Assessment of Pediatric Telemedicine Using Remote Physical Examinations With a Mobile Medical Device: A Nonrandomized Controlled Trial

Abstract

Importance: The number of innovations in health care based on the use of platforms, digital devices, apps, and artificial intelligence has grown exponentially in recent years. When used correctly, these technologies allow inequities in access to health care to be addressed by optimizing care and reducing social and geographic barriers. However, most of the technological health care solutions proposed have not undergone rigorous clinical studies. Objective: To assess the concordance between measurements from a remote physical examination using a mobile medical device and measurements from a conventional in-person physical examination. Design, Setting, and Participants: This nonrandomized controlled trial was conducted from January 1 to December 31, 2020. The clinical parameters compared were heart rate; body temperature; heart, lung, and abdominal auscultation; otoscopy; throat and oral examination; and skin examination. A total of 690 patients with clinical stability and various symptoms who were seen in the emergency department of 2 Brazilian pediatric hospitals were eligible to enter this study. Main Outcomes and Measures: The primary outcome was concordance between measurements from a telemedicine physical examination using a mobile medical device and measurements from a conventional in-person physical examination. The secondary outcome was the specificity and sensitivity of the digital device, considering the conventional in-person consultation as the gold standard. Results: Among 690 patients, the median (IQR) age at study entry was 5 (1-9) years; 348 (50.4%) were female, and 331 (48.0%) presented with a chronic disease. Regarding the primary outcome, the concordance values were 90% or greater for skin examination (94% for rash, 100% for hemorrhagic suffusion, and 95% for signs of secondary infection), characteristics of the mucosa (98% for hydration and 97% for coloring), and heart (95% for murmur, 97% for rhythms, and 98% for sounds), lung (91% for adventitious sounds, 97% for vesicular sounds, and 90% for fever), and abdominal (92% for abdominal sounds) auscultations. Concordance values were lower for otoscopy (72% for the ear canal and 86% for the tympanic membrane), throat and oral examination (72%), and rhinoscopy (79% for mucosa and 81% for secretion). The specificity was greater than 70% (ranging from 74.5% for the ear canal to 99.7% for hemorrhagic suffusion) for all variables. The sensitivity was greater than 52% for skin examination (58.0% for rash and 54.8% for signs of secondary infection), throat and oral examination (52.7%), and otoscopy (66.1% for the ear canal and 64.4% for the tympanic membrane). Conclusions and Relevance: In this study, measurements from remote physical examination with a mobile medical device had satisfactory concordance with measurements from in-person physical examination for otoscopy, throat and oral examination, skin examination, and heart and lung auscultation, with limitations regarding heart and lung auscultation in infants and abdominal auscultation in children of all ages. Measurements from remote physical examination via a mobile medical device were comparable with those from in-person physical examination in children older than 2 years. These findings suggest that telemedicine may be an alternative to in-person examination in certain contexts and may help to optimize access to health care services and reduce social and geographic barriers.