A multicentric study for early occlusal caries diagnosis

Method: After a signed informs consent, voluntary participants were randomly recruited in each respectively preventive department of Marseille, Montpellier, Rennes, Lyon and Paris V universities. Occlusal molar surfaces were observed (n=540). The investigation was running by two experimented dentists in each department, blind each other and pre-calibrated to ICDAS classification. 5 groups were used: G1-visual scoring with loupe (3.5x) (ICDAS), G2-radiographic scoring, G3-Soprolife® mode daylight, G4-Soprolife® mode fluorescence, G5-DiagnoPen®. Air polishing cleaning of the occlusal surfaces with sodium-bicarbonate powder (30s) was performed before any observations. Sensitivity, specificity of each measurement method and carious prevalence were measured. ROC and AUC curves were analysed. Visual inspection with ICDAS classification was used as gold standard. (Ethics committee approval, CPP Marseille)  

Result: Intra and inter-investigators kappa values were 0.87 and 0.79. Sensitivity and specificity values (%) were respectively  G1-60.67; 99.56. G2-NS. G3-86.2; 85. G4-85.7; 80.9. G5- 92.13; 53.64. Caries prevalence increased when using the fluorescence camera (G1-54%. G3-73%. G4-78%. G5-54%) and the specificity highly decreased with the laser system. Radiographic (G2) examination was the worst method for early occlusal caries diagnosis. 

Conclusion: With the limitation of this experimentation, Soprolife® seemed to be a reliable visual aid in the diagnosis of early occlusal caries.