Rewarming From Hypothermic Cardiac Arrest Applying Extracorporeal Life Support: A Systematic Review and Meta-Analysis

Abstract

Introduction: This systematic review and meta-analysis aims at comparing outcomes of rewarming after accidental hypothermic cardiac arrest (HCA) with cardiopulmonary bypass (CPB) or/and extracorporeal membrane oxygenation (ECMO). Material and Methods: Literature searches were limited to references with an abstract in English, French or German. Additionally, we searched reference lists of included papers. Primary outcome was survival to hospital discharge. We assessed neurological outcome, differences in relative risks (RR) of surviving, as related to the applied rewarming technique, sex, asphyxia, and witnessed or unwitnessed HCA. We calculated hypothermia outcome prediction probability score after extracorporeal life support (HOPE) in patients in whom we found individual data. P < 0.05 considered significant. Results: Twenty-three case observation studies comprising 464 patients were included in a meta-analysis comparing outcomes of rewarming with CPB or/and ECMO. One-hundred-and-seventy-two patients (37%) survived to hospital discharge, 76 of 245 (31%) after CPB and 96 of 219 (44 %) after ECMO; 87 and 75%, respectively, had good neurological outcomes. Overall chance of surviving was 41% higher (P = 0.005) with ECMO as compared with CPB. A man and a woman had 46% (P = 0.043) and 31% (P = 0.115) higher chance, respectively, of surviving with ECMO as compared with CPB. Avalanche victims had the lowest chance of surviving, followed by drowning and people losing consciousness in cold environments. Assessed by logistic regression, asphyxia, unwitnessed HCA, male sex, high initial body temperature, low pH and high serum potassium (s-K⁺) levels were associated with reduced chance of surviving. In patients displaying individual data, overall mean predictive surviving probability (HOPE score; n = 134) was 33.9 ± 33.6% with no significant difference between ECMO and CPB-treated patients. We also surveyed 80 case reports with 96 victims of HCA, who underwent resuscitation with CPB or ECMO, without including them in the meta-analysis. Conclusions: The chance of surviving was significantly higher after rewarming with ECMO, as compared to CPB, and in patients with witnessed compared to unwitnessed HCA. Avalanche victims had the lowest probability of surviving. Male sex, high initial body temperature, low pH, and high s-K⁺ were factors associated with low surviving chances.

Keywords: cardiac arrest; cardiopulmonary bypass; extracorporeal life support; extracorporeal membrane oxygenation; hypothermia; hypothermic cardiac arrest; resuscitation; rewarming.

Figure 1

Search strategy. Systematic literature search of August 2020 in the following databases: Ovid MEDLINE^(R) In-Process & Other Non-Indexed Citations. Ovid MEDLINE^(R) Daily and Ovid MEDLINE^(R) 1946 to Present and Embase Classic+Embase 1947 to Present. Medline subject headings; ti, titles; ab, abstracts; kw, keywords.

Figure 2

Prisma flow-chart displaying the literature search history. Of totally 1,538 references, we included 23 observational studies of patients with hypothermic cardiac arrest, who underwent attempted resuscitation with extracorporeal life support (ECLS) in the meta-analysis. Fifty-one articles lead up to the present review and meta-analysis and/or were used for discussion of our findings. Finally, we addressed 80 case reports.

Figure 3

Odds ratios of different variables from univariate logistic regression with death as outcome. From left to right, results from all patients from whom we had access to individual data, who underwent attempts on resuscitation by means of cardiopulmonary bypass (CPB) or extracorporeal membrane oxygenation (ECMO), respectively. Male sex and initial body temperature are per unit increase, while the rest are per standard deviation increase.

Figure 4

Predictive ability of surviving attempted resuscitation from hypothermic cardiac arrest assessed by ROC curves and AUC by HOPE score and serum K⁺. (A) displays Hope score (red) %. AUC 0.85 (CI 95% 0.78–0.91). (B) shows serum concentration of K⁺ (blue) predicting probability of surviving. AUC 0.79 (CI 95% 0.72–0.86). (C) depicts the difference between the AUC areas in curves (A) (red) and (B) (blue) of 0.056 (P = 0.0426).

Figure 5

Flowchart surveying 80 case reports with 96 victims of hypothermic cardiac arrest (HCA) rewarmed with ECLS. Thirty-nine patients underwent resuscitation with cardiopulmonary bypass (CPB) and 57 patients with extracorporeal membrane oxygenation (ECMO). We distinguished between witnessed and not witnessed hypothermic cardiac arrest (HCA), whether hypothermia was associated with asphyctic incidents, as drowning or avalanche, or non-asphyctic incidents, as immersion in cold water or exposure to cold environments.