Electrocardiographic deep learning for predicting post-procedural mortality: a model development and validation study.

BACKGROUND: Preoperative risk assessments used in clinical practice are insufficient in their ability to identify risk for postoperative mortality. Deep-learning analysis of electrocardiography can identify hidden risk markers that can help to prognosticate postoperative mortality. We aimed to develop a prognostic model that accurately predicts postoperative mortality in patients undergoing medical procedures and who had received preoperative electrocardiographic diagnostic testing. METHODS: In a derivation cohort of preoperative patients with available electrocardiograms (ECGs) from Cedars-Sinai Medical Center (Los Angeles, CA, USA) between Jan 1, 2015 and Dec 31, 2019, a deep-learning algorithm was developed to leverage waveform signals to discriminate postoperative mortality. We randomly split patients (8:1:1) into subsets for training, internal validation, and final algorithm test analyses. Model performance was assessed using area under the receiver operating characteristic curve (AUC) values in the hold-out test dataset and in two external hospital cohorts and compared with the established Revised Cardiac Risk Index (RCRI) score. The primary outcome was post-procedural mortality across three health-care systems. FINDINGS: 45 969 patients had a complete ECG waveform image available for at least one 12-lead ECG performed within the 30 days before the procedure date (59 975 inpatient procedures and 112 794 ECGs): 36 839 patients in the training dataset, 4549 in the internal validation dataset, and 4581 in the internal test dataset. In the held-out internal test cohort, the algorithm discriminates mortality with an AUC value of 0·83 (95% CI 0·79-0·87), surpassing the discrimination of the RCRI score with an AUC of 0·67 (0·61-0·72). The algorithm similarly discriminated risk for mortality in two independent US health-care systems, with AUCs of 0·79 (0·75-0·83) and 0·75 (0·74-0·76), respectively. Patients determined to be high risk by the deep-learning model had an unadjusted odds ratio (OR) of 8·83 (5·57-13·20) for postoperative mortality compared with an unadjusted OR of 2·08 (0·77-3·50) for postoperative mortality for RCRI scores of more than 2. The deep-learning algorithm performed similarly for patients undergoing cardiac surgery (AUC 0·85 [0·77-0·92]), non-cardiac surgery (AUC 0·83 [0·79-0·88]), and catheterisation or endoscopy suite procedures (AUC 0·76 [0·72-0·81]). INTERPRETATION: A deep-learning algorithm interpreting preoperative ECGs can improve discrimination of postoperative mortality. The deep-learning algorithm worked equally well for risk stratification of cardiac surgeries, non-cardiac surgeries, and catheterisation laboratory procedures, and was validated in three independent health-care systems. This algorithm can provide additional information to clinicians making the decision to perform medical procedures and stratify the risk of future complications. FUNDING: National Heart, Lung, and Blood Institute.

Recurrent Myocarditis Treated with Intravenous Immune Globulin and Steroids.

BACKGROUND Myocarditis is an inflammatory process that can present as acute or chronic with either focal or diffuse involvement of the myocardium. Its incidence is approximately 1.5 million cases per year worldwide. In the United States, viral infection is the most common cause of myocarditis. Most of the reported cases are singular and self-limiting in nature. We present the case of severe recurrent myocarditis in a young adult who was transferred to the Intensive Care Unit. CASE REPORT An 18-year-old man presented with chest pressure and troponin I 33 ng/mL. He had presented to another hospital with similar symptoms 3 months prior and was diagnosed with myocarditis that had resolved with colchicine. As part of his workup during this admission, coronary angiogram was normal and biopsy obtained without evidence of an inflammatory process; however, cardiac magnetic resonance imaging (MRI) was consistent with myocarditis and Coxsackie B titers indicated prior infection, leading to a diagnosis of clinically suspected recurrent viral myocarditis. He was treated with intravenous immunoglobulin (IV Ig) and a steroid taper, with rapid improvement in symptoms over the ensuing weeks without evidence of further recurrence or sequelae. CONCLUSIONS We present a case of recurrent Coxsackie B myocarditis based on presentation and imaging. Myocarditis is an important diagnosis to consider when a young, healthy individual presents with chest pain mimicking acute coronary syndrome, especially during the COVID pandemic. If there is evidence of myocarditis on MRI or endomyocardial biopsy, immunosuppressive therapy should be considered in patients with recurrent and severe presentations.

Premature coronary disease, in-stent restenosis and vascular complications in a young man with Behçet syndrome.

A young man with a history of early-onset coronary disease presented with an ST-elevation myocardial infarction at the age of 38. He subsequently had recurrent in-stent restenosis requiring repeating interventions and ultimately bypass surgery. After 4 years, he presents with systemic symptoms, new skin lesions and a femoral artery pseudoaneurysm. He is diagnosed with Behçet syndrome, a rare systemic vasculitis characterised by the triad of oral aphthous ulcers, genital ulcers and ocular involvement. Behçet is not associated with premature coronary disease but can have a variety of cardiac complications. Additionally, pathergy, an exaggerated inflammatory response to local injury, is characteristic. We hypothesise that in retrospect, subclinical inflammation and a vascular pathergy likely predisposed him to his cardiac and vascular complications. Here, we review risk factors and presentation of premature coronary artery disease and review the literature on the cardiovascular complications of Behçet syndrome.

Revision rates and time to revision following endoscopic sinus surgery: A large database analysis.

OBJECTIVES/HYPOTHESIS: Endoscopic sinus surgery (ESS) is performed for patients with chronic rhinosinusitis (CRS) that have failed maximal medical therapy. This study seeks to determine the prevalence of revision surgery and factors predicting the need for revision after ESS using a large statewide surgery database. STUDY DESIGN: Large retrospective cohort study using the State Ambulatory Surgery Database for the state of California between 2005 and 2011. METHODS: We identified over 61,000 patients with CRS who underwent ESS, determined by Current Procedural Terminology code. We identified which patients underwent a repeat surgery, and performed multivariable modeling to determine which factors (nasal polyps, age, gender, insurance, hospital setting, ethnicity) predicted the need for revision. Adjusted odds ratios (AOR) and 95% confidence intervals are presented. RESULTS: Of 61,339 patients who underwent ESS, 4,078 (6.65%) returned for revision ESS during the time period investigated. In a multivariable logistic regression model, positive predictors of revision were a diagnosis of nasal polyps (AOR: 1.20, 95% CI: 1.11-1.29, P < .001) and female gender (AOR: 1.20, 95% CI: 1.11-1.29, P < .001); public insurance was marginally predictive of increased reoperation (AOR: 1.10, 95% CI: 1.00-1.21, P = .048). Patients of Hispanic ethnicity were less likely to have revision surgery (AOR: 0.86, 95% CI: 0.77-0.97, P = .011). Age, income, and hospital setting were not significant predictors. CONCLUSIONS: A minority of patients with CRS who undergo ESS will have a revision surgery. This likelihood is increased in female patients and those with nasal polyps, and decreased in patients of Hispanic ethnicity, even when controlling for income, insurance, and hospital setting. LEVEL OF EVIDENCE: 4. Laryngoscope, 128:31-36, 2018.

Electroencephalographic inverse localization of brain activity in acute traumatic brain injury as a guide to surgery, monitoring and treatment.

OBJECTIVE: To inverse-localize epileptiform cortical electrical activity recorded from severe traumatic brain injury (TBI) patients using electroencephalography (EEG). METHODS: Three acute TBI cases were imaged using computed tomography (CT) and multimodal magnetic resonance imaging (MRI). Semi-automatic segmentation was performed to partition the complete TBI head into 25 distinct tissue types, including 6 tissue types accounting for pathology. Segmentations were employed to generate a finite element method model of the head, and EEG activity generators were modeled as dipolar currents distributed over the cortical surface. RESULTS: We demonstrate anatomically faithful localization of EEG generators responsible for epileptiform discharges in severe TBI. By accounting for injury-related tissue conductivity changes, our work offers the most realistic implementation currently available for the inverse estimation of cortical activity in TBI. CONCLUSION: Whereas standard localization techniques are available for electrical activity mapping in uninjured brains, they are rarely applied to acute TBI. Modern models of TBI-induced pathology can inform the localization of epileptogenic foci, improve surgical efficacy, contribute to the improvement of critical care monitoring and provide guidance for patient-tailored treatment. With approaches such as this, neurosurgeons and neurologists can study brain activity in acute TBI and obtain insights regarding injury effects upon brain metabolism and clinical outcome.