Abstract Diarrheal diseases are the among the leading cause of death in children worldwide, most of which occur in low- and middle-income countries (LMICs). The cornerstone for management of diarrhea is rehydration, though antimicrobials are beneficial in some instances. Unfortunately, given that treatment is frequently empiric, based mostly on clinical suspicion for bacterial causes, antimicrobials are overused in management of diarrheal illness worldwide. Thus, there is a need for clinical decision support tools to inform clinical management and promote antimicrobial stewardship. In particular, frontline healthcare providers in rural areas, such as village doctors in Bangladesh, often have inadequate knowledge for the basis of antimicrobial use, and may benefit greatly from guidance. We have recently developed a mobile phone based electronic clinical decision-support tool (eCDST), and shown the impact of etiological prediction on physician antibiotic prescription practices for pediatric diarrhea in two LMICs. In this proposal, our overarching goal is to customize our eCDST into a comprehensive mHealth application (Accessible Diarrhea Etiology Prediction Tool, ADEPT) to support rural health care providers in the management of pediatric diarrhea (R21 phase), and to measure its impact through the performance of a pilot before-after feasibility study (R33 phase). Our work will contribute to an evidence base for mHealth-enabled antimicrobial stewardship, with potential to extend beyond diarrheal illness, and into other syndromes, including respiratory illness and febrile illness.

Project Narrative Diarrheal diseases are a leading cause of morbidity and mortality in children worldwide. Appropriate management of childhood diarrhea depends on the type of pathogen responsible, but in many cases testing is not affordable or feasible. We propose to develop and validate a tool for rural healthcare workers that predicts etiology for pediatric diarrhea, and supports clinical decision making.