Resuscitation 88 (2015) 143–149
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Correlation between initial serum levels of lactate after return of spontaneous circulation and survival and neurological outcomes in patients who undergo therapeutic hypothermia after cardiac arrest夽 Dong Hoon Lee a,1 , In Soo Cho b,∗ , Sun Hwa Lee c,2 , Yong Il Min d,3 , Jin Hong Min e,4 , Soo Hyun Kim f,5 , Young Hwan Lee g,6 , the Korean Hypothermia Network Investigators a
Department of Emergency Medicine, College of Medicine, Chung-Ang University, Seoul, Republic of Korea Department of Emergency Medicine, KEPCO Medical Center, Seoul, Republic of Korea c Department of Emergency Medicine, Ewha Womans University, Seoul, Republic of Korea d Department of Emergency Medicine, Chonnam National University, Gwangju, Republic of Korea e Department of Emergency Medicine, College of Medicine Chungbuk National University, Cheongju-si, Republic of Korea f Department of Emergency Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea g Departments of Emergency Medicine, Hallym University Sacred Heart Hospital, Anyang-si, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 25 July 2014 Received in revised form 6 November 2014 Accepted 9 November 2014 Keywords: Lactate Outcome Therapeutic hypothermia Cardiac arrest
a b s t r a c t Objectives: We analysed the relationship between serum levels of lactate within 1 h of return of spontaneous circulation (ROSC) and survival and neurological outcomes in patients who underwent therapeutic hypothermia (TH). Methods: This was a multi-centre retrospective and observational study that examined data from the first Korean Hypothermia Network (KORHN) registry from 2007 to 2012. The inclusion criteria were out-of-hospital cardiac arrest (OHCA) and examination of serum levels of lactate within 1 h after ROSC, taken from KORHN registry data. The primary endpoint was survival outcome at hospital discharge, and the secondary endpoint was poor neurological outcome (Cerebral Performance Category, CPC, 3–5) at hospital discharge. Initial lactate levels and other variables collected within 1 h of ROSC were analysed via multivariable logistic regression. Results: Data from 930 cardiac arrest patients who underwent TH were collected from the KORHN registry. In a total of 443 patients, serum levels of lactate were examined within 1 h of ROSC. In-hospital mortality was 289/443 (65.24%), and 347/443 (78.33%) of the patients had CPCs of 3–5 upon hospital discharge. The odds ratios of lactate levels for CPC and in-hospital mortality were 1.072 (95% confidence interval (CI) 1.026–1.121) and 1.087 (95% CI = 1.031–1.147), respectively, based on multivariate ordinal logistic regression analyses. Conclusion: High levels of lactate in serum measured within 1 h of ROSC are associated with hospital mortality and high CPC scores in cardiac arrest patients treated with TH. © 2014 Elsevier Ireland Ltd. All rights reserved.
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.11.005. ∗ Corresponding author. E-mail addresses: [email protected] (D.H. Lee), [email protected] (I.S. Cho), [email protected] (S.H. Lee), [email protected] (Y.I. Min), [email protected] (J.H. Min), [email protected] (S.H. Kim), [email protected] (Y.H. Lee). 1 Department of Emergency Medicine, Chung-Ang University Hospital, 224-1 Heoukseok-dong, Dongjak-gu, Seoul, Republic of Korea 2 Department of Emergency Medicine, Ewha Womans University Medical center 911-1 Mok-dong, Yang chun-Gu, Republic of Korea 3 Department of Emergency Medicine, Chonnam National University Hospital, 42 Jebong-ro, Dong-gu, Gwangju, Republic of Korea 4 Department of Emergency Medicine, College of Medicine Chungbuk National University Hospital, Gaesin-dong, Heungdeok-gu, Cheongju-si, Chungbuk, Republic of Korea 5 Emergency Medicine, Seoul St. Mary’s hospital, 222 Banpo-daero, Seocho-Gu, Seoul, Republic of Korea 6 Department of Emergency Medicine, Hallym University Sacred Heart Hospital, Hallym University Medical Center, 896 Pyeongchon-dong, Dongan-gu, Anyang-si, Gyeonggido, Republic of Korea http://dx.doi.org/10.1016/j.resuscitation.2014.11.005 0300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.
144
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
1. Introduction Therapeutic hypothermia (TH) has improved the survival and neurological outcomes of patients who survive cardiac arrest 1,2 . The therapy is commonly used for comatose survivors with cardiac arrest as part of a series of survival treatments for cardiopulmonary resuscitation 3 . During TH, physicians need to be aware of the neurological and survival outcomes of comatose patients. Thus, several studies have been performed to develop tools to predict the neurological prognosis and survival outcomes. However, no single effective method for the prediction of outcome has been found. Several biological markers, such as the S-100B protein and neuron-specific enolase, have been demonstrated to be useful, and electroencephalography and neuroimaging, including brain computed tomography and magnetic resonance imaging, have been used by physicians to predict neurological outcomes 4–12 . However, these methods require serial check-ups and longer times to determine results. An effective method for predicting survival and neurological outcomes that can be applied early on in the treatment process would be useful for planning treatment strategies for post-cardiac arrest care. However, there is a lack of objective data on such patients collected within 1 h after return of spontaneous circulation (ROSC) in the emergency department (ED). Such data would comprise pre-hospital data, including data from emergency medical service (EMS); initial laboratory findings taken after ROSC; and the results of physical and neurological examinations performed in the ED. Serum lactate levels are one of these initial laboratory findings. Early and effective clearance of serum lactate (i.e., lower levels) are associated with lower mortality in patients with many critical illnesses, including sepsis and severe trauma 13–18 . For cardiac arrest survivors, the relationships between clearance of serum lactate and neurological and survival outcomes have been studied 19–23 . However, few studies have addressed the potential relationship between initial serum levels of lactate and outcomes in patients who have been treated with TH. Thus, in this study, we evaluated whether initial levels of lactate in the serum at ROSC were correlated with neurological and survival outcomes of patients undergoing TH.
(http://starcrf.com) by the clinical research coordinator. A data manager and three clinical research associates regularly monitored and reviewed data quality. The site’s principal investigators or clinical research coordinators could be contacted through the query function in the system or directly by phone to clarify any questions about the data.
2.2. Data collection Data that could be collected within 1 h after ROSC were used. The inclusion criteria from the KORHN registry data were OHCA and examination of serum levels of lactate within 1 h after ROSC (Fig. 1). Specifically, the data were demographics (gender, age), cause of arrest, pre-hospital resuscitation variables (witness, bystander CPR, ECG rhythm in EMS, pre-hospital time), in-hospital resuscitation variables (ECG rhythm at ED arrival, amount of injected epinephrine, hospital resuscitation time), neurological examination after ROSC (corneal reflex, light reflex, myoclonus), serum levels of lactate within 1 h after ROSC, and neurological and survival outcomes at discharge.
2.3. Outcome measures The primary endpoint of the study was survival outcome at hospital discharge. A secondary endpoint was poor neurological outcome at hospital discharge, defined as a Cerebral Performance Category (CPC) of 3–5. CPC 3 indicates severe cerebral disability: conscious, dependent on others for daily support because of impaired brain function. It ranges from an ambulatory state to severe dementia or paralysis. CPC 4 indicates a coma or a vegetative state: any degree of coma without the presence of brain death criteria. Unawareness, even if appears awake (vegetative states) without interaction with environment; may have spontaneous eye opening and sleep/awake cycles. CPC 5 indicates cerebral unresponsiveness, indicates brain death: apnoea, areflexia, EEG silence.
2.4. Statistical analysis 2. Methods 2.1. Registry data collection This was a multi-centre, retrospective, observational, registrybased study that used data from the first Korean Hypothermia Network (KORHN) registry. KORHN, a multi-centre clinical research consortium for TH in South Korea, was organised in 2011, and a multi-centre retrospective, registry project was performed in 2012. The KORHN investigators collected data regarding post-cardiac arrest TH in 24 teaching hospitals around South Korea from 2007 to 2012. The institutional review board of each institution approved the study protocol before data collection. The need for informed consent was waived because of the retrospective nature of the study. Adults (≥18 years) who suffered out-of-hospital cardiac arrest (OHCA), treated with TH after ROSC achieved pre-hospital and in the ED, were included in the study. Traumatic cardiac arrest patients were excluded. In a case report form, standard definitions of 87 variables and an investigator manual were developed from a literature review and consensus of the study investigators. The registry data were collected from manual medical charts or electronic medical record reviews. Data collected in each hospital were verified for completeness by the site’s principal investigator, and were recorded on a web-based data registration system
Baseline descriptive characteristics of the subjects are presented with basic statistical analyses including means and standard deviations (SDs) for continuous variables, and numbers and percentages for categorical variables. Patient data were analysed according to survival and neurological outcomes at hospital discharge. Univariate and multivariate linear regression were used to evaluate the relationship between lactate levels and variables before hospital arrival. To compare lactate levels and neurological outcomes (CPC scores) as continuous variables, ordinal univariate and multivariate logistic regressions were used. All of the parameters were included in the multivariable logistic analysis to determine independent predictors for neurologic outcomes. Binary univariate and multivariate logistic regression analyses were used to estimate correlations between lactate levels and poor neurological outcome. CPC scores of 3, 4, and 5 were considered to be poor neurological outcomes. For correlation between lactate levels and mortality, binary univariate and multivariate logistic regression analyses were used. Odds ratios (ORs) and 95% confidence intervals (CIs) are reported as estimates of variability and effect size. p-Values ≤0.05 were considered statistically significant. p-Values were corrected for multiple comparisons and within-method comparisons using the Holm–Bonferroni method. All of the statistical analyses were performed using the SPSS (ver. 20.0 for Windows; Chicago, IL) and SAS (SAS Institute, Cary, NC) software.
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
145
Fig. 1. Flow chart of participants. CPC: cerebral performance category.
3. Results Data on 930 cardiac arrest patients with TH were collected from the KORHN registry; 443 patients who met the inclusion criteria were enrolled. Of these patients, 14 arrived at the ED with ROSC, and 429 patients were resuscitated at the ED. The general characteristics of these patients are listed in Table 1 according to hospital mortality and neurological outcomes at hospital discharge. The mean age of the patients was 57.70 ± 16.06 years. There were 295 (66.6%) males and 148 (33.4%) females. The cardiac rhythm, confirmed by EMS, was shockable in 100 patients (22.6%), unshockable in 239 (54.0%), and unknown in 104 (23.5%). Initial electrocardiography (ECG) on ED arrival with a cardiac arrest state was ventricular fibrillation (Vf) or pulse-less ventricular tachycardia (VT) in 85 patients (19.2%), pulse-less electrical activity (PEA) in 95 (21.4%), and asystole in 249 (56.2%). In total, 14 patients (3.2%) arrived at the ED in a ROSC state. 254 patients (57.3%) survived and were discharged. The neurological outcomes at hospital discharge were 81 (18.3%) in CPC 1, 15 (3.4%) in CPC 2, 37 (8.4%) in CPC 3, 121 (27.3%) in CPC 4, and 189 in CPC 5 (41.7%). 3.1. Initial lactate levels after ROSC The level of lactate within 1 h after ROSC was 10.32 ± 4.52 mmol L−1 in all of the enrolled patients. In comparing survival and neurological outcomes, there were statistically significant differences between surviving and expired patients, and between those with good and poor neurological outcomes (Fig. 2). Lactate levels were analysed with other variables in univariate simple linear regression and multivariate linear regression. In the simple linear regression analysis, the OR of age was 0.958 (95% CI = −0.069 to 0.017), and the OR of pre-hospital time was 1.062 (95% CI = 0.028–0.092). In multiple regression analysis, age greater than 40 years and a pre-hospital time of more than 20 min were statistically significantly associated with lactate levels (Table 2).
1.656–3.470), pre-hospital time (1.034: 1.018–1.050), unshockable ECG rhythm in EMS (4.971: 3.187–7.752), no defibrillation by the EMS (2.888: 1.876–4.444), PEA and asystole at ED arrival (3.309: 1.929–5.677 and 5.612: 3.511–8.972, respectively), no defibrillation in the ED (2.848: 1.981–4.097), and no light reflex and no corneal reflex after ROSC (4.265: 2.910–6.251 and 4.005: 2.608–6.151, respectively). In multivariate regression analysis, CPC score was associated with lactate levels, age, unwitnessed cardiac arrest, unshockable rhythm in the EMS, no defibrillation in the ED, and no light reflex after ROSC (Table 3). 3.3. Prediction of hospital mortality The difference in lactate levels between survivors and non-survivors at hospital discharge was statistically significant (9.55 ± 4.33 mmol L−1 in survivors, 11.36 ± 4.58 mmol L−1 in non-survivors; Fig. 1). In univariate logistic regression analysis for hospital mortality, the OR of lactate was 1.096 (95% CI = 1.049–1.146). Other statistically significant variables were male gender (OR = 1.631: 95% CI = 1.095–2.430), age (1.026: 1.013–1.039), unwitnessed cardiac arrest (2.198: 1.479–3.267), pre-hospital time (1.032: 1.015–1.049), unshockable ECG rhythm in EMS (3.591: 2.125–6.067), no defibrillation by the EMS (1.985: 1.204–3.274), PEA and asystole on ED arrival (0.242: 1.258–4.652 and 3.675: 2.083–6.482, respectively), no defibrillation in the ED (2.271: 1.503–3.432), no light reflex and no corneal reflex after ROSC (2.545: 1.649–3.929 and 2.974: 1.770–4.999, respectively), and no seizure and no myoclonus (2.168: 1.314–3.577 and 2.593: 1.499–4.486, respectively). In multivariate logistic regression analysis, initial lactate levels were statistically significantly associated with non-survival at hospital discharge (1.087: 1.031–1.147). Moreover, age over 50, unwitnessed cardiac arrest, no bystander CPR, pre-hospital time >27 min, and no defibrillation in the ED were associated with mortality after hospital admission (Table 3). 4. Discussion
3.2. Prediction of neurological outcome at hospital discharge The OR of lactate levels for CPC was 1.072 (95% CI = 1.026–1.121) in multivariate ordinal logistic regression analysis. In univariate regression analysis, lactate was associated with CPC score (OR = 1.090, 95% CI = 1.048–1.135), and there were statistically significant correlations with male gender (1.590: 1.101–2.297), age (1.033: 1.022–1.045), unwitnessed cardiac arrest (2.399:
In resuscitated patients from OHCA undergoing TH, initial serum levels of lactate within 1 h after ROSC were associated with survival and neurological outcomes at hospital discharge (Table 3). Many investigators have sought to develop effective tools that can predict survival and neurological outcomes in post-cardiac arrest patients. Thus, many biomarkers, brain imaging methods, and neurophysiological examinations have been studied as
146
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
Table 1 Baseline characteristics of study subjects (n = 443). Survival outcome at discharge
Sex Male Female
Neurologic outcome
Survival (n = 254)
No survival (n = 289)
Good (n = 96)
Poor (n = 347)
181 (71.3) 73 (28.7)
114 (60.3) 75 (39.7)
69 (71.9) 27 (28.1)
226 (65.1) 121 (34.9)
55.02 ± 15.84 77 (30.3) 78 (30.7) 54 (21.3) 45 (17.7)
Age (year) 18–40 41–50 51–64 >65
61.31 ± 15.68 38 (20.1) 37 (19.6) 53 (28.0) 61 (32.3)
49.17 ± 15.26 41 (42.7) 32 (33.3) 14 (14.6) 9 (9.4)
60.06 ± 15.49 74 (21.3) 83 (23.9) 93 (26.8) 97 (28.0)
Cause of arrest Cardiac Submersion Drug Asphyxia Exanguiantion Other non-cardiac Hanging
177 (69.7) 59 (2.0) 2 (0.8) 27 (10.6) 0 (0.0) 25 (9.8) 18 (7.1)
81 (42.9) 9 (4.8) 10 (5.3) 32 (16.9) 2 (1.1) 33 (17.5) 22 (11.6)
83 (86.5) 1 (1.0) 2 (2.1) 2 (2.1) 0 (0.0) 6 (6.3) 2 (2.1)
175 (50.4) 13 (3.7) 10 (2.9) 57 (16.4) 2 (0.6) 52 (15.0) 38 (11.0)
Witness Yes No
183 (72.0) 71 (28.0)
102 (54.0) 87 (46.0)
79 (82.3) 17 (17.7)
206 (59.4) 141 (40.6)
Bystander CPR Yes No
72 (28.7) 179 (71.3)
60 (31.9) 128 (68.1)
35 (36.8) 60 (63.2)
97 (28.2) 247 (71.8)
ECG rhythm in EMS Shockable Unshockable Unknown
76 (29.9) 112 (44.1) 66 (26.0)
24 (12.7) 127 (67.2) 38 (20.1)
46 (47.9) 20 (20.8) 30 (31.3)
54 (15.6) 219 (63.1) 74 (21.3)
19.61 ± 10.71 83 (32.7) 72 (28.3) 54 (21.3) 45 (17.7)
Prehospital time (min) 28 ECG rhythm at ED arrival Vf/Pulseless VT PEA Asystole ROSC before arrival
24.48 ± 15.04 40 (21.2) 28 (14.8) 59 (31.2) 62 (32.8)
66 (26.0) 56 (22.0) 121 (47.6) 11 (4.3)
19 (10.1) 39 (20.6) 128 (67.7) 3 (1.6)
13.96 ± 12.67 90 (35.4) 46 (18.1) 58 (22.8) 60 (23.6) 9.55 ± 4.33
Hospital resuscitation time (min) 21 Lactate (mmol L−1 )
17.30 ± 8.93 36 (37.5) 33 (34.4) 19 (19.8) 8 (8.3)
41 (42.7) 20 (20.8) 25 (26.0) 10 (10.4)
15.70 ± 13.67 46 (24.5) 45 (23.9) 47 (25.0) 50 (26.6) 11.36 ± 4.58
22.90 ± 13.62 87 (25.1) 67 (19.3) 94 (27.1) 99 (28.5)
44 (12.7) 75 (21.6) 224 (64.6) 4 (1.2)
12.25 ± 11.34 37 (38.5) 17 (17.7) 25 (26.0) 17 (17.7) 8.99 ± 4.17
15.38 ± 13.50 99 (28.6) 74 (21.4) 80 (23.1) 93 (26.9) 10.70 ± 4.55
Values are expressed as number (%) or mean ± standard deviation as appropriate. Age, prehospital time, and hospital resuscitation time are divided by quartile. (CPR: cardiopulmonary resuscitation, ECG: electrocardiogram, EMS; emergency medical service, ED: emergency department, Vf: ventricular fibrillation, VT: ventricular tachycardia, PEA: pulseless electrical activity, ROSC: return of spontaneous circulation).
Table 2 Variables associated with lactate level. Multivariate linear regression lactate level
OR
95% CI
5.01 4.64 4.15 3.78
0.309 0.278 0.097
−2.34 −2.492 −3.555
−0.008 −0.071 −1.117
9.2 ± 3.98 9.9 ± 4.12 11.0 ± 4.75 11.3 ± 4.90
1.728 4.707 5.995
−0.636 0.369 0.61
1.73 2.729 2.972
Age (years) 18–40 41–50 51–64 >65
11.5 10.1 10.1 9.45
Prehospital time (min) 28 OR: odds ratio, CI: confidence interval.
p Value
± ± ± ±
0.048 0.038 65
1.140 2.454 3.213
0.593 1.256 1.589
2.190 4.795 6.499
0.695 0.009 0.001
1.496 2.120 2.620
0.081 0.419 0.619
0.725 1.084 1.307
0.015
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical paper
Correlation between initial serum levels of lactate after return of spontaneous circulation and survival and neurological outcomes in patients who undergo therapeutic hypothermia after cardiac arrest夽 Dong Hoon Lee a,1 , In Soo Cho b,∗ , Sun Hwa Lee c,2 , Yong Il Min d,3 , Jin Hong Min e,4 , Soo Hyun Kim f,5 , Young Hwan Lee g,6 , the Korean Hypothermia Network Investigators a
Department of Emergency Medicine, College of Medicine, Chung-Ang University, Seoul, Republic of Korea Department of Emergency Medicine, KEPCO Medical Center, Seoul, Republic of Korea c Department of Emergency Medicine, Ewha Womans University, Seoul, Republic of Korea d Department of Emergency Medicine, Chonnam National University, Gwangju, Republic of Korea e Department of Emergency Medicine, College of Medicine Chungbuk National University, Cheongju-si, Republic of Korea f Department of Emergency Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea g Departments of Emergency Medicine, Hallym University Sacred Heart Hospital, Anyang-si, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 25 July 2014 Received in revised form 6 November 2014 Accepted 9 November 2014 Keywords: Lactate Outcome Therapeutic hypothermia Cardiac arrest
a b s t r a c t Objectives: We analysed the relationship between serum levels of lactate within 1 h of return of spontaneous circulation (ROSC) and survival and neurological outcomes in patients who underwent therapeutic hypothermia (TH). Methods: This was a multi-centre retrospective and observational study that examined data from the first Korean Hypothermia Network (KORHN) registry from 2007 to 2012. The inclusion criteria were out-of-hospital cardiac arrest (OHCA) and examination of serum levels of lactate within 1 h after ROSC, taken from KORHN registry data. The primary endpoint was survival outcome at hospital discharge, and the secondary endpoint was poor neurological outcome (Cerebral Performance Category, CPC, 3–5) at hospital discharge. Initial lactate levels and other variables collected within 1 h of ROSC were analysed via multivariable logistic regression. Results: Data from 930 cardiac arrest patients who underwent TH were collected from the KORHN registry. In a total of 443 patients, serum levels of lactate were examined within 1 h of ROSC. In-hospital mortality was 289/443 (65.24%), and 347/443 (78.33%) of the patients had CPCs of 3–5 upon hospital discharge. The odds ratios of lactate levels for CPC and in-hospital mortality were 1.072 (95% confidence interval (CI) 1.026–1.121) and 1.087 (95% CI = 1.031–1.147), respectively, based on multivariate ordinal logistic regression analyses. Conclusion: High levels of lactate in serum measured within 1 h of ROSC are associated with hospital mortality and high CPC scores in cardiac arrest patients treated with TH. © 2014 Elsevier Ireland Ltd. All rights reserved.
夽 A Spanish translated version of the summary of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.11.005. ∗ Corresponding author. E-mail addresses: [email protected] (D.H. Lee), [email protected] (I.S. Cho), [email protected] (S.H. Lee), [email protected] (Y.I. Min), [email protected] (J.H. Min), [email protected] (S.H. Kim), [email protected] (Y.H. Lee). 1 Department of Emergency Medicine, Chung-Ang University Hospital, 224-1 Heoukseok-dong, Dongjak-gu, Seoul, Republic of Korea 2 Department of Emergency Medicine, Ewha Womans University Medical center 911-1 Mok-dong, Yang chun-Gu, Republic of Korea 3 Department of Emergency Medicine, Chonnam National University Hospital, 42 Jebong-ro, Dong-gu, Gwangju, Republic of Korea 4 Department of Emergency Medicine, College of Medicine Chungbuk National University Hospital, Gaesin-dong, Heungdeok-gu, Cheongju-si, Chungbuk, Republic of Korea 5 Emergency Medicine, Seoul St. Mary’s hospital, 222 Banpo-daero, Seocho-Gu, Seoul, Republic of Korea 6 Department of Emergency Medicine, Hallym University Sacred Heart Hospital, Hallym University Medical Center, 896 Pyeongchon-dong, Dongan-gu, Anyang-si, Gyeonggido, Republic of Korea http://dx.doi.org/10.1016/j.resuscitation.2014.11.005 0300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.
144
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
1. Introduction Therapeutic hypothermia (TH) has improved the survival and neurological outcomes of patients who survive cardiac arrest 1,2 . The therapy is commonly used for comatose survivors with cardiac arrest as part of a series of survival treatments for cardiopulmonary resuscitation 3 . During TH, physicians need to be aware of the neurological and survival outcomes of comatose patients. Thus, several studies have been performed to develop tools to predict the neurological prognosis and survival outcomes. However, no single effective method for the prediction of outcome has been found. Several biological markers, such as the S-100B protein and neuron-specific enolase, have been demonstrated to be useful, and electroencephalography and neuroimaging, including brain computed tomography and magnetic resonance imaging, have been used by physicians to predict neurological outcomes 4–12 . However, these methods require serial check-ups and longer times to determine results. An effective method for predicting survival and neurological outcomes that can be applied early on in the treatment process would be useful for planning treatment strategies for post-cardiac arrest care. However, there is a lack of objective data on such patients collected within 1 h after return of spontaneous circulation (ROSC) in the emergency department (ED). Such data would comprise pre-hospital data, including data from emergency medical service (EMS); initial laboratory findings taken after ROSC; and the results of physical and neurological examinations performed in the ED. Serum lactate levels are one of these initial laboratory findings. Early and effective clearance of serum lactate (i.e., lower levels) are associated with lower mortality in patients with many critical illnesses, including sepsis and severe trauma 13–18 . For cardiac arrest survivors, the relationships between clearance of serum lactate and neurological and survival outcomes have been studied 19–23 . However, few studies have addressed the potential relationship between initial serum levels of lactate and outcomes in patients who have been treated with TH. Thus, in this study, we evaluated whether initial levels of lactate in the serum at ROSC were correlated with neurological and survival outcomes of patients undergoing TH.
(http://starcrf.com) by the clinical research coordinator. A data manager and three clinical research associates regularly monitored and reviewed data quality. The site’s principal investigators or clinical research coordinators could be contacted through the query function in the system or directly by phone to clarify any questions about the data.
2.2. Data collection Data that could be collected within 1 h after ROSC were used. The inclusion criteria from the KORHN registry data were OHCA and examination of serum levels of lactate within 1 h after ROSC (Fig. 1). Specifically, the data were demographics (gender, age), cause of arrest, pre-hospital resuscitation variables (witness, bystander CPR, ECG rhythm in EMS, pre-hospital time), in-hospital resuscitation variables (ECG rhythm at ED arrival, amount of injected epinephrine, hospital resuscitation time), neurological examination after ROSC (corneal reflex, light reflex, myoclonus), serum levels of lactate within 1 h after ROSC, and neurological and survival outcomes at discharge.
2.3. Outcome measures The primary endpoint of the study was survival outcome at hospital discharge. A secondary endpoint was poor neurological outcome at hospital discharge, defined as a Cerebral Performance Category (CPC) of 3–5. CPC 3 indicates severe cerebral disability: conscious, dependent on others for daily support because of impaired brain function. It ranges from an ambulatory state to severe dementia or paralysis. CPC 4 indicates a coma or a vegetative state: any degree of coma without the presence of brain death criteria. Unawareness, even if appears awake (vegetative states) without interaction with environment; may have spontaneous eye opening and sleep/awake cycles. CPC 5 indicates cerebral unresponsiveness, indicates brain death: apnoea, areflexia, EEG silence.
2.4. Statistical analysis 2. Methods 2.1. Registry data collection This was a multi-centre, retrospective, observational, registrybased study that used data from the first Korean Hypothermia Network (KORHN) registry. KORHN, a multi-centre clinical research consortium for TH in South Korea, was organised in 2011, and a multi-centre retrospective, registry project was performed in 2012. The KORHN investigators collected data regarding post-cardiac arrest TH in 24 teaching hospitals around South Korea from 2007 to 2012. The institutional review board of each institution approved the study protocol before data collection. The need for informed consent was waived because of the retrospective nature of the study. Adults (≥18 years) who suffered out-of-hospital cardiac arrest (OHCA), treated with TH after ROSC achieved pre-hospital and in the ED, were included in the study. Traumatic cardiac arrest patients were excluded. In a case report form, standard definitions of 87 variables and an investigator manual were developed from a literature review and consensus of the study investigators. The registry data were collected from manual medical charts or electronic medical record reviews. Data collected in each hospital were verified for completeness by the site’s principal investigator, and were recorded on a web-based data registration system
Baseline descriptive characteristics of the subjects are presented with basic statistical analyses including means and standard deviations (SDs) for continuous variables, and numbers and percentages for categorical variables. Patient data were analysed according to survival and neurological outcomes at hospital discharge. Univariate and multivariate linear regression were used to evaluate the relationship between lactate levels and variables before hospital arrival. To compare lactate levels and neurological outcomes (CPC scores) as continuous variables, ordinal univariate and multivariate logistic regressions were used. All of the parameters were included in the multivariable logistic analysis to determine independent predictors for neurologic outcomes. Binary univariate and multivariate logistic regression analyses were used to estimate correlations between lactate levels and poor neurological outcome. CPC scores of 3, 4, and 5 were considered to be poor neurological outcomes. For correlation between lactate levels and mortality, binary univariate and multivariate logistic regression analyses were used. Odds ratios (ORs) and 95% confidence intervals (CIs) are reported as estimates of variability and effect size. p-Values ≤0.05 were considered statistically significant. p-Values were corrected for multiple comparisons and within-method comparisons using the Holm–Bonferroni method. All of the statistical analyses were performed using the SPSS (ver. 20.0 for Windows; Chicago, IL) and SAS (SAS Institute, Cary, NC) software.
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
145
Fig. 1. Flow chart of participants. CPC: cerebral performance category.
3. Results Data on 930 cardiac arrest patients with TH were collected from the KORHN registry; 443 patients who met the inclusion criteria were enrolled. Of these patients, 14 arrived at the ED with ROSC, and 429 patients were resuscitated at the ED. The general characteristics of these patients are listed in Table 1 according to hospital mortality and neurological outcomes at hospital discharge. The mean age of the patients was 57.70 ± 16.06 years. There were 295 (66.6%) males and 148 (33.4%) females. The cardiac rhythm, confirmed by EMS, was shockable in 100 patients (22.6%), unshockable in 239 (54.0%), and unknown in 104 (23.5%). Initial electrocardiography (ECG) on ED arrival with a cardiac arrest state was ventricular fibrillation (Vf) or pulse-less ventricular tachycardia (VT) in 85 patients (19.2%), pulse-less electrical activity (PEA) in 95 (21.4%), and asystole in 249 (56.2%). In total, 14 patients (3.2%) arrived at the ED in a ROSC state. 254 patients (57.3%) survived and were discharged. The neurological outcomes at hospital discharge were 81 (18.3%) in CPC 1, 15 (3.4%) in CPC 2, 37 (8.4%) in CPC 3, 121 (27.3%) in CPC 4, and 189 in CPC 5 (41.7%). 3.1. Initial lactate levels after ROSC The level of lactate within 1 h after ROSC was 10.32 ± 4.52 mmol L−1 in all of the enrolled patients. In comparing survival and neurological outcomes, there were statistically significant differences between surviving and expired patients, and between those with good and poor neurological outcomes (Fig. 2). Lactate levels were analysed with other variables in univariate simple linear regression and multivariate linear regression. In the simple linear regression analysis, the OR of age was 0.958 (95% CI = −0.069 to 0.017), and the OR of pre-hospital time was 1.062 (95% CI = 0.028–0.092). In multiple regression analysis, age greater than 40 years and a pre-hospital time of more than 20 min were statistically significantly associated with lactate levels (Table 2).
1.656–3.470), pre-hospital time (1.034: 1.018–1.050), unshockable ECG rhythm in EMS (4.971: 3.187–7.752), no defibrillation by the EMS (2.888: 1.876–4.444), PEA and asystole at ED arrival (3.309: 1.929–5.677 and 5.612: 3.511–8.972, respectively), no defibrillation in the ED (2.848: 1.981–4.097), and no light reflex and no corneal reflex after ROSC (4.265: 2.910–6.251 and 4.005: 2.608–6.151, respectively). In multivariate regression analysis, CPC score was associated with lactate levels, age, unwitnessed cardiac arrest, unshockable rhythm in the EMS, no defibrillation in the ED, and no light reflex after ROSC (Table 3). 3.3. Prediction of hospital mortality The difference in lactate levels between survivors and non-survivors at hospital discharge was statistically significant (9.55 ± 4.33 mmol L−1 in survivors, 11.36 ± 4.58 mmol L−1 in non-survivors; Fig. 1). In univariate logistic regression analysis for hospital mortality, the OR of lactate was 1.096 (95% CI = 1.049–1.146). Other statistically significant variables were male gender (OR = 1.631: 95% CI = 1.095–2.430), age (1.026: 1.013–1.039), unwitnessed cardiac arrest (2.198: 1.479–3.267), pre-hospital time (1.032: 1.015–1.049), unshockable ECG rhythm in EMS (3.591: 2.125–6.067), no defibrillation by the EMS (1.985: 1.204–3.274), PEA and asystole on ED arrival (0.242: 1.258–4.652 and 3.675: 2.083–6.482, respectively), no defibrillation in the ED (2.271: 1.503–3.432), no light reflex and no corneal reflex after ROSC (2.545: 1.649–3.929 and 2.974: 1.770–4.999, respectively), and no seizure and no myoclonus (2.168: 1.314–3.577 and 2.593: 1.499–4.486, respectively). In multivariate logistic regression analysis, initial lactate levels were statistically significantly associated with non-survival at hospital discharge (1.087: 1.031–1.147). Moreover, age over 50, unwitnessed cardiac arrest, no bystander CPR, pre-hospital time >27 min, and no defibrillation in the ED were associated with mortality after hospital admission (Table 3). 4. Discussion
3.2. Prediction of neurological outcome at hospital discharge The OR of lactate levels for CPC was 1.072 (95% CI = 1.026–1.121) in multivariate ordinal logistic regression analysis. In univariate regression analysis, lactate was associated with CPC score (OR = 1.090, 95% CI = 1.048–1.135), and there were statistically significant correlations with male gender (1.590: 1.101–2.297), age (1.033: 1.022–1.045), unwitnessed cardiac arrest (2.399:
In resuscitated patients from OHCA undergoing TH, initial serum levels of lactate within 1 h after ROSC were associated with survival and neurological outcomes at hospital discharge (Table 3). Many investigators have sought to develop effective tools that can predict survival and neurological outcomes in post-cardiac arrest patients. Thus, many biomarkers, brain imaging methods, and neurophysiological examinations have been studied as
146
D.H. Lee et al. / Resuscitation 88 (2015) 143–149
Table 1 Baseline characteristics of study subjects (n = 443). Survival outcome at discharge
Sex Male Female
Neurologic outcome
Survival (n = 254)
No survival (n = 289)
Good (n = 96)
Poor (n = 347)
181 (71.3) 73 (28.7)
114 (60.3) 75 (39.7)
69 (71.9) 27 (28.1)
226 (65.1) 121 (34.9)
55.02 ± 15.84 77 (30.3) 78 (30.7) 54 (21.3) 45 (17.7)
Age (year) 18–40 41–50 51–64 >65
61.31 ± 15.68 38 (20.1) 37 (19.6) 53 (28.0) 61 (32.3)
49.17 ± 15.26 41 (42.7) 32 (33.3) 14 (14.6) 9 (9.4)
60.06 ± 15.49 74 (21.3) 83 (23.9) 93 (26.8) 97 (28.0)
Cause of arrest Cardiac Submersion Drug Asphyxia Exanguiantion Other non-cardiac Hanging
177 (69.7) 59 (2.0) 2 (0.8) 27 (10.6) 0 (0.0) 25 (9.8) 18 (7.1)
81 (42.9) 9 (4.8) 10 (5.3) 32 (16.9) 2 (1.1) 33 (17.5) 22 (11.6)
83 (86.5) 1 (1.0) 2 (2.1) 2 (2.1) 0 (0.0) 6 (6.3) 2 (2.1)
175 (50.4) 13 (3.7) 10 (2.9) 57 (16.4) 2 (0.6) 52 (15.0) 38 (11.0)
Witness Yes No
183 (72.0) 71 (28.0)
102 (54.0) 87 (46.0)
79 (82.3) 17 (17.7)
206 (59.4) 141 (40.6)
Bystander CPR Yes No
72 (28.7) 179 (71.3)
60 (31.9) 128 (68.1)
35 (36.8) 60 (63.2)
97 (28.2) 247 (71.8)
ECG rhythm in EMS Shockable Unshockable Unknown
76 (29.9) 112 (44.1) 66 (26.0)
24 (12.7) 127 (67.2) 38 (20.1)
46 (47.9) 20 (20.8) 30 (31.3)
54 (15.6) 219 (63.1) 74 (21.3)
19.61 ± 10.71 83 (32.7) 72 (28.3) 54 (21.3) 45 (17.7)
Prehospital time (min) 28 ECG rhythm at ED arrival Vf/Pulseless VT PEA Asystole ROSC before arrival
24.48 ± 15.04 40 (21.2) 28 (14.8) 59 (31.2) 62 (32.8)
66 (26.0) 56 (22.0) 121 (47.6) 11 (4.3)
19 (10.1) 39 (20.6) 128 (67.7) 3 (1.6)
13.96 ± 12.67 90 (35.4) 46 (18.1) 58 (22.8) 60 (23.6) 9.55 ± 4.33
Hospital resuscitation time (min) 21 Lactate (mmol L−1 )
17.30 ± 8.93 36 (37.5) 33 (34.4) 19 (19.8) 8 (8.3)
41 (42.7) 20 (20.8) 25 (26.0) 10 (10.4)
15.70 ± 13.67 46 (24.5) 45 (23.9) 47 (25.0) 50 (26.6) 11.36 ± 4.58
22.90 ± 13.62 87 (25.1) 67 (19.3) 94 (27.1) 99 (28.5)
44 (12.7) 75 (21.6) 224 (64.6) 4 (1.2)
12.25 ± 11.34 37 (38.5) 17 (17.7) 25 (26.0) 17 (17.7) 8.99 ± 4.17
15.38 ± 13.50 99 (28.6) 74 (21.4) 80 (23.1) 93 (26.9) 10.70 ± 4.55
Values are expressed as number (%) or mean ± standard deviation as appropriate. Age, prehospital time, and hospital resuscitation time are divided by quartile. (CPR: cardiopulmonary resuscitation, ECG: electrocardiogram, EMS; emergency medical service, ED: emergency department, Vf: ventricular fibrillation, VT: ventricular tachycardia, PEA: pulseless electrical activity, ROSC: return of spontaneous circulation).
Table 2 Variables associated with lactate level. Multivariate linear regression lactate level
OR
95% CI
5.01 4.64 4.15 3.78
0.309 0.278 0.097
−2.34 −2.492 −3.555
−0.008 −0.071 −1.117
9.2 ± 3.98 9.9 ± 4.12 11.0 ± 4.75 11.3 ± 4.90
1.728 4.707 5.995
−0.636 0.369 0.61
1.73 2.729 2.972
Age (years) 18–40 41–50 51–64 >65
11.5 10.1 10.1 9.45
Prehospital time (min) 28 OR: odds ratio, CI: confidence interval.
p Value
± ± ± ±
0.048 0.038 65
1.140 2.454 3.213
0.593 1.256 1.589
2.190 4.795 6.499
0.695 0.009 0.001
1.496 2.120 2.620
0.081 0.419 0.619
0.725 1.084 1.307
0.015
