From the Journals

Study identifies GI bleed risk factors in von Willebrand disease



Researchers have identified several risk factors associated with gastrointestinal (GI) bleeding in adult patients with von Willebrand disease (VWD), based on findings from a retrospective analysis.

“[We] evaluated prevalence and risk factors of GIB among individuals with and without VWD, using a large national database,” wrote Anastasia Tsagianni, MD, of the University of Pittsburgh, and colleagues. The findings were published in Thrombosis Research.

The researchers retrospectively reviewed discharge data from the National Inpatient Sample database. The team analyzed correlates of GI bleeding among patients with VWD and estimated prevalence rates using measures in the database. Risk factors for GI bleeding in VWD were correlated via multivariable logistic regression.

Between Jan. 1, 2009, and Dec. 31, 2014, there were 16,640 admissions with VWD and 618 were admitted with GI bleeding, the researchers reported. After analysis, the researchers found that the prevalence of GI bleeding was 3.7% and 1.49% in VWD and non-VWD patients, respectively – a 2.5-fold greater rate in VWD patients.

“Comorbidities associated with greater [GI bleeding] risk in individuals with VWD include past surgery, hypertension, hyperlipidemia, smoking, renal disease, hepatitis C, thrombocytopenia, or liver disease,” the researchers wrote.

In the multivariable analysis, the factors associated with GI bleeding were smoking status (odds ratio, 1.40), African American race (OR, 1.80), male gender (OR, 1.61), angiodysplasia (OR, 104.06), colonic diverticulitis (OR, 16.66), and hepatitis C (OR 2.17). These variables were similar to risk factors identified in the non-VWD group, the researchers noted.

“Although significant, age did not appear to be a strong risk factor for either group,” they wrote. “Steroids were not associated with increased risk for [GI bleeding] in either group.”

A key limitation of the study was the use of discharge diagnosis codes as an inclusion method, the researchers noted. As a result, misclassification bias could be present, they added.

The Pennsylvania Department of Health and the Health Resources and Services Administration funded the study. The authors reported having no conflicts of interest.

SOURCE: Tsagianni A et al. Thromb Res. 2019 Apr 17. doi: 10.1016/j.thromres.2019.04.017.

Next Article: