Validity of scoring caries and primary molar hypomineralization on intraoral photographs

AIM: The aims of this study were to assess whether intraoral photographs could be used to score caries and hypomineralization on primary molars (Using adapted Molar Incisor Hypomineralization (MIH)-criteria), and also to assess the reliability and validity in 3-7 year-old Dutch children of these scores by comparing them to direct clinical scorings. STUDY DESIGN: Cross-sectional study. METHODS: In this study 62 children (38.7% girls) with a mean age of 4.96 years (SD 1.27) participated. The children were rated clinically by their own dentist (authors JV or ME) for caries reaching the dentine in their primary molars (WHO criteria) and also for primary molar hypomineralization using the adapted MIH-criteria. For the intraoral photographs, a digital intraoral camera was used. The two paediatric dentists rated all the intraoral photographs on caries and hypomineralizations on the second primary molars, using the same criteria for the clinical scoring as for the scoring of the photographs. They scored independently, at least 2 weeks after the initial clinical scoring to avoid observational bias with the clinical scoring. STATISTICS: This clinical observation was used as the gold standard from which sensitivity, specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV) and the Positive and Negative Likelihood Ratio were computed. To test the intra-observer agreement 25 % of the photographs were scored again, at least 2 weeks after the initial scoring of the images. Inter- and intra-observer agreement were tested using Cohen's Kappa. RESULTS: The mean prevalence of clinically detected caries at tooth level was 46.7% and the mean prevalence of clinically detected hypomineralizations in second primary molars at tooth level was 21.8%. The sensitivity of assessing caries using intraoral photographs was 85.5%, the specificity 83.6%, the positive likelihood ratio 5.2 and the negative likelihood ratio was 0.17. For Primary Molar Hypomineralization (DMH) the sensitivity was 72.3%, the specificity 92.8%, the positive likelihood ratio 10.1 and the negative likelihood ratio was 0.30. The inter-observer agreement yielded the following Cohen's Kappa scores: for caries 0.76 and for DMH 0.62. The intra-observer agreement was for caries 0.80 (ME) and 0.72 (JV) and for DMH 0.95 (both ME and JV). CONCLUSIONS: From this investigation it was concluded that the sensitivity, specificity and the likelihood ratio of scoring caries and DMH on photographs made with an i