Reproducibility among examiners of the ICDAS-radiographic caries assessment in primary molar teeth.

Objectives: To evaluate inter- and intra-reproducibility among three examiners for the caries detection/assessment using the ICDAS-radiographic criteria in primary molar teeth.
Methods: Three members of the Caries-Research-Unit UNICA familiarized with the ICDAS system participated. Initially, a 2-day training course in the ICDAS-radiographic criteria was conducted by an expert (AC). The training included a 1-day theoretical/pre-clinical session with bite-wing radiograph Power-Point images, and a 1-day clinical session assessing pre-scored surfaces in radiographs. The radiolucency was classified according to the ICDAS radiographic-lesion severity classification in: 0: no radiolucency, 1: radiolucency in outer-enamel half; 2: radiolucency in inner-enamel half ± EDJ; 3: radiolucency until outer-dentine third; 4: radiolucency in middle-dentine third; 5: radiolucency in inner-dentine third; 6: radiolucency into the pulp. The examiners assessed the radiolucency of occlusal/proximal surfaces of 1^st/2^nd primary-molar teeth in 125 bite-wing radiographs (12 surfaces/x-ray; 1500 surfaces) of 4 to-6 yr. olds, using a Mattson’s magnifying glass (Dental X-ray, Copenhagen, Denmark) and adequate light. Filled/non-assessable surfaces and due-to-caries missing teeth were discarded. Codes were included in an Excel data base and weighted-Kappa and perfect agreement values were calculated against the expert examiner (AC). After 1 week a second examination was conducted in similar conditions in order to determine examiners intra-reproducibility.
Results: Out of the 1500 examined surfaces, 178 were discarded (filled-surfaces: 45; due-to-caries missing tooth: 24, and not-assessable-surfaces: 109), for a total of 1322 scored surfaces per examiner. Expert examiner codes distribution was: ICDAS-0: 943, ICDAS-1: 116, ICDAS-2: 106, ICDAS-3: 85, ICDAS-4: 45, ICDAS-5: 15, ICDAS-6: 9. Examiners inter-reproducibility weighted-Kappa and perfect agreement values were: Examiner-1: 0.96, 98.2%; Examiner-2: 0.83, 92.4%; Examiner-3: 0.70, 85.5%; respectively. Correspondent intra-reproducibility figures were: 0.91, 99.8%; 0.78, 91.6%, and 0.70, 89.2%.
Conclusions: This study shows a substantial to almost perfect inter- intra-reproducibility among examiners for the ICDAS-radiographic criteria in primary dentition.