Survey of Medical Applications of Federated Learning.

OBJECTIVES: Medical artificial intelligence (AI) has recently attracted considerable attention. However, training medical AI models is challenging due to privacy-protection regulations. Among the proposed solutions, federated learning (FL) stands out. FL involves transmitting only model parameters without sharing the original data, making it particularly suitable for the medical field, where data privacy is paramount. This study reviews the application of FL in the medical domain. METHODS: We conducted a literature search using the keywords "federated learning" in combination with "medical," "healthcare," or "clinical" on Google Scholar and PubMed. After reviewing titles and abstracts, 58 papers were selected for analysis. These FL studies were categorized based on the types of data used, the target disease, the use of open datasets, the local model of FL, and the neural network model. We also examined issues related to heterogeneity and security. RESULTS: In the investigated FL studies, the most commonly used data type was image data, and the most studied target diseases were cancer and COVID-19. The majority of studies utilized open datasets. Furthermore, 72% of the FL articles addressed heterogeneity issues, while 50% discussed security concerns. CONCLUSIONS: FL in the medical domain appears to be in its early stages, with most research using open data and focusing on specific data types and diseases for performance verification purposes. Nonetheless, medical FL research is anticipated to be increasingly applied and to become a vital component of multi-institutional research.

Using Real-Time Interaction Analysis to Explore Human-Robot Interaction.

Despite the increasing presence of social robots (SRs) in Human-Robot Interaction, there are few studies that quantify these interactions and explore children's attitudes by analyzing real-time data as they communicate with SRs. Therefore, we attempted to explore the interaction between pediatric patients and SRs by analyzing the interaction log collected from real-time. This study is a retrospective analysis of data collected in a prospective study conducted on 10 pediatric cancer patients at tertiary hospitals in Korea. Using the Wizard of Oz method, we collected the interaction log during the interaction between pediatric cancer patients and the robot. Out of the collected data, 955 sentences from the robot and 332 sentences from the children were available for analysis, except for the logs that were missing due to environmental errors. we analyzed the delay time from saving the interaction log and the sentence similarity of the interaction log. The interaction log delay time between robot and child was 5.01 seconds. And the child's delay time averaged 7.2 seconds, which was longer than the robot's delay time of 4.29 seconds. Additionally, as a result of analyzing the sentence similarity of the interaction log, the robot (97.2%) was higher than the children (46.2%). The results of the sentiment analysis of the patient's attitude toward the robot were 73% neutral, 13.59% positive, and 12.42% negative. The observational evaluations of pediatric psychological experts identified curiosity (n=7, 70.0%), activity (n=5, 50.0%), passivity (n=5, 50.0%), sympathy (n=7, 70.0%), concentration (n=6, 60.0%), high interest (n=5, 50.0%), positive attitude (n=9, 90.0%), and low interaction initiative (n=6, 60.0%). This study made it possible to explore the feasibility of interaction with SRs and to confirm differences in attitudes toward robots according to child characteristics. To increase the feasibility of human-robot interaction, measures such as improving the completeness of log records by enhancing the network environment are required.

Clinical factors predicting return emergency department visits in chemotherapy-induced febrile neutropenia patients.

BACKGROUND: Although chemotherapy-induced febrile neutropenia (FN) is the most common and life-threatening oncologic emergency, the characteristics and outcomes associated with return visits to the emergency department (ED) in these patients are uncertain. Hence, we aimed to investigate the predictive factors and clinical outcomes of chemotherapy-induced FN patients returning to the ED. METHOD: This single-center, retrospective observational study spanning 14 years included chemotherapy-induced FN patients who visited the ED and were discharged. The primary outcome was a return visit to the ED within five days. We conducted logistic regression analyses to evaluate the factors influencing ED return visit. RESULTS: This study included 1318 FN patients, 154 (12.1%) of whom revisited the ED within five days. Patients (53.3%) revisited the ED owing to persistent fever (56.5%), with no intensive care unit admission and only one mortality case who was discharged hopelessly. Multivariable analysis revealed that shock index >0.9 (odds ratio [OR]: 1.45, 95% confidence interval [CI], 1.01-2.10), thrombocytopenia (<100 × 103/uL) (OR: 1.64, 95% CI, 1.11-2.42), and lactic acid level > 2 mmol/L (OR: 1.51, 95% CI, 0.99-2.25) were associated with an increased risk of a return visit to the ED, whereas being transferred into the ED from other hospitals (OR: 0.08; 95% CI, 0.005-0.38) was associated with a decreased risk of a return visit to the ED. CONCLUSION: High shock index, lactic acid, thrombocytopenia, and ED arrival type can predict return visits to the ED in chemotherapy-induced FN patients.

Deep-learning-based personalized prediction of absolute neutrophil count recovery and comparison with clinicians for validation.

Neutropenia and its complications are major adverse effects of cytotoxic chemotherapy. The time to recovery from neutropenia varies from patient to patient, and cannot be easily predicted even by experts. Therefore, we trained a deep learning model using data from 525 pediatric patients with solid tumors to predict the day when patients recover from severe neutropenia after high-dose chemotherapy. We validated the model with data from 99 patients and compared its performance to those of clinicians. The accuracy of the model at predicting the recovery day, with a 1-day error, was 76%; its performance was better than those of the specialist group (58.59%) and the resident group (32.33%). In addition, 80% of clinicians changed their initial predictions at least once after the model's prediction was conveyed to them. In total, 86 prediction changes (90.53%) improved the recovery day estimate.

Identifying the Risk of Sepsis in Patients With Cancer Using Digital Health Care Records: Machine Learning-Based Approach.

BACKGROUND: Sepsis is diagnosed in millions of people every year, resulting in a high mortality rate. Although patients with sepsis present multimorbid conditions, including cancer, sepsis predictions have mainly focused on patients with severe injuries. OBJECTIVE: In this paper, we present a machine learning-based approach to identify the risk of sepsis in patients with cancer using electronic health records (EHRs). METHODS: We utilized deidentified anonymized EHRs of 8580 patients with cancer from the Samsung Medical Center in Korea in a longitudinal manner between 2014 and 2019. To build a prediction model based on physical status that would differ between sepsis and nonsepsis patients, we analyzed 2462 laboratory test results and 2266 medication prescriptions using graph network and statistical analyses. The medication relationships and lab test results from each analysis were used as additional learning features to train our predictive model. RESULTS: Patients with sepsis showed differential medication trajectories and physical status. For example, in the network-based analysis, narcotic analgesics were prescribed more often in the sepsis group, along with other drugs. Likewise, 35 types of lab tests, including albumin, globulin, and prothrombin time, showed significantly different distributions between sepsis and nonsepsis patients (P<.001). Our model outperformed the model trained using only common EHRs, showing an improved accuracy, area under the receiver operating characteristic (AUROC), and F1 score by 11.9%, 11.3%, and 13.6%, respectively. For the random forest-based model, the accuracy, AUROC, and F1 score were 0.692, 0.753, and 0.602, respectively. CONCLUSIONS: We showed that lab tests and medication relationships can be used as efficient features for predicting sepsis in patients with cancer. Consequently, identifying the risk of sepsis in patients with cancer using EHRs and machine learning is feasible.

Use of Gallbladder Width Measurement by Computed Tomography in the Diagnosis of Acute Cholecystitis.

This study aimed to evaluate the diagnostic value of gallbladder width measurement with computed tomography (CT) in patients with acute cholecystitis. This retrospective case−control study was conducted between March 2016 and March 2020 at a tertiary emergency department. Of 310 patients, 254 patients with acute cholecystitis confirmed by surgery were compared with 254 patients diagnosed with other diseases (controls). In the acute cholecystitis group, the number of older patients with underlying illnesses was much higher (64% of men). Upon CT, the median (interquartile range [IQR]) gallbladder width was significantly longer in patients with acute cholecystitis (2.26 [1.82−2.78] cm vs. 3.73 [3.32−4.16] cm, p < 0.001). The optimal cut-off value of gallbladder width for differentiating acute cholecystitis was 3.12 cm, showing a sensitivity of 88% and specificity of 86%. In a multivariable analysis using a logistic regression model for diagnosing acute cholecystitis with CT findings (gallbladder width, length, stone, wall thickening, and pericholecystic fluid), a gallbladder width of ≥3.12 cm was significantly meaningful, even when adjusting for other variables (odds ratio 37.9; p < 0.001). Therefore, an increase in gallbladder width (≥3.12 cm) measured with CT can be a simple and sensitive diagnostic sign of acute cholecystitis, supporting the underlying pathophysiology of bile outflow obstruction.

Echocardiographic Assessment of Patients with Pulmonary Tumor Thrombotic Microangiopathy First Diagnosed in the Emergency Department.

Pulmonary tumor thrombotic microangiopathy (PTTM) is a fatal disease that obstructs pulmonary vessels, leading to pulmonary hypertension (PH) and right-sided heart failure causing rapid progressive dyspnea in patients with cancer. This retrospective chart review involved nine patients with PTTM who were first clinically diagnosed in a tertiary emergency department (ED) between January 2015 and June 2021. They underwent laboratory tests, chest radiography, chest computed tomography (CT), and echocardiography. All patients presented with severe and rapidly progressive dyspnea within a few days, a high oxygen demand. The right ventricle (RV): left ventricle ratio was >1 on chest CT, and no life-threatening pulmonary thromboembolism (PTE) was observed. Echocardiographic findings indicated that all patients had moderate-to-severe RV dilatation with a D-shaped LV. The median tricuspid regurgitation maximum velocity was 3.8 m/s, and the median RV systolic pressure was 63 mmHg, indicating severe PH. The median value of tricuspid annular plane systolic excursion was 15 mm, showing a decrease in RV systolic function, and McConnell's sign was observed in five patients. Two patients immediately underwent chemotherapy and are currently alive. PTTM should be suspected and evaluated using echocardiography in patients with cancer presenting to the ED with acute dyspnea and RV failure without PTE.

Correction: Candidemia Risk Prediction (CanDETEC) Model for Patients With Malignancy: Model Development and Validation in a Single-Center Retrospective Study.

[This corrects the article DOI: 10.2196/24651.].

Candidemia Risk Prediction (CanDETEC) Model for Patients With Malignancy: Model Development and Validation in a Single-Center Retrospective Study.

BACKGROUND: Appropriate empirical treatment for candidemia is associated with reduced mortality; however, the timely diagnosis of candidemia in patients with sepsis remains poor. OBJECTIVE: We aimed to use machine learning algorithms to develop and validate a candidemia prediction model for patients with cancer. METHODS: We conducted a single-center retrospective study using the cancer registry of a tertiary academic hospital. Adult patients diagnosed with malignancies between January 2010 and December 2018 were included. Our study outcome was the prediction of candidemia events. A stratified undersampling method was used to extract control data for algorithm learning. Multiple models were developed-a combination of 4 variable groups and 5 algorithms (auto-machine learning, deep neural network, gradient boosting, logistic regression, and random forest). The model with the largest area under the receiver operating characteristic curve (AUROC) was selected as the Candida detection (CanDETEC) model after comparing its performance indexes with those of the Candida Score Model. RESULTS: From a total of 273,380 blood cultures from 186,404 registered patients with cancer, we extracted 501 records of candidemia events and 2000 records as control data. Performance among the different models varied (AUROC 0.771- 0.889), with all models demonstrating superior performance to that of the Candida Score (AUROC 0.677). The random forest model performed the best (AUROC 0.889, 95% CI 0.888-0.889); therefore, it was selected as the CanDETEC model. CONCLUSIONS: The CanDETEC model predicted candidemia in patients with cancer with high discriminative power. This algorithm could be used for the timely diagnosis and appropriate empirical treatment of candidemia.

A Mobile Phone-Based Self-Monitoring Tool for Perioperative Gastric Cancer Patients With Incentive Spirometer: Randomized Controlled Trial.

BACKGROUND: An incentive spirometer (IS) is a medical device used to help patients improve the functioning of their lungs. It is provided to patients who have had any surgery that might jeopardize respiratory function. An incentive spirometer plays a key role in the prevention of postoperative complications, and the appropriate use of an IS is especially well known for the prevention of respiratory complications. However, IS utilization depends on the patient's engagement, and information and communication technology (ICT) can help in this area. OBJECTIVE: This study aimed to determine the effect of mobile ICT on the usage of an IS (Go-breath) app by postoperative patients after general anesthesia. METHODS: For this study, we recruited patients from April to May 2018, who used the Go-breath app at a single tertiary hospital in South Korea. The patients were randomly classified into either a test or control group. The main function of the Go-breath app was to allow for self-reporting and frequency monitoring of IS use, deep breathing, and active coughing in real time. The Go-breath app was identical for both the test and control groups, except for the presence of the alarm function. The test group heard an alarm every 60 min from 9 am to 9 pm for 2 days. For the test group alone, a dashboard was established in the nurse's station through which a nurse could rapidly assess the performance of multiple patients. To evaluate the number of performances per group, we constructed an incentive spirometer index (ISI). RESULTS: A total of 44 patients were recruited, and 42 of them completed the study protocol. ISI in the test group was 20.2 points higher than that in the control group (113.5 points in the test group and 93.2 points in the control group, P=.22). The system usability scale generally showed almost the same score in the 2 groups (79.3 points in the test group and 79.4 points in the control group, P=.94). We observed that the performance rates of IS count, active coughing, and deep breathing were also higher in the test group but with no statistically significant difference between the groups. For the usefulness "yes or no" question, over 90% (38/42) of patients answered "yes" and wanted more functional options and information. CONCLUSIONS: The use of the Go-breath app resulted in considerable differences between the test group and control group but with no statistically significant differences. TRIAL REGISTRATION: ClinicalTrials.gov NCT03569332; https://clinicaltrials.gov/ct2/show/NCT03569332 (Archived by WebCite at http://www.webcitation.org/74ihKmQIX).

Variability in Doctors' Usage Paths of Mobile Electronic Health Records Across Specialties: Comprehensive Analysis of Log Data.

BACKGROUND: With the emergence of mobile devices, mobile electronic health record (mEHR) systems have been utilized by health care professionals (HCPs), including doctors, nurses, and other practitioners, to improve efficiency at the point of care. Although several studies on mEHR systems were conducted, including those analyzing their effects and HCPs' usage frequency, only a few considered the specific workflows of doctors based on their specialties in which the work process differs greatly. OBJECTIVE: This study aimed to investigate the differences in mEHR usage paths across clinical specialties. METHODS: We collected the log data of 974 doctors who worked from August 2016 to August 2017 and used the mEHR system at the Samsung Medical Center, one of the biggest hospitals in South Korea. The doctors were classified into 3 groups based on their specialty: the physician, the surgeon, and other hospital-based physician (OHBP) groups. We used various descriptive and visualization methods to understand and compare doctors' usage paths of mEHRs. First, the average numbers of log-ins per day and features used per log-in were examined over different specialties and positions. Second, the number of features used by each doctor was visualized via a heat map to provide an overview of mEHR usage across feature types and doctors' specialties. Third, we conducted a path analysis via a Sankey diagram to describe main usage paths and association rule mining to find frequent paths in mEHR usage. RESULTS: The physician group logged on most frequently, whereas the OHBP group logged on least frequently. In fact, the number of log-ins per day of residents in the physician group was 4.4 times higher than that of staff members in the other groups. The heat map visualization showed a visible difference among specialty groups. The physician group used more consultation-related features, whereas the surgeon group used more surgery-related features. Generally, 50% of the doctors spent about 15 seconds at a time when using mEHRs. In the Sankey diagram, the physician group showed diverse usage patterns with higher complexity compared with the other 2 groups; in particular, their paths contained more loops, which reflected repetitive checks on multiple patients. The most frequent path included inpatient summary, which means that most users stopped at the point of summary and did not proceed to view more details. CONCLUSIONS: The usage paths of mEHRs showed considerable differences among the specialty groups. Such differences can be accommodated into an mEHR design to enhance the efficiency of care.

Usefulness of C-MAC video laryngoscope in direct laryngoscopy training in the emergency department: A propensity score matching analysis.

OBJECTIVES: We determined the usefulness of C-MAC video laryngoscope (C-MAC) as a safe training tool for the direct laryngoscopy technique in the emergency department. METHODS: We retrospectively analyzed an institutional airway registry of adult (≥18 years old) patients from April 2014 through October 2016. In this study, the operator used C-MAC as a direct laryngoscope (DL) with limited access to the screen, and the supervisor instructed the operator via verbal feedback while watching the screen. Patients were categorized into the DL group if a conventional DL was used and the C-DL group if a C-MAC used as a DL. RESULTS: Of 744 endotracheal intubations, 163 propensity score-matched pairs were generated (1-to-n matching: C-DL group, 163 vs. DL group, 428). For the propensity-matched groups, the overall first pass success rate was 69%, while those in the C-DL and DL groups were 79% and 65%, respectively. Overall, multiple attempts were required in 8% of patients, with 4% in the C-DL group and 9% in the DL group. The overall complication rate was 11%, with 4% in the C-DL group and 14% in the DL group. In multivariable analysis, the adjusted odds ratios of C-DL use for first pass success, multiple attempts, and complications were 2.05 (95% confidence interval [CI] 1.18-2.87, p < 0.01), 0.38 (95% CI 0.15-0.94; p < 0.01), and 0.28 (95% CI 0.12-0.63; p < 0.01), respectively. CONCLUSIONS: Our study suggests that the C-MAC could be useful for training residents in the direct laryngoscopy while ensuring patient safety in the emergency department.

Experience of Emergency Department Patients With Using the Talking Pole Device: Prospective Interventional Descriptive Study.

BACKGROUND: Patient engagement is important. However, it can be difficult in emergency departments (EDs). OBJECTIVE: The aim of this study was to evaluate the satisfaction of ED patients using a patient-friendly health information technology (HIT) device, the "Talking Pole," and to assess the factors relevant to their satisfaction. METHODS: This study was conducted in May 2017 at the ED of a tertiary hospital. The "Talking Pole" is a smartphone-based device attached to a intravenous infusion pole with sensors. It is capable of sensing patient movement and fluid dynamics. In addition, it provides clinical information from electronic medical records to patients and serves as a wireless communication tool between patients and nurses. Patients and caregivers who entered the observation room of the ED were selected for the study. The "Talking Pole" devices were provided to all participants, regardless of their need for an intravenous pole upon admittance to the ED. After 2 hours, each participant was given an 18-item questionnaire created for this research, measured on a 5-point Likert scale, regarding their satisfaction with "Talking Pole." RESULTS: Among 52 participants recruited, 54% (28/52) were patients and the remaining were caregivers. In total, 38% (20/52) were male participants; the average age was 54.6 (SD 12.9) years, and 63% (33/52) of the participants were oncology patients and their caregivers. The overall satisfaction rate was 4.17 (SD 0.79 ) points. Spearman correlation coefficient showed a strong association of "overall satisfaction" with "comparison to the previous visit" (ρ=.73 ), "perceived benefit" (ρ=.73), "information satisfaction" (ρ=.70), and "efficiency" (ρ=.70). CONCLUSIONS: In this study, we introduced a patient-friendly HIT device, the "Talking Pole." Its architecture focused on enhancing information delivery, which is regarded as a bottleneck toward achieving patient engagement in EDs. Patient and caregiver satisfaction with the "Talking Pole" was positive in the ED environment. In particular, correlation coefficient results improved our understanding about patients' satisfaction, HIT devices, and services used in the ED.

Development and Validation of a Multidisciplinary Mobile Care System for Patients With Advanced Gastrointestinal Cancer: Interventional Observation Study.

BACKGROUND: Mobile health apps have emerged as supportive tools in the management of advanced cancers. However, only a few apps have self-monitoring features, and they are not standardized and validated. OBJECTIVE: This study aimed to develop and validate a multidisciplinary mobile care system with self-monitoring features that can be useful for patients with advanced gastrointestinal cancer. METHODS: The development of the multidisciplinary mobile health management system was divided into 3 steps. First, the service scope was set up, and the measurement tools were standardized. Second, the service flow of the mobile care system was organized. Third, the mobile app (Life Manager) was developed. The app was developed to achieve 3 major clinical goals: support for quality of life, nutrition, and rehabilitation. Three main functional themes were developed to achieve clinical goals: a to-do list, health education, and in-app chat. Thirteen clinically oriented measures were included: the modified Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events questionnaire, Scored Patient-Generated Subjective Global Assessment (PG-SGA), distress, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, International Physical Activity Questionnaire-Short Form, Low anterior resection syndrome score, satisfaction rate, etc. To validate the system, a prospective observational study was conducted. Patients with gastric cancer or colon cancer undergoing chemotherapy were recruited. We followed the subjects for 12 weeks, and selected clinical measures were taken online and offline. RESULTS: After the development process, a multidisciplinary app, the Life Manager, was launched. For evaluation, 203 patients were recruited for the study, of whom 101 (49.8%) had gastric cancer, and 102 (50.2%) were receiving palliative care. Most patients were in their fifties (35.5%), and 128 (63.1%) were male. Overall, 176 subjects (86.7%) completed the study. Among subjects who dropped out, the most common reason was the change of patient's clinical condition (51.9%). During the study period, subjects received multiple health education sessions. For the gastric cancer group, the "general gastric cancer education" was most frequently viewed (322 times), and for the colon cancer group, the "warming-up exercise" was most viewed (340 times). Of 13 measurements taken from subjects, 9 were taken offline (response rate: 52.0% to 90.1%), and 3 were taken online (response rate: 17.6% to 57.4%). The overall satisfaction rate among subjects was favorable and ranged from 3.93 (SD 0.88) to 4.01 (SD 0.87) on the 5-point Likert scale. CONCLUSIONS: A multidisciplinary mobile care system for patients with advanced gastrointestinal cancer was developed with clinically oriented measures. A prospective study was performed for its evaluation, which showed favorable satisfaction.

Efficacy of Mobile Health Care Application and Wearable Device in Improvement of Physical Performance in Colorectal Cancer Patients Undergoing Chemotherapy.

BACKGROUND: The use of a mobile health care application, the delivery of health care or health care-related services through the use of portable devices, to manage functional loss, treatment-related toxicities, and impaired quality of life in cancer patients during chemotherapy through supervised self-management has been increasing. The aim of the present study was to evaluate the efficacy and feasibility of comprehensive mobile health care using a tailored rehabilitation program for colorectal cancer patients undergoing active chemotherapy. PATIENTS AND METHODS: A total of 102 colorectal cancer patients undergoing chemotherapy underwent 12 weeks of smartphone aftercare through provision of a mobile application and wearable device that included a rehabilitation exercise program and information on their disease and treatment. The grip strength test, 30-second chair stand test, 2-minute walk test, amount of physical activity (International Physical Activity Questionnaire short-form), quality of life (European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C30), and nutritional status (Patient-generated Subjective Global Assessment) were assessed and measured at baseline, at mid-intervention (6 weeks), and at completion of the intervention (12 weeks). The rehabilitation exercise intensity was adjusted by the test results at every assessment and through real-time communication between the patients and clinicians. RESULTS: Of the 102 patients, 75 completed all 12 weeks of the smartphone aftercare rehabilitation program. The lower extremity strength (P < .001) and cardiorespiratory endurance (P < .001) was significantly improved. Fatigue (P < .007) and nausea/vomiting (P < .040) symptoms were significantly relieved after the program. CONCLUSION: A tailored rehabilitation exercise program provided through a comprehensive mobile health care application was effective in improving patients' physical capacity and treatment-related symptoms even during active chemotherapy.

Upper airway obstruction resulting from acute mucosal injury induced by direct ingestion of sodium picosulfate/magnesium citrate powder.

A 59-year-old man presented to the emergency department with a chief complaint of sore throat after swallowing sodium picosulfate/magnesium citrate powder for bowel preparation, without first dissolving it in water. The initial evaluation showed significant mucosal injury involving the oral cavity, pharynx, and epiglottis. Endotracheal intubation was performed for airway protection in the emergency department, because the mucosal swelling resulted in upper airway compromise. After conservative treatment in the intensive care unit, he underwent tracheostomy because stenosis of the supraglottic and subglottic areas was not relieved. The tracheostomy tube was successfully removed after confirming recovery, and he was discharged 3 weeks after admission.

Successful fibrinolytic and therapeutic hypothermic management of cardiac arrest following massive pulmonary embolism.

Massive pulmonary embolism (MPE) with hemodynamic instability is a clinical condition with a poor prognosis and high mortality rates. There are no definitive treatment options for cardiac arrest due to MPE. A 52-year-old female presented at our emergency department with cardiac arrest, and a 62-year-old female presented after achieving return of spontaneous circulation of cardiac arrest from a local hospital, respectively. In each case, computed tomographic pulmonary angiography after return of spontaneous circulation demonstrated heavy burdens of pulmonary embolism in the pulmonary arteries. We immediately started therapeutic hypothermia and fibrinolytic therapy. They were transferred to the thoracic surgery and cardiology departments respectively, and then discharged with a cerebral performance categories scale score of 1. In summary, we report two cases of out-of-hospital cardiac arrest due to MPE in which fibrinolytic therapy was successfully combined with therapeutic hypothermia.