A hybrid approach based on dynamic trajectories to predict mortality in COVID-19 patients upon steroids administration

Since the World Health Organization (WHO) has declared Artificial Intelligence (AI) as a powerful tool in the fight against COVID-19, multiple studies have been launched aiming to shed light into risk factors for ICU admission and mortality. None of the existing studies, however, have captured the dynamic trajectories of hospitalized COVID-19 patients who receive steroids nor have explored trajectory-based mortality indicators. In this work, we present a novel, hybrid approach to address this need. Latent Growth Mixture Modelling (LGMM) was used to analyze the trajectories of patients who received steroids. The patients were then grouped into clusters based on the similarity of their dynamic trajectories. State-of-the art machine learning classifiers are trained on the original dataset with and without dynamic trajectories to assess whether their inclusion can enhance the prediction of mortality. Our results highlight the importance of trajectories for predicting mortality in patients who receive steroids yielding 4% and 5% increase in the sensitivity (0.84) and specificity (0.85). The FiO2 and percentage of neutrophils at day 5, along with the percentage of lymphocytes at day 7, were identified as the main causes for mortality in patients who receive steroids, where the SatO2 levels showed significant alterations in the dynamic trajectories. © 2022 IEEE.

Association of patient characteristics with clinical outcomes in a cohort of hospitalised patients with SARS-CoV-2 infection in a Greek referral centre for COVID-19.

Patient-important outcomes related to coronavirus disease 2019 (COVID-19) continue to drive the pandemic response across the globe. Various prognostic factors for COVID-19 severity have emerged and their replication across different clinical settings providing health services is ongoing. We aimed to describe the clinical characteristics and their association with outcomes in patients hospitalised with COVID-19 in the University Hospital of Ioannina. We assessed a cohort of 681 consecutively hospitalised patients with COVID-19 from January 2020 to December 2021. Demographic data, underlying comorbidities, clinical presentation, biochemical markers, radiologic findings, COVID-19 treatment and outcome data were collected at the first day of hospitalisation and up to 90 days. Multivariable Cox regression analyses were performed to investigate the associations between clinical characteristics (hazard ratios (HRs) per standard deviation (s.d.)) with intubation and/or mortality status. The participants' mean age was 62.8 (s.d., 16.9) years and 57% were males. The most common comorbidities were hypertension (45%), cardiovascular disease (19%) and diabetes mellitus (21%). Patients usually presented with fever (81%), cough (50%) and dyspnoea (27%), while lymphopenia and increased inflammatory markers were the most common laboratory abnormalities. Overall, 55 patients (8%) were intubated, and 86 patients (13%) died. There were statistically significant positive associations between intubation or death with age (HR: 2.59; 95% CI 1.52-4.40), lactate dehydrogenase (HR: 1.44; 95% CI 1.04-1.98), pO2/FiO2 ratio < 100 mmHg (HR: 3.52; 95% CI 1.14-10.84), and inverse association with absolute lymphocyte count (HR: 0.54; 95% CI 0.33-0.87). These data might help to identify points for improvement in the management of COVID-19 patients.

Baseline characteristics and outcome predictors in patients hospitalized with COVID-19 in the University General Hospital of Ioannina, Greece

INTRODUCTION The novel Severe Acute Respiratory Syndrome Coronavirus-2, which causes the coronavirus disease COVID-19, is a highly infectious viral pathogen that is responsible for the ongoing pandemic. The aim of the present study was to illustrate the pre-hospitalization baseline characteristics and comorbidities of patients admitted with COVID-19 and their association with patient outcomes. METHODS This was a retrospective observational study of consecutive patients who were admitted to the COVID-19 departments of the University General Hospital of Ioannina, Greece (March 2020 - August 2021). Patients' demographic data, chronic disease medication use, and comorbidities were recorded upon their admission. RESULTS A total of 627 patients were hospitalized with mean age 62.5 years, 65.2% with at least one comorbidity, and 43.1% female. The median hospitalization duration was 11 days;554 (88.4%) of the patients were discharged and the mortality rate was 11.6%. Older age, admission during the second pandemic wave, arterial hypertension, and diabetes mellitus were associated with longer hospitalization. In multivariate analyses, cardiovascular disease was an independent predictor of hospitalization length (OR=1.834;95% CI: 1.039-3.228), whereas age (HR=1.079;95% CI: 1.045-1.115), history of malignancy (HR=1.246;95% CI: 1.002-1.595), and a diagnosis of COPD (HR=1.989;95% CI: 1.025-7.999) remained the independent mortality predictors. CONCLUSIONS Our data highlight the effect of COPD and malignancy on mortality risk in COVID-19 patients and the association of cardiovascular disease with a longer hospitalization.