JMIR Publications

ABSTRACT

Background:

Portugal faces persistent challenges in ensuring timely access to paediatric care, particularly for families without a designated family physician or with barriers to in-person consultations. These challenges contribute to the inappropriate use of emergency departments and increased pressure on primary care services. Telemedicine has emerged as a viable alternative to increase access to pediatric care, particularly for families facing geographic, economic, or system-related challenges. Incorporating artificial intelligence (AI) into asynchronous telemedicine platforms could improve clinical decision-making, lessen administrative work, and make better use of doctors' time. However, real-world evidence on AI-supported asynchronous models in pediatric primary care remains sparse, notably within the Portuguese National Health Service (NHS) environment.

Objective:

The primary objective of this study is to evaluate satisfaction and perceived quality associated with the delivery of paediatric care through an AI-supported asynchronous telemedicine platform (Usawa Care) in a primary care context. Secondary objectives include assessing perceived advantages in access to care; exploring perceived impacts on the use of emergency departments and primary care consultations; evaluating operational efficiency, including clinician active time per interaction; exploring user and clinician experiences with asynchronous communication and AI support.

Methods:

This is a mixed-methods study conducted in ULS Almada-Seixal. The intervention comprises an asynchronous messaging platform akin to WhatsApp (Usawa Care) that incorporates AI algorithms to organize patient-submitted data and propose healthcare treatments. All AI-generated recommendations are reviewed, validated, and changed as needed by paediatricians before being delivered to families. The project will enroll 50-100 families with children aged 0-18 years who are currently utilizing the platform as part of a municipal pilot program. Quantitative data will be collected using original questionnaires provided to parents and physicians, platform utilization metrics (number of interactions, physician active time per case, response times), and self-reported averted healthcare visits. Primary outcomes include user satisfaction scores, perceived quality and safety of care, physician time efficiency, and number of avoided in-person visits. Secondary outcomes include platform adoption rates, technical feasibility, and barriers/facilitators to implementation. The ethics committees of ULS AS that is taking part have given its’ assent.

Results:

The project was founded in July 2025. Ethics approval for both sites was received in August and October 2025. Recruitment at the Almada-Seixal site is completed, with roughly 50 families enrolled. Complete results, including quantitative and qualitative findings, are expected to be published in late 2026.

Conclusions:

This study will offer real-world data on the role of AI-supported asynchronous telemedicine in paediatric primary care. Findings may impact policy decisions about digital health deployment, lead to the establishment of clinical standards for telemedicine in paediatrics, and contribute to understanding how AI can safely improve clinical decision-making while retaining human oversight. The work tackles a key gap in knowledge regarding the acceptability, safety, and efficiency of asynchronous AI-enhanced care models in resource-constrained healthcare systems.