Aim: Accurate detection and diagnosis of caries in primary molars is challenging, especially for proximal lesions where direct visual examination is difficult. Therefore, the aim of this in-vivo and in-vitro study was to assess the validity, reproducibility and acceptability of a laser fluorescence pen and compare these outcomes with those of conventional methods of proximal caries detection in primary molar teeth.
Methods: Eighty-two children (aged 5-10 yrs) were recruited. Initially 1030 proximal surfaces were clinically examined using meticulous visual examination (ICDAS), bitewing radiographs, and a laser fluorescence device (LF pen). Temporary tooth separation (TTS) was achieved for 447 surfaces and these surfaces were re-examined visually (ICDAS) and by the LF pen. The teeth were subsequently extracted and serially sectioned for histological validation. Proximal surfaces were further assessed in-vitro using direct visual examination and the LF pen. The validity of all diagnostic methods was assessed. Results of both in-vivo and in-vitro assessments were compared. Intra- and inter-examiner reproducibility were assessed, the second examiner re-examined 10% of surfaces. Patient acceptability of the different diagnostic methods was measured using self-completed questionnaires.
Results: At D₁ (enamel and dentine caries) diagnostic threshold, the sensitivity of ICDAS visual examination, TTS, radiographic examination and LF pen examination was 0.52, 0.75, 0.14, 0.58 and the specificity at this threshold was 0.89, 0.88, 0.97, 0.85 respectively. At D₃ (dentine caries) diagnostic threshold, the sensitivity of the ICDAS examination, TTS, radiographic examination, and LF pen examination was 0.42, 0.49, 0.71, 0.63 respectively, while the specificity was 0.93 for both ICDAS examination and TTS, and 0.98 and 0.87 for radiographic and LF pen examinations respectively. ROC comparison of the different methods showed the radiographic examination to be superior at D₃ level. Intra-examiner reproducibility was ‘substantial’ to ‘almost perfect’ for all examinations, with the Kappa coefficient varying from K=0.75 at D₁ to K=0.95 at D₃. Inter-examiner reproducibility for ICDAS and radiographic examinations also demonstrated ‘substantial’ to ‘almost perfect’ agreement which varied from K=0.73 at D₁ to K=0.0.85 at D₃. The LF pen had significantly higher validity in-vitro than in-vivo. However, in-vitro LF pen readings were significantly different from the in-vivo readings (P<0.05). Regarding acceptability of these different approaches, children found TTS to be significantly less acceptable than the other methods.
Conclusions: Meticulous visual examination should be supported by radiographs. The LF pen did provide additional diagnostic information particularly at the D₁ threshold but not as much as
