Bridging evidence-to-care gaps with mHealth: Designing a symptom checker for parents accessing knowledge translation resources on acute children’s illnesses in a smartphone application

Abstract

Background: Smartphone applications offer a novel platform for delivering health information to parents. This study created and evaluated an app-based symptom checker that recommends educational tools to parents based on their child’s symptoms. Methods: Symptoms extracted from 23 knowledge translation (KT) tools for 10 children’s illnesses comprised a set of plain-language symptoms. The symptom checker works by producing confusion matrices evaluating a child’s reported symptoms against possible illnesses, comparing precision scores to examine how well each illness matched reported symptoms, and ordering possible illnesses by performance score. Performance was evaluated by extracting symptoms from 8 clinical vignettes, and examining correct first-try matches. Results: We created a final list of 54 plain-language symptoms. Visualizations of the symptom set creation process and logic mapping are presented, as well as images of the working symptom checker. The symptom checker matched 100% (8/8) of tested clinical vignettes to the appropriate illness resource. Discussion: Symptom checkers are a potentially useful tool to integrate into apps that parents use for their children’s health. The design of these systems has the potential to change parents’ relationship with technology, affecting both their adoption and acceptance of symptom checkers. Our design choices contribute to addressing current barriers to the adoption of symptom checkers, reducing functional, critical, and interactive literacy requirements for parents.

Keywords

• Recommender system
• Knowledge translation
• Patient-oriented research
• Pediatric
• Health
• Parent