In the context of pediatric dentistry, a critical question is how different health care professionals, namely dentists and pediatricians, interpret and act on teledental data. Comparing the accuracy, reliability, and effectiveness of teledental diagnoses made by both groups is essential to understanding the potential of teledentistry in this interdisciplinary setting. Sensitivity (true positive/all positive findings) and specificity (true negative/all negative findings) are traditional measures used to evaluate diagnostic quality, and these will be critical in determining whether dentists and pediatricians can make equally reliable diagnostic and management decisions using teledentistry17. The following null hypotheses were tested: 1. There is no significant difference between digital teledental findings (IOS) and the gold standard of analog visual examination (VIS). 2. There is no significant difference in the diagnostic reliability between dentists and pediatricians. 3. Sensitivity and specificity of digital teledental assessments (IOS) do not significantly differ from those of analog visual examination (VIS), regardless of whether assessments are conducted by dentists or pediatricians.
The study was conducted according to the guidelines of the Declaration of Helsinki, approved by the Ethics Committee of the Medical Faculty of the Justus Liebig University (JLU, ref. no. 46/20).
Sample size calculation was done using the formula described by Bujang et al. based on the expected diagnostic performance for the IOS. With a target power of 90% and an alpha value of 0.05, 70 participants were calculated as necessary.
All children aged 4-17 years who underwent regular dental examination at the Department of Pediatric Dentistry at JLU (Germany) between August 2022 and February 2023 participated in this study. Therefore, the participants comprised a representative cross-section of patients treated at the pediatric dentistry department of a university hospital. Common treatment measures included the management of dental caries and structural anomalies, preventive interventions, trauma care, treatment under general anesthesia, specialized care for children with special healthcare needs, and thorough diagnostic assessments.
In addition to a thorough verbal explanation of the study procedures, written informed consent was obtained from all parents/guardians.
To ensure comparable test conditions and obtain reproducible data based on predefined criteria, each patient was initially examined by two experienced dentists (N.S.-W. and M.A.S.) using an analog method, with the visual examination (VIS) defined as the reference method and gold standard for subsequent analysis. The IOS required for teledental diagnosis were also performed by the same examiner. Subsequently, the teledental digital findings of the IOS were evaluated by two independent examiners: one dentist (DEN, E.S.) and one pediatrician (PED, T.F.).
Prior to the study, E.S. and T.F. were trained and calibrated by senior researchers (N.S.-W. and M.A.S.). The training period lasted two weeks and consisted of both theoretical and practical parts. The theoretical part included structured instruction on the clinical features and diagnostic criteria of the conditions under study, using standardized materials and case examples (e.g., following the approach described by Kühnisch et al.). The practical part involved supervised clinical assessments of ten pediatric patients under the guidance of N.S.-W. and M.A.S. To ensure diagnostic consistency, readiness for independent assessment was evaluated based on each examiner's performance in a benchmark assessment, including both intra-rater reliability (agreement within the same examiner) and inter-rater reliability (agreement with expert diagnoses). According to the classification by Landis and Koch, the level of agreement was almost perfect, with a Cohen's kappa coefficient (κ) of 0.95. Only after achieving this benchmark were E.S. and T.F. permitted to conduct independent assessments within the study. While E.S. received extended training aimed at reaching a diagnostic level comparable to that of the senior dental researchers (M.A.S. and N.S.-W.), T.F. as pediatrician, received tailored instruction focused on core dental diagnostic skills. This included basic knowledge of caries recognition, structural dental anomalies (e.g., MIH), differentiation between primary and permanent dentition, and identification of soft tissue signs such as swelling, redness, and plaque accumulation. It is important to acknowledge that, despite this structured theoretical and practical training, a pediatrician cannot attain the diagnostic proficiency of a trained dentist. Nonetheless, this tailored training allowed for a realistic evaluation of the potential role of pediatricians in supporting teledentistry-based screening or triage in pediatric patients.
The presence of caries was verified as yes/no decision (dentin caries) under standardized lighting (25,000 lx), with a mouth mirror and an air syringe. Both primary dentition (pD) and permanent dentition (PD) were included in the evaluation.
The IOS were captured by Trios 4 (software version 20.1.4, 3Shape, Copenhagen, Denmark) that was used to scan the entire upper and lower jaw following a standardized scanning path, which included the occlusal, oral, and buccal surfaces. Prior to scanning, the intraoral scanner was calibrated according to the manufacturer's instructions. Figure 1 displays an example of IOS.
To ensure standardized comparability between the established analog VIS of the oral cavity, considered the gold standard and reference method, and the new digital teledental diagnosis based on IOS, a diagnostic sheet was developed to record the following aspects of oral health: overall dental status, including caries (yes/no), restorations (yes/no) and if yes, type of restorations (fillings materials (composite/glass ionomer cements), fissure sealing, steel crown), and molar incisor hypomineralization (MIH). Additionally, urgency of dental intervention (urgent/not urgent) and treatment recommendations (no treatment/prophylaxis/check-up for suspected caries lesions/immediate dental intervention required) were also recorded. Figure 2 shows the study procedure.
To compare the digital (IOS) and analog (VIS) findings, Gwet's AC1 (κ) was used. Kappa values were categorized as poor agreement (< 0.00), slight agreement (0.00-0.20), fair agreement (0.21-0.40), moderate agreement (0.41-0.60), substantial agreement (0.61-0.80), and almost perfect agreement (≥ 0.81). Additionally, Cohen's d, a measurement of effect size (d = 0.2: small, d = 0.5: medium, d = 0.8: large), was used to assess significant differences between the two digital examiners, DEN and PED in terms of caries detection (pD and PD), restorations in general and type of restorations (pD and PD), MIH, urgency of dental intervention and treatment recommendation.
To assess the diagnostic quality of the digital findings (IOS) in comparison to the analog findings (VIS), sensitivity (SE), specificity (SP), and area under the curve (AUC) values were calculated using receiver operating characteristic (ROC) analysis for caries detection and restorations (pD and PD), MIH and urgency of dental intervention. Higher AUC values indicate better diagnostic quality.
SPSS Statistics (version 28, IBM, Armonk, NY, USA) was used for the statistical analysis. Due to the study design of an examination on a tooth level, data was further analyzed separately for pD and PD.