Deep learning of longitudinal chest X-ray and clinical variables predicts duration on ventilator and mortality in COVID-19 patients.

OBJECTIVES: To use deep learning of serial portable chest X-ray (pCXR) and clinical variables to predict mortality and duration on invasive mechanical ventilation (IMV) for Coronavirus disease 2019 (COVID-19) patients. METHODS: This is a retrospective study. Serial pCXR and serial clinical variables were analyzed for data from day 1, day 5, day 1-3, day 3-5, or day 1-5 on IMV (110 IMV survivors and 76 IMV non-survivors). The outcome variables were duration on IMV and mortality. With fivefold cross-validation, the performance of the proposed deep learning system was evaluated by receiver operating characteristic (ROC) analysis and correlation analysis. RESULTS: Predictive models using 5-consecutive-day data outperformed those using 3-consecutive-day and 1-day data. Prediction using data closer to the outcome was generally better (i.e., day 5 data performed better than day 1 data, and day 3-5 data performed better than day 1-3 data). Prediction performance was generally better for the combined pCXR and non-imaging clinical data than either alone. The combined pCXR and non-imaging data of 5 consecutive days predicted mortality with an accuracy of 85 ± 3.5% (95% confidence interval (CI)) and an area under the curve (AUC) of 0.87 ± 0.05 (95% CI) and predicted the duration needed to be on IMV to within 2.56 ± 0.21 (95% CI) days on the validation dataset. CONCLUSIONS: Deep learning of longitudinal pCXR and clinical data have the potential to accurately predict mortality and duration on IMV in COVID-19 patients. Longitudinal pCXR could have prognostic value if these findings can be validated in a large, multi-institutional cohort.

Prone Chest Radiographs: Distinguishing Features and Identification of Support Devices.

PURPOSE: Prone position is known to improve acute lung injury, and chest radiographs are often necessary to monitor disease and confirm support device placement. However, there is a paucity of literature regarding radiographs obtained in this position. We evaluated prone radiographs for distinguishing features and ability to identify support devices. METHODS: Pairs of prone and supine radiographs obtained during the COVID-19 pandemic were assessed retrospectively. IRB approval and waiver of informed consent were obtained. Radiographs were assessed for imaging adequacy, distinguishing features, and support device identification (endotracheal tube, enteric tube, or central line). Radiographs were reviewed by ≥ 2 cardiothoracic radiologists. RESULTS: Radiographs from 81 patients (63yo ± 13, 30% women) were reviewed. Prone and supine radiographs were comparable for imaging the lung bases (81% vs. 90%, p = 0.35) and apices (93% vs. 94%, p = 1); prone radiographs more frequently had significant rotation (36% vs. 19%, p = 0.021). To identify prone technique, scapula tip located beyond the rib border was 89% sensitive (95%CI 80-95%) and 85% specific (76-92%), and a fundal stomach bubble was 44% sensitive (33-56%) and 90% specific (81-96%). For women, displaced breast shadow was 46% sensitive (26-67%) and 92% specific (73-99%). Prone and supine radiographs each identified > 99% of support devices. Prone exams trended toward increased rate of malpositioned device (12% vs. 6%, p = 0.07). CONCLUSION: Scapula position reliably distinguishes prone from supine position; fundal stomach bubble or displaced breast shadow is specific for prone position. Prone radiographs reliably identify line and tube position, which is particularly important as prone patients appear at increased risk for malpositioned devices.

Epidemiology, Risk Factors and Outcomes of Pneumomediastinum in Patients with Coronavirus Disease 2019: A Case-Control Study.

Background: Since the beginning of the ongoing Coronavirus Disease 2019 (COVID-19) pandemic, pneumomediastinum has been reported in patients with COVID-19 pneumonia and acute respiratory distress syndrome. It has been suggested that pneumomediastinum may portend a worse outcome in such patients although no investigation has established this association definitively. Research Question: We hypothesized that the finding of pneumomediastinum in the setting of COVID-19 disease may be associated with a worse clinical outcome. The purpose of this study was to determine if the presence of pneumomediastinum was predictive of increased mortality in patients with COVID-19. Study Design and Methods: A retrospective case-control study utilizing clinical data and imaging for COVID-19 patients seen at our institution from 3/7/2020 to 5/20/2020 was performed. 87 COVID-19 positive patients with pneumomediastinum were compared to 87 COVID-19 positive patients without pneumomediastinum and to a historical group of patients with pneumomediastinum during the same time frame in 2019. Results: The incidence of pneumomediastinum was increased more than 6-fold during the COVID-19 pandemic compared to 2019 (P = <.001). 1.5% of all COVID-19 patients and 11% of mechanically ventilated COVID-19 patients at our institution developed pneumomediastinum. Patients who developed pneumomediastinum had a significantly higher PEEP and lower P/F ratio than those who did not (P = .002 and .033, respectively). Pneumomediastinum was not found to be associated with increased mortality (P = .16, confidence interval [CI]: 0.89-2.09, 1.37). The presence of concurrent pneumothorax at the time of pneumomediastinum diagnosis was associated with increased mortality (P = .013 CI: 1.15-3.17, 1.91). Conclusion: Pneumomediastinum is not independently associated with a worse clinical prognosis in COVID-19 positive patients. The presence of concurrent pneumothorax was associated with increased mortality.

Incidence of new-onset atrial fibrillation in COVID-19 is associated with increased epicardial adipose tissue.

PURPOSE: Coronary artery calcium (CAC) and epicardial adipose tissue (EAT) can predict AF in the general population. We aimed to determine if CAC and EAT measured by computed tomographic (CT) scanning can predict new-onset AF in patients admitted with COVID-19 disease. METHODS: We performed a retrospective, post hoc analysis of all patients admitted to Montefiore Medical Center with a confirmed COVID-19 diagnosis from March 1st to June 23rd, 2020, who had a non-contrast CT of the chest within 5 years prior to admission. We determined ordinal CAC scores and quantified the EAT volume and examined their relationship with inpatient mortality. RESULTS: A total of 379 patients were analyzed. There were 16 events of new-onset AF (4.22%). Patients who developed AF during the index admission were more likely to be male (75 vs 47%, p < 0.001) and had higher EAT (129.5 [76.3-197.3] vs 91.0 [60.0-129.0] ml, p = 0.049). There were no differences on age (68 [56-71] vs 68 [58-76] years; p = 0.712), BMI (28.5 [25.3-30.8] vs 26.9 [23.1-31.8] kg/m2; p = 0.283), ordinal CAC score (3 [1-6] vs 2 [0-4]; p = 0.482), or prevalence of diabetes (56.3 vs 60.1%; p = 0.761), hypertension (75.0 vs 87.3%, p = 0.153), or coronary artery disease (50.0 vs 39.4%, p = 0.396). Patients with new-onset AF had worse clinical outcomes (death/intubation/vasopressors) (87.5 vs 44.1%; p = 0.001). CONCLUSION: Increased EAT measured by non-contrast chest CT identifies patients hospitalized with COVID-19 at higher risk of developing new-onset AF. Patients with new-onset AF have worse clinical outcomes.

Coronary artery calcification and epicardial adipose tissue as independent predictors of mortality in COVID-19.

Recent epidemiological studies have demonstrated that common cardiovascular risk factors are strongly associated with adverse outcomes in COVID-19. Coronary artery calcium (CAC) and epicardial fat (EAT) have shown to outperform traditional risk factors in predicting cardiovascular events in the general population. We aim to determine if CAC and EAT determined by Computed Tomographic (CT) scanning can predict all-cause mortality in patients admitted with COVID-19 disease. We performed a retrospective, post-hoc analysis of all patients admitted to Montefiore Medical Center with a confirmed COVID-19 diagnosis from March 1st, 2020 to May 2nd, 2020 who had a non-contrast CT of the chest within 5 years prior to admission. We determined ordinal CAC scores and quantified the epicardial (EAT) and thoracic (TAT) fat volume and examined their relationship with inpatient mortality. A total of 493 patients were analyzed. There were 197 deaths (39.95%). Patients who died during the index admission had higher age (72, [64-80] vs 68, [57-76]; p < 0.001), CAC score (3, [0-6] vs 1, [0-4]; p < 0.001) and EAT (107, [70-152] vs 94, [64-129]; p = 0.023). On a competing risk analysis regression model, CAC ≥ 4 and EAT ≥ median (98 ml) were independent predictors of mortality with increased mortality of 63% (p = 0.003) and 43% (p = 0.032), respectively. As a composite, the group with a combination of CAC ≥ 4 and EAT ≥ 98 ml had the highest mortality. CAC and EAT measured from chest CT are strong independent predictors of inpatient mortality from COVID-19 in this high-risk cohort.

Echocardiography Abnormal Findings and Laboratory Operations during the COVID-19 Pandemic at a High Volume Center in New York City.

(1) Background: This study sought to explore how the novel coronavirus (COVID-19) pandemic affected the echocardiography (TTE) laboratory operations at a high volume medical center in New York City. Changes in cardiac imaging study volume, turn-around time, and abnormal findings were analyzed and compared to a pre-pandemic period. (2) Methods: Volume of all cardiac imaging studies and TTE reports between 11 March 2020 to 5 May 2020 and the same calendar period in 2019 were retrospectively identified and compared. (3) Results: During the pandemic, our center experienced a 46.72% reduction in TTEs, 82.47% reduction in transesophageal echocardiograms, 83.16% reduction in stress echo, 70.32% reduction in nuclear tests, 46.25% reduction in calcium score, 73.91% reduction in coronary computed tomography angiography, and 87.23% reduction in cardiac magnetic resonance imaging. TTE findings were overall similar between 2020 and 2019 (all p ≥ 0.05), except for a significantly higher right ventricular systolic pressure in 2020 (39.8 ± 14.2 vs. 34.6 ± 11.2 mmHg, p = 0.012). (4) Conclusions: Despite encountering an influx of critically ill patients, our hospital center experienced a reduction in the number of cardiac imaging studies, which likely represents a change in both patient mindset and physician management approach.