Assessment of different methods for diagnosing dental caries in epidemiological surveys

The aims of the study were: (i) to assess different clinical diagnostic methods of dental caries during epidemiological surveys; (ii) to determine which combinations of methods and diagnostic adjuncts show the best performances in epidemiological surveys when compared with examinations performed in a traditional dental setting (standard); (iii) to evaluate the influence of including noncavitated (NC) lesions in dental caries estimation.

Forty 12-year-old children were divided into low and moderate caries prevalence groups. The individuals were submitted to 12 epidemiological examinations (in an outdoor setting), which combined three methods (blade, mirror and mirror + CPI (Community Periodontal Index) dental probe) with or without diagnostic adjuncts (previous dental brushing and dental drying). The last examination was performed in a traditional dental setting (standard examination). The unit of measure was the DMFS (decayed, missing and filled surfaces) index according to WHO criteria. The variance analysis, Dunnet's and Tukey's tests were applied.

For the DMFS analysis, the visual/tactile method, with or without diagnostic adjuncts, was the best method for both groups, presenting a performance higher than 90% when compared with the standard examination, except for the examinations without previous dental brushing for the low caries prevalence group. Previous dental brushing was more relevant than dental drying (P = 0.0054). All of the epidemiological examinations underestimated the NC diagnosis even with the association of diagnostic adjuncts when compared with the standard examination.

The visual-tactile (for both groups) and the visual (mirror) methods plus dental brushing (for the moderate group) are appropriate for diagnosing cavitated lesions, but not NC lesions.