Developing and Validating Multi-Modal Models for Mortality Prediction in COVID-19 Patients: a Multi-center Retrospective Study.

The unprecedented global crisis brought about by the COVID-19 pandemic has sparked numerous efforts to create predictive models for the detection and prognostication of SARS-CoV-2 infections with the goal of helping health systems allocate resources. Machine learning models, in particular, hold promise for their ability to leverage patient clinical information and medical images for prediction. However, most of the published COVID-19 prediction models thus far have little clinical utility due to methodological flaws and lack of appropriate validation. In this paper, we describe our methodology to develop and validate multi-modal models for COVID-19 mortality prediction using multi-center patient data. The models for COVID-19 mortality prediction were developed using retrospective data from Madrid, Spain (N = 2547) and were externally validated in patient cohorts from a community hospital in New Jersey, USA (N = 242) and an academic center in Seoul, Republic of Korea (N = 336). The models we developed performed differently across various clinical settings, underscoring the need for a guided strategy when employing machine learning for clinical decision-making. We demonstrated that using features from both the structured electronic health records and chest X-ray imaging data resulted in better 30-day mortality prediction performance across all three datasets (areas under the receiver operating characteristic curves: 0.85 (95% confidence interval: 0.83-0.87), 0.76 (0.70-0.82), and 0.95 (0.92-0.98)). We discuss the rationale for the decisions made at every step in developing the models and have made our code available to the research community. We employed the best machine learning practices for clinical model development. Our goal is to create a toolkit that would assist investigators and organizations in building multi-modal models for prediction, classification, and/or optimization.

Health-related quality of life in nonvalvular atrial fibrillation patients with controlled or uncontrolled anticoagulation status.

BACKGROUND: There is a dearth of evidence regarding Health-Related Quality of Life (HRQoL) in nonvalvular atrial fibrillation (NVAF) patients undergoing oral anticoagulation therapy. Our objective was to describe HRQoL in NVAF patients on oral anticoagulation, focusing on uncontrolled patients on vitamin K antagonists (VKAs) versus controlled patients on VKAs or non-vitamin K antagonist oral anticoagulants (NOACs), in a real-world setting. Additionally, we assessed the clinical characteristics of patients with uncontrolled anticoagulation. METHODS: An observational, multicentre, and cross-sectional study, enrolling 38 Spanish Hospitals' Internal Medicine Departments. HRQoL was assessed using the validated Spanish version of the Sawicki questionnaire. High self-perceived HRQoL was indicated by high scores in the general treatment satisfaction and self-efficacy dimensions, and by low scores in the strained social network, daily hassles and distress dimensions. RESULTS: Five hundred and one patients were included for assessment. Mean scores ± SD were closer to a high perceived HRQoL in controlled than uncontrolled patients for the five dimensions of the questionnaire: 4.9 ± 1.0 versus 3.6 ± 1.3 for general treatment satisfaction; 4.3 ± 1.0 versus 3.6 ± 1.0 for self-efficacy, 3.1 ± 0.9 versus 3.9 ± 1.1 for strained social network, 2.1 ± 0.8 versus 3.0 ± 1.0 for daily hassles and 1.8 ± 0.9 versus 2.6 ± 1.2 for distress. CONCLUSIONS: HRQoL in patients with controlled anticoagulant status treated with NOACs or VKAs was better than in patients with uncontrolled anticoagulant status. This seems to indicate that anticoagulation control status influences perception of HRQoL, highlighting the importance of its evaluation when assessing HRQoL in NVAF patients.

Deficiency in the production of antibodies to lipids correlates with increased lipid metabolism in severe COVID-19 patients.

Background: Antibodies to lipids are part of the first line of defense against microorganisms and regulate the pro/anti-inflammatory balance. Viruses modulate cellular lipid metabolism to enhance their replication, and some of these metabolites are proinflammatory. We hypothesized that antibodies to lipids would play a main role of in the defense against SARS-CoV-2 and thus, they would also avoid the hyperinflammation, a main problem in severe condition patients. Methods: Serum samples from COVID-19 patients with mild and severe course, and control group were included. IgG and IgM to different glycerophospholipids and sphingolipids were analyzed using a high-sensitive ELISA developed in our laboratory. A lipidomic approach for studying lipid metabolism was performed using ultra-high performance liquid chromatography coupled to electrospray ionization and quadrupole time-of-flight mass spectrometry (UHPLC-ESI-QTOF-MS). Results: Mild and severe COVID-19 patients had higher levels of IgM to glycerophosphocholines than control group. Mild COVID-19 patients showed higher levels of IgM to glycerophosphoinositol, glycerophosphoserine and sulfatides than control group and mild cases. 82.5% of mild COVID-19 patients showed IgM to glycerophosphoinositol or glycerophosphocholines plus sulfatides or glycerophosphoserines. Only 35% of severe cases and 27.5% of control group were positive for IgM to these lipids. Lipidomic analysis identify a total of 196 lipids, including 172 glycerophospholipids and 24 sphingomyelins. Increased levels of lipid subclasses belonging to lysoglycerophospholipids, ether and/or vinyl-ether-linked glycerophospholipids, and sphingomyelins were observed in severe COVID-19 patients, when compared with those of mild cases and control group. Conclusion: Antibodies to lipids are essential for defense against SARS-CoV-2. Patients with low levels of anti-lipid antibodies have an elevated inflammatory response mediated by lysoglycerophospholipids. These findings provide novel prognostic biomarkers and therapeutic targets.

Carotid Plaque Burden by 3-Dimensional Vascular Ultrasound as a Risk Marker for Patients with Metabolic Syndrome.

Our aim was to analyse the associations between carotid plaque burden (CPB), cardiovascular risk factors (CVRF), and surrogate markers of CV risk in subjects with metabolic syndrome (MetS). We consecutively included 75 asymptomatic outpatients with MetS components, <60 years old and non-smokers. We determined the presence of CVRF, left ventricular hypertrophy (LVH), carotid intima-media thickness (cIMT), albumin-creatinine ratio (ACR), coronary artery calcium score (CACS) and CPB by 3-dimensional vascular ultrasound (3DVUS) for comparison. A total of 50 (67%) subjects had MetS defined by harmonized criteria. A CPB >0 mm3 and a CACS >0 AU were the risk biomarkers most frequently observed (72% and 77%, respectively), followed by LVH (40%). CPB and CACS revealed association with cardiovascular risk (r = 0.308; p = 0.032 and r = 0.601 p < 0.01, respectively), and CPB also showed association with the burden of CVRF (r = 0.349; p = 0.014). CPB by 3DVUS was a prevalent CV risk marker, directly associated with CVRF and cardiovascular risk in MetS subjects.

Primary pulmonary lymphoma: diagnosis and follow-up of 6 cases and review of an uncommon entity.

AIMS AND BACKGROUND: Primary pulmonary lymphoma is an uncommon disease with a poorly defined management. We reviewed and followed the cases of primary pulmonary lymphoma in our institution to gather an estimation of this entity in our population. DESIGN AND METHODS: We reviewed the records of all patients with biopsy-proven lymphoma of the lung. The main diagnostic criterion for primary pulmonary lymphoma was the absence of extrapulmonary involvement. RESULTS: We identified 6 cases of primary pulmonary lymphoma among 33 patients with biopsy-proven lymphoma of the lung evaluated in our center in a 12-year period. A radiological abnormality in an asymptomatic patient was the most common clinical presentation. Four cases were low-grade and two cases high-grade non-Hodgkin PPL. Histopathologic analyses of lung specimens obtained by transbronchial biopsy were sufficient for a diagnosis in 5 of the 6 cases and avoided invasive surgical maneuvers. Most patients followed an indolent course, but with a tendency to relapse. CONCLUSIONS: Although clinical management of this entity is undefined, we feel bronchoscopic study, which is less aggressive than surgery, may be an adequate procedure for a diagnosis. Mono-chemotherapy using alkylating agents and careful clinical observation may be the best therapeutic approach for these patients, since most of them have favorable outcomes, whatever the treatment modalities.

[Stroke in young adults]. / Ictus en el adulto joven.

Stroke in young adults (15-45 years) is a rare condition. Up to 10% of patients with a first stroke admitted to the hospital belong in this age group. Stroke in the young patient is different from stroke in the elderly in several aspects such as etiology and prognosis. Usually, the management of stroke in young adults warrants an exhaustive etiological work-up. In this article, we review the most relevant issues in the study of young adults who suffer from stroke.

Ictus en el adulto joven / Stroke in young adults

El ictus en el paciente adulto joven (15-45 años) es una entidad poco frecuente pero no excepcional, hasta el 10 por ciento de todos los ictus evaluados en un hospital. Presenta una serie de peculiaridades que lo diferencian del ictus del paciente de más edad desde un punto de vista tanto etiopatogénico y etiológico como del pronóstico. Requiere, por tanto, un abordaje específico e individualizado, generalmente más exhaustivo que el practicado en el ictus del paciente de edad más avanzada. En el presente artículo se revisan los aspectos más relevantes en el tratamiento de los pacientes adultos jóvenes que sufren un ictus (AU)