A locally optimised machine learning approach to early prognostication of long-term neurological outcomes after out-of-hospital cardiac arrest.

Background: Out-of-hospital cardiac arrest (OHCA) represents a major burden for society and health care, with an average incidence in adults of 67 to 170 cases per 100,000 person-years in Europe and in-hospital survival rates of less than 10%. Patients and practitioners would benefit from a prognostication tool for long-term good neurological outcomes. Objective: We aim to develop a machine learning (ML) pipeline on a local database to classify patients according to their neurological outcomes and identify prognostic features. Methods: We collected clinical and biological data consecutively from 595 patients who presented OHCA and were routed to a single regional cardiac arrest centre in the south of France. We applied recursive feature elimination and ML analyses to identify the main features associated with a good neurological outcome, defined as a Cerebral Performance Category score less than or equal to 2 at six months post-OHCA. Results: We identified 12 variables 24 h after admission, capable of predicting a six-month good neurological outcome. The best model (extreme gradient boosting) achieved an AUC of 0.96 and an accuracy of 0.92 in the test cohort. Conclusion: We demonstrated that it is possible to build accurate, locally optimised prediction and prognostication scores using datasets of limited size and breadth. We proposed and shared a generic machine-learning pipeline which allows external teams to replicate the approach locally.

Comparing COVID-19 and influenza: Epidemiology, clinical characteristics, outcomes and mortality in the ICU.

RATIONALE: Several authors have compared COVID-19 infection with influenza in the ICU. OBJECTIVE: This study aimed to compare the baseline clinical profiles, care procedures, and mortality outcomes of patients admitted to the intensive care unit, categorized by infection status (Influenza vs. COVID-19). METHODS: Retrospective observational study. Data were extracted from the Toulouse University Hospital from March 2014 to March 2021. To compare survival curves, we plotted the survival at Day-90 using the Kaplan-Meier curve and conducted a log-rank test. Additionally, we performed propensity score matching to adjust for confounding factors between the COVID-19 and influenza groups. Furthermore, we use the CART model for multivariate analysis. RESULTS: The study included 363 patients admitted to the ICU due to severe viral pneumonia: 152 patients (41.9 %) with influenza and 211 patients (58.1 %) with COVID-19. COVID-19 patients exhibited a higher prevalence of cardiovascular risk factors, whereas influenza patients had significantly higher severity scores (SOFA: 10 [6-12] vs. 6 [3-9], p<0.01 and SAPS II: 51 [35-67] vs. 37 [29-50], p<0.001). Overall mortality rates were comparable between the two groups (27.6 % (n = 42) in the influenza group vs. 21.8 % (n = 46) in the COVID-19 group, p=NS). Mechanical ventilation was more commonly employed in the influenza group (76.3 % (n = 116) vs. 59.7 % (n = 126), p<0.001); however, COVID-19 patients required longer durations of mechanical ventilation (18 [9-29] days vs. 13 [5-24] days, p<0.006) and longer hospital stays (23 [13-34] days vs. 18.5 [9-34.5] days, p = 0.009). The CART analysis revealed that the use of extra renal replacement therapy was the most influential prognostic factor in the influenza group, while the PaO2/FiO2-PEEP ratio played a significant role in the COVID-19 group. CONCLUSIONS: Despite differences in clinical presentation and prognostic factors, the mortality rates at 90 days, after adjusting for confounding factors, were similar between COVID-19 and influenza patients.

Role of Campylobacter jejuni gamma-glutamyl transpeptidase on epithelial cell apoptosis and lymphocyte proliferation.

BACKGROUND: A gamma-glutamyl transpeptidase (GGT) is produced by up to 31% of strains of Campylobacter jejuni isolates. C. jejuni GGT is close to Helicobacter pylori GGT suggesting a conserved activity but unlike the latter, C. jejuni GGT has not been studied extensively. In line with the data available for H. pylori, our objectives were to purify C. jejuni GGT from the bacteria, and to evaluate its inhibitory and proapoptotic activities on epithelial cells and human lymphocytes. METHODS: C. jejuni GGT was purified from culture supernatants by chromatography. After verification of the purity by using mass spectrometry of the purified enzyme, its action on two epithelial cell lines and human lymphocytes was investigated. Cell culture as well as flow cytometry experiments were developed for these purposes. RESULTS: This study demonstrated that C. jejuni GGT is related to Helicobacter GGTs and inhibits the proliferation of epithelial cells with no proapoptotic activity. C. jejuni GGT also inhibits lymphocyte proliferation by causing a cell cycle arrest in the G0/G1 phase. These effects are abolished in the presence of a specific pharmacological inhibitor of GGT. CONCLUSION: C. jejuni GGT activity is comparable to that of other Epsilonproteobacteria GGTs and more generally to Helicobacter bilis (inhibition of epithelial cell and lymphocyte proliferation, however with no proapoptotic activity). It could therefore be considered as a pathogenicity factor and promote, via the inhibition of lymphocyte proliferation, the persistence of the bacteria in the host. These observations are consistent with a role of this enzyme in the pathophysiology of chronic infections associated with C. jejuni.