A critical evaluation of some problems associated with clinical caries trials by computer simulation.

Unbiased examiners' errors have little effect on the outcome of a clinical trial, provided that they are not larger than 10%. Unbiased error greater than 10% would tend to make the variance larger than expected. Demonstration schemes as discussed in this paper may show spurious positive effects, due to attempts by examiners to avoid reversals in diagnosis. Professional intervention is mostly responsible for the SU component - the direct transition from sound surface to DMF. These surfaces are lost for reasons other than caries activities per se. It should not be included in the measure of caries experience in clinical trials. The component UU represents the surfaces that show incipient caries at two consecutive examinations. It represents mostly past history and has little bearing on current caries activities. Since the DMFS = IU + SU + UU, and the contributions of IU as a rule are small, the use of DMFS as a measure of caries experience in clinical trials is unjustified. The NCI = SI + IU deals directly with the stage where the cariostatic agent is supposed to be working. There is a definite advantage to using this index as the caries experience measure in clinical trials.