Clinical Predictors of Shigella and Campylobacter Infection in Children in the United States

Abstract Background Infectious gastroenteritis is a major cause of morbidity and mortality among children worldwide. While most episodes are self-limiting, for select pathogens such as Shigella and Campylobacter, etiological diagnosis may allow effective antimicrobial therapy and aid public health interventions. Unfortunately, clinical predictors of such pathogens are not well established and are based on small studies using bacterial culture for identification. Methods We used prospectively collected data from a multi-center study of pediatric gastroenteritis employing multi-pathogen molecular diagnostics to determine clinical predictors associated with 1) Shigella and 2) Shigella or Campylobacter infection. We used machine learning algorithms for clinical predictor identification, then performed logistic regression on features extracted plus pre-selected variables of interest. Results Of 993 children enrolled with acute diarrhea, we detected Shigella spp. in 56 (5.6%) and Campylobacter spp. in 24 (2.4%). Compared with children who had neither pathogen detected (of whom, >70% had ≥1 potential pathogen identified), bloody diarrhea (odds ratio 4.0), headache (OR 2.2), fever (OR 7.1), summer (OR 3.3), and sick contact with GI illness (OR 2.2), were positively associated with Shigella, and out-of-state travel (OR 0.3) and vomiting and/or nausea (OR 0.4) were negatively associated (Table). For Shigella or Campylobacter, predictors were similar but season was no longer significantly associated with infection. Conclusion These results can create prediction models and assist clinicians with identifying patients who would benefit from diagnostic testing and earlier antibiotic treatment. This may curtail unnecessary antibiotic use, and help to direct and target appropriate use of stool diagnostics. Disclosures A. Leber, BioFIre Diagnostics: Research Contractor and Scientific Advisor, Research support, Speaker honorarium and Travel expenses J. Daly, Biofire: Grant Investigator, Grant recipient R. Selvarangan, BioFire Diagnostics: Board Member and Investigator, Consulting fee and Research grant Luminex Diagnostics: Investigator, Research grant J. Dien Bard, BioFire: Consultant and Investigator, Research grant and Speaker honorarium K. Holmberg, BioFire Diagnostics: Employee, Salary K. Bourzac, BioFire Diagnostics: Employee, Salary K. C. Chapin, BioFire Diagnstics: Investigator, Research support A. Pavia, BioFire Diagnostics: Grant Investigator, Research grant