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
performed, arrest witnessed, initial rhythm, TTM, GCS and response time as covariates. The multivariate logistic regression analysis was then carried out to investigate the association between CAG with or without PCI, and survival with propensity score included as a covariate in the model along with the other covariates (adjusted unpaired analysis). Next, patients were categorized into two groups based on their CAG status (CAG vs no CAG) and were matched by propensity scores using a 1:1 matching procedure without replacement. Among the matched patients, the standardized difference for each of the covariates between the two groups was < 0.1 suggesting a negligible difference in the prevalence of the covariates between the two groups and, hence, successful matching [19]. Finally, we compared the survival and favorable neurologic outcomes between the CAG and no CAG groups for the matched pairs using the multivariable logistic regression model (propensity match adjusted analysis). Data were analyzed using Statistical Analysis System (SAS version 9.3, SAS Institute, Inc. Cary, NC). P value of < 0.05 was considered statistically significant.
initial rhythm VF/VT and to receive TTM. They were more likely to be comatose upon arrival to the ED (Table 3). To account for indication bias as the three groups differ significantly on most of the covariates, we estimated the propensity of receiving CAG for each of the non-STE patients using a multivariate logistic regression model. The adjusted unpaired logistic regression analysis with propensity score as a covariate showed that CAG was associated with improved survival and neurologic outcomes whether or not PCI was performed. The aOR for survival with CAG alone was 2.34 (95% CI 1.69–3.24) and for CAG with PCI was 1.98 (95% CI 1.26–3.09). The aOR for CPC 1/2 with CAG alone was 6.89 (95% CI 3.99–11.91) and for CAG with PCI was 2.95 (95% CI 1.59–5.47) (Table 4). After matching on propensity score, the survival benefit of CAG persisted. CAG conferred an aOR for survival of 2.10 (95% CI 1.30–3.55) and for CPC 1/2 of 5.06 (95% CI 2.29–11.19). Results were similar when including only comatose (GCS ≤ 8) patients and when analyzing the subgroup without initial STE but with an initial shockable rhythm.
Results
Discussion
There were 11,976 OHCA cases during the study period. After exclusions, the study population was 1881 (Fig. 1), 1007 (54%) of which underwent CAG. Of those undergoing CAG 482 (48%) had PCI (Table 1). Overall, 63% of cases survived to hospital discharge and 81% of survivors had a CPC 1/2. Ninety percent of (450/499) STE cases underwent CAG, 70% of which had PCI. Of the 1230 cases without initial STE 43% underwent CAG, 157 (30%) of which had PCI performed (Table 2). Fig. 2 presents a histogram depicting the timing of CAG. Overall survival to hospital discharge was 53% for those without CAG and 72% for those who underwent CAG whether or not PCI was performed (p < 0.0001) (Table 2). In the 1230 cases without initial STE survival was again better in those who underwent CAG whether or not PCI was performed (78.3% and 81.5%) compared to the group without CAG (56.0%, p < 0.0001). Table 2 also presents the results by CPC score which demonstrates favorable neurologic outcomes in those undergoing CAG, but no difference between CAG with PCI or without PCI. After restricting the analyses to only patients who did not have STE (1230), those without CAG were less likely to be witnessed, to have
The AHA Guidelines recommend that OHCA with initial STE undergo emergent CAG whether or not they are comatose [2]. It remains less clear whether OHCA patients without initial STE should be taken for early CAG. While observational data regularly show associations between CAG and improved outcomes even without STE, authors have suggested that indication bias has impacted these findings [2,5,6,20]. The fact that CAG with PCI has not consistently been associated with better outcomes when compared to CAG alone lends further support to the argument that there is bias [6,8]. In this statewide analysis encompassing 40 CRCs and over 100 EMS agencies we used logistic regression and propensity matching in an effort to control for confounding. CAG was widely utilized across the state with 90% of OHCA with STE and 43% of OHCA without initial STE undergoing CAG. These numbers are strikingly similar to previous observational studies reporting rates of CAG for OHCA (93% with STE and 42% without STE) [4]. We did not assess the impact of CAG in STE because only a small minority did not undergo CAG. In cases without STE, CAG was associated with better survival to discharge and neurologic outcomes, but PCI did not confer additional benefit. (Table 2) As shown in Table 3, subjects who underwent CAG we more likely to have a witnessed arrest, to have an initial shockable rhythm, and to receive TTM. They were also less likely to be comatose upon arrival to the ED. Because, all of these factors have been associated with better outcomes, a propensity analysis was performed to assess whether indication bias played a role in our findings. After matching, CAG was still associated with a doubling of survival to hospital discharge and an even stronger association with a positive neurologic outcome (Table 4). Previous studies have been conflicting regarding whether OHCA without STE should undergo early CAG. In a recent meta-analysis of 11 articles Millin et al reported that 32.2% of OHCA without STE taken to CAG had an acute culprit lesion requiring intervention [4]. Consistently, in the PROCAT II study, Dumas et al brought all OHCA without an obvious non-cardiac cause to immediate CAG and found lesions amenable to PCI in 29% (199/695) [7]. There was a favorable outcome in 43% of those undergoing PCI and 33% without PCI. Conversely, Hollenbeck et al reviewed 269 OHCA due to ventricular arrhythmia without STE who were treated with TTM and found that early CAG was strongly associated with survival (65.6% vs. 48.6%; p = 0.017), but that PCI did not confer an additional benefit over CAG alone [6]. Finally, Casella et al reviewed 278 comatose OHCA treated with TTM and found that CAG, but not PCI, was associated with improved outcomes [8]. The propensity analysis combined with the large size of this study, including multiple EMS agencies and hospitals, makes these findings as strong and generalizable as any previous study evaluating CAG and PCI
Fig. 1. Study population profile. 23
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
Table 2 Survival and neurologic outcomes by coronary angiography (CAG) and percutaneous coronary intervention (PCI). Overall (N = 1881)
No-CAG CAG without PCI CAG with PCI P-value
No STEMI (n = 1230)
No STEMI VF/VT (n = 557)
n
Survival
CPC Scale (1/2)
n
Survival
CPC Scale (1/2)
n
Survival
CPC Scale (1/2)
874 525 482
% (N) 53.1% (464) 72.2% (379) 72.4% (349) < 0.0001
30.6% (268) 64.0% (336) 66.0% (318) < 0.0001
706 367 157
% (N) 56.0% (395) 81.5% (299) 78.3% (123) < 0.0001
34.1% (241) 75.2% (276) 68.2% (107) < 0.0001
209 233 115
% (N) 72.7% (152) 87.1% (203) 82.6% (95) < 0.001
49.3% (103) 82.4% (192) 73.0% (84) < 0.0001
findings. For example, we chose to include all cases whether or not they remained comatose and we controlled for GCS. The vast majority of cases, however, were comatose and, as such, when we ran the analysis on only the comatose cases the findings were nearly identical. We also chose to include all cases with CAG during the index hospitalization rather than only those with CAG in the first 2, 12 or 24 h. The histogram in Fig. 2 shows that the vast majority of CAG was in the first 12 h. Clearly late CAG requires survival at least until that point, however it is also true that late CAG is not the same as no CAG. Finally, multiple imputation was required because we were not able to match all CRC cases with their out-of-hospital data. As such, there was a significant amount of “missing” prehospital data.
after OHCA without STE. Our findings add to the body of literature suggesting that CAG is a beneficial component of post-cardiac arrest care even without initial STE. Our results, however, do not make clear that revascularization is the primary driver of the association between CAG and improved outcomes. As such, it is possible that the association with improved outcomes is a result of other factors such as a higher intensity of care with early CAG or even that CAG facilitates making alternative diagnoses (such as pulmonary embolism) or informs other treatment decisions. In our study, CAG patients were more likely to be treated with TTM suggesting a higher intensity of care. Hollenbeck et al found that patients treated with CAG were more likely to receive mechanical support with intraaortic balloon counterpulsation and were also treated with more aggressive anticoagulation [6]. It has also been suggested that CAG might be associated with timely venous access, invasive hemodynamic monitoring and rapid titration of vasoactive medications [6]. The resuscitation community is anxiously awaiting the results of the randomized DISCO and PEARL trials to help determine which and/or whether OHCA patients without initial STE should go to early CAG. Until that time, authors such as Rab et al have attempted to create risk stratification tools for comatose cardiac arrest patient to determine which non-STE patients are most likely to benefit early CAG [21].
Conclusions In conclusion, we found that coronary angiography was widely utilized after out-of-hospital cardiac arrest in this statewide cardiac receiving system. In cases without initial ST-elevation, a significant minority of cases had a lesion amenable to percutaneous coronary intervention and coronary angiography was associated with improved outcomes with or without PCI. After propensity matching, coronary angiography remained significantly associated with higher survival to hospital discharge and positive neurologic outcomes.
Limitations As with any observational study, this analysis has several limitations. Despite the large size of this study and the use of propensity matching, unaccounted for bias still remains a possibility. We also selected specific inclusion criteria which could potentially impact the
Role of funding None.
Fig. 2. Emergency department arrival to coronary angiography time. 24
Resuscitation 127 (2018) 21–25
T.F. Vadeboncoeur et al.
Table 3 Demographics and event characteristics by type of intervention: No-STEMI patients. No-STEMI patients (N = 1230)
No CAG (n = 706)
Male Age, yrs-median (Q1, Q3) Arrest witnessed Bystander CPR performed Initial rhythm VF/VT GCS < = 8 Targeted temperature management Survival Response Interval, minutes- median (Q1, Q3)
CAG with PCI (n = 157)
N (%)
Missing (%)
N (%)
429 (60.8) 62 (49–74) 296 (58.2) 258 (51.7) 209 (31.3) 562 (79.6) 221 (31.3) 395 (56.0) 5 (4–6)
0 0 197 (27.9) 207 (29.3) 38 (5.4) 76 (10.8) 0 0 196 (27.8)
123 (78.3) 65 (55–72) 83 (70.3) 62 (57.4) 115 (75.2) 119 (75.8) 59 (37.6) 123 (78.3) 5 (3–6)
Table 4 Adjusted and propensity match adjusted association of coronary angiography (CAG) with survival and good neurologic outcome. No-STEMI patients
Adjusted unpaired analysis No CAG (Ref) CAG without PCI CAG with PCI Propensity match adjusted No CAG (Ref) CAG
Survival 95% CI
OR
95% CI
1.00 2.34 1.98
– 1.69–3.24 1.26–3.09
1.00 6.89 2.95
– 3.99–11.91 1.59–5.47
1.00 2.10
– 1.30–3.39
1.00 5.06
– 2.29–11.19
39 49 4 18
42
0 0 (24.8) (31.2) (2.5) (11.5) 0 0 (26.8)
N (%)
Missing (%)
P value
259 (70.6) 63 (52–71) 181 (67.3) 130 (50.6) 233 (66.8) 254 (69.2) 166 (45.2) 299 (81.5) 5 (4–6)
0 0 98 (26.7) 110 (30.0) 18 (4.9) 52 (14.2) 0 0 98 (26.7)
< 0.0001 0.2978 < 0.05 0.4754 < 0.0001 < 0.001 < 0.0001 < 0.0001 0.6082
[7] Dumas F, Bougouin W, Geri G, Lamhaut L, Rosencher J, Pene F, et al. Emergency percutaneous Coronary intervention in post-cardiac arrest patients without STsegment elevation pattern: insights from the PROCAT II registry. JACC Cardiovasc Interv 2016;9:1011–8. [8] Casella G, Carinci V, Cavallo P, Guastaroba P, Pavesi PC, Pallotti MG, et al. Combining therapeutic hypothermia and emergent coronary angiography in out-ofhospital cardiac arrest survivors: optimal post-arrest care for the best patient. Eur Heart J Acute Cardiovasc Care 2015;4:579–88. [9] http://quickfacts.Census.Gov/qfd/states/04000.html, Accessed 10-14-15. [10] Bobrow BJ, Vadeboncoeur TF, Clark L, Chikani V. Establishing Arizona's statewide cardiac arrest reporting and educational network. Prehosp Emerg Care 2008;12:381–7. [11] Bobrow BJ, Clark LL, Ewy GA, Chikani V, Sanders AB, Berg RA, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA 2008;299:1158–65. [12] Bobrow BJ, Spaite DW, Berg RA, Stolz U, Sanders AB, Kern KB, et al. Chest compression-only CPR by lay rescuers and survival from out-of-hospital cardiac arrest. JAMA 2010;304:1447–54. [13] Spaite DW, Stiell IG, Bobrow BJ, de Boer M, Maloney J, Denninghoff K, et al. Effect of transport interval on out-of-hospital cardiac arrest survival in the OPALS study: implications for triaging patients to specialized cardiac arrest centers. Ann Emerg Med 2009;54:248–55. [14] Spaite DW, Bobrow BJ, Stolz U, Berg RA, Sanders AB, Kern KB, et al. Statewide regionalization of postarrest care for out-of-hospital cardiac arrest: association with survival and neurologic outcome. Ann Emerg Med 2014;64:496–506. [15] Jacobs I, Nadkarni V, Bahr J, Berg RA, Billi JE, Bossaert L, et al. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare 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. [16] Haukoos JS, Newgard CD. Advanced statistics: missing data in clinical research–part 1: an introduction and conceptual framework. Acad Emerg Med 2007;14:662–8. [17] Newgard CD, Haukoos JS. Advanced statistics: missing data in clinical research–part 2: multiple imputation. Acad Emerg Med 2007;14:669–78. [18] Rubin DB. Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons; 1987. [19] Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar Behav Res 2011;46:399–424. [20] Patel N, Patel NJ, Macon CJ, Thakkar B, Desai M, Rengifo-Moreno P, et al. Trends and outcomes of coronary angiography and percutaneous coronary intervention after out-of-hospital cardiac arrest associated with ventricular fibrillation or pulseless ventricular tachycardia. JAMA Cardiol 2016;1:890–9. [21] Rab T, Kern KB, Tamis-Holland JE, Henry TD, McDaniel M, Dickert NW, et al. Cardiac arrest: a treatment algorithm for emergent invasive cardiac procedures in the resuscitated comatose patient. J Am Coll Cardiol 2015;66:62–73.
CPC 1 or 2
OR
Missing (%)
CAG without PCI (n = 367)
Model is adjusted for age, gender, bystander CPR performed, arrest witnessed, initial rhythm, targeted temperature management, GCS and response time.
Conflicts of interest None. References [1] Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. American Heart Association statistics C and stroke statistics S. heart disease and stroke statistics–2015 update: a report from the American Heart Association. Circulation 2015;131:e29–322. [2] Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132:S465–82. [3] Lotun K, Kern KB. How much is enough… what more is needed? Circ Cardiovasc Interv 2015:8. [4] Millin MG, Comer AC, Nable JV, Johnston PV, Lawner BJ, Woltman N, et al. Patients without ST elevation after return of spontaneous circulation may benefit from emergent percutaneous intervention: a systematic review and meta-analysis. Resuscitation 2016;108:54–60. [5] Kern KB, Lotun K, Patel N, Mooney MR, Hollenbeck RD, McPherson JA, et al. Outcomes of comatose cardiac arrest survivors with and without ST-segment elevation myocardial infarction: importance of coronary angiography. JACC Cardiovasc Interv 2015;8:1031–40. [6] Hollenbeck RD, McPherson JA, Mooney MR, Unger BT, Patel NC, McMullan Jr. PW, et al. Early cardiac catheterization is associated with improved survival in comatose survivors of cardiac arrest without STEMI. Resuscitation 2014;85:88–95.
25