Novel Model to Predict Gastrointestinal Bleeding During Left Ventricular Assist Device Support

Area of Science:

• Cardiology
• Medical Devices
• Gastroenterology

Background:

• Gastrointestinal bleeding (GIB) is a significant complication in continuous-flow left ventricular assist device (CF-LVAD) patients.
• Accurate GIB risk assessment is crucial for patient selection, informed consent, and post-implant care.
• Developing a predictive model for GIB in CF-LVAD recipients is essential.

Purpose of the Study:

• To derive and validate a predictive model for gastrointestinal bleeding (GIB) in patients supported by continuous-flow left ventricular assist devices (CF-LVADs).

Main Methods:

• A cohort of 351 CF-LVAD recipients from 2004-2017 was analyzed.
• GIB was defined as a hemoglobin decrease of ≥2 g/dL.
• A weighted score, the Utah bleeding risk score, was developed using pre-implant variables and internally validated.

Main Results:

• GIB occurred in 120 (34%) patients after a median follow-up of 196 days.
• Independent predictors of GIB included age >54, prior bleeding, coronary artery disease, chronic kidney disease, severe RV dysfunction, low pulmonary artery pressure, and elevated fasting glucose.
• The Utah bleeding risk score effectively stratified patients: low risk (4.8%), intermediate risk (39.8%), and high risk (83.8%).

Conclusions:

• The Utah bleeding risk score is a novel, simple tool for personalized GIB risk estimation in CF-LVAD patients.
• This scoring system can aid clinicians in developing tailored, risk-based strategies to mitigate GIB.
• Implementing this score may reduce the burden of GIB on individual patients and healthcare systems.