JMIR Publications

ABSTRACT

Background:

Medication dosing errors remain a leading cause of preventable harm in pediatric emergency care. Although large language models (LLMs) are increasingly explored as clinical support tools, their potential as developmental instruments for rapid prototyping of digital decision aids has been underexplored.

Objective:

To describe the expedited design and dissemination of WebbCalc, a free, open-access, guideline-based pediatric antibiotic and sepsis calculator developed using LLM-assisted coding.

Methods:

WebbCalc was conceived and launched over a 9-day period (May 23–31, 2025) using LLM-assisted programming with physician oversight. The tool was deployed on a cloud-based platform and designed with a single-page interface for streamlined emergency department (ED) workflow. User inputs include clinical indication, patient age, weight, and allergy status. Outputs provide antibiotic options, dosing regimens, formulation volumes, and guideline-supported treatment durations. Additional modules include pediatric fever workup, sepsis antibiotics, fluid resuscitation, and vasopressor recommendations.

Results:

As a niche medical calculator, WebbCalc is beginning to cultivate a global user base. Recent analytics (August 9 to September 8, 2025) show initial traction, with the platform attracting approximately 800 page views from 500 unique visitors. This organic growth, driven by search engine optimization, points to a developing audience across North America, Europe, and Asia.

Conclusions:

This study provides proof-of-concept that LLMs can be leveraged to accelerate the development and launch of open-access, guideline-based clinical decision aids within days. Such tools have potential to reduce cognitive load, standardize pediatric antibiotic prescribing, and improve patient safety in emergency medicine.