Internal and emergency medicine
-
Multicenter Study
Machine learning versus traditional methods for the development of risk stratification scores: a case study using original Canadian Syncope Risk Score data.
Artificial Intelligence and machine learning (ML) methods are promising for risk-stratification, but the added benefit over traditional statistical methods remains unclear. We compared predictive models developed using machine learning (ML) methods to the Canadian Syncope Risk Score (CSRS), a risk-tool developed with logistic regression for predicting serious adverse events (SAE) after emergency department (ED) disposition for syncope. We used the prospective multicenter cohort data collected for CSRS development at 11 Canadian EDs over an 8-year period to develop four ML models to predict 30-day SAE (death, arrhythmias, MI, structural heart disease, pulmonary embolism, hemorrhage) after ED disposition. ⋯ The AUCs and calibration slopes for the ML models and CSRS were similar. Two ML models used fewer predictors than the CSRS but matched its performance. Overall, the ML models matched the CSRS in performance, with some models using fewer predictors.
-
Randomized Controlled Trial
Modelling the risk of hospital admission of lab confirmed SARS-CoV-2-infected patients in primary care: a population-based study.
The objectives of this study are to develop a predictive model of hospital admission for COVID-19 to help in the activation of emergency services, early referrals from primary care, and the improvement of clinical decision-making in emergency room services. The method is the retrospective cohort study of 49,750 patients with microbiological confirmation of SARS-CoV-2 infection. The sample was randomly divided into two subsamples, for the purposes of derivation and validation of the prediction rule (60% and 40%, respectively). ⋯ Based on the risk score, five risk groups were derived with hospital admission ranging from 2.94 to 51.87%. In conclusion, we propose a classification system for people with COVID-19 with a higher risk of hospitalization, and indirectly with it a greater severity of the disease, easy to be completed both in primary care, as well as in emergency services and in hospital emergency room to help in clinical decision-making. Registration: ClinicalTrials.gov Identifier: NCT04463706.
-
Enterococcal bloodstream infections (E-BSI) constitute the second cause of Gram-positive bacterial BSI in Europe with a high rate of in-hospital mortality. Furthermore, E-BSI treatment is still challenging because of intrinsic and acquired antibiotic resistances. We conducted a retrospective, 2-year, observational, single-centre study to evaluate clinical outcome and risk factors for E-BSI mortality in internal medicine wards. 201patients with E-BSI were included in the analysis. ⋯ No difference in 28-day survival was observed between appropriate or inappropriate treatment, except for endocarditis. However, E-BSI sources in clinical practices are not always properly investigated, including the rule-out of intracardiac vegetations. We did not demonstrate a difference in mortality for inappropriate therapy in the absence of endocarditis in comorbid patients with a long history of medicalization.