Predictive factors of upper gastrointestinal bleeding in children: a cross-sectional analysis

Upper gastrointestinal bleeding (UGIB) in children presents diagnostic challenges. Geographic variations in etiology and the role of prolapse gastropathy as a distinct entity remain poorly characterized.

To identify endoscopic findings and clinical predictors of pediatric UGIB, with an emphasis on prolapse gastropathy, and to develop a predictive model for risk stratification.

This cross-sectional study enrolled 100 children (aged 2 months to 17 years) with UGIB. Demographic, clinical, laboratory, and endoscopic data were collected. Univariate and multivariate logistic regression identified predictors.

The mean age of the children was 7.2 ± 4.1 years; 62% presented with hematemesis. In 41% of cases, the endoscopic findings were normal; prolapse gastropathy was the leading pathological finding (27%). Male gender was associated with prolapse gastropathy (OR = 3.87, p = 0.007); NSAID use predicted peptic ulcer (OR = 127.3, p < 0.001). Multivariate analysis identified abdominal pain (OR = 111.6, p < 0.001) and platelet count (p = 0.006) as independent predictors, with 97.0% model accuracy.

Prolapse gastropathy represents the predominant pathological cause of UGIB in children. Male gender, abdominal pain, and platelet count are significant predictors enabling evidence-based risk stratification.