Machine learning methods to predict mechanical ventilation and mortality in patients with COVID-19.

BACKGROUND: The Coronavirus disease 2019 (COVID-19) pandemic has affected millions of people across the globe. It is associated with a high mortality rate and has created a global crisis by straining medical resources worldwide. OBJECTIVES: To develop and validate machine-learning models for prediction of mechanical ventilation (MV) for patients presenting to emergency room and for prediction of in-hospital mortality once a patient is admitted. METHODS: Two cohorts were used for the two different aims. 1980 COVID-19 patients were enrolled for the aim of prediction ofMV. 1036 patients' data, including demographics, past smoking and drinking history, past medical history and vital signs at emergency room (ER), laboratory values, and treatments were collected for training and 674 patients were enrolled for validation using XGBoost algorithm. For the second aim to predict in-hospital mortality, 3491 hospitalized patients via ER were enrolled. CatBoost, a new gradient-boosting algorithm was applied for training and validation of the cohort. RESULTS: Older age, higher temperature, increased respiratory rate (RR) and a lower oxygen saturation (SpO2) from the first set of vital signs were associated with an increased risk of MV amongst the 1980 patients in the ER. The model had a high accuracy of 86.2% and a negative predictive value (NPV) of 87.8%. While, patients who required MV, had a higher RR, Body mass index (BMI) and longer length of stay in the hospital were the major features associated with in-hospital mortality. The second model had a high accuracy of 80% with NPV of 81.6%. CONCLUSION: Machine learning models using XGBoost and catBoost algorithms can predict need for mechanical ventilation and mortality with a very high accuracy in COVID-19 patients.

Novel Case Report: A Previously Reported, but Pathophysiologically Unexplained, Association Between Collagenous Colitis and Protein-Losing Enteropathy May Be Explained by an Undetected Link with Collagenous Duodenitis.

Collagenous colitis (CC) is associated with non-bloody, watery diarrhea, which is pathophysiologically reasonable because normal colonic absorption (or excretion) of water and electrolytes can be blocked by the abnormally thick collagen layer in CC. However, CC has also been associated with six previous cases of protein-losing enteropathy (PLE), with no pathophysiologic explanation. The colon does not normally absorb (or excrete) amino acids/proteins, which is primarily the function of the small bowel. Collagenous duodenitis (CD) has not been associated with PLE. This work reports a novel case of CD (and CC) associated with PLE; a pathophysiologically reasonable mechanism for CD causing PLE (by the thick collagen layer of CD blocking normal intestinal amino acid absorption); and a novel association of PLE with severe COVID-19 infection (attributed to relative immunosuppression from hypoproteinemia, hypoalbuminemia, hypogammaglobulinemia, and malnutrition from PLE).

Fatal central nervous system co-infection with SARS-CoV-2 and tuberculosis in a healthy child.

BACKGROUND: Central and peripheral nervous system symptoms and complications are being increasingly recognized among individuals with pandemic SARS-CoV-2 infections, but actual detection of the virus or its RNA in the central nervous system has rarely been sought or demonstrated. Severe or fatal illnesses are attributed to SARS-CoV-2, generally without attempting to evaluate for alternative causes or co-pathogens. CASE PRESENTATION: A five-year-old girl with fever and headache was diagnosed with acute SARS-CoV-2-associated meningoencephalitis based on the detection of its RNA on a nasopharyngeal swab, cerebrospinal fluid analysis, and magnetic resonance imaging findings. Serial serologic tests for SARS-CoV-2 IgG and IgA showed seroconversion, consistent with an acute infection. Mental status and brain imaging findings gradually worsened despite antiviral therapy and intravenous dexamethasone. Decompressive suboccipital craniectomy for brain herniation with cerebellar biopsy on day 30 of illness, shortly before death, revealed SARS-CoV-2 RNA in cerebellar tissue using the Centers for Disease Control and Prevention 2019-nCoV Real-Time Reverse Transcriptase-PCR Diagnostic Panel. On histopathology, necrotizing granulomas with numerous acid-fast bacilli were visualized, and Mycobacterium tuberculosis complex DNA was detected by PCR. Ventricular cerebrospinal fluid that day was negative for mycobacterial DNA. Tracheal aspirate samples for mycobacterial DNA and culture from days 22 and 27 of illness were negative by PCR but grew Mycobacterium tuberculosis after 8 weeks, long after the child's passing. She had no known exposures to tuberculosis and no chest radiographic findings to suggest it. All 6 family members had normal chest radiographs and negative interferon-γ release assay results. The source of her tuberculous infection was not identified, and further investigations by the local health department were not possible because of the State of Michigan-mandated lockdown for control of SARS-CoV-2 spread. CONCLUSION: The detection of SARS-CoV-2 RNA in cerebellar tissue and the demonstration of seroconversion in IgG and IgA assays was consistent with acute SARS-CoV-2 infection of the central nervous infection. However, the cause of death was brain herniation from her rapidly progressive central nervous system tuberculosis. SARS-CoV-2 may mask or worsen occult tuberculous infection with severe or fatal consequences.