CADTH Health Technology Review

Temperature Management in Patients After Cardiac Arrest

Rapid Review

Authors: Diksha Kumar, Khai Tran, Zahra Premji

Key Messages

• Two systematic reviews (1 with a network meta-analysis and 1 with a meta-analysis), 1 randomized controlled trial, and 7 non-randomized studies were identified about the comparative clinical effectiveness of normothermia versus hypothermia in adult patients after cardiac arrest.

• Normothermia was found to be similar to hypothermia for several clinical- and patient-related outcomes, such as survival, hospital mortality, and quality of life. There was limited evidence to suggest that either type of targeted temperature management was more efficacious, with findings suggesting that normothermia may be associated with greater protocol adherence and decreased prescription medication use coming from low-quality non-randomized studies.

• Four evidence-based guidelines were identified regarding targeted temperature management (normothermia or hypothermia) in adult patients after cardiac arrest. All guidelines strongly recommend targeted temperature management for eligible patients, particularly for patients resuscitated following out-of-hospital cardiac arrest. Identified guidelines from the Canadian Cardiovascular Society and American Academy of Neurology present strong recommendations for hypothermic targeted temperature management.

Context and Policy Issues

An estimated 35,000 cardiac arrests occur annually in Canada, with the majority of them occurring outside of hospital settings.¹ The abrupt loss of cardiac function is associated with poor clinical outcomes; the rate of survival to hospital discharge for out-of-hospital cardiac arrest (OCHA) ranges between 3.9% and 7.1%.² Survival after cardiac arrest not only requires the coordination of emergency services to facilitate immediate resuscitation and return of circulation, but also critical care over the next 24 hours to ensure that the brain is adequately supplied with oxygenated blood.³ Fever is common among patients following cardiac arrest and resuscitation, and is thought to worsen brain damage.⁴ Brain injury is a major cause of both mortality and decreased function, therefore its prevention is an imperative aspect of post–cardiac arrest care.⁵

Hypothermic targeted temperature management (TTM) has been widely practised within this patient population to control body temperature.⁴ Also known as therapeutic hypothermia, the intervention involves the use of feedback-controlled intravascular cooling techniques in comatose patients to achieve a target body temperature typically in the range of 32°C to 34°C.³ Canadian guidelines have recommended therapeutic hypothermia for unresponsive patients following OCHA; however, the evidence base has been largely inconclusive.² Results from a recently published randomized controlled trial (RCT)⁶ showed that hypothermic TTM may not provide significantly clinical improvements over targeted normothermia at 37°C, which raises the question of whether existing hypothermic TTM protocols in critical care settings are of clinical value to this patient population.

The objective of this report is to identify and review current evidence regarding the comparative clinical effectiveness of normothermia versus hypothermia in adult patients after cardiac arrest and summarize identified evidence-based guidelines and recommendations regarding TTM.

Research Questions

1. What is the comparative clinical effectiveness of normothermia versus hypothermia for targeted temperature management in adult patients after cardiac arrest?

2. What are the evidence-based guidelines regarding targeted temperature management (normothermia or hypothermia) in adult patients after cardiac arrest?

Methods

Literature Search Methods

A limited literature search was conducted by an information specialist on key resources including MEDLINE, the Cochrane Library, the University of York Centre for Reviews and Dissemination (CRD) databases, the websites of Canadian and major international health technology agencies, as well as a focused internet search. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. For question 1, the main search concepts were cardiac arrest, hypothermia, and normothermia. No filters were applied to limit the retrieval by study type. For question 2, the main search concepts were cardiac arrest and temperature management. Search filters were applied to this search to limit retrieval to guidelines. When possible, retrieval was limited to the human population. The search was also limited to English-language documents published between January 1, 2016, and December 16, 2021.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1, were duplicate publications, or were published before 2016. Systematic reviews (SRs) in which all relevant studies were captured in other more recent or more comprehensive systematic reviews were excluded. Primary studies retrieved by the search were excluded if they were captured in 1 or more included SR. Guidelines with unclear methodology were also excluded.

Critical Appraisal of Individual Studies

The included publications were critically appraised by 1 reviewer using the following tools as a guide: A MeaSurement Tool to Assess systematic Reviews 2 (AMSTAR 2)⁷ for SR, the Professional Society for Health Economics and Outcomes Research (ISPOR) questionnaire⁸ for network meta-analyses, the Downs and Black checklist⁹ for randomized and non-randomized studies, and the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument¹⁰ for guidelines. Summary scores were not calculated for the included studies; rather, the strengths and limitations of each included publication were described narratively.

Summary of Evidence

Quantity of Research Available

A total of 280 citations were identified in the literature search. Following screening of titles and abstracts, 228 citations were excluded and 52 potentially relevant reports from the electronic search were retrieved for full-text review. Two potentially relevant publications were retrieved from the grey literature search for full-text review. Of these potentially relevant articles, 40 publications were excluded for various reasons and 14 publications met the inclusion criteria and were included in this report. These comprised 2 SRs, 1 RCT, 7 non-randomized studies, and 4 evidence-based guidelines. Appendix 1 presents the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA)¹¹ flow chart of the study selection. Additional references of potential interest are provided in Appendix 6.

Summary of Study Characteristics

Additional details regarding the characteristics of included publications are provided in Appendix 2.

Study Design

Two SRs were identified, both published in 2021.^5,12 The authors of 1 SR, which included RCTs published up to June 21, 2021, conducted a network meta-analysis (NMA) of 10 RCTs.⁵ Similarly, the other SR included RCTs that were published up to June 17, 2021, and included a meta-analysis of 9 RCTs.¹² A table outlining the overlap between included primary studies within the 2 SRs included in this report is provided in Appendix 5.

The included RCT, published in 2016, was of non-blinded, parallel design.¹³ Seven non-randomized studies published between 2016 and 2021 were also identified, including 6 retrospective cohort studies^14-19 and 1 observational post hoc analysis of an RCT.²⁰

Additionally, 4 guidelines with recommendations regarding post–cardiac arrest TTM were included in this report.^2,21-23 The most recent guideline, published in 2018, was developed by the French Intensive Care Society (FICS) and French Society of Anesthesia and Intensive Care Medicine.²¹ Two of the included guidelines were published in 2017, 1 by the American Academy of Neurology (AAN),²² and the other developed jointly by the Canadian Cardiovascular Society (CCS), Canadian Cardiovascular Critical Care Society, and Canadian Association of Interventional Cardiology.² The final guideline from 2016 was developed by the Canadian Critical Care Society (CCCS), Canadian Neurocritical Care Society, and Canadian Critical Care Trials Group.²³ Methods for the collection, selection, and synthesis of evidence informing the recommendations were reported in 3 guidelines, stating that a comprehensive literature search was performed^2,22,23 and that the evidence was reviewed by committee members.^2,22 Two guidelines classified the level of evidence for a recommendation from “very low” or “low” to “high.”^21,23 The AAN guideline graded evidence as Class I, II, III, or IV.²² The authors of the CCS guideline used Grading of Recommendations, Assessment, Development and Evaluations (GRADE) methodology to assess the quality of evidence.² Final recommendations were made using Delphi methodology,^21,23 GRADE methodology,^2,21 and nominal group technique.²³ The strength of recommendations was classified in various ways; 2 of the included guidelines referred to them as either “strong” or “conditional,”^2,23 1 graded them as A, B, C, or U (insufficient evidence),²² and the other organized them as positive or negative and as strong or weak.²¹

Country of Origin

One of the SRs listed first authors from Canada,⁵ and the other listed first authors from Denmark.¹² The included primary clinical studies were conducted in Israel,¹⁷ Singapore,^13,18 South Korea,¹⁶ Switzerland,¹⁴ the Netherlands,²⁰ and Turkey.¹⁵ Two of the included guidelines were Canadian,^2,23 1 was from the US,²² and 1 was from France.²¹

Patient Population

Both SRs^5,12 included RCTs with adult patients after cardiac arrest, with the SR by Fernando et al. (2021)⁵ limiting inclusion to patients with OCHA who remained unresponsive following signs of return of spontaneous circulation (ROSC). The SR by Fernando et al. (2021)⁵ included 4,218 participants across 10 RCTs, and the SR by Granfeldt et al. (2021)¹² included 2,968 participants from 9 RCTs.¹² Patients’ mean age was similar between the 2 SRs (approximately 63 years⁵ and 64¹² years). The majority of the included population in both SRs were male, ranging from 60% to 88.9% in 1 SR⁵ and from 56% to 100% in the other.¹²

There was some variation in the patient population among the 8 primary studies included in this report. All studies included adult patients hospitalized following cardiac arrest.^13-20 The mean age of the patients ranged between 52.5 years and 65 years.^13-20 Sex was reported in 7 of the 8 primary studies, and the majority of patients across the studies were male, ranging from 53.4% to 81.5%. One retrospective cohort study¹⁶ and 1 observational post hoc analysis²⁰ limited inclusion to patients with OCHA, whereas the others did not specify between in-hospital and out-of-hospital cardiac arrest. The RCT¹³ and retrospective cohort study¹⁸ by Pang et al. included primarily patients who had an in-hospital cardiac arrest (90.5% to 92.4% of participants) and specifically comprised patients on extracorporeal life support. Three of the included retrospective cohort studies^15,17,19 reviewed patient data from patients in the intensive care unit (ICU).

The relevant target population of the 4 included guidelines^2,21-23 in this report was adult patients with cardiac arrest. One guideline²¹ contained recommendations for a broader population: both adult and pediatric patients with cardiac arrest, traumatic brain injury, stroke or other brain injuries, and shock. Authors of the CCS guideline² specified that the recommendations pertained to patients with OCHA. The guideline developed by FICS²¹ was broadly intended for clinicians who provide neurologic care to critically ill patients. The intended users of the 3 other included guidelines^2,22,23 were clinicians who provide care for patients with cardiac arrest; the AAN guideline²² is specifically for nontraumatic cardiac arrest and the CCS is guideline for OCHA.²

Interventions and Comparators

The SR with NMA by Fernando et al. (2021)⁵ compared normothermia (37°C to 37.8°C) with 3 temperature ranges of hypothermia: 31°C to 32°C (which they referred to as “deep hypothermia”), 33°C to 34°C (“moderate hypothermia”), and 35°C to 36°C (“mild hypothermia”). The SR with meta-analysis by Granfeldt et al. (2021)¹² compared normothermia, whether it involved active cooling as part of TTM or not, with hypothermic TTM (32°C to 34°C).

Normothermia was defined in various ways in the included primary clinical studies. Therapeutic normothermia was described as a body temperature maintained at 36°C,^14,19 37°C,¹³ 36°C to 37°C,²⁰ or higher than 35.5°C.¹⁶ Three retrospective studies^15,17,18 considered normothermia as a comparison group but did not specify a temperature range; however, 1¹⁵ of the studies considered normothermia as a non-TTM intervention. Targeted normothermia was specified to be maintained for 28 hours in the retrospective cohort study by Casamento et al. (2016).¹⁹

Hypothermia was also defined in various ways in the included primary studies. Targeted hypothermia was described as body temperature maintained at 33°C,¹⁴ 34°C,¹³ 32°C to 34°C,^17,19,20 33°C to 35°C,¹⁵ or lower than 35.5°C.¹⁶ Hypothermia was specified as maintained for 24 hours in the studies by Pang et al.^13,18 and Casamento et al. (2016).¹⁹

All included guidelines presented recommendations regarding TTM as part of in-hospital care following cardiac arrest.^2,21-23 The AAN guideline defined therapeutic hypothermia, or hypothermic TTM, as cooling to a core body temperature of 32°C to 34°C.²² The CCS guideline broadly defined the intervention considered, TTM, as a strategy of intentional temperature management of a patient after cardiac arrest, which consists of active patient cooling, subsequent rewarming, and extended fever control.²

Outcomes

Several clinical outcomes of relevance were described in the studies included in this review. These outcomes included the following: survival with good neurologic outcome,^5,12,13,18 overall survival,^{5,12,13,15-20} mortality,^13-15,17,19 ICU or hospital length of stay,^13-15,18,19 neurologic status,^13-16,20 protocol adherence,¹⁴ quality of life,²⁰ time spent in temperature target range,¹⁹ medication prescription,^14,19 and adverse events.^{5,13,14,17-19}

The modified Rankin Scale, Cerebral Performance Categories scale, and Pittsburgh Cardiac Arrest Category scale were the validated measures used to describe neurologic outcomes in 2 SRs,^5,12 1 RCT,¹³ and 5 non-randomized studies.^14-16,18,20 The modified Rankin Scale is commonly used to measure the degree of disability of people who have suffered a stroke or other causes or neurologic disability.²⁴ The scale ranges from 0 to 6, with 0 for no symptoms, 3 for moderate disability (requires some help but able to walk without assistance), and 6 for dead.²⁴ The Cerebral Performance Categories are the gold standard for describing the level of function and impairment after cardiac arrest.²⁵ The categories, which rank from 1 to 5, are as follows: good cerebral performance, moderate cerebral disability, severe cerebral disability, coma or vegetative state, and brain death.²⁵ Finally, the Pittsburgh Cardiac Arrest Category is a 4-level illness severity score with the following rankings: awake, coma with mild cardiopulmonary failure, coma with severe cardiopulmonary failure, and deep coma.²⁶ Studies that investigated survival with good neurologic outcome defined a favourable neurologic outcome as any of the following: modified Rankin Scale score of 0 to 3,^5,12 Cerebral Performance Categories scale score of 1 or 2,^5,12,13,18 Pittsburgh Cardiac Arrest Category score of 1 or 2,⁵ or blinded assessment from a health professional indicating no, mild, or moderate disability.⁵

The authors of 1 retrospective cohort study¹⁴ reported protocol adherence as a primary outcome, and patients were considered to have adhered to the protocol if their body core temperature was within a range of 1°C around the target temperatures of 33°C and 36°C for the hypothermia and normothermia groups, respectively.

The authors of 2 retrospective cohort studies^14,19 reported several outcomes related to the prescription and dose of catecholamines, sedatives, and relaxants administered to patients in the targeted normothermia and hypothermia groups.

The 4 guidelines included in this report regarded survival with good neurologic function as an outcome for consideration when developing recommendations.^2,21-23

Summary of Critical Appraisal

Additional details regarding the strengths and limitations of the included publications are provided in Appendix 2.

Systematic Reviews and Network Meta-Analyses

Both SRs^5,12 used comprehensive search and selection strategies, which is evident because most RCTs relevant to this report were excluded on the basis of inclusion within the SRs. Both SRs by Fernando et al. (2021)⁵ and Granfeldt et al. (2021)¹² had registered published protocols with detailed search strategies. Fernando et al. (2021)⁵ searched 6 databases (MEDLINE, PubMed, Embase, Scopus, Web of Science, and the Cochrane Database of Systematic Reviews), conducted further surveillance searches, and performed a grey literature search, including the reference lists of all included studies and existing SR on TTM. Granfeldt et al. (2021) searched 3 databases (PubMed, Embase, and the Cochrane Central Register of Controlled Trials) as well as the International Clinical Trials Registry Platform and ClinicalTrials.gov for registered ongoing and unpublished trials. However, it is unclear whether content experts were consulted during the review process of either of the SRs^5,12 and, in the case of Fernando et al. (2021),⁵ it is unclear whether trial registries were searched as part of the grey literature search outlined in the published protocol. Research questions and inclusion criteria were clearly reported and consisted of components of population, intervention, comparison, and outcomes.^5,12 Although Fernando et al. (2021)⁵ explained their selection of RCTs for inclusion in the review, Granfeldt et al. (2021)^5,12 did not justify their selection of randomized and non-randomized controlled studies for inclusion. The study selection and data extraction processes were performed in duplicate, and the included studies were described in adequate detail in supplementary materials.^5,12 Reasons for exclusion cannot be easily verified because neither of the SRs published a list of excluded studies.

In terms of conducting meta-analyses, authors of both SRs^5,12 justified combining the data in a meta-analysis or NMA and used assessment tools by Cochrane to assess the risk of bias of included studies. Statistical heterogeneity among trials were assessed using visual inspection of forest plots, the I² statistic, and the Chi-square test.^5,12 Heterogeneity was observed in the results of the reviews, but the authors investigated the potential sources of any heterogeneity in the results and discussed its impact in the findings.^5,12

Fernando et al. (2021)⁵ initially performed pairwise MA using the random-effects model, and then evaluated the feasibility of conducting an NMA by checking the availability of evidence, transitivity assumption, connectivity, and agreement between direct and indirect evidence. With the assumption of a common heterogeneity parameter, the frequentist random-effects model was chosen to combine the included data.⁵ However, the authors did not provide a justification about why the analyses were performed with a frequentist instead of a Bayesian approach. Due to insufficient data, the authors were unable to conduct subgroup analyses or meta-regression to explain statistical heterogeneity.⁵ The authors presented a network graph of included studies, and given the geometry of the NMA, the comparisons between normothermia and deep and moderate hypothermia were direct, and the comparison between normothermia and mild hypothermia was indirect.⁵ In addition, Fernando et al. (2021)⁵ provided a clear summary of the main findings but did not perform any sensitivity analyses.

Randomized Controlled Trial

Several aspects of the included RCT¹³ were clearly described, such as the study objective, primary and secondary outcomes, patient characteristics, interventions, potential confounders, main findings, estimates of random variability, and adverse events. However, in reviewing the results section, it was unclear whether 1 of the 12 participants in the normothermia group was lost to follow-up given the discrepancy between the total number of participants and the number of patients discharged or deceased. The single-centre study was conducted in a tertiary referral centre, which may not be representative of where most patients seek immediate care after a cardiac arrest. In addition, due to the non-blinded design, investigators, outcomes assessors, and patients were aware of the treatment administered, which may have increased the risk of performance bias and detection bias. It is unclear whether the study was adequately powered because of its sample size of 21 participants. This RCT was also published as a letter to the editor, thus it is unclear whether the methodology and findings have been subject to peer review.

Non-Randomized Studies

Overall, the reporting was clear across the 7 non-randomized studies^14-20 in terms of study objectives, patient characteristics, potential confounders, interventions, and estimates of random variability. In 3 studies,^14,19,20 the characteristics of the patients lost to follow-up were not described. The authors of 2 studies^17,18 did not specify a temperature range for the normothermia group. In Kocayigit et al. (2021),¹⁵ it was unclear when the survival and good neurologic results outcomes were measured. The external validity of the study findings to other similar settings was generally considered poor because 4 of the included non-randomized studies^14,15,17,18 were single-centre, retrospective cohort studies. The other 3 studies consisted of patient data derived from either multiple centres^19,20 or national registries,¹⁶ which may have increased generalizability. None of the included studies were conducted in Canada, which may limit relevance of the results in the Canadian context. Threats to the internal validity of the studies were primarily due to the non-randomized nature of the studies: 6 of the included studies^14-19 were of retrospective cohort design, which means the investigators, outcome assessors, and patients may have been aware of the intervention. Additionally, there is uncertainty about whether there was a high level of compliance with the intervention and whether the recorded patient data are accurate. Participants across the intervention and comparator groups appeared to be recruited at the same sites or from the same population and over the same periods of time. Some studies^16,17,19,20 considered potential confounders when adjusting main outcomes, whereas other studies^15,18 did not take into account baseline differences between the groups. Only 1 included study¹⁹ provided power calculations, but sample size was noted to be a limitation in 3 studies.^14,17,18

Guidelines

All included guidelines described a clear scope and purpose.^2,21-23 Evidence of some stakeholder involvement was evident in the development of all guidelines; however, it was unclear whether the views and preferences of the target patient population were sought.^2,21-23 In terms of rigour of development, all included guidelines conducted a comprehensive search to collect evidence to support recommendations and did not specify whether other systematic methods, such as screening and study selection in duplicate, were undertaken.^2,21-23 The guidelines also did not provide a procedure for future updates.^2,21-23 Nonetheless, the supporting evidence was clearly appraised and used to support recommendations, and there was an evident process and methodology for the development of recommendations.^2,21-23 There was an explicit link between the recommendations and the supporting evidence, with varying strengths of recommendations associated with different levels of evidence.^2,21-23 All relevant recommendations were easily identifiable and clearly presented.^2,21-23 The authors of the guidelines adequately described potential facilitators and barriers to application of recommendations but did not provide intended users with additional guidance on how to implement the recommendations and monitor the implementation of relevant interventions.^2,21-23 Finally, all the included guidelines appeared to have editorial independence, with clear disclosures of competing interests.^2,21-23

Summary of Findings

Appendix 4 presents the main study findings and authors’ conclusions.

Clinical Effectiveness of Normothermia Versus Hypothermia

Survival

Two SRs,^5,12 1 RCT,¹³ and 1 retrospective cohort study¹⁸ evaluated the effect of TTM on survival rate with good neurologic or functional outcome. A favourable neurologic or functional outcome was defined as any of the following: modified Rankin Scale score of 0 to 3,^5,12; Cerebral Performance Categories scale score of 1 or 2,^5,12,13,18; Pittsburgh Cardiac Arrest Category score of 1 or 2,⁵ or blinded assessment from a health professional indicating no, mild, or moderate disability.⁵ No statistically significant differences were identified in the survival rate with good neurologic outcome between the normothermia and hypothermia groups.^5,12,13,18 The SRs with NMA and meta-analysis reported odds ratios⁵ and risk ratios,¹² respectively; however, the estimates were not statistically significant, suggesting no differences between interventions.

Two SRs, 1 RCT, 5 retrospective cohort studies, and 1 observational post hoc analysis of an RCT also evaluated overall survival, recorded at an unspecified time point,^15,17 at hospital discharge,^{5,12,16,18,19} or within 30 days,¹² 90 days,^5,12,20 6 months,^5,12,13 or 1 year after discharge.²⁰ The authors of 1 retrospective cohort study¹⁹ found a significantly greater proportion of survivors at discharge in the normothermia group compared with the hypothermia (TTM 32°C to 34°C) group, but no statistically significant differences were reported in the other studies.^{5,12,13,15-18,20}

Mortality

Hospital and ICU mortality rates were evaluated in 1 RCT¹⁸ and 4 of the retrospective cohort studies^14,15,17,19 included in this report. There were no statistically significant differences between the normothermia and hypothermia groups.

ICU or Hospital Length of Stay

One RCT¹³ and 4 retrospective cohort studies^14,15,18,19 reported the average length of ICU^14,19 or hospital^13,15,18 stay as either a mean with standard deviation¹⁴ or median with interquartile range.^13,15,18,19 The authors of 1 retrospective cohort study found that patients in the normothermia group who received standard supportive therapy experienced significantly longer hospital stays compared with patients in the hypothermia (TTM at 33°C to 35°C) group.¹⁵ There were no significant differences between the groups reported in the other studies.^13,14,18,19

Neurologic Status

Neurologic status and function were measured in the studies included in this report using Cerebral Performance Categories scale scores^13-16,20 and modified Rankin Scale scores.¹⁴ The authors of 1 retrospective cohort study¹⁵ found that a significantly greater proportion of patients in the hypothermia (TTM at 33°C to 35°C) group had good neurologic results, defined as a Cerebral Performance Categories scale score of either 1 or 2, compared with patients in normothermia (non-TTM) group. Although this finding was in favour of targeted hypothermia, no statistically significant differences were identified among the groups in other studies that also investigated this outcome.^13,14,16,20

Protocol Adherence

One retrospective cohort study by Duggelin et al. (2021)¹⁴ investigated the proportion of patients adherent to the TTM protocol at 24 hours. Using a Cox proportional hazard model, it was demonstrated that a significantly higher proportion patients in the normothermia (36°C) group were adherent to the TTM protocol compared with patients in the hypothermia (33°C) group.¹⁴

Quality of Life

Quality of life at 1-year follow-up, measured using the 36-Item Short Form Survey, was reported in 1 included observational post hoc analysis.²⁰ Patients in both groups reported similar physical and mental component scores.²⁰ The risk ratio estimates were not statistically significant, suggesting no differences between interventions.

Time Spent in Temperature Target Range

The authors of 1 retrospective cohort study¹⁹ evaluated the median proportion of time spent in the prescribed temperature range among patients in the hypothermia (TTM 32 to 34°C) versus normothermia (TTM 36°C) groups and found no statistically significant differences.

Medication Prescription

The proportion of patients receiving sedatives and relaxants and the mean or median doses prescribed were reported in 2 retrospective cohort studies.^14,19 There were no statistically significant differences in the proportion of patients receiving propofol, morphine, midazolam, fentanyl, and pethidine for sedation, but 1 of the studies found that a significantly lower proportion of patients in the normothermia (TTM 36°C) group received at least 1 dose of muscle relaxant compared with patients in the hypothermia (TTM 32 to 34°C) group. Patients in the normothermia (TTM 36°C) group were reported to have received significantly lower doses of midalozam^14,19 and fentanyl¹⁹ compared with patients in the hypothermia (TTM 32 to 34°C) group.

Adverse Events

Several adverse events were reported in the studies included in this report: arrhythmia,^5,13,14,19 bleeding,^{5,13,14,18,19} pneumonia,^5,13,14 sepsis,⁵ seizures,^5,17 limb ischemia,^13,18 acute renal failure,^13,18 continuous renal replacement therapy,^13,18,19 thrombosis,¹⁴ bradycardia,¹⁴ hypoxic ischemia encephalopathy,¹⁸ stroke,¹⁸ ischemic hepatitis,¹⁸ cardiovascular instability,¹⁹ and mechanical ventilation.¹⁹ The authors of Duggelin et al. (2021)¹⁴ found that there was a significantly lower incidence of bradycardia among patients in the normothermia (TTM 36°C) group compared with patients in the hypothermia (TTM 33°C) group. However, no statistically significant differences were identified among the groups in other studies or for other adverse events.^5,13,17-19

Guidelines

Among the 4 included guidelines, all strongly recommended the use of TTM to improve survival and neurologic outcome in patients resuscitated from OCHA with a shockable cardiac rhythm (ventricular tachycardia or ventricular fibrillation).^2,21-23 The CCCS guideline²³ extended the strong recommendation to patients presenting with pulseless ventricular tachycardia. In terms of patients with in-hospital cardiac arrest, the FICS guideline issued a positive recommendation for TTM at the level of expert opinion.²¹ However, the CCCS guideline contained a strong recommendation supported by low-level evidence in favour of TTM for patients with in-hospital cardiac arrest.²³

The CCS guideline strongly recommended hypothermia (between 33°C and 36°C) for eligible patients who undergo TTM. The AAN guideline²² similarly specified in their recommendation that the TTM intervention should be hypothermia (32°C to 34°C for 24 hours); however, they also issued a lower-level recommendation supported by the same level of evidence for normothermia (36°C for 24 hours followed by 8 hours of rewarming to 37°C and temperature maintenance below 37.5°C until 72 hours) in patients who are comatose with either a shockable or non-shockable initial cardiac rhythm. The FICS guideline²¹ contains a weak recommendation for hypothermic (32°C to 36°C) TTM. Recommendations in the CCCS guideline²³ do not state target temperature ranges for TTM.

With regard to specific technical recommendations, the CCS issued a lower-level recommendation supporting the use of surface cooling or intravascular cooling techniques.²

Limitations

Although there was a high quantity of eligible publications identified in this report for the clinical effectiveness question, the quality of the included primary studies^5,12-20 was low. Some limitations were explicitly acknowledged by authors of the individual studies included in this review, such as the lack of generalizability of a single-centre retrospective cohort study and small sample sizes. Moreover, the critical appraisal performed as part of this review identified some important limitations of the included studies. For example, some of the included non-randomized studies represented outcomes with unclear time points and follow-up periods and lacked detail on important adverse events. With the exception of 1 RCT included in the SR by Fernando et al. (2021),⁵ none of the studies included in this report were conducted in a Canadian setting, which could be a limitation.

Furthermore, there was a significant amount of overlap identified across the 2 SRs included in this report; much of the available evidence identified in this review is based on the same trials. The 1 RCT included in this report that did not overlap with the SRs is a preliminary study with 21 participants. This limits the available evidence base to address research questions regarding the clinical effectiveness of targeted and non-targeted normothermia for patients in critical care after cardiac arrest.

The SRs included in this report had several strengths in methodology, but authors of both SRs^5,12 reported heterogeneity in study characteristics, such as patient populations and interventions, among the included RCTs and did not perform subsequent meta-regression or subgroup analyses to investigate the heterogeneity due to insufficient data. The NMA by Fernando et al. (2021)⁵ also lacked direct comparative evidence evaluating the differences between normothermia and mild hypothermia (35°C to 36°C).

Although the guidelines included in this report were generally of high quality, some relevant recommendations regarding TTM were categorized as conditional and supported by low-quality evidence, as stated by guideline authors. The guidelines did not contain any recommendations pertaining to the management of fever and shivering or specifics on the timing of TTM therapy.

Conclusions and Implications for Decision- or Policy-Making

This report identified 2 SRs (1 with an NMA⁵ and 1 with a meta-analysis¹²), 1 RCT,¹³ and 7 non-randomized studies^14-20 published since 2016 informing the comparative clinical effectiveness of normothermia versus hypothermia for adult patients after cardiac arrest. Overall, there were no statistically significant differences across all the studies for several clinical effectiveness and patient-related outcomes: survival with good neurologic outcome, overall survival, hospital mortality, quality of life, and time spent in the temperature target range. There was an indication that targeted normothermia at 36°C may offer slight benefits over hypothermic TTM. Outcomes in favour of normothermia, such as greater proportion of survivors at discharge, protocol adherence, and decreased dose and use of prescribed sedatives and relaxants, were identified in 2 retrospective cohort studies.^14,19 However, the studies^14,19 were characterized to be of low quality and with potentially limited generalizability to the Canadian context. Therefore, there is limited evidence to suggest that normothermic TTM may provide more clinical benefits over hypothermic TTM.

In addition, 4 evidence-based guidelines published since 2016 were identified regarding the use of TTM for adult patients after cardiac arrest. In general, TTM was strongly recommended by all included guidelines, particularly for adult patients resuscitated from OCHA. Patients presenting with shockable initial cardiac rhythm versus non-shockable cardiac rhythm were also more strongly recommended for TTM. Unlike the clinical evidence identified in this report, guidelines — particularly those developed by Canadian and American organizations — strongly recommended the use of hypothermia for eligible patients, either in the 33°C to 36°C or 32°C to 34°C ranges. The guidelines included in this report were found to be of moderate to high quality. It is important to acknowledge that these guidelines are older compared with some of the clinical studies included in this report, with the latest 1 published in 2018, which may be the cause of discrepancies between the evidence used to inform the recommendations and the current literature on the topic.

Baseline characteristics, such as age, location of cardiac arrest (in hospital vs. out of hospital), and initial cardiac rhythm, may play a role in determining which patients are mostly likely to benefit from either modality of TTM. Larger RCTs and additional SRs with less heterogeneity may allow for the necessary means to identify such subgroups.

References

1.Heart and Stroke Foundation of Canada. Saving lives. 2022; https://www.heartandstroke.ca/what-we-do/our-impact/saving-lives. Accessed 2022 Jan 19.

2.Wong GC, van Diepen S, Ainsworth C, et al. Canadian Cardiovascular Society/Canadian Cardiovascular Critical Care Society/Canadian Association of Interventional Cardiology Position Statement on the Optimal Care of the Postarrest Patient. Can J Cardiol. 2017;33(1):1-16. PubMed

3.Schafer A, Bauersachs J, Akin M. Therapeutic Hypothermia following Cardiac Arrest after the TTM2 trial - more Questions Raised than Answered. Curr Probl Cardiol. 2021:101046. PubMed

4.Kikutani K, Nishikimi M, Shimatani T, Kyo M, Ohshimo S, Shime N. Differential Effectiveness of Hypothermic Targeted Temperature Management According to the Severity of Post-Cardiac Arrest Syndrome. J Clin Med. 2021;10(23):5643. PubMed

5.Fernando SM, Di Santo P, Sadeghirad B, et al. Targeted temperature management following out-of-hospital cardiac arrest: a systematic review and network meta-analysis of temperature targets. Intensive Care Med. 2021;47(10):1078-1088. PubMed

6.Dankiewicz J, Friberg H, Belohlavek J, et al. Time to start of cardiopulmonary resuscitation and the effect of target temperature management at 33degreeC and 36degreeC. Resuscitation. 2016;99:44-49. PubMed

7.Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. PubMed

8.Jansen JP, Trikalinos T, Cappelleri JC, et al. Indirect treatment comparison/network meta-analysis study questionnaire to assess relevance and credibility to inform health care decision making: an ISPOR-AMCP-NPC Good Practice Task Force report. Value Health. 2014;17(2):157-173. PubMed

9.Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. PubMed

10.Agree Next Steps Consortium. The AGREE II Instrument. Hamilton (ON): AGREE Enterprise; 2017: https://www.agreetrust.org/wp-content/uploads/2017/12/AGREE-II-Users-Manual-and-23-item-Instrument-2009-Update-2017.pdf. Accessed 2022 Jan 27.

11.Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. PubMed

12.Granfeldt A, Holmberg MJ, Nolan JP, Soar J, Andersen LW, International Liaison Committee on Resuscitation Advanced Life Support Task F. Targeted temperature management in adult cardiac arrest: Systematic review and meta-analysis. Resuscitation. 2021;167:160-172. PubMed

13.Pang PY, Wee GH, Hoo AE, et al. Therapeutic hypothermia in adult patients receiving extracorporeal life support: early results of a randomized controlled study. J Cardiothorac Surg. 2016;11:43. PubMed

14.Duggelin R, Maggiorini M, Voigtsberger S, Schwarz U, Hilty MP. Increased protocol adherence and safety during controlled normothermia as compared to hypothermia after cardiac arrest. J Crit Care. 2021;63:146-153. PubMed

15.Kocayigit H, Suner KO, Kaya B, Tomak Y, Tuna AT, Erdem AF. Neurological Outcomes of Normothermia versus Targeted Temperature Management in Post-cardiac Arrest Syndrome. J Coll Physicians Surg Pak. 2021;30(5):497-501. PubMed

16.Kim JH, Park JH, Shin SD, et al. Effects of moderate hypothermia versus normothermia on survival outcomes according to the initial body temperature in out-of-hospital cardiac patients: A nationwide observational study. Resuscitation. 2020;151:157-165. PubMed

17.Eilam A, Samogalskyi V, Bregman G, Eliner-Avishai S, Gilad R. Occurrence of overt seizures in comatose survivor patients treated with targeted temperature. Brain Behav. 2017;7(11):e00842. PubMed

18.Pang PYK, Wee GHL, Huang MJ, et al. Therapeutic Hypothermia May Improve Neurological Outcomes in Extracorporeal Life Support for Adult Cardiac Arrest. Heart Lung Circ. 2017;26(8):817-824. PubMed

19.Casamento A, Minson A, Radford S, et al. A comparison of therapeutic hypothermia and strict therapeutic normothermia after cardiac arrest. Resuscitation. 2016;106:83-88. PubMed

20.Spoormans EM, Lemkes JS, Janssens GN, et al. Targeted Temperature Management in Out-of-Hospital Cardiac Arrest With Shockable Rhythm: A Post Hoc Analysis of the Coronary Angiography After Cardiac Arrest Trial. Crit Care Med. 2021;22:22. PubMed

21.Cariou A, Payen JF, Asehnoune K, et al. Targeted temperature management in the ICU: Guidelines from a French expert panel. Anaesth Crit Care Pain Med. 2018;37(5):481-491. PubMed

22.Geocadin RG, Wijdicks E, Armstrong MJ, et al. Practice guideline summary: Reducing brain injury following cardiopulmonary resuscitation: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology. 2017;88(22):2141-2149. PubMed

23.Howes D, Gray SH, Brooks SC, et al. Canadian Guidelines for the use of targeted temperature management (therapeutic hypothermia) after cardiac arrest: A joint statement from The Canadian Critical Care Society (CCCS), Canadian Neurocritical Care Society (CNCCS), and the Canadian Critical Care Trials Group (CCCTG). Resuscitation. 2016;98:48-63. PubMed

24.Banks JL, Marotta CA. Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke. 2007;38(3):1091-1096. PubMed

25.Balouris SA, Raina KD, Rittenberger JC, Callaway CW, Rogers JC, Holm MB. Development and validation of the Cerebral Performance Categories-Extended (CPC-E). Resuscitation. 2015;94:98-105. PubMed

26.Coppler PJ, Elmer J, Calderon L, et al. Validation of the Pittsburgh Cardiac Arrest Category illness severity score. Resuscitation. 2015;89:86-92. PubMed

Appendix 1: Selection of Included Studies

Figure 1: Selection of Included Studies

Appendix 3: Critical Appraisal of Included Publications

Note that this appendix has not been copy-edited.

Table 5: Strengths and Limitations of Systematic Reviews and Network Meta-Analysis Using AMSTAR 2⁷ and the ISPOR Questionnaire⁸

AMSTAR 2 = A MeaSurement Tool to Assess systematic Reviews 2; CI = confidence interval; GRADE = Grading of Recommendations, Assessment, Development and Evaluations; ISPOR = International Society for Pharmacoeconomics and Outcomes Research; MA = meta-analysis; NMA = network meta-analysis; NR = not reported; RCT = randomized controlled trial.

Table 6: Strengths and Limitations of Clinical Studies Using the Downs and Black Checklist⁹

APACHE = Acute Physiologic Assessment and Chronic Health Evaluation; COACT = Coronary Angiography after Cardiac Arrest; ECMO = extracorporeal membrane oxygenation; ICU = intensive care unit; IQR = interquartile range; NR = not reported; OCHA = out-of-hospital cardiac arrest; ROSC = return of spontaneous circulation; SD = standard deviation.

Table 7: Strengths and Limitations of Guidelines Using AGREE II¹⁰

AGREE II = Appraisal of Guidelines for Research and Evaluation II.

Appendix 4: Main Study Findings and Authors’ Conclusions

Note that this appendix has not been copy-edited.

Table 8: Summary of Findings by Outcomes for Survival

CI = confidence interval; ECLS = extracorporeal life support; HR = hazard ratio; ICU = intensive care unit; MA = meta-analysis; NMA = network meta-analysis; OHCA = out-of-hospital cardiac arrest; OR = odds ratio; RCT = randomized controlled trial; ROSC = return of spontaneous circulation; RR = risk ratio; SR = systematic review; TTM = targeted temperature management.

Table 9: Summary of Findings by Outcomes for Mortality

CPR = cardiopulmonary resuscitation; ECLS = extracorporeal life support; ICU = intensive care unit; RCT = randomized controlled trial; TTM = targeted temperature management.

Table 10: Summary of Findings by Outcomes for ICU and Hospital Length of Stay

ECLS = extracorporeal life support; ICU = intensive care unit; SD = standard deviation; RCT = randomized controlled trial; TTM = targeted temperature management.

Table 11: Summary of Findings by Outcomes for Neurological Status

CPC = cerebral performance category; ECLS = extracorporeal life support; ICU = intensive care unit; OR = odds ratio; RCT = randomized controlled trial; TTM = targeted temperature management.

Table 12: Summary of Findings by Outcomes for Protocol Adherence

TTM = targeted temperature management.

Table 13: Summary of Findings by Outcomes for Quality of Life

CI = confidence interval; OHCA = out-of-hospital cardiac arrest; RCT = randomized controlled trial; RR = risk ratio; TTM = targeted temperature management.

Table 14: Summary of Findings by Outcomes for Time Spent in Temperature Target Range

ICU = intensive care unit; TTM = targeted temperature management.

Table 15: Summary of Findings by Outcomes for Medication Prescription

ICU = intensive care unit; TTM = targeted temperature management.

Table 16: Summary of Findings by Outcomes for Adverse Events

AE = adverse event; CI = confidence interval; ECLS = extracorporeal life support; HR = hazard ratio; ICU = intensive care unit; MA = meta-analysis; NMA = network meta-analysis; OHCA = out-of-hospital cardiac arrest; OR = odds ratio; RCT = randomized controlled trial; ROSC = return of spontaneous circulation; RR = risk ratio; SR = systematic review; TTM = targeted temperature management.

Table 17: Summary of Recommendations in Included Guidelines

CA = cardiac arrest; MA = meta-analysis; OHCA = out-of-hospital cardiac arrest; PEA = pulseless electrical activity; RCT = randomized controlled trial; ROSC = return of spontaneous circulation; SR = systematic reviews; TH = therapeutic hypothermia; TTM = targeted temperature management; VF = ventricular fibrillation; VT = ventricular tachycardia.

Appendix 5: Overlap Between Included Systematic Reviews

Note that this appendix has not been copy-edited.

Table 18: Overlap in Relevant Primary Studies Between Included Systematic Reviews

Appendix 6: References of Potential Interest

Note that this appendix has not been copy-edited.

Previous CADTH Reports

Temperature settings for therapeutic hypothermia: guidelines. (CADTH rapid response report: summary of abstracts). Ottawa (ON): CADTH; 2016: https://www.cadth.ca/sites/default/files/pdf/htis/june-2016/RB0993%20Therapeutic%20Hypothermia%20Final.pdf. Accessed 2022 Jan 27.

Normothermia versus therapeutic hypothermia for adult patients after cardiac arrest: clinical evidence. (CADTH rapid response report: summary of abstracts). Ottawa (ON): CADTH; 2014: https://www.cadth.ca/media/pdf/htis/dec-2014/RB0727%20Normothermia%20Final.pdf. Accessed 2022 Jan 27.

Additional References

National Institute for Health and Care Excellence. Arctic Sun 5000 for therapeutic hypothermia after cardiac arrest. (Medtech innovation briefing MIB112) 2017; https://www.nice.org.uk/advice/mib112. Accessed 2022 Jan 27.