COVID-19 is responsible for high mortality, but robust machine learning-based predictors of mortality are lacking. To generate a model for predicting mortality in patients hospitalized with COVID-19 ...using Gradient Boosting Decision Trees (GBDT). The Spanish SEMI-COVID-19 registry includes 24,514 pseudo-anonymized cases of patients hospitalized with COVID-19 from 1 February 2020 to 5 December 2021. This registry was used as a GBDT machine learning model, employing the CatBoost and BorutaShap classifier to select the most relevant indicators and generate a mortality prediction model by risk level, ranging from 0 to 1. The model was validated by separating patients according to admission date, using the period 1 February to 31 December 2020 (first and second waves, pre-vaccination period) for training, and 1 January to 30 November 2021 (vaccination period) for the test group. An ensemble of ten models with different random seeds was constructed, separating 80% of the patients for training and 20% from the end of the training period for cross-validation. The area under the receiver operating characteristics curve (AUC) was used as a performance metric. Clinical and laboratory data from 23,983 patients were analyzed. CatBoost mortality prediction models achieved an AUC performance of 84.76 (standard deviation 0.45) for patients in the test group (potentially vaccinated patients not included in model training) using 16 features. The performance of the 16-parameter GBDT model for predicting COVID-19 hospital mortality, although requiring a relatively large number of predictors, shows a high predictive capacity.
Abstract
Background
Advanced age is a well-known risk factor for poor prognosis in COVID-19. However, few studies have specifically focused on very old inpatients with COVID-19. This study aims to ...describe the clinical characteristics of very old inpatients with COVID-19 and identify risk factors for in-hospital mortality at admission.
Methods
We conducted a nationwide, multicenter, retrospective, observational study in patients ≥ 80 years hospitalized with COVID-19 in 150 Spanish hospitals (SEMI-COVID-19) Registry (March 1–May 29, 2020). The primary outcome was in-hospital mortality. A uni- and multivariate logistic regression was performed to assess predictors of mortality at admission.
Results
A total of 2772 consecutive patients (49.4% men, median age 86.3 years) were analyzed. Rates of atherosclerotic cardiovascular disease, diabetes mellitus, dementia, and Barthel Index < 60 were 30.8%, 25.6%, 30.5%, and 21.0%, respectively. The overall case-fatality rate was 46.9% (n: 1301) and increased with age (80–84 years: 41.6%; 85–90 years: 47.3%; 90–94 years: 52.7%; ≥95 years: 54.2%). After analysis, male sex and moderate-to-severe dependence were independently associated with in-hospital mortality; comorbidities were not predictive. At admission, independent risk factors for death were: oxygen saturation < 90%; temperature ≥ 37.8°C; quick sequential organ failure assessment (qSOFA) score ≥ 2; and unilateral–bilateral infiltrates on chest x-rays. Some analytical findings were independent risk factors for death, including estimated glomerular filtration rate < 45 mL/min/1.73 m2; lactate dehydrogenase ≥ 500 U/L; C-reactive protein ≥ 80 mg/L; neutrophils ≥ 7.5 × 103/μL; lymphocytes < 0.8 × 103/μL; and monocytes < 0.5 × 103/μL.
Conclusions
This first large, multicenter cohort of very old inpatients with COVID-19 shows that age, male sex, and poor preadmission functional status—not comorbidities—are independently associated with in-hospital mortality. Severe COVID-19 at admission is related to poor prognosis.
Background:
The emergence of new SARS-CoV-2 variants with significant immune-evasiveness, the relaxation of measures for reducing the number of infections, the waning of immune protection ...(particularly in high-risk population groups), and the low uptake of new vaccine boosters, forecast new waves of hospitalizations and admission to intensive care units. There is an urgent need for easily implementable and clinically effective Early Warning Scores (EWSs) that can predict the risk of complications within the next 24–48 hr. Although EWSs have been used in the evaluation of COVID-19 patients, there are several clinical limitations to their use. Moreover, no models have been tested on geographically distinct populations or population groups with varying levels of immune protection.
Methods:
We developed and validated COVID-19 Early Warning Score (COEWS), an EWS that is automatically calculated solely from laboratory parameters that are widely available and affordable. We benchmarked COEWS against the widely used NEWS2. We also evaluated the predictive performance of vaccinated and unvaccinated patients.
Results:
The variables of the COEWS predictive model were selected based on their predictive coefficients and on the wide availability of these laboratory variables. The final model included complete blood count, blood glucose, and oxygen saturation features. To make COEWS more actionable in real clinical situations, we transformed the predictive coefficients of the COEWS model into individual scores for each selected feature. The global score serves as an easy-to-calculate measure indicating the risk of a patient developing the combined outcome of mechanical ventilation or death within the next 48 hr.
The discrimination in the external validation cohort was 0.743 (95% confidence interval CI: 0.703–0.784) for the COEWS score performed with coefficients and 0.700 (95% CI: 0.654–0.745) for the COEWS performed with scores. The area under the receiver operating characteristic curve (AUROC) was similar in vaccinated and unvaccinated patients. Additionally, we observed that the AUROC of the NEWS2 was 0.677 (95% CI: 0.601–0.752) in vaccinated patients and 0.648 (95% CI: 0.608–0.689) in unvaccinated patients.
Conclusions:
The COEWS score predicts death or MV within the next 48 hr based on routine and widely available laboratory measurements. The extensive external validation, its high performance, its ease of use, and its positive benchmark in comparison with the widely used NEWS2 position COEWS as a new reference tool for assisting clinical decisions and improving patient care in the upcoming pandemic waves.
Funding:
University of Vienna.
The aim of this study was to analyze whether the coronavirus disease 2019 (COVID‐19) vaccine reduces mortality in patients with moderate or severe COVID‐19 disease requiring oxygen therapy. A ...retrospective cohort study, with data from 148 hospitals in both Spain (111 hospitals) and Argentina (37 hospitals), was conducted. We evaluated hospitalized patients for COVID‐19 older than 18 years with oxygen requirements. Vaccine protection against death was assessed through a multivariable logistic regression and propensity score matching. We also performed a subgroup analysis according to vaccine type. The adjusted model was used to determine the population attributable risk. Between January 2020 and May 2022, we evaluated 21,479 COVID‐19 hospitalized patients with oxygen requirements. Of these, 338 (1.5%) patients received a single dose of the COVID‐19 vaccine and 379 (1.8%) were fully vaccinated. In vaccinated patients, mortality was 20.9% (95% confidence interval CI: 17.9–24), compared to 19.5% (95% CI: 19–20) in unvaccinated patients, resulting in a crude odds ratio (OR) of 1.07 (95% CI: 0.89–1.29; p = 0.41). However, after considering the multiple comorbidities in the vaccinated group, the adjusted OR was 0.73 (95% CI: 0.56–0.95; p = 0.02) with a population attributable risk reduction of 4.3% (95% CI: 1–5). The higher risk reduction for mortality was with messenger RNA (mRNA) BNT162b2 (Pfizer) (OR 0.37; 95% CI: 0.23–0.59; p < 0.01), ChAdOx1 nCoV‐19 (AstraZeneca) (OR 0.42; 95% CI: 0.20–0.86; p = 0.02), and mRNA‐1273 (Moderna) (OR 0.68; 95% CI: 0.41–1.12; p = 0.13), and lower with Gam‐COVID‐Vac (Sputnik) (OR 0.93; 95% CI: 0.6–1.45; p = 0.76). COVID‐19 vaccines significantly reduce the probability of death in patients suffering from a moderate or severe disease (oxygen therapy).
Background
The WHO ordinal severity scale has been used to predict mortality and guide trials in COVID-19. However, it has its limitations.
Objective
The present study aims to compare three ...classificatory and predictive models: the WHO ordinal severity scale, the model based on inflammation grades, and the hybrid model.
Design
Retrospective cohort study with patient data collected and followed up from March 1, 2020, to May 1, 2021, from the nationwide SEMI-COVID-19 Registry. The primary study outcome was in-hospital mortality. As this was a hospital-based study, the patients included corresponded to categories 3 to 7 of the WHO ordinal scale. Categories 6 and 7 were grouped in the same category.
Key Results
A total of 17,225 patients were included in the study. Patients classified as high risk in each of the WHO categories according to the degree of inflammation were as follows: 63.8% vs. 79.9% vs. 90.2% vs. 95.1% (
p
<0.001). In-hospital mortality for WHO ordinal scale categories 3 to 6/7 was as follows: 0.8% vs. 24.3% vs. 45.3% vs. 34% (
p
<0.001). In-hospital mortality for the combined categories of ordinal scale 3a to 5b was as follows: 0.4% vs. 1.1% vs. 11.2% vs. 27.5% vs. 35.5% vs. 41.1% (
p
<0.001). The predictive regression model for in-hospital mortality with our proposed combined ordinal scale reached an AUC=0.871, superior to the two models separately.
Conclusions
The present study proposes a new severity grading scale for COVID-19 hospitalized patients. In our opinion, it is the most informative, representative, and predictive scale in COVID-19 patients to date.
Current research on the association between dietary patterns and subclinical atherosclerotic disease (SAD) is still limited, and published results are inconsistent and often consist of small ...population sizes. We aimed to evaluate the association between the Mediterranean diet (MDiet) and SAD in a large cohort of Mediterranean individuals.
This was a cross-sectional study that included 8116 subjects from the ILERVAS cohort. The presence of atherosclerotic plaques (AP) was assessed by ultrasound examination. Adherence to the MDiet was assessed using the 14-item Mediterranean Diet Adherence Score (MEDAS). Inclusion criteria were subjects with at least one cardiovascular risk factor. Exclusion criteria were a clinical history of diabetes, chronic kidney disease, or a prior cardiovascular event. Bivariable and multivariable models were performed.
Compared with subjects without SAD, participants with SAD were older and had a higher frequency of smoking habit, hypertension, dyslipidemia, HbA1c and waist circumference. The adjusted multivariable analysis showed that a higher MEDAS was associated with a lower risk of AP (incidence rate ratios IRR 0.97, 95% CI 0.96–0.98; p<0.001). Furthermore, moderate or high adherence to the MDiet was associated with a lower number of AP compared with a low MDiet adherence (IRR 0.90, 95% CI 0.87–0.94; p<0.001). In both models, female sex was associated with a lower risk of AP.
Our findings point to a potentially protective role of MDiet for SAD in a Mediterranean population with low-to-moderate cardiovascular risk. Further research is needed to establish a causal relationship between both variables.
Display omitted
•Higher adherence to a Mediterranean diet is related to fewer atherosclerotic plaques.•Subjects with multiple plaques had a more unfavorable lipid and metabolic profile.•Women presented a lower frequency and number of atherosclerotic plaques.
To analyze the ability of medical students to be integrated in the teaching of basic abdominal ultrasound using a peer-mentoring design.
Thirty medical students previously trained in basic abdominal ...ultrasound (mentors) had to teach all fourth-year students (n = 136) from a single academic year the same training they had received. There were 3 stages to the ultrasound teaching: theoretical (online course); basic training (3 practical sessions in which students were guaranteed to have had a minimum of 15 hours of practical experience with ultrasound and performed at least 20 basic abdominal ultrasound studies); and evaluation (objective structured clinical examination in which students had to obtain the basic abdominal views and to identify 17 structures).
The mean grade ± SD obtained was 8.71 ± 1.53 of a possible 10 points. Only 2 students (1.56%) obtained a grade lower than 5, and 14 students (10.86%) obtained a grade lower than 7. A total of 33 students (25.5%) achieved the maximum grade. The structures most easily identified were the liver, the right kidney, and the urinary bladder, with 97.7% of correct answers. Students obtained the poorest results when trying to identify the left and right cardiac cavities (subxiphoid view), with only 53.5% and 55.8% of correct answers, respectively.
Teaching based on peer mentoring achieved an adequate level of training in basic abdominal ultrasound. The students acquired these skills in a relatively short training period. These results suggest that peer mentoring can facilitate the large-scale implementation of ultrasound teaching in undergraduate students.
Introduction and Objectives
Pulmonary congestion (PC) is associated with an increased risk of hospitalization and death in patients with heart failure (HF). Lung ultrasound has shown to be highly ...sensitive for detecting PC in HF. The aim of this study is to evaluate whether lung ultrasound–guided therapy improves 6-month outcomes in patients with HF compared with conventional treatment.
Materials and Methods
Randomized, multicenter, single-blind clinical trial in patients discharged from Internal Medicine Departments after hospitalization for decompensated HF. Participants will be assigned 1:1 to receive treatment guided according to the presence of lung ultrasound signs of congestion (semi-quantitative evaluation of B lines and the presence of pleural effusion) versus clinical assessment of congestion. The primary outcome is the combination of cardiovascular death and readmission for HF at 6 months.
Conclusions
The results of this study will provide more evidence about the impact of lung ultrasound on treatment monitoring in patients with chronic HF.
To assess the efficacy of sodium-glucose cotransporter-2 inhibitor (SGLT2i) and glucagon-like peptide-1 receptors agonist (GLP-1RA) therapy on liver steatosis measured by fatty liver index (FLI) and ...hepatic steatosis index (HSI) at 26 weeks in outpatients with diabetes and obesity.
Observational, prospective, multicenter study. Patients with steatosis determined by FLI (values <30 rule out and >60 indicate steatosis) and HIS (values <30 rule out and >36 indicate steatosis) who received combination therapy were included. Patients were stratified into three groups according to the sequential order of treatment. We used robust statistical methods.
In our final report we included 174 patients (58.6% males), mean age 61.9 (10) years. Baseline body mass index, waist circumference and weight were 36.5 (6.8) kg/m
2
, 117.5 (15.1) cm and 99.4 (20.5) kg, respectively. One hundred percent of patients had altered biomarkers of fatty liver scores (FLI 96 13 and HSI 49.2 8.5). At 26 weeks, significant reductions in FLI (−4.5 95% CI 3.5-5.9 p < .001) and HSI (−2.4 95% CI 1.6-3.2 p < .001) were found in the total sample and pre-specified treatment and FLI cut-off point subgroups.
Our results show a beneficial effect of the combination of GLP-1RAs plus SGLT2is on liver steatosis that goes beyond glucose control, and it is related mainly to weight loss, a decline in biomarkers and reductions in abdominal circumference. For many patients, early detection is essential to improving outcomes in NAFLD and could allow us to select the most efficient treatment options.