Long-term outcomes of hospitalized SARS-CoV-2/COVID-19 patients with and without neurological involvement: 3-year follow-up assessment (preprint)

Background and ObjectivesAcute neurological manifestations are a common complication of acute COVID-19 disease. This study investigated the 3-year outcomes of patients with and without significant neurological manifestations during initial COVID-19 hospitalization. MethodsPatients infected by SARS-CoV-2 between March 1 and April 16, 2020 and hospitalized in the Montefiore Health System in the Bronx, an epicenter of the early pandemic, were included. Follow-up data was captured up to January 23, 2023 (3 years post COVID-19). This cohort consisted of 414 COVID-19 patients with significant neurological manifestations and 1199 propensity-matched COVID- 19 patients without neurological manifestations. Primary outcomes were mortality, stroke, heart attack, major adverse cardiovascular events (MACE), reinfection, and hospital readmission post-discharge. Secondary outcomes were clinical neuroimaging findings (hemorrhage, active stroke, prior stroke, mass effect, and microhemorrhage, white-matter changes, microvascular disease, and volume loss). Predictive models were used to identify risk factors of mortality post-discharge. ResultsMore patients in the neurological cohort were discharged to acute rehabilitation (10.54% vs 3.68%, p<0.0001), skilled nursing facilities (30.67% vs 20.78%, p=0.0002) and fewer to home (55.27% vs 70.21%, p<0.0001) compared to the matched controls. Incidence of readmission for any medical reason (65.70% vs 60.72%, p=0.036), stroke (6.28% vs 2.34%, p<0.0001), and MACE (20.53% vs 16.51%, p=0.032) was higher in the neurological cohort post-discharge. Neurological patients were more likely to die post-discharge (58 (14.01%) vs 94 (7.84%), p=0.0001) compared to controls (HR=2.346, 95% CI=(1.586, 3.470), p<0.0001). The major causes of death post-discharge were heart disease (14.47%), sepsis (13.82%), influenza and pneumonia (11.18%), COVID-19 (8.55%) and acute respiratory distress syndrome (7.89%). Factors associated with mortality after leaving the hospital were belonging to the neurological cohort (OR=1.802 (1.237, 2.608), p=0.002), discharge disposition (OR=1.508, 95% CI=(1.276, 1.775), p<0.0001), congestive heart failure (OR=2.281 (1.429, 3.593), p=0.0004), higher COVID-19 severity score (OR=1.177 (1.062, 1.304), p=0.002), and older age (OR=1.027 (1.010, 1.044), p=0.002). There were no group differences in gross radiological findings, except the neurological cohort showed significantly more age-adjusted brain volume loss (p<0.05) compared to controls. DiscussionCOVID-19 patients with neurological manifestations have worse long-term outcomes compared to matched controls. These findings raise awareness and the need for closer monitoring and timely interventions for COVID-19 patients with neurological manifestations.

Incidence, characteristics, risk factors, and outcomes of diabetic ketoacidosis in COVID-19 patients without prior insulin dependence: comparison with influenza (preprint)

Whether SARS-CoV-2 infection triggers diabetic ketoacidosis (DKA) is unclear. This study characterized incidence, patient profiles, risk factors, and outcomes of in-hospital DKA in COVID-19 patients without prior insulin dependence and compared with influenza. This cohort consisted of 13,383 hospitalized COVID-19 patients (March 2020 to July 2022) and 19,165 hospitalized influenza patients (January 2018 to July 2022) in Bronx, NY. Patients with prior DKA and prior insulin use were excluded. Primary outcomes were in-hospital mortality and new-insulin use 3-month post-infection. The incidence of DKA in hospitalized COVID-19 patients was significantly higher than hospitalized influenza patients (1.4% vs. 0.8%, p < 0.05). COVID-19 patients with DKA were more likely to be intubated, receive steroid treatment, and die (mortality OR = 6.178, p < 0.05) than those without DKA. DKA patients without pre-existing diabetes were more likely to die than DKA patients with pre-existing diabetes (OR = 7.56, p < 0.05). Steroid use, pre-existing type-2 diabetes, and male sex were risk factors for DKA. Patients with DKA had a higher rate of insulin use 3 months post SARS-CoV-2 infection compared to those without DKA (8.2% vs. 1.6%, p < 0.05), suggesting SARS-CoV-2 infection could trigger new insulin dependence. Identification of risk factors for DKA and new insulin-dependency could enable careful monitoring and timely intervention.

COVID-19 patients in the COVID-19 Recovery and Engagement (CORE) Clinics in the Bronx (preprint)

BACKGROUND: Early in the pandemic, we established COVID-19 Recovery and Engagement (CORE) Clinics in the Bronx and implemented a detailed evaluation protocol was implemented to assess physical, emotional, and cognitive function, pulmonary function tests, and imaging for COVID-19 survivors. Here we report our findings five months post-acute COIVD-19. METHODS: Main outcomes and measures included pulmonary function tests, imaging tests, and a battery of symptom, physical, emotional, and cognitive assessments 5 months post-acute COVID-19. FINDINGS: Dyspnea, fatigue, decreased exercise tolerance, brain fog, and shortness of breath were the most common symptoms but there were generally no significant differences between hospitalized and non-hospitalized cohorts (p>0.05). Many patients had abnormal physical, emotional, and cognitive scores, but most functioned independently; there were no significant differences between hospitalized and non-hospitalized cohorts (p>0.05). Six-minute walk test, lung ultrasound, and diaphragm excursion were abnormal but only in the hospitalized cohort. Pulmonary function tests showed moderately restrictive pulmonary function only in the hospitalized cohort but no obstructive pulmonary function. Newly detected major neurological events, microvascular disease, atrophy, and white-matter changes were rare, but lung opacity and fibrosis-like findings were common after acute COVID-19. INTERPRETATION: Many COVID-19 survivors experienced moderately restrictive pulmonary function, and significant symptoms across the physical, emotional, and cognitive health domains. Newly detected brain imaging abnormalities were rare, but lung imaging abnormalities were common. This study provides insights into post-acute sequelae following SARS-CoV-2 infection in neurological and pulmonary systems which may be used to support at-risk patients, develop effective screening methods and interventions.

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

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 5-fold 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 significant prognostic value if these findings can be validated in a large, multi-institutional cohort.

In-Hospital Rehabilitation and Functional Status of Non-Critically ill COVID- 19 Survivors (preprint)

This study investigated in-hospital rehabilitation and functional status at discharge of non-critically ill COVID-19 survivors with respect to pre-admission dependency status, discharge durable medical equipment, discharge medical follow-up recommendation, hospitalization duration, demographics, comorbidities, laboratory tests, and vital signs at hospital discharge. Comparisons were made between COVID-19 survivors who received rehab (N=155) and those who did not (N=162). Functional scores were obtained using the “Mental Status”, ICU Mobility, and modified Barthel Index scores at hospital discharge. Relative to the non-rehab patients, rehab patients were older, had more comorbidities, had worse pre-admission dependency status (p<0.05), were discharged with more assistive equipment and supplemental oxygen (p<0.001), spent more days in the hospital (p<0.001), had more follow-up referrals (p<0.05) with cardiology, vascular medicine, urology, and endocrinology being the top referrals, and had more secondary in-hospital diagnosis of AKI and acute respiratory failure. Functional scores of non-critically ill COVID-19 survivors were impaired at discharge and were associated with pre-admission dependency. Some functional scores were negatively correlated with age, hypertension, coronary artery disease, chronic kidney disease, psychiatric disease, anemia, and neurological disorders (p<0.05). These findings warrant follow up of COVID-19 survivors as many survivors will likely have significant post-acute COVID-19 sequela.

COVID-19 Image Data Collection: Prospective Predictions Are the Future (preprint)

Across the world's coronavirus disease 2019 (COVID-19) hot spots, the need to streamline patient diagnosis and management has become more pressing than ever. As one of the main imaging tools, chest X-rays (CXRs) are common, fast, non-invasive, relatively cheap, and potentially bedside to monitor the progression of the disease. This paper describes the first public COVID-19 image data collection as well as a preliminary exploration of possible use cases for the data. This dataset currently contains hundreds of frontal view X-rays and is the largest public resource for COVID-19 image and prognostic data, making it a necessary resource to develop and evaluate tools to aid in the treatment of COVID-19. It was manually aggregated from publication figures as well as various web based repositories into a machine learning (ML) friendly format with accompanying dataloader code. We collected frontal and lateral view imagery and metadata such as the time since first symptoms, intensive care unit (ICU) status, survival status, intubation status, or hospital location. We present multiple possible use cases for the data such as predicting the need for the ICU, predicting patient survival, and understanding a patient's trajectory during treatment. Data can be accessed here: https://github.com/ieee8023/covid-chestxray-dataset

Predicting COVID-19 Pneumonia Severity on Chest X-ray with Deep Learning (preprint)

Purpose: The need to streamline patient management for COVID-19 has become more pressing than ever. Chest X-rays provide a non-invasive (potentially bedside) tool to monitor the progression of the disease. In this study, we present a severity score prediction model for COVID-19 pneumonia for frontal chest X-ray images. Such a tool can gauge severity of COVID-19 lung infections (and pneumonia in general) that can be used for escalation or de-escalation of care as well as monitoring treatment efficacy, especially in the ICU. Methods: Images from a public COVID-19 database were scored retrospectively by three blinded experts in terms of the extent of lung involvement as well as the degree of opacity. A neural network model that was pre-trained on large (non-COVID-19) chest X-ray datasets is used to construct features for COVID-19 images which are predictive for our task. Results: This study finds that training a regression model on a subset of the outputs from an this pre-trained chest X-ray model predicts our geographic extent score (range 0-8) with 1.14 mean absolute error (MAE) and our lung opacity score (range 0-6) with 0.78 MAE. Conclusions: These results indicate that our model's ability to gauge severity of COVID-19 lung infections could be used for escalation or de-escalation of care as well as monitoring treatment efficacy, especially in the intensive care unit (ICU). A proper clinical trial is needed to evaluate efficacy. To enable this we make our code, labels, and data available online at https://github.com/mlmed/torchxrayvision/tree/master/scripts/covid-severity and https://github.com/ieee8023/covid-chestxray-dataset