https://orcid.org/0000-0003-3571-9448
Figures
Abstract
We investigated the association of insulin administration method with the achievement of mean glucose ≤ 180 mg/dL and neurological outcomes in out-of-hospital cardiac arrest (OHCA) survivors who had hyperglycemia after the return of spontaneous circulation. From a multicenter prospective registry, we extracted the data of adult OHCA survivors who underwent targeted temperature management (TTM) between 2015 and 2018. Blood glucose levels every 4 h after initiating TTM were obtained for 72 h. We divided insulin administration methods into three categories: subcutaneous (SQI), intravenous bolus (IBI), and continuous intravenous (CII). We calculated the mean glucose and standard deviation (SD) of glucose. The primary outcome was the achievement of mean glucose ≤ 180 mg/dL. The secondary outcomes were the 6-month neurological outcome based on the Cerebral Performance Category (CPC) scale (good, CPC 1–2; poor, CPC 3–5), mean glucose, and SD of glucose. Of the 549 patients, 296 (53.9%) achieved mean glucose ≤ 180 mg/dL, and 438 (79.8%) had poor neurological outcomes, 134 (24.4%), 132 (24.0), and 283 (51.5%) were in the SQI, IBI, and CII groups, respectively. The SQI (adjusted odds ratio [aOR], 0.848; 95% confidence intervals [CIs], 0.493–1.461) and IBI (aOR, 0.673; 95% CIs, 0.415–1.091) groups were not associated with mean glucose ≤ 180 mg/dL and the SQI (aOR, 0.660; 95% CIs, 0.335–1.301) and IBI (aOR, 1.757; 95% CIs, 0.867–3.560) groups were not associated with poor neurological outcomes compared to the CII group. The CII (168 mg/dL [147–202]) group had the lowest mean glucose than the SQI (181 mg/dL [156–218]) and IBI (184 mg/dL [162–216]) groups. The CII (45.0[33.9–63.5]) group had a lower SD of glucose than the IBI (50.8 [39.1–72.0]) group. The insulin administration method was not associated with achieving mean glucose ≤ 180 mg/dL and 6-month neurological outcomes.
Citation: Lee DH, Lee BK, Cho YS, Jung YH, Lee HY, Jeung KW, et al. (2022) Association between insulin administration method and six-month neurological outcome in survivors of out-of-hospital cardiac arrest who underwent targeted temperature management. PLoS ONE 17(12): e0279776. https://doi.org/10.1371/journal.pone.0279776
Editor: Simone Savastano, Fondazione IRCCS Policlinico San Matteo, ITALY
Received: June 23, 2022; Accepted: December 8, 2022; Published: December 30, 2022
Copyright: © 2022 Lee et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This study was supported by a grant from National Research Foundation of Korea (NRF-2021R1A2C1005800). BK Lee received this fund. The funders had no role in the study design, data collection and analysis, publication decision, or manuscript preparation. There was no additional external funding received for this study.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Cardiac arrest leads to various metabolic derangements due to ischemic-reperfusion injury even after the return of spontaneous circulation (ROSC). Hyperglycemia is one of the common abnormalities following cardiac arrest [1]. A well-known scientific observation is the association between hyperglycemia and poor neurological outcomes or increased mortality in cardiac arrest survivors [2–5]. Therefore, the target glucose level is generally recommended as < 180 mg/dL, this parameter is applied to critically ill patients [6], as in comatose cardiac arrest survivors, as protection against potential neurological injury, even if the ultimate target glycemic range has not been elucidated [7, 8].
In a historic comparison study, although the target glucose range of the continuous intravenous insulin (CII) method changed from conventional range (150 to 200 mg/dL) to intensive range (100 to 150 mg/dL) in patients with coronary artery bypass graft, the CII method has been proven to reduce mortality compared to the subcutaneous insulin (SQI) method [9]. Therefore, intravenous insulin infusion is the preferred route for administering insulin in critically ill patients [10]. A randomized controlled trial has revealed that conventional glucose control (< 180 mg/dL) was associated with lower mortality than intensive glucose control (81 to 108 mg/dL) in critically ill patients [6]. However, a randomized controlled trial that compared strict and conventional glucose control in comatose cardiac arrest survivors failed to find the optimal target glucose range [11]. Furthermore, although insulin is administrated subcutaneously or intravenously (bolus or continuous), the clinical difference according to the insulin administration in cardiac arrest survivors has not been investigated.
To address this question, we hypothesized that the insulin administration method would be related to blood glucose levels and, thus, to neurological outcomes. To examine this hypothesis, we used a multicenter registry of out-of-hospital cardiac arrest (OHCA) who underwent targeted temperature management (TTM) with blood glucose recordings for 72 h after the initiation of TTM.
Materials and methods
Study design and population
The Korean Hypothermia Network Prospective Registry (KORHN-PRO) has been gathering data on comatose adult (age ≥ 18 years) OHCA survivors who underwent TTM at 20 participating hospitals since October 2015 (KORHN-PRO; NCT02827422) [12]. The KORHN-PRO collects data on blood glucose after ROSC and every 4 h from the initiation of TTM to 72 h. We performed a retrospective analysis of the KORHN-PRO data between October 2015 and December 2018. The ethics and institutional review board of all participating hospitals approved KORHN-PRO. Written informed consent was obtained from all patients or patients’ proxies per national requirements and the principle of the Declaration of Helsinki [13]. Independent researchers assessed the neurological outcomes at 1 and 6 months after ROSC and recorded these as the Cerebral Performance Category (CPC) scale [14].
We included adult OHCA survivors who had hyperglycemia (> 180 mg/dL) within 24 h following ROSC. We excluded patients who had: insufficient data on glucose (glucose measurement less than six times during the 72 h after the initiation of TTM); no insulin for glucose control; no data on insulin treatment; no hyperglycemia within 24 h following ROSC; no data on 6-month neurological outcomes; died within 24 h after ROSC.
TTM and glucose control
After ROSC, the target temperature range of 33–36°C was achieved as soon as possible. Sedatives were administered, and if needed, neuromuscular blockade to control shivering. Patients were rewarmed after completing the maintenance phase at 0.2–0.5°C/h. Sedatives and neuromuscular blockade were discontinued as the patient achieved normothermia. Blood glucose level was monitored and controlled to avoid hyperglycemia or hypoglycemia throughout post-cardiac arrest care following ROSC. Hyperglycemia was managed with SQI, intravenous bolus insulin (IBI), or CII according to each hospital protocol and the attending physician (S1 File). Moderate (< 70 mg/dL) or severe hypoglycemia (< 40 mg/dL) was managed with glucose-containing solution.
Data collection
We extracted the following data from the KORHN-PRO: age; sex; body mass index (BMI); pre-existing illness; a witness of collapse; bystander cardiopulmonary resuscitation (CPR); first monitored rhythm (shockable or non-shockable); etiology of cardiac arrest (cardiac or non-cardiac); time from collapse to ROSC; epinephrine dose; serum lactate level after ROSC, PaO2 and PaCO2 after ROSC; glucose levels after ROSC and every 4 h during the 72 h from the initiation of TTM; sequential organ failure assessment (SOFA) within the first day following ROSC [15]; target temperature (33–34°C or 35–36°C); insulin administration method (SQI, IBI, or CII); CPC 6-months after ROSC.
The primary outcome was the achievement of mean blood glucose ≤ 180 mg/dL. The secondary outcomes were neurological outcomes assessed using CPC 6 months after ROSC, maximum glucose, mean glucose, a standard deviation (SD) of glucose, minimum glucose, moderate hypoglycemia, and severe hypoglycemia. The neurological outcomes were defined as good (CPC 1 or 2) or poor (CPC 3–5).
Statistical analysis
We report continuous variables as median with interquartile ranges because all continuous variables had a non-normal distribution and categorical variables as the frequency with percentile. We used chi-square or Fisher’s exact test to compare categorical variables, as appropriate. We used the Mann–Whitney U test to compare continuous variables between two groups and the Kruskal–Wallis test to compare continuous variables among three groups. We performed posthoc analysis using a pair-wise Mann–Whitney U test with Bonferroni correction. We conducted logistic regression analyses to investigate the association between the insulin administration method and the achievement of mean blood glucose ≤ 180 mg/dL and the association between insulin administration methods and neurological outcomes. We selected the variables with a p-value < 0.05 in comparisons among insulin administration methods as covariates for the association between insulin administration methods and mean blood glucose ≤ 180 mg/dL. Additionally, we selected the covariates after performing the multivariable logistic regression analysis with the variables with a p-value < 0.2 in comparisons between groups of mean glucose ≤ 180 mg/dL and > 180 mg/dL. We also selected the covariates for the association between insulin administration methods and neurological outcomes through the multivariable logistic regression analysis with variables with a p-value < 0.2 in comparisons between neurological outcome groups. We selected the variables with a p-value < 0.05 in the multivariate logistic regression analyses as final covariates. We performed the Hosmer–Lemeshow test to test the goodness-of-fit of the logistic model. We also performed the multivariate logistic regression analysis to examine the association between glucose variables and neurological outcomes after adjusting covariates. We report the logistic regression analysis results as an adjusted odds ratio (aOR) with 95% confidence interval (CIs). We used IBM SPSS Statistics 26.0 for Windows (IBM Corp., Armonk, NY). A two-sided p-value < 0.05 was used to indicate statistical significance.
Results
Study population
Of 1,373 OHCA survivors who were recorded in the registry, 289 patients had no hyperglycemia during 24 h after ROSC; 251 patients were either not administered insulin or lacked insulin data; 152 patients had insufficient blood glucose measurement (< 6 times) data during the 72 h after ROSC; 97 patients died or transferred within 24 h; 35 patients had no available data on 6-month CPC. Finally, 549 patients were included in the study (Fig 1).
OHCA, out-of-hospital cardiac arrest; TTM, targeted temperature management; ROSC, return of spontaneous circulation.
Characteristics according to the insulin administration method
Table 1 shows the baseline and clinical characteristics stratified by the insulin administration method. The SQI, IBI, and CII groups comprised 134 (24.4%), 132 (24.0%), and 283 (51.5%) patients, respectively. Pre-existing diabetes mellitus and renal disease, a witness of collapse, and etiology of cardiac arrest were different among the three groups. Time from collapse to ROSC and PaCO2 were different among the three groups. However, subgroup analyses found no difference in time from collapse to ROSC and PaCO2 between paired two groups. Pre-TTM shock and target temperature were different among the three groups.
Table 2 shows the glucose variables according to the insulin administration method. Achievement of mean glucose ≤ 180 mg/dL, mean glucose, SD of glucose, and minimum glucose differed among the three groups. The CII group (168 mg/dL [147–202]) had lower mean glucose levels than the SQI (181 mg/dL [156–218]) and IBI (184 mg/dL [162–216]) groups. The IBI group (50.8 [39.1–72.0]) had higher SDs of glucose than the SQI (47.0 [33.2–61.1]) and CII (45.0 [33.9–63.5]) groups. The CII group (98 mg/dL [82–112]) had lower minimum glucose levels than the SQI group (106 mg/dL [91–122]). However, the IBI group had similar minimum glucose levels as the SQI and CII groups. Severe hypoglycemia was different among the three groups.
Association between insulin administration method and mean glucose under 180 mg/dL
Table 3 shows the demographic and clinical characteristics stratified by mean glucose of 180 mg/dL. A total of 296 (53.9%) achieved mean glucose ≤ 180 mg/dL. Patients who had mean glucose ≤ 180 mg/dL had lower BMI (23.0 kg/m2 [20.7–25.4]) vs. 23.9 kg/m2 [21.3–26.1]) than those had mean glucose > 180 mg/dL. Patients with mean glucose ≤ 180 mg/dL had fewer pre-existing illnesses (hypertension and diabetes mellitus). Patient who had mean glucose ≤ 180 mg/dL had fewer witnesses of collapse (208/296 vs. 154/253), lower epinephrine dose (2 mg [0–4] vs. 2 mg [1–5]), lower serum lactate (9.3 mg/dL [6.0–12.6] vs. 11.3 mg/dL [8.0–14.0]), higher PaO2 (117.5 mmHg [76.0–221.0] vs. 100.0 mmHg [71.0–160.0]), lower pre-TTM shock (155/296 vs. 176/253), lower SOFA scores (11 [9–13] vs. 12 [10–14]), and a higher proportion of target temperature of 33°C–34°C (255/296 vs. 192/253).
Multivariable logistic regression analysis found hypertension (aOR, 0.609; 95% CIs, 0.399–0.930), diabetes mellitus (aOR, 0.273; 95% CIs, 0.175–0.425), witnessed (aOR, 1.992; 95% CIs, 1.289–3.078), PaO2 (aOR, 1.002; 95% CIs, 1.000–1.004), pre-TTM shock (aOR, 0.541; 95% CIs, 0.355–0.826), target temperature of 35–36°C (aOR, 0.500; 95% CIs, 0.297–0.842), glucose after ROSC (aOR, 0.995; 95% CIs, 0.993–0.997) as covariates. The insulin administration method was not associated with the achievement of mean glucose ≤ 180 mg/dL after adjusting covariates (Table 4).
Association between insulin administration method and neurological outcomes
Table 5 shows the demographic and clinical characteristics according to the neurological outcomes. A total of 438 (79.8%) patients had poor neurological outcomes. Patients in the poor neurological outcome group were older (62.0 years [51.8–72.0] vs. 58.0 years [48.0–67.0]) than those in the good neurological outcome group, although BMI and pre-existing illness were not different between neurological outcome groups. Those in the poor neurological outcome group had fewer witnesses of collapse (268/438 vs. 94/111), a lower proportion of bystander CPR (256/438 vs. 82/111), a lower proportion of shockable rhythm (83/438 vs. 80/111), fewer cardiac etiology (242/438 vs. 103/111), longer time from collapse to ROSC (36.0 min [23.0–47.0] vs. 18.0 min [12.0–27.0]), and received higher epinephrine dose (2 mg [1–4] vs. 0 mg [0–2]) than those in the good neurological outcome group. Those in the poor neurological outcome group had higher serum lactate levels after ROSC (10.9 mg/dL [7.5–13.7] vs. 7.5 mg/dL [4.1–11.2]), higher PaCO2 (53.7 mmHg [39.0–75.1] vs. 39.0 mmHg [32.2–52.4]), more pre-TTM shock (282/438 vs. 49/111), and higher SOFA scores (12 [10–13] vs. 10 [8–12]).
Multivariate logistic regression analysis found that age (aOR, 1.031; 95% CIs, 1.008–1.054), shockable rhythm (aOR, 0.183; 95% CIs, 0.096–0.349), cardiac etiology (aOR, 0.196; 95% CIs, 0.076–0.506), time from collapse to ROSC (aOR, 1.074; 95% CIs, 1.048–1.102), PaO2 (aOR, 1.003; 95% CIs, 1.000–1.006), PaCO2 (aOR, 1.015; 95% CIs, 1.001–1.029), and pre-TTM shock (aOR, 1.855; 95% CIs, 1.037–3.319) were associated with poor neurological outcome. The insulin administration method was not associated with the 6-month neurological outcomes after adjusting covariates (Table 6).
Association between glucose variables and neurological outcomes
Table 7 shows the glucose variables stratified by neurological outcome groups. The poor neurological outcome group had higher maximum glucose (343 mg/dL [284–410] vs. 308 mg/dL [250–366]), mean glucose (179 mg/dL [154–216] vs. 164 mg/dL [145–189]), SD (48.2 [35.9–67.0] vs. 41.1 [29.9–52.9]), and more frequent moderate hypoglycemia (52/438 vs. 6/111) than the good neurological outcome group (Table 7).
Table 8 shows the association between glucose variables during 72 h and neurological outcomes. Maximum glucose (aOR, 1.004; 95% CIs, 1.001–1.007), mean glucose (aOR, 1.009; 95% CIs, 1.002–1.016), andSD (aOR, 1.019; 95% CIs, 1.005–1.032) of glucose were independently associated with poor neurological outcomes.
Adjusted for age, first monitored rhythm, etiology of cardiac arrest, time from collapse to ROSC, PaO2 after ROSC, PaCO2 after ROSC, and pre-targeted temperature management shock
Discussion
This retrospective analysis found that the insulin administration method had no association with the achievement of mean glucose ≤ 180 mg/dL and the 6-month neurological outcome. Nevertheless, the CII group had the lowest mean glucose levels, and the IBI group had the highest SD of glucose levels. The CII and IBI groups had lower minimum glucose levels than the SQI group. The IBI group had a higher incidence of severe hypoglycemia. Higher maximum glucose, mean glucose, and SD of glucose were independently associated with poor neurological outcomes.
CII is the preferred route and delivery method of insulin in critically ill patients with hyperglycemic crises, such as perioperative care of cardiac surgery, cardiogenic shock, myocardial infarction, and acute ischemic stroke [16]. Such situations may require a rapid change in the insulin level or be associated with poor perfusion of subcutaneous tissue; for these reasons, guidelines recommend CII in critically ill patients instead of SQI [17]. A study comparing CII to SQI in patients with diabetes undergoing coronary artery bypass grafting demonstrated that CII was associated with lower postoperative blood glucose and reduced mortality [9]. Nevertheless, SQI is still used to control blood glucose in critically ill patients [18]. Likewise, IBI has a comparable effect on blood glucose to CII without any adverse effect [19]. Although CII was used the most, SQI and IBI were also used in about half of the patients in the present study. We found that the SQI and IBI groups had higher mean blood glucose levels during the 72 h after TTM initiation than the CII group. Higher mean blood glucose level was associated with the 6-month poor neurological outcomes in the present study. However, SQI or IBI had no association with the achievement of mean glucose ≤ 180 mg/dL and the 6-month neurological outcomes compared to CII. It might be postulated that SQI or IBI is not inferior to CII regarding controlling hyperglycemia, considering the independent association between high mean glucose and poor neurological outcome.
Regardless of the relationship of the insulin administration method with clinical outcomes, regarding glycemic control, CII seems to be the best measure among the three insulin administration methods. Hyperglycemia early after ROSC, at admission, or during 36 h after admission in survivors of cardiac arrest who underwent therapeutic hypothermia or TTM was associated with poor neurological outcomes [2, 4, 5]. The blood glucose level after ROSC was not associated with the neurological outcomes in the present study. We postulate that because we excluded the patients without hyperglycemia within 24 h after ROSC, it might affect the insignificant relationship between glucose after ROSC and neurological outcomes; this is because the patients with no hyperglycemia after ROSC would be injured less than those with hyperglycemia due to ischemic insult. However, consistent with the previous study reports, high maximum and mean glucose levels during the 72 h following initiation of TTM were associated with poor neurological outcomes in the present study. Our results strengthen the importance of glucose control for a more extended period following ROSC.
Blood glucose level and glucose variability have been assessed in comatose cardiac arrest survivors, and high glucose variability is associated with poor neurological outcomes [2, 3, 5]. We calculated the SD of glucose for 72 h after the initiation of TTM to determine glucose variability. We also found that high SD of glucose was associated with poor neurological outcomes in the present study. Although there was no association between the insulin administration method and neurological outcomes, lower mean glucose levels and lower SD of glucose associated with favorable neurological outcomes emphasize that the CII may be the best measure to control glycemic status in cardiac arrest survivors.
A landmark trial regarding glucose control in critically ill patients revealed that intensive glucose control (81 to 110 mg/dL) increased mortality in the intensive care unit, and the high mortality in the intensive glucose control group was associated with iatrogenic hypoglycemia [6, 20]. Therefore, avoiding hypoglycemia is as crucial as avoiding hyperglycemia when controlling blood glucose. Although we found no association between neurological outcomes and moderate or severe hypoglycemia, the IBI and CII groups had frequent severe hypoglycemia compared to the SQI group in the present study. Frequent blood glucose monitoring is required more in the IBI and CII groups than in the SQI group to avoid hypoglycemia. Conventional glucose control (target ≤ 180 mg/dL) resulted in total 16.3% of hypoglycemia (moderate hypoglycemia, 15.8%; severe hypoglycemia, 0.5%) in a randomized control trial [6]. We reported 13.7% of hypoglycemia (moderate hypoglycemia, 10.6%; severe hypoglycemia, 3.1%). In the randomized control group, patients in the conventional glucose control group had a history of diabetes mellitus in about 20%, whereas we reported that 35.5% of the patients had a history of diabetes mellitus [6]. Diabetes mellitus does not seem to play a role in hypoglycemia during insulin treatment.
This study has several limitations. Although we used the data from the prospective multicenter registry, we can only demonstrate an association because this is a retrospective analysis. Of the total registry, about 60% of the patients were excluded, which might have caused selection bias. The registry is configured to record blood glucose every 4 h during TTM. Therefore, glucose characteristics might be different from the actual values. Although diabetes mellitus was not associated with neurological outcomes, pre-existing diabetes mellitus differed among the insulin treatment groups. Glycated hemoglobin has been reported to be associated with neurological outcomes in comatose cardiac arrest survivors [21], which means that glucose control status before cardiac arrest rather than the history of diabetes mellitus might contribute to neurological outcomes. However, we could not analyze the data according to glycated hemoglobin level due to the limitation of data. Future studies need to address the interaction of the insulin administration method with glycemic control status. The target temperature differed among the insulin treatment groups because the target temperature and the insulin administration method depended on the post-cardiac arrest care protocol at each hospital rather than the attending physician. However, the target temperature was not associated with neurological outcomes in the present study. The targeted temperature management trial also showed no difference in blood glucose and glucose variability between the target temperature groups [2]. Even between CII and IBI, the detailed administration method is different. The total insulin dosage may vary depending on the insulin administration method. Due to data limitations, it was impossible to compare the total amount of insulin administered in the present study.
Conclusions
The insulin administration method was not associated with the achievement of mean glucose ≤ 180 mg/dL and the 6-month neurological outcomes in OHCA survivors who underwent TTM. The CII method had lower mean glucose and lower SD of glucose rather than the IBI and SQI methods.
Supporting information
S1 File. Insulin administration protocols.
The participating hospitals in Korean Hypothermia Network have their insulin administration protocol. The protocols are a bit different from each other.
https://doi.org/10.1371/journal.pone.0279776.s001
(PDF)
Acknowledgments
The following investigators participated in the Korean Hypothermia Network. Chair: Kyung Woon Jeung (Chonnam National University Hospital, E-mail: neoneti@hanmail.net). Principal investigators of each hospital: Kyu Nam Park (The Catholic University of Korea, Seoul St. Mary’s Hospital); Minjung Kathy Chae (Ajou University Medical Center); Won Young Kim (Asan Medical Center); Byung Kook Lee (Chonnam National University Hospital); Dong Hoon Lee (Chung-Ang University Hospital); Tae Chang Jang (Daegu Catholic University Medical Center); Jae Hoon Lee (Dong-A University Hospital); Yoon Hee Choi (Ewha Womans University Mokdong Hospital); Je Sung You (Gangnam Severance Hospital); Young Hwan Lee (Hallym University Sacred Heart Hospital); In Soo Cho (Hanil General Hospital); Su Jin Kim (Korea University Anam Hospital); Jong-Seok Lee (Kyung Hee University Medical Center); Yong Hwan Kim (Samsung Changwon Hospital); Min Seob Sim (Samsung Medical Center); Jonghwan Shin (Seoul Metropolitan Government Seoul National University Boramae Medical Center); Yoo Seok Park (Severance Hospital); Hyung Jun Moon (Soonchunhyang University Hospital Cheonan); Won Jung Jeong (The Catholic University of Korea, St. Vincent’s Hospital); Joo Suk Oh (The Catholic University of Korea, Uijeongbu St. Mary’s Hospital); Seung Pill Choi (The Catholic University of Korea, Yeouido St. Mary’s Hospital); Kyoung-Chul Cha (Wonju Severance Christian Hospital).
References
- 1. Skrifvars MB, Pettila V, Rosenberg PH, Castren M. A multiple logistic regression analysis of in-hospital factors related to survival at six months in patients resuscitated from out-of-hospital ventricular fibrillation. Resuscitation. 2003; 59: 319–328. pmid:14659601
- 2. Borgquist O, Wise MP, Nielsen N, Al-Subaie N, Cranshaw J, Cronberg T, et al. Dysglycemia, Glycemic Variability, and Outcome After Cardiac Arrest and Temperature Management at 33 degrees C and 36 degrees C. Crit Care Med. 2017; 45: 1337–1343.
- 3. Cueni-Villoz N, Devigili A, Delodder F, Cianferoni S, Feihl F, Rossetti AO, et al. Increased blood glucose variability during therapeutic hypothermia and outcome after cardiac arrest. Crit Care Med. 2011; 39: 2225–2231. pmid:21705888
- 4. Kim SH, Choi SP, Park KN, Lee SJ, Lee KW, Jeong TO, et al. Association of blood glucose at admission with outcomes in patients treated with therapeutic hypothermia after cardiac arrest. Am J Emerg Med. 2014; 32: 900–904. pmid:24928411
- 5. Lee BK, Lee HY, Jeung KW, Jung YH, Lee GS, You Y. Association of blood glucose variability with outcomes in comatose cardiac arrest survivors treated with therapeutic hypothermia. Am J Emerg Med. 2013; 31: 566–572. pmid:23380125
- 6. Investigators N-SS, Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009; 360: 1283–1297. pmid:19318384
- 7. Nolan JP, Sandroni C, Bottiger BW, Cariou A, Cronberg T, Friberg H, et al. European Resuscitation Council and European Society of Intensive Care Medicine Guidelines 2021: Post-resuscitation care. Resuscitation. 2021; 161: 220–269. pmid:33773827
- 8. Panchal AR, Bartos JA, Cabanas JG, Donnino MW, Drennan IR, Hirsch KG, et al. Part 3: Adult Basic and Advanced Life Support: 2020 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation. 2020; 142: S366–S468. pmid:33081529
- 9. Furnary AP, Gao G, Grunkemeier GL, Wu Y, Zerr KJ, Bookin SO, et al. Continuous insulin infusion reduces mortality in patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg. 2003; 125: 1007–1021. pmid:12771873
- 10. American Diabetes A. Standards of medical care in diabetes—2010. Diabetes Care. 2010; 33:S11–S61. pmid:20042772
- 11. Oksanen T, Skrifvars MB, Varpula T, Kuitunen A, Pettila V, Nurmi J, et al. Strict versus moderate glucose control after resuscitation from ventricular fibrillation. Intensive Care Med. 2007; 33: 2093–2100. pmid:17928994
- 12. Kim SH, Park KN, Youn CS, Chae MK, Kim WY, Lee BK, et al. Outcome and status of postcardiac arrest care in Korea: results from the Korean Hypothermia Network prospective registry. Clin Exp Emerg Med. 2020; 7: 250–258. pmid:33440102
- 13. World Medical A. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013; 310: 2191–2194. pmid:24141714
- 14. Booth CM, Boone RH, Tomlinson G, Detsky AS. Is this patient dead, vegetative, or severely neurologically impaired? Assessing outcome for comatose survivors of cardiac arrest. JAMA. 2004; 291: 870–879. pmid:14970067
- 15. Jones AE, Trzeciak S, Kline JA. The Sequential Organ Failure Assessment score for predicting outcome in patients with severe sepsis and evidence of hypoperfusion at the time of emergency department presentation. Crit Care Med. 2009; 37: 1649–1654. pmid:19325482
- 16. Clement S, Braithwaite SS, Magee MF, Ahmann A, Smith EP, Schafer RG, et al. Management of diabetes and hyperglycemia in hospitals. Diabetes Care. 2004; 27: 553–591. pmid:14747243
- 17. Moghissi ES, Korytkowski MT, DiNardo M, Einhorn D, Hellman R, Hirsch IB, et al. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care. 2009; 32: 1119–1131. pmid:19429873
- 18. Baker L, Maley JH, Arevalo A, DeMichele F 3rd, Mateo-Collado R, Finkelstein S, et al. Real-world characterization of blood glucose control and insulin use in the intensive care unit. Sci Rep. 2020; 10: 10718. pmid:32612144
- 19. Hemmerling TM, Schmid MC, Schmidt J, Kern S, Jacobi KE. Comparison of a continuous glucose-insulin-potassium infusion versus intermittent bolus application of insulin on perioperative glucose control and hormone status in insulin-treated type 2 diabetics. J Clin Anesth. 2001; 13: 293–300. pmid:11435055
- 20. Investigators N-SS, Finfer S, Liu B, Chittock DR, Norton R, Myburgh JA, et al. Hypoglycemia and risk of death in critically ill patients. N Engl J Med. 2012;367: 1108–1118. pmid:22992074
- 21. Lee BK, Lee DH, Jeung KW, Yun SW, Callaway CW, Rittenberger JC. Glycated Hemoglobin is Associated with Glycemic Control and 6-Month Neurologic Outcome in Cardiac Arrest Survivors Undergoing Therapeutic Hypothermia. Neurocrit Care. 2020;32: 448–458. pmid:31187435