Evaluation of the Medentum Diagnostik Platform for Pediatric ENT Conditions and Asthma

Overview

This study evaluates the diagnostic performance, usability, and clinical utility of the Medentum Diagnostik platform, a multisensor device with integrated software and artificial intelligence, in pediatric patients with asthma and ENT conditions. Participants will be assigned to either home-based use of the Diagnostik platform with telehealth-supported clinician evaluation or standard in-person care. The primary objective is to assess agreement between AI-generated diagnoses and clinician diagnoses obtained during study encounters, including telehealth evaluations in the Diagnostik arm and in-person evaluations in the standard-of-care arm. Secondary objectives include evaluating agreement between AI-generated diagnoses and independent research team diagnoses, as well as assessing usability, patient and caregiver satisfaction, healthcare utilization, and clinical outcomes such as asthma control and symptom resolution. This study will also evaluate differences in healthcare utilization, including in-person visits, emergency care use, and time to diagnosis and treatment, between study groups.

Study Design and Methodology This study is a randomized, controlled clinical trial designed to evaluate the diagnostic performance, usability, and clinical utility of the Medentum Diagnostik platform in pediatric patients with asthma and ear, nose, and throat (ENT) conditions, including acute otitis media and pharyngitis . Participants will be randomized to one of two groups: 1. home-based use of the Diagnostik platform with telehealth-supported clinician evaluation (intervention arm), or 2. standard-of-care clinical evaluation via in-person or telehealth visits without use of the Diagnostik platform (control arm). The study duration is 6 months. Intervention participants will use the Diagnostik platform during symptomatic episodes and for monthly monitoring (asthma patients), while control participants will receive routine clinical care. Follow-up data will be collected through platform data, medical record review, and patient-reported outcomes. Intervention and Data Collection Participants in the intervention arm will use the Diagnostik platform, which integrates a smartphone application with a multifunctional medical device, including a digital stethoscope, otoscope, thermometer, and pulse oximeter, to collect clinical data at home . During symptomatic episodes, participants will: Record lung sounds using the digital stethoscope Capture images of the ear or throat Enter symptom data and complete relevant questionnaires (e.g., Asthma Control Test) Measure vital signs, including temperature, heart rate, and oxygen saturation The platform's AI algorithms will analyze these data to generate a preliminary diagnostic assessment (e.g., presence or absence of wheeze, acute otitis media, pharyngitis). Intervention participants will then undergo a telehealth consultation with a clinician who will review the collected data, consider the AI-generated assessment, and establish a clinical diagnosis. Participants in the control arm will receive standard clinical care, including in-person or telehealth evaluations, during which clinicians will make diagnoses based on routine clinical assessment and testing. Primary Outcome Assessment The primary objective is to assess agreement between AI-generated diagnoses and clinician diagnoses obtained during study encounters. Clinician diagnoses will be established via telehealth evaluation in the intervention arm and via standard-of-care evaluation (in-person or telehealth) in the control arm. Diagnostic agreement will be evaluated separately for asthma and ENT conditions. Lung sounds will be assessed for the presence or absence of wheeze, and ear/throat images will be evaluated for acute otitis media and pharyngitis. Agreement between AI and clinician diagnoses will be quantified using percent agreement and Cohen's kappa. Secondary and Exploratory Assessments Secondary analyses will include evaluation of agreement between AI-generated diagnoses and research team diagnoses. Research diagnoses will be established by independent clinicians, including pediatric pulmonologists, based on blinded review of lung sound recordings and ear/throat images collected using the Diagnostik platform. Additional secondary outcomes will include usability, patient and caregiver satisfaction, healthcare utilization, and clinical outcomes such as symptom days, asthma control, rescue inhaler use, and time to symptom resolution. Healthcare utilization measures will include in-person visits, emergency department visits, and time to diagnosis and treatment. Exploratory analyses will assess the performance of the AI system across clinical conditions and care settings.