Desarrollo de un modelo por inteligencia artificial con hemodinamia no invasiva para predecir preeclampsia en embarazos de alto riesgo / Artificial Intelligence Modeling with Non-Invasive Hemodynamics to Predict Preeclampsia in High-Risk Pregnancy

RESUMEN Introducción: la preeclampsia (PE) es la principal causa de morbimortalidad materno-fetal en nuestro país. Alteraciones hemodinámicas precoces durante el embarazo podrían predecir la evolución a PE. El machine learning (ML) permite el hallazgo de patrones ocultos que podrían detectar precozmente el desarrollo de PE. Objetivos: desarrollar un árbol de clasificación con variables de hemodinamia no invasiva para predecir precozmente desarrollo de PE. Material y métodos: estudio observacional prospectivo con embarazadas de alto riesgo (n=1155) derivadas del servicio de Obstetricia desde enero 2016 a octubre 2022 para el muestreo de entrenamiento por ML con árbol de clasificación j48. Se seleccionaron 112 embarazadas entre semanas 10 a 16, sin tratamiento farmacológico y que completaron el seguimiento con el término de su embarazo con evento final combinado (PE): preeclampsia, eclampsia y síndrome HELLP. Se evaluaron simultáneamente con cardiografía de impedancia y velocidad de onda del pulso y con monitoreo ambulatorio de presión arterial de 24 hs (MAPA). Resultados: presentaron PE 17 pacientes (15,18%). Se generó un árbol de clasificación predictivo con las siguientes variables: índice de complacencia arterial (ICA), índice cardíaco (IC), índice de trabajo sistólico (ITS), cociente de tiempos eyectivos (CTE), índice de Heather (IH). Se clasificaron correctamente el 93,75%; coeficiente Kappa 0,70, valor predictivo positivo (VPP) 0,94 y negativo (VPN) 0,35. Precisión 0,94, área bajo la curva ROC 0,93. Conclusión: las variables ICA, IC, ITS, CTE e IH predijeron en nuestra muestra el desarrollo de PE con excelente discriminación y precisión, de forma precoz, no invasiva, segura y con bajo costo.

ABSTRACT Background: Preeclampsia (PE) is the main cause of maternal-fetal morbidity and mortality in our country. Early hemodynamic changes during pregnancy could predict progression to PE. Machine learning (ML) enables the discovery of hidden patterns that could early detect PE development. Objectives: The aim of this study was to build a classification tree with non-invasive hemodynamic variables for the early prediction of PE occurrence. Results: Seventeen patients (15.18%) presented PE. A predictive classification tree was generated with arterial compliance index (ACI), cardiac index (CI), cardiac work index (CWI), ejective time ratio (ETR), and Heather index (HI). A total of 93.75% patients were correctly classified (Kappa 0.70, positive predictive value 0.94 and negative predictive value 0.35; accuracy 0.94, and area under the ROC curve 0.93). Conclusion: ACI, CI, CWI, ETR and HI variables predicted the early development of PE in our sample with excellent discrimination and accuracy, non-invasively, safely and at low cost.

Timing of convalescent plasma administration and 28-day mortality in COVID-19 pneumonia.

This is a multicenter cohort study including consecutive, hospitalized patients ≥18 years, with moderate to severe COVID-19, carried out to evaluate the relationship between the timing of convalescent plasma administration and 28-day mortality. Data were prospectively collected between May 14, 2020 and October 31, 2020. Patients were grouped according to the timing of administration of convalescent plasma as <3 days, between 3 and 7 days, and >7 days. The main outcome variable was 28-day mortality. Independent predictors of mortality were identified by logistic regression. Of 4719 patients receiving convalescent plasma, 3036 (64.3%) were in the general ward, 1171 (24.8%) in the intensive care unit (ICU), and 512 (10.8%) in the ICU on mechanical ventilation. Convalescent plasma was administered to 3113 (66%) patients within the first 3 days of hospital admission, to 1380 (29.2%) between 3 and 7 days, and to 226 after 7 days; 28-day mortality was, respectively, 18.1%, 30.4% and 38.9% (p<0.001). In the regression model, convalescent plasma administration within the first 3 days of admission was associated with reduced 28-day mortality, compared with the administration after 7 days (OR 0.40, 95% CI 0.30 to 0.53). Early convalescent plasma administration was associated to a significant decreased mortality in patients in the general ward (OR 0.45, 95% CI 0.29 to 0.69) and in the ICU (OR 0.35, 95% CI 0.19 to 0.64), but not in those requiring mechanical ventilation (OR 0.52, 95% CI 0.27 to 1.01). In conclusion, this study suggests that early administration of convalescent plasma to patients with COVID-19 pneumonia is critical to obtain therapeutic benefit.

Risk factors for COVID-19 mortality: The effect of convalescent plasma administration.

BACKGROUND: Convalescent plasma, widely utilized in viral infections that induce neutralizing antibodies, has been proposed for COVID-19, and preliminary evidence shows that it might have beneficial effect. Our objective was to determine the risk factors for 28-days mortality in patients who received convalescent plasma for COVID-19 compared to those who did not, who were admitted to hospitals in Buenos Aires Province, Argentina, throughout the pandemic. METHODS: This is a multicenter, retrospective cohort study of 2-month duration beginning on June 1, 2020, including unselected, consecutive adult patients with diagnosed COVID-19, admitted to 215 hospitals with pneumonia. Epidemiological and clinical variables were registered in the Provincial Hospital Bed Management System. Convalescent plasma was supplied as part of a centralized, expanded access program. RESULTS: We analyzed 3,529 patients with pneumonia, predominantly male, aged 62±17, with arterial hypertension and diabetes as main comorbidities; 51.4% were admitted to the ward, 27.1% to the Intensive Care Unit (ICU), and 21.7% to the ICU with mechanical ventilation requirement (ICU-MV). 28-day mortality was 34.9%; and was 26.3%, 30.1% and 61.4% for ward, ICU and ICU-MV patients. Convalescent plasma was administered to 868 patients (24.6%); their 28-day mortality was significantly lower (25.5% vs. 38.0%, p<0.001). No major adverse effects occurred. Logistic regression analysis identified age, ICU admission with and without MV requirement, diabetes, and preexistent cardiovascular disease as independent predictors of 28-day mortality, whereas convalescent plasma administration acted as a protective factor. CONCLUSIONS: Our study suggests that the administration of convalescent plasma in COVID-19 pneumonia admitted to the hospital might be associated with improved outcomes.