Creating a Parent-Informed Pediatric Emergency Department Wait Time App: Human-Centered Design Approach to Creating an AI Health Care Tool

Editorial reading guide · Collaboration

How can our family’s experience help a care service decide what information an AI tool should provide?

Parent accounts of uncertainty, feeling forgotten, confusing triage, and lack of progress shaped an AI-supported wait-time app. Parent and nurse preferences sometimes diverged: nurses wanted elapsed waiting time displayed, while parents considered that unnecessary or offensive. The deployed app used predictive ML and NLP for personalized estimates/education, but the study evaluated design needs rather than predictive accuracy or clinical benefit.

Study context

64 parents/guardians waiting with a child in a Toronto pediatric emergency department and five triage nurses; 24 parents tested wireframe prototypes

Age / stage: Children under 1–18 years; median seven years

Read with three questions

1. Which parent concerns became concrete design requirements, and where did parent and staff preferences conflict?
2. What did prototype feedback establish before the predictive model was deployed?
3. Which evidence is still needed to show accurate estimates, accessible information, or improved care rather than appeal and use?

Keep the limits in view

Convenience interviews took place during summer weekday daytime waits under two hours; longer and more stressful waits were not represented. All parents owned smartphones, most preferred English, and most had higher education. The report supplied no predictive-accuracy metrics, app-data analysis, controlled satisfaction effect, or health/cost outcome evidence; later use by thousands is uptake, not validated benefit.