Predicting Out-of-Hospital Cardiac Arrest in the General Population Using Electronic Health Records.

BACKGROUND: The majority of out-of-hospital cardiac arrests (OHCAs) occur among individuals in the general population, for whom there is no established strategy to identify risk. In this study, we assess the use of electronic health record (EHR) data to identify OHCA in the general population and define salient factors contributing to OHCA risk. METHODS: The analytical cohort included 2366 individuals with OHCA and 23 660 age- and sex-matched controls receiving health care at the University of Washington. Comorbidities, electrocardiographic measures, vital signs, and medication prescription were abstracted from the EHR. The primary outcome was OHCA. Secondary outcomes included shockable and nonshockable OHCA. Model performance including area under the receiver operating characteristic curve and positive predictive value were assessed and adjusted for observed rate of OHCA across the health system. RESULTS: There were significant differences in demographic characteristics, vital signs, electrocardiographic measures, comorbidities, and medication distribution between individuals with OHCA and controls. In external validation, discrimination in machine learning models (area under the receiver operating characteristic curve 0.80-0.85) was superior to a baseline model with conventional cardiovascular risk factors (area under the receiver operating characteristic curve 0.66). At a specificity threshold of 99%, correcting for baseline OHCA incidence across the health system, positive predictive value was 2.5% to 3.1% in machine learning models compared with 0.8% for the baseline model. Longer corrected QT interval, substance abuse disorder, fluid and electrolyte disorder, alcohol abuse, and higher heart rate were identified as salient predictors of OHCA risk across all machine learning models. Established cardiovascular risk factors retained predictive importance for shockable OHCA, but demographic characteristics (minority race, single marital status) and noncardiovascular comorbidities (substance abuse disorder) also contributed to risk prediction. For nonshockable OHCA, a range of salient predictors, including comorbidities, habits, vital signs, demographic characteristics, and electrocardiographic measures, were identified. CONCLUSIONS: In a population-based case-control study, machine learning models incorporating readily available EHR data showed reasonable discrimination and risk enrichment for OHCA in the general population. Salient factors associated with OCHA risk were myriad across the cardiovascular and noncardiovascular spectrum. Public health and tailored strategies for OHCA prediction and prevention will require incorporation of this complexity.

Inflammation, endothelial injury, and the acute respiratory distress syndrome after out-of-hospital cardiac arrest.

Background: Acute respiratory distress syndrome (ARDS) is often seen in patients resuscitated from out-of-hospital cardiac arrest (OHCA). We aim to test whether inflammatory or endothelial injury markers are associated with the development of ARDS in patients hospitalized after OHCA. Methods: We conducted a prospective, cohort, pilot study at an urban academic medical center in 2019 that included a convenience sample of adults with non-traumatic OHCA. Blood and pulmonary edema fluid (PEF) were collected within 12 hours of hospital arrival. Samples were assayed for cytokines (interleukin [IL]-1, tumor necrosis factor-α [TNF-α], tumor necrosis factor receptor1 [TNFR1], IL-6), epithelial injury markers (pulmonary surfactant-associated protein D), endothelial injury markers (Angiopoietin-2 [Ang-2] and glycocalyx degradation products), and other proteins (matrix metallopeptidase-9 and myeloperoxidase). Patients were followed for 7 days for development of ARDS, as adjudicated by 3 blinded reviewers, and through hospital discharge for mortality and neurological outcome. We examined associations between biomarker concentrations and ARDS, hospital mortality, and neurological outcome using multivariable logistic regression. Latent phase analysis was used to identify distinct biological classes associated with outcomes. Results: 41 patients were enrolled. Mean age was 58 years, 29% were female, and 22% had a respiratory etiology for cardiac arrest. Seven patients (17%) developed ARDS within 7 days. There were no significant associations between individual biomarkers and development of ARDS in adjusted analyses, nor survival or neurologic status after adjusting for use of targeted temperature management (TTM) and initial cardiac arrest rhythm. Elevated Ang-2 and TNFR-1 were associated with decreased survival (RR = 0.6, 95% CI = 0.3-1.0; RR = 0.5, 95% CI = 0.3-0.9; respectively), and poor neurologic status at discharge (RR = 0.4, 95% CI = 0.2-0.8; RR = 0.4, 95% CI = 0.2-0.9) in unadjusted associations. Conclusion: OHCA patients have markedly elevated plasma and pulmonary edema fluid biomarker concentrations, indicating widespread inflammation, epithelial injury, and endothelial activation. Biomarker concentrations were not associated with ARDS development, though several distinct biological phenotypes warrant further exploration. Latent phase analysis demonstrated that patients with low biomarker levels aside from TNF-α and TNFR-1 (Class 2) fared worse than other patients. Future research may benefit from considering other tools to predict and prevent development of ARDS in this population.

Pediatric Out-of-Hospital Cardiac Arrest: The Role of the Telecommunicator in Recognition of Cardiac Arrest and Delivery of Bystander Cardiopulmonary Resuscitation.

BACKGROUND: Telecommunicator CPR (T-CPR), whereby emergency dispatch facilitates cardiac arrest recognition and coaches CPR over the telephone, is an important strategy to increase early recognition and bystander CPR in adult out-of-hospital cardiac arrest (OHCA). Little is known about this treatment strategy in the pediatric population. We investigated the role of T-CPR and related performance among pediatric OHCA. METHODS AND RESULTS: This study was a retrospective cohort investigation of OHCA among individuals <18 years in King County, Washington, from April 1, 2013, to December 31, 2019. We reviewed the 911 audio recordings to determine if and how bystander CPR was delivered (unassisted or T-CPR), key time intervals in recognition of arrest, and key components of T-CPR delivery. Of the 185 eligible pediatric OHCAs, 23% (n=43) had bystander CPR initiated unassisted, 59% (n=109) required T-CPR, and 18% (n=33) did not receive CPR before emergency medical services arrival. Among all cases, cardiac arrest was recognized by the telecommunicator in 89% (n=165). Among those receiving T-CPR, the median (interquartile range) interval from start of call to OHCA recognition was 59 seconds (38-87) and first CPR intervention was 115 seconds (94-162). When stratified by age (≤8 versus >8), the older age group was less likely to receive CPR before emergency medical services arrival (88% versus 69%, P=0.002). For those receiving T-CPR, bystanders spent a median of 207 seconds (133-270) performing CPR. The median compression rate was 93 per minute (82-107) among those receiving T-CPR. CONCLUSIONS: T-CPR is an important strategy to increase early recognition and early CPR among pediatric OHCA.

A retrospective study on machine learning-assisted stroke recognition for medical helpline calls.

Advanced stroke treatment is time-dependent and, therefore, relies on recognition by call-takers at prehospital telehealth services to ensure fast hospitalisation. This study aims to develop and assess the potential of machine learning in improving prehospital stroke recognition during medical helpline calls. We used calls from 1 January 2015 to 31 December 2020 in Copenhagen to develop a machine learning-based classification pipeline. Calls from 2021 are used for testing. Calls are first transcribed using an automatic speech recognition model and then categorised as stroke or non-stroke using a text classification model. Call-takers achieve a sensitivity of 52.7% (95% confidence interval 49.2-56.4%) with a positive predictive value (PPV) of 17.1% (15.5-18.6%). The machine learning framework performs significantly better (p < 0.0001) with a sensitivity of 63.0% (62.0-64.1%) and a PPV of 24.9% (24.3-25.5%). Thus, a machine learning framework for recognising stroke in prehospital medical helpline calls may become a supportive tool for call-takers, aiding in early and accurate stroke recognition.