Study setting and design

We conducted a prospective, observational, multicenter registry of hypothermia: the Hypothermia study 2018 and 2019. This study was carried out from December 2018 to February 2019 and December 2019 to February 2020; 87 institutions participated in 2018 and 89 in 2019 from various regions in Japan.

This study was approved by the Ethics Review Board of Teikyo University Hospital, Japan (Approval No: 17-090-2). The requirement for informed consent was waived by the Ethics Review Board of Teikyo University Hospital because the study was an observational study. In addition, the review boards of each hospital listed in Appendix S1 approved this study.

Patient selection and data collection

Consecutive patients whose body temperature was below 35°C as measured by emergency medical services (EMSs) or at the emergency department (ED) were included in the study. This study also included patients with cardiopulmonary arrest (CPA) on arrival at the hospital. We excluded patients aged <18 years. The following data were collected: age, sex, Charlson comorbidity index, Sequential Organ Failure Assessment (SOFA) score,13 activities of daily living (ADLs), lifestyle, location, causes underlying the hypothermia (acute medical illness, trauma [submersion, distress], alcohol intoxication, drugs), geographic information, prehospital data, temperature, Glasgow Coma Scale (GCS) score, laboratory data, blood pressure, heart rate, respiratory rate, cardiac arrest during prehospital, tracheal intubation, length of hospital stay, mortality, Cerebral Performance Category (CPC) score on day 30 after admission, and complications. The core temperature from the rectum, bladder, and esophagus was used to record the body temperature, if available; otherwise, the peripheral temperature from the axilla and ears was recorded. We classified the severity of hypothermia into mild (35–32°C), moderate (32–28°C), and severe (<28°C) according to body temperature.

The geographical region was divided into four areas: Northern, Eastern, Western, and Southern areas of Japan. The area was defined by the definition of the Japan Meteorological Agency.14

The prehospital data included the mode of arrival at hospital, body temperature at prehospital, presence or absence of shivering, and rewarming method used at the prehospital site.

The laboratory data measured at the ED consisted of the pH value, potassium level, lactate level, platelet count, prothrombin time-international normalized ratio (PT-INR) level, creatine phosphokinase level, blood urea nitrogen level, and creatinine level. As a rule, the pH value assessed with an arterial blood gas analysis and measured with venous blood gas was adjusted as described in a previous study.15

Complications were categorized as arrhythmia, pneumonia, pancreatitis, electrolyte abnormality, coagulopathy, or other. Pneumonia was defined as an evident shadow on a chest radiograph or computed tomography. Pancreatitis was defined if patients presented with at least two of the following conditions: (i) abdominal or back pain, (ii) elevated pancreatic enzyme levels in the blood, and (iii) pancreatic edema or peripancreatic effusion on ultrasound/computed tomography.

Rewarming methods were divided into active external rewarming (warmed blanket, forced warm air, heating pad, and warmed bath) and active internal rewarming (warmed fluid infusion, lavage, hemodialysis, intravascular catheter, and ECMO).

Outcome measures

The primary outcome in this study was the survival rate after 30 days of admission. The secondary outcomes were the length of intensive care unit stay and hospital stay, CPC at 30 days after admission, and complications. A favorable outcome was defined as a CPC of 1 or 2, whereas an unfavorable outcome was defined as a CPC of 3–5.

Statistical analyses

Categorical variables were expressed as the n (%), and continuous variables were expressed as the median (interquartile range). Comparisons between groups were made by means of Fisher’s exact test for categorical data and Mann–Whitney U test for continuous data. The comparisons between three groups were made by means of the Kruskal–Wallis test for continuous data. Prognostic factors were analyzed using multivariable logistic regression analyses. The following covariates were included in the multivariable model based on the relevant literature,5, 16-18 or the clinically important variables, which included age, sex, ADL, potassium level, causes underlying the hypothermia, GCS, location, temperature, systolic blood pressure, pH value, and PT-INR value. All tests were two-sided, and P values of <0.05 were considered statistically significant. All statistical analyses were performed with the R software program (version, 4.0.3; R Foundation for Statistical Computing, Vienna, Austria).