Gamified Training and AI Support for Improving Ear Infection Diagnosis in Primary Care

N/ANot Yet RecruitingNCT07246551

Vastra Gotaland Region200 enrolled

Overview

Acute otitis media (AOM), or middle ear infection, is one of the most common childhood infections and a leading cause of antibiotic prescribing in primary care. Diagnosing AOM can be challenging, as ear symptoms and eardrum appearances often overlap with mild or transient conditions. This uncertainty may lead to both unnecessary antibiotic use and missed cases requiring treatment, affecting patient safety and contributing to antibiotic resistance. This study evaluates two digital tools designed to support more accurate diagnosis and treatment decisions in primary care: 1. AOM Dx \[diagnosis\] Trainer (Otitspelet) - a gamified digital training program for physicians that provides interactive exercises using eardrum images and patient cases, with direct feedback to improve diagnostic accuracy and adherence to guidelines. 2. AI-based diagnostic support - a system that analyses tympanic membrane images, with and without symptom and tympanometry data, to evaluate its potential for future diagnostic use. The trial is coordinated by the Västra Götaland Region (VGR) in collaboration with Umeå University and conducted across four Swedish regions: Västra Götaland, Västerbotten, Östergötland, and Skåne. VGR leads the evaluation of the AOM Dx Trainer intervention, while Umeå University leads the AI development and retrospective diagnostic analyses. The study is carried out as a multicentre, cluster-randomised controlled trial in primary care, where participating primary care centres are randomised to either the training intervention or standard care. Physicians are the research participants under evaluation. At intervention centres, physicians complete training with the AOM Diagnosis (Dx) Trainer before study start; at control centres, no training is provided. Each participating centre then conducts an 8-week observation period, during which physicians diagnose and manage patients with new-onset ear symptoms. Patients are included only to allow evaluation of physicians' diagnostic and treatment decisions and to provide data for AI analysis. Estimated patient enrollment is \~200. Depending on centre size and recruitment success, up to 20 primary care centres across four Swedish regions - Västra Götaland, Västerbotten, Östergötland, and Skåne - will participate. After each consultation, research nurses collect tympanic membrane images and tympanometry data from patients who have given informed consent. These data are used for expert panel reference diagnoses and retrospective AI analysis; no information is shared with treating physicians. The primary outcome is diagnostic accuracy of tympanic membrane assessment by physicians trained with the AOM Dx Trainer compared with untrained physicians, using expert consensus as the reference standard. Secondary outcomes include adherence to treatment guidelines and antibiotic prescribing rates. The AI system's diagnostic performance will also be benchmarked against the expert panel and physician groups. By combining educational and technological innovation, this study aims to enhance diagnostic precision, improve guideline adherence, and reduce unnecessary antibiotic use in primary care-strengthening antimicrobial stewardship and providing a scalable model for future infection management.

This multicentre, cluster-randomised controlled trial evaluates two digital innovations for improving diagnostic accuracy and treatment quality in acute otitis media (AOM) management in Swedish primary care: a gamified training tool for physicians (AOM Diagnosis \[Dx\] Trainer) and an AI-based diagnostic support system. The study assesses diagnostic accuracy, adherence to evidence-based treatment guidelines, and comparative AI performance. The trial is coordinated by the Västra Götaland Region (VGR) in collaboration with Umeå University and conducted across four Swedish regions: Västra Götaland, Västerbotten, Östergötland, and Skåne. Up to 20 primary care centres will participate, depending on centre size and recruitment rates. Centres are randomised to either the AOM Dx Trainer intervention or standard care (control). Study design and participants: Both physicians and patients are research participants. Physicians in the intervention arm complete the AOM Dx Trainer before patient inclusion begins, while control physicians receive no training. Each centre recruits consecutive patients with new-onset ear symptoms (≤1 month) during an 8-week inclusion period (intervention or control). Consultations are conducted according to standard clinical practice. After each visit, research nurses ensure eligibility and consent and collect tympanic membrane images and tympanometry data solely for research purposes. Interventions: The AOM Dx Trainer is a gamified, case-based training tool designed to improve recognition of AOM and related conditions. Physicians classify anonymised tympanic images with symptom vignettes into diagnostic categories and receive immediate feedback. Training continues until a predefined performance threshold is reached. The AI diagnostic tool, developed at Umeå University, uses convolutional neural networks (CNNs) to analyse tympanic images, with or without symptom and tympanometry data. AI analyses will be performed retrospectively in a laboratory setting and will not influence patient care. Data collection: After each visit, patients meet a research nurse who ensures eligibility and consent, and records demographics, symptom severity (AOM-SOS v5), and potential complicating factors (e.g., severe pain despite analgesics, immunosuppression, previous ear surgery, cochlear implant). The physician's diagnosis and any antibiotic prescription (drug and duration) are documented for later comparison with guideline recommendations and expert panel consensus. Physician characteristics are recorded under pseudonymised IDs. Tympanic membrane images are captured using CE-marked video otoscopes (EarPenguin) and tympanometry devices. All data are entered into case report forms (CRFs) and stored securely but are not shared with treating physicians, ensuring real-world diagnostic conditions. Data handling and ethics: All data are pseudonymised. Code keys are securely stored within each region. The national coordinating centre in Gothenburg oversees data management and quality assurance. Data are stored on GDPR-compliant servers. Ethical approval has been granted by the Swedish Ethical Review Authority (Ref. 2025-03523-01). Expert panel reference diagnoses: Tympanic images and tympanometry results are reviewed retrospectively by an expert panel (two ENT specialists and one senior GP). Consensus diagnoses serve as the reference standard for assessing diagnostic accuracy among physicians and for benchmarking AI performance. Primary outcome: Diagnostic accuracy of tympanic membrane classification (normal, AOM, erythematous membrane without effusion, or otitis media with effusion) compared with expert consensus. Secondary outcomes: Adherence to national treatment guidelines; antibiotic prescribing rates and duration; and comparative diagnostic performance between physicians and AI configurations. Sample size and power: The sample size calculation targets the primary research question. Using a bivariate logistic regression as a proxy (outcome: correct vs. incorrect diagnosis; predictor: intervention status), unpublished data from our group suggest approximately 50% diagnostic accuracy among physicians without AOM Dx Trainer training. We hypothesize an improvement to around 75% among trained physicians, which is considered a clinically relevant threshold. Assuming a two-sided alpha of 0.05 and 95% power, a minimum of 195 patients is required (G\*Power 3.1.9.7). Therefore, approximately 200 patients will be recruited, with roughly equal numbers in the intervention and control arms. This target aligns with the planned multicentre cluster-randomised design and is feasible within the established Swedish primary care research network.

Eligibility

Sex: ALL

Medical Language ↔ Plain English

Inclusion Criteria Patients: * Newly developed ear discomfort or symptoms within the past month. The patient seeks care at a participating primary care centre and is assessed by a physician because of these ear-related symptoms. Alternatively, for young children: respiratory tract infection with concurrent signs or symptoms suggesting possible ear involvement, even if the child cannot clearly express ear pain. * Age: Children and adults of all ages are eligible to participate. * Informed consent: The patient (or guardian for minors) agrees to participate and provides written informed consent. For children under 15 years, consent is obtained from both guardians before any data storage. * The patient must have been managed by a physician who has consented to participate in the study (as a physician participant) and who is a GP specialist, GP resident, or junior physician (intern or basic training physician) working at the primary care centre. * At intervention centres, the managing physician must have completed the AOM Diagnosis (Dx) Trainer and reached the required score threshold ("diploma") before including patients. Inclusion Criteria Physicians: * General practitioners (GP specialists), GP residents, or junior physicians (interns or basic training physicians) working clinically at participating primary care centres. * Have provided written informed consent to participate in the study as physician participants. * At intervention centres, must have completed the AOM Dx Trainer and achieved the required diploma level before patient inclusion. * At control centres, receive no access to the AOM Dx Trainer during the study period. Exclusion Criteria: * Withdrawal of consent by the participant (physician or patient).

Outcomes

Primary Outcomes

Diagnostic accuracy of tympanic membrane assessment by primary care physicians

Diagnostic accuracy of tympanic membrane classification (normal, acute otitis media, erythematous membrane without effusion, or otitis media with effusion) by primary care physicians trained with the AOM Dx Trainer compared with untrained physicians (control group). Accuracy is defined as correct versus incorrect diagnosis relative to the expert panel consensus. Analysed using mixed-effects logistic regression adjusted for physician sex, age, and training level (GP specialist, resident, or junior physician), with physician ID included as a random effect to account for clustering at the physician level.

Time frame: During the 8-week inclusion period at each participating centre (per patient consultation).

Secondary Outcomes

Adherence to national AOM treatment guidelines

Proportion of correct vs. incorrect treatment decisions (antibiotic vs. watchful waiting, drug choice, duration) according to national guidelines, comparing trained vs. untrained physicians. Analysed using mixed-effects logistic regression with the same covariates and random effect as for the primary outcome.

Time frame: During the 8-week inclusion period at each participating centre (per patient consultation).

Diagnostic performance across groups

Sensitivity, specificity, predictive values, and ROC analyses (with 95% confidence intervals) will be calculated for classification of the four tympanic membrane categories. Performance will be compared against the expert panel, separately for: 1. physicians without training, 2. physicians with AOM Dx Trainer training, 3. the AI diagnostic tool, evaluated with (i) image only, (ii) image + symptoms, (iii) image + tympanometry, and (iv) all modalities combined. In this project, the AI will be evaluated retrospectively in a laboratory setting and will not influence clinical consultations; all patient care remains the responsibility of the treating physician.

Time frame: Retrospective analysis after completion of data collection (expected within 12 months of final patient inclusion).

Gamified Training and AI Support for Improving Ear Infection Diagnosis in Primary Care

Overview

Conditions

Interventions

Eligibility

Outcomes

Primary Outcomes

Secondary Outcomes

Central Contacts