OBJECTIVES: Acute bloody diarrhea (hematochezia) in children is traditionally associated with invasive bacterial infection, but its etiologic spectrum is increasingly recognized as heterogeneous. We aimed to characterize the etiologic distribution of acute bloody diarrhea in a pediatric emergency department and to compare the diagnostic yields of multiplex gastrointestinal PCR and conventional stool testing. METHODS: In this prospective observational study conducted between July 2018 and July 2019, consecutive children aged 0 to 18 years presenting with visible blood in stool (hematochezia) to a tertiary pediatric emergency department were enrolled. Disease severity was assessed using the Clinical Dehydration Scale (CDS) and the modified Vesikari score. All stool samples were tested with conventional methods (culture, antigen for rotavirus/adenovirus, microscopy) and a 22-target commercial multiplex gastrointestinal PCR panel. The contribution of multiplex PCR was analyzed as a descriptive diagnostic yield rather than as a predictor in a regression model, and a multivariable logistic regression model containing clinical variables only was used to test whether clinical severity independently predicted bacterial etiology. A sensitivity analysis used a stricter culture-confirmed definition of bacterial etiology. RESULTS: Eighty-three children were enrolled (median age 48 mo, IQR 21 to 84; 55% male). At least 1 enteropathogen was detected in 62 patients (74.7%). A bacterial pathogen was identified in 37 of 83 patients (44.6%), a viral-only infection in 14 (16.9%), a parasitic infection in 11 (13.3%), and a mixed viral-bacterial co-detection in 4 (4.8%). Of the 37 bacterial detections, 10 (27.0%) were identified by stool culture alone, 22 (59.5%) by multiplex PCR alone, 1 (2.7%) by both culture and PCR, and 4 (10.8%) by combined antigen and PCR in mixed infections. In the multivariable logistic regression model, none of CDS ≥5, fever, continuous modified Vesikari score, or recent antibiotic use was independently associated with bacterial etiology. Culture and PCR were complementary: 27.0% of bacterial pathogens were detected only by culture despite a negative bacterial PCR target. CONCLUSIONS: Pediatric acute bloody diarrhea is etiologically heterogeneous: fewer than half of the presentations had a bacterial etiology, and clinical severity alone did not reliably identify those with bacterial infection. Multiplex PCR expanded-but did not replace-stool culture, with more than a quarter of bacterial pathogens detected only by culture. Clinical management of children with bloody diarrhea should combine maintained clinical vigilance with comprehensive microbiological testing (stool culture plus multiplex PCR).
Seçilmiş et al. (Wed,) studied this question.