m+Medelevator.RU
← Поиск публикаций
iGIE : innovation, investigation and insights · 2023

GI bleeding in patients with left ventricular assist device: endoscopic approach and prediction model using supervised machine learning.

СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…

История · было → стало

Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.

McEntire Dan, Gow-Lee Benjamin, Kucharski Kimberly, Ali Hassam, Fang John, Mohan Babu P

Аннотация

BACKGROUND AND AIMS: GI bleeding is a common adverse event in patients with left ventricular assist devices (LVADs). Data are limited regarding the optimal endoscopic management approach and quantifiable risk of bleeding.

METHODS: This retrospective analysis included adult patients who underwent LVAD implantation at the University of Utah between February 1996 and September 2021. The endoscopic management of GI bleeding events was analyzed, along with clinical predictors of bleeding, using a supervised machine learning model.

RESULTS: A total of 557 LVADs were implanted during the study period. Of these, 132 (23.7%) had at least 1 GI bleed, with a total of 252 GI bleeds. Gastric (17.2%) and non-duodenum small intestinal (13.1%) bleeding were the most common sites. A total of 429 procedures were performed; EGD was the most common (45.5%). Video capsule endoscopy exhibited the highest diagnostic yield overall (68.9%); 203 (80.6%) GI bleeds were treated with endoscopy (EGD, 45.5%; colonoscopy, 26.8%). A hemostatic intervention was possible in 32.8% of EGDs, 22.6% of colonoscopies, and 26.2% of push enteroscopies. For the model predicting GI bleeding risk after LVAD transplantation, 3 predictors were retained: destination therapy (score = 1.5), warfarin use (score = 5), and antiplatelet use (score = 10). A total score ≥10 was considered high risk.

CONCLUSIONS: Although EGD was the most performed procedure, push enteroscopy rivaled both EGD and colonoscopy on diagnostic yield and percentage of interventions performed. Clinical variables of destination therapy, warfarin use, and antiplatelet use were retained by the machine learning model to help quantify risk of GI bleed.

ИП Урсолов Владлен Сергеевич

ИНН 500118482352 · ОГРНИП 314500117700049

Реквизиты и контакты