There are no clearly superior methods for diagnosing, predicting, and noninvasively treating dental caries.

Selection Criteria: To be included in the systematic review, all articles were required to meet predetermined criteria: the results of the studies should be relevant to the questions posed by the project, ie, have appropriate outcome measures and an appropriate follow-up period and study design.

Key Study Factor: With respect to diagnosis, the efficacy of the various methods was compared with reference methods that included studies of extracted teeth. For prediction, the material comprised mainly prospective studies, where the reliability of risk assessment could be evaluated. With respect to treatment of early caries lesions, it was a requirement for inclusion that the study had a control group.

Main Outcome Measure: For caries detection of accessible surfaces, visual-tactile testing was used. Radiographic diagnosis was used for approximal surfaces. For caries risk prediction, modeling was used to determine the influence of specific variables (eg, sugar intake). For noninvasive treatment assessment, size of lesion growth or shrinkage after treatment was used.

Main Results: Caries detection: Visual-tactile detection for accessible surfaces carried the best, but far from optimal, accuracy. Although specificity tends to be high with this method, sensitivity is low. Caries risk prediction: Streptococcus mutans count, salivary Lactobacillus count, salivary buffering capacity, visible plaque levels, and sugar intake have low accuracy. The best, but still far from optimal, predictor was past caries history. Noninvasive methods of caries treatment: Too few studies exist for meaningful analysis.

Conclusions: (1) Past caries is the best predictor of future caries. (2) A combination of visual-tactile and radiographic examination is better than either method alone. (3) There is insufficient evidence to support noninvasive treatment of early caries. STUDY OBJECTIVE AND METHODS: To perform a systematic review of published literature to (1) determine the evidence for the effectiveness of different methods (visual-tactile versus radiographic) for detecting caries; (2) identify the most prescient predictors of dental caries; and (3) evaluate the evidence supporting the use of noninvasive methods for the treatment of caries, particularly early caries.

Main Results And Conclusions: With regard to the questions posed by the study team: (1) Past caries is the best predictor of future caries; (2) a combination of visual- tactile and radiographic examination is better than either method alone; and (3) there is insufficient evidence to support noninvasive treatment of early caries.