JMIR Publications

ABSTRACT

Background:

Telemedicine could bring pediatric expertise to children seen in primary emergency care without a pediatrician on site, but children seen by direct-to-consumer telemedicine receive antibiotics more often, and less often in line with guidelines, than children seen in person, a difference attributed to the clinician’s inability to examine the child. Examining a young child remotely requires watching the whole room, including caregivers, while viewing close-up images and hearing auscultation. Single-session setups oblige on-site staff to switch cameras and devices, and videoconferencing audio is processed for speech, not for heart and breath sounds.

Objective:

This study aimed to describe an always-on parallel multichannel telemedicine system linking a tertiary children’s hospital with a community after-hours primary emergency center and to report its early feasibility.

Methods:

Three channels stay open throughout each session: a conversation channel with a fixed wide-angle room camera; an examination channel carrying pen-type otoscope and pharyngeal cameras, a close-up camera, and radiograph display, with digital-stethoscope sound and electrocardiogram live-streamed separately; and a secure messaging service linking the remote pediatrician directly to reception, nursing, and pharmacy. The pediatrician sees and hears all channels at once, whereas the nurse positions the devices and selects examination images with a single button. We retrospectively reviewed the first 4 evening sessions (July to September 2026). Outcomes were patient characteristics, modalities used, disposition, technical events affecting clinical information, adverse events, prescribing, and same-night return for emergency care.

Results:

Of 14 children who attended, 10 completed a telemedicine visit; 4 were seen in person, 3 by caregiver preference and 1 after triage. The median age of the 10 children was 4.8 years (range 1 year 7 months to 14 years), and 5 (50%) were aged 4 years or younger. Auscultation was performed in real time in 9 children; other examinations included otoscopy, chest radiograph interpretation, rapid testing, and nurse-assisted abdominal and hip examination. Every visit, including dispensing and billing, was completed on site, 1 with on-site physician confirmation of abdominal findings. An antibiotic was prescribed in 1 visit, clarithromycin for an adolescent with suspected pneumonia on radiography; ear pain with a normal tympanic membrane on otoscopy, and suspected Mycoplasma infection with a nearly clear chest and a negative rapid test, were managed without antibiotics. No serious adverse events were reported, and no child returned for emergency care on the same night.

Conclusions:

An always-on parallel multichannel configuration was feasible for nurse-assisted pediatric primary emergency telemedicine, including in children younger than 5 years. Auscultation that conveys the fast, shallow breathing of young children is central to such a system. In this small series, antibiotic decisions rested on examination findings; whether such a configuration improves the appropriateness of prescribing requires comparative evaluation.