Abstract
AIM: To determine whether the combined use of periapical radiography (PR) and cone beam computed tomography (CBCT) improves agreement among three endodontic experts compared with using either diagnostic tool alone. METHODOLOGY: Sixty clinical records with CBCT scans were selected from cases performed by postgraduate and master's students in endodontics at the University of Barcelona Dental Hospital. Three endodontic experts independently evaluated each case using a standardised questionnaire. Initially, they assessed the PR images. After 4 months, they reassessed the cases using CBCT. One year later, they reviewed both PR and CBCT together. Agreement levels were analysed for pulpal and periradicular diagnoses, treatment recommendations and aetiological assessments. RESULTS: Inter-observer agreement improved across the study phases, peaking in phase 3-particularly for treatment recommendations. In this final phase, where both PR and CBCT were evaluated together, the agreement rate for pulpal diagnosis was 73.33%, with substantial reliability (κ = 0.67). Periradicular diagnosis showed a 40.00% agreement rate, with moderate inter-observer concordance (κ = 0.44). The lowest agreement was observed in the aetiology category (28.33%), with moderate concordance (κ = 0.50). Agreement on treatment recommendations reached 43.33% (κ = 0.51) and was the only category to show statistically significant differences between phases (p = 0.01). No significant differences were found for pulpal diagnosis (p = 0.39), periradicular diagnosis (p = 0.19) or aetiology (p = 0.36). Intra-observer agreement varied among experts, with some showing lower consistency in their responses when evaluating PR, CBCT and the combination of both, highlighting the influence of imaging modality on clinical decision-making. CONCLUSIONS: The combination of PR and CBCT improved inter-observer agreement across all categories compared to the use of a single imaging modality. However, this improvement reached statistical significance only for treatment recommendations. These findings suggest that, whereas combining PR and CBCT may support more consistent therapeutic decision-making, its impact on diagnostic agreement may vary depending on the specific diagnostic domain.