Endoscopy of the upper gastrointestinal tract as a diagnostic tool for children with human immunodeficiency virus infection

OBJECTIVE: The purpose of this study was to determine the prevalence of upper gastrointestinal tract lesions in children with human immunodeficiency virus (HIV) infection who undergo endoscopy of the upper gastrointestinal tract and to identify important clinical predictors of abnormal endoscopic results. METHODS: All HIV-infected children who underwent endoscopy and were followed at Children's Hospital, Boston, from January 1985 to August 1994 were studied. The main outcome measure was endoscopic results, which were categorized into observational, histologic, and microbiologic findings. Potential predictors included height, weight, nutritional interventions, HIV disease stage, CD4 T-lymphocyte count, medications, active infections, and indications for endoscopy. RESULTS: Forty-three endoscopies in unique patients are reported. Most children had advanced HIV infection (67% acquired immunodeficiency syndrome, mean CD4 T-lymphocyte count z score = -2.71, weight z score = -2.04). An abnormal endoscopic finding was discovered in 93% of children and confirmed by histologic, microbiologic, or a combination of these studies in 72% of children. Thirty-five percent of children had an opportunistic pathogen identified endoscopically; 65% of these pathogens were previously undiagnosed. Observational findings often were poor indicators of histologic and microbiologic abnormalities. Independent predictors of abnormal histologic findings include younger age at endoscopy (odds ratio (OR) = 1.16 per year, 95% confidence interval (CI) (1.02, 1.33)) and guaiac-negative stools (OR = 16.7, 95% CI (1.92, 142.9)). Independent predictors of finding a pathogen at the time of endoscopy include a greater number of indications for endoscopy (OR = 2.6 per indication, 95% CI (1.3, 5.3)) and diagnosis of acquired immunodeficiency syndrome (OR = 16.4, 95% CI (1.3, 213)). No other gastrointestinal, nutritional, or immunologic parameters were significantly predictive of endoscopic outcomes. Medical management was changed in 70% of children because of the endoscopic findings. CONCLUSIONS: Endoscopy is a useful tool to direct therapy against peptic and infectious disorders of the upper gastrointestinal tract in children with HIV infection. Specific gastrointestinal symptoms are not useful predictors of abnormal results.