Grant: Travel Awards - Institute Community Support

Acute gastroenteritis is a common illness among children. The ability to identify children who would benefit from diagnostic stool testing is crucial to patient care and public health. Existing stool testing guidelines offer inconsistent recommendations, and their performance characteristics are unknown. Our objective was to determine performance characteristics of six leading gastroenteritis guidelines that recommend when to test children's stool for bacterial pathogens. We recruited 2,447 children <18 years old who presented for emergency department care with 3 or more episodes of diarrhea and/or vomiting in a 24-hour period at two children's hospitals in Alberta, Canada, from December 2014 to February 2018. As part of the study protocol, all participants were tested for nine bacterial pathogens. Additionally, clinicians could request stool for testing as per usual practice. Demographic and clinical data gathered at the index visit were used to determine whether guidelines would recommend testing. The primary outcome was positivity for any of the nine bacterial pathogens. Outcome data were available for 2,391 (97.7%) participants. Six percent (144) of participants tested positive for a bacterial pathogen. Guidelines identified 25.8% to 66.9% of children with a bacterial pathogen, and clinician-requested testing identified 31.3%. Guidelines varied in their ability to identify children with individual pathogens, such as Salmonella (25-78%), Escherichia coli O157 (48-72%), and non-O157 Shiga toxin-producing E. coli (STEC) (15-56%). Guidelines correctly identified children without a bacterial pathogen in 63.6% to 96.5% of cases, and clinician-requested testing correctly identified 94.7%. The guideline that did the poorest job identifying children without a bacterial pathogen would have increased the number of tests ordered over 5-fold, from 165 to 914 tests. No guidelines did a good job of identifying both children with and without bacterial pathogens.