Procedural-Related Bleeding in Hospitalized Patients With Liver Disease (PROC-BLeeD): An International, Prospective, Multicenter Observational Study

Nicolas M. Intagliata, Robert S. Rahimi, Fatima Higuera-de-la-Tijera, Douglas A. Simonetto, Alberto Queiroz Farias, Daniel F. Mazo, Justin R. Boike, Jonathan G. Stine, Marina Serper, Gustavo Pereira, Angelo Z. Mattos, Sebastian Marciano, Jessica P.E. Davis, Carlos Benitez, Ryan Chadha, Nahum Méndez-Sánchez, Andrew S. deLemos, Arpan Mohanty, Melisa Dirchwolf, Brett E. FortunePatrick G. Northup, James T. Patrie, Stephen H. Caldwell

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

Background & Aims: Hospitalized patients with cirrhosis frequently undergo multiple procedures. The risk of procedural-related bleeding remains unclear, and management is not standardized. We conducted an international, prospective, multicenter study of hospitalized patients with cirrhosis undergoing nonsurgical procedures to establish the incidence of procedural-related bleeding and to identify bleeding risk factors. Methods: Hospitalized patients were prospectively enrolled and monitored until surgery, transplantation, death, or 28 days from admission. The study enrolled 1187 patients undergoing 3006 nonsurgical procedures from 20 centers. Results: A total of 93 procedural-related bleeding events were identified. Bleeding was reported in 6.9% of patient admissions and in 3.0% of the procedures. Major bleeding was reported in 2.3% of patient admissions and in 0.9% of the procedures. Patients with bleeding were more likely to have nonalcoholic steatohepatitis (43.9% vs 30%) and higher body mass index (BMI; 31.2 vs 29.5). Patients with bleeding had a higher Model for End-Stage Liver Disease score at admission (24.5 vs 18.5). A multivariable analysis controlling for center variation found that high-risk procedures (odds ratio [OR], 4.64; 95% confidence interval [CI], 2.44–8.84), Model for End-Stage Liver Disease score (OR, 2.37; 95% CI, 1.46–3.86), and higher BMI (OR, 1.40; 95% CI, 1.10–1.80) independently predicted bleeding. Preprocedure international normalized ratio, platelet level, and antithrombotic use were not predictive of bleeding. Bleeding prophylaxis was used more routinely in patients with bleeding (19.4% vs 7.4%). Patients with bleeding had a significantly higher 28-day risk of death (hazard ratio, 6.91; 95% CI, 4.22–11.31). Conclusions: Procedural-related bleeding occurs rarely in hospitalized patients with cirrhosis. Patients with elevated BMI and decompensated liver disease who undergo high-risk procedures may be at risk to bleed. Bleeding is not associated with conventional hemostasis tests, preprocedure prophylaxis, or recent antithrombotic therapy.

Original languageEnglish (US)
Pages (from-to)717-732
Number of pages16
JournalGastroenterology
Volume165
Issue number3
DOIs
StatePublished - Sep 2023
Externally publishedYes

Keywords

  • Cirrhosis
  • Hemostasis
  • Obesity
  • Prophylaxis
  • Risk

ASJC Scopus subject areas

  • Hepatology
  • Gastroenterology

Fingerprint

Dive into the research topics of 'Procedural-Related Bleeding in Hospitalized Patients With Liver Disease (PROC-BLeeD): An International, Prospective, Multicenter Observational Study'. Together they form a unique fingerprint.

Cite this