RESUMO
BACKGROUND: For optimal management of acute infectious diarrhoeal diseases, it is necessary to utilize a screening process to distinguish between invasive and non-invasive diarrhoeas. The aim of this study was to compare the diagnostic utilities of clinical features, faecal microscopy (FM), and faecal occult blood testing (FOBT) in distinguishing invasive diarrhoeas from non-invasive ones. METHODS: A total of 1008 patients with acute diarrhoea were evaluated. Rectal swabs were cultured for Salmonella, Shigella, and Vibrio species; rectal swabs from 109 of these patients were also examined for Campylobacter, enterotoxigenic Escherichia coli, and rotavirus species. Isolation of faecal enteropathogens served as the gold standard. FOBT was performed with a commercial modified guaiac test. Specificity, sensitivity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and likelihood ratio were compared. RESULTS: Among the 1008 patients 402 with a single identified enteropathogen were available for analysis. Invasive and non-invasive enteropathogens were isolated from 262 (65.2%) and 140 (34.8%) cases, respectively. The presence of visible blood in faeces was almost a pathognomonic sign of invasive diarrhoea but had poor sensitivity. Clinical features were useful but inadequate in differentiating patients with non-bloody diarrhoea (74% of patients) into invasive and non-invasive categories. The sensitivities, specificities, PPVs, and NPVs of FM and FOBT were 75%, 77%, 58%, 88%, and 85%, 68%, 53%, and 91%, respectively. CONCLUSION: The presence of visible blood in faeces is a highly specific clinical feature of invasive diarrhoea but suffers from low sensitivity. In non-bloody diarrhoea FOBT is a valuable screening test and is comparable to FM, particularly when interpreted in the clinical context.
