Machine learning for predicting mortality in adult critically ill patients with Sepsis: A systematic review.

INTRODUCTION: Various Machine Learning (ML) models have been used to predict sepsis-associated mortality. We conducted a systematic review to evaluate the methodologies employed in studies to predict mortality among patients with sepsis. METHODS: Following a pre-established protocol registered at the International Prospective Register of Systematic Reviews, we performed a comprehensive search of databases from inception to February 2024. We included peer-reviewed articles reporting predicting mortality in critically ill adult patients with sepsis. RESULTS: Among the 1822 articles, 31 were included, involving 1,477,200 adult patients with sepsis. Nineteen studies had a high risk of bias. Among the diverse ML models, Logistic regression and eXtreme Gradient Boosting were the most frequently used, in 22 and 16 studies, respectively. Nine studies performed internal and external validation. Compared with conventional scoring systems such as SOFA, the ML models showed slightly higher performance in predicting mortality (AUROC ranges: 0.62-0.90 vs. 0.47-0.86). CONCLUSIONS: ML models demonstrate a modest improvement in predicting sepsis-associated mortality. The certainty of these findings remains low due to the high risk of bias and significant heterogeneity. Studies should include comprehensive methodological details on calibration and hyperparameter selection, adopt a standardized definition of sepsis, and conduct multicenter prospective designs along with external validations.

Pre-hospital Aspirin Use and Patient Outcomes in COVID-19: Results from the International Viral Infection and Respiratory Illness Universal Study (VIRUS).

INTRODUCTION: The goal of this investigation is to assess the association between prehospital use of aspirin (ASA) and patient-centered outcomes in a large global cohort of hospitalized COVID-19 patients. METHODS: This study utilizes data from the Society of Critical Care Medicine Discovery Viral Infection and Respiratory Illness Universal Study (VIRUS) Registry. Adult patients hospitalized from February 15th, 2020, to September 30th, 2021, were included. Multivariable regression analyses were utilized to assess the association between pre-hospital use of ASA and the primary outcome of overall hospital mortality. RESULTS: 21,579 patients were included from 185 hospitals (predominantly US-based, 71.3%), with 4691 (21.7%) receiving pre-hospital ASA. Patients receiving ASA, compared to those without pre-admission ASA use, were generally older (median 70 vs. 59 years), more likely to be male (58.7 vs. 56.0%), caucasian (57.4 vs. 51.6%), and more commonly had higher rates of medical comorbidities. In multivariable analyses, patients receiving pre-hospital ASA had lower mortality (HR: 0.89, 95% CI 0.82-0.97, p=0.01) and reduced hazard for progression to severe disease or death (HR: 0.91, 95% CI 0.84-0.99, p=0.02) and more hospital free days (1.00 days, 95% CI 0.66-1.35, p=0.01) compared to those without pre-hospital ASA use. The overall direction and significance of the results remained the same in sensitivity analysis, after adjusting the multivariable model for time since pandemic. CONCLUSIONS: In this large international cohort, pre-hospital use of ASA was associated with a lower hazard for death in hospitalized patients with COVID-19. Randomized controlled trials may be warranted to assess the utility of pre-hospital use of ASA.

Differential mortality with COVID-19 and invasive mechanical ventilation between high-income and low-and middle-income countries: a systematic review, meta-analysis, and meta-regression.

The COVID-19 pandemic has markedly affected the health care of patients in low- and middle-income countries (LMICs), but no systematic study to corroborate this effect has been undertaken. In addition, the survival outcomes of patients with COVID-19 who received invasive mechanical ventilation (IMV) have not been well established. We pooled evidence from all available studies and did a systematic review and meta-analysis to assess and compare mortality outcomes between LMICs and high-income countries (HICs). We searched MEDLINE and the University of Michigan Library according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines from December 1, 2019, to July 15, 2021, for case-control studies, cohort studies, and brief reports that discussed mortality ratios and survival outcomes among patients with SARS-CoV-2 who received IMV. We excluded studies and case reports without comparison groups, narrative reviews, and preprints. A random-effects estimate of the arcsine square root transformation (PAS) of each outcome was generated with the DerSimonian-Laird method. Seven eligible studies, consisting of 243,835 patients with COVID-19, were included. We identified a significantly higher mortality rate (i.e., a larger PAS) among the patients receiving IMV in LMICs (PAS, 0.754; 95% CI, 0.569-0.900; P<.001) compared to patients in HICs (PAS, 0.588; 95% CI, 0.263-0.876; P<.001). Considerable heterogeneity was present within the individual subgroups possibly because of the extent of the included studies, which had data from specific countries and states but not from individual hospitals or health care centers. Moreover, the sample population in each study was diverse. Meta-regression showed that a higher mortality rate among patients with COVID-19 who received IMV in both HICs (P<.001) and LMICs (P=.04) was associated with chronic pulmonary disease. Our study suggests that chronic pulmonary diseases and poor demographics lead to a worse prognosis among patients with COVID-19 who received IMV. Moreover, the survival outcome is worse in LMICs, where health care systems are usually understaffed and poorly financed.

Infection risk with the use of interleukin inhibitors in hospitalized patients with COVID-19: A narrative review.

INTRODUCTION: To date, only corticosteroids and interleukin-6 (IL-6) inhibitors have been shown to reduce mortality of hospitalized patients with COVID-19. In this literature review, we aimed to summarize infection risk of IL inhibitors, with or without the use of corticosteroids, used to treat hospitalized patients with COVID-19. METHODS: A literature search was conducted using the following evidence-based medicine reviews: Cochrane Central Register of Controlled Trials; Cochrane Database of Systematic Reviews; Embase; Ovid Medline; and Epub Ahead of Print, In-Process, In-Data-Review & Other Non-Indexed Citations, Daily and Versions 1946 to April 28, 2021. All relevant articles were identified using the search terms COVID-19 or SARS-coronavirus-2, infections, interleukins, inpatients, adults, and i ncidence. RESULTS: We identified 36 studies of which 2 were meta-analyses, 5 were randomized controlled trials, 9 were prospective studies, and 20 were retrospective studies. When anakinra was compared with control, 2 studies reported an increased risk of infection, and 3 studies reported a similar or decreased incidence of infection. Canakinumab had a lower associated incidence of infection compared with placebo in one study. When sarilumab was compared with placebo, one study reported an increased risk of infection. Nine studies comparing tocilizumab with placebo reported decreased or no difference in infection risk (odds ratio [OR] for the studies ranged from 0.39-1.21). Fourteen studies comparing tocilizumab with placebo reported an increased risk of infection, ranging from 9.1% to 63.0% (OR for the studies ranged from 1.85-5.04). Infection most commonly presented as bacteremia. Of the 6 studies comparing tocilizumab and corticosteroid use with placebo, 4 reported a nonsignificant increase toward corticosteroids being associated with bacterial infections (OR ranged from 2.76-3.8), and 2 studies reported no increased association with a higher infection risk. CONCLUSIONS: Our literature review showed mixed results with variable significance for the association of IL-6 inhibitors with risk of infections in patients with COVID-19.