TO THE EDITOR:

The call for caries diagnosis by caries lesion activity assessment is growing. Lesion progression monitoring is difficult and requires careful and time-consuming record-keeping, whereas a single time-point activity assessment makes immediate treatment decisions, mainly involving preventive care, possible. Nyvad and co-workers presented a formal visual/tactile scoring system for lesion activity (1999). They also recently presented a clinical validation of this method for non-cavitated enamel lesions (2003), an excellent effort. Using their data, we would like to extend their analysis.

Nyvad and co-workers analyzed the results from a previous study involving two groups: a group that received daily supervised toothbrushing with a fluoride-containing toothpaste, and a control group that received neither supervised brushing nor fluoride. Baseline and three-year lesion assessments were available. The authors present their case as involving both construct validity and predictive validity. Most important is the case the authors present for the predictive validity of the assessment. The ‘proof of the pudding’ of caries activity measurement must surely lie in the validation through observed caries progression. The authors conclude that, in the control group, lesions diagnosed at baseline as active, non-cavitated (ANC) have a relative risk (RR) of 1.24 to progress in a three-year period to a cavity or filling or be extracted (CFE) as compared with lesions that, at baseline, were diagnosed as inactive, non-cavitated (INC). For the fluoride group, this RR was 1.04.

We were interested in what happened to the non-cavitated lesions that did not progress. Therefore, we transformed the percentages in Table 1 of Nyvad et al.(2003) to numbers, using the totals given. The results may deviate from the original ones by a few units. From these results, we calculated the numbers and fractions of ANC and INC lesions that progressed to CFE, that remained non-cavitated, and that regressed to sound. From those, we determined the relative risks of ANC lesions compared with INC lesions. These calculations are presented in the Table.

We find a RR for progression of 1.01 and 1.24 for the F-group and the control group, respectively, which agrees with the values in the paper. The discrepancy between 1.01 and the value of 1.04 in the paper can be ascribed to our necessarily inaccurate determination of the numbers in the categories. In addition, we find RR values for regression to sound of 1.45 in the F-group and 1.25 in the control group. This does imply that, in the control group (the non-F group), the RR of regression is as large as the RR of progression. In the fluoride group, the RR of regression is even higher. This implies that so-called active lesions are not specifically caries-active (suggesting only progression), but rather are more prone to change in general than are so-called inactive lesions. This is supported by data included in Table 1 of Nyvad and co-workers (2003), where, in the fluoride group, 53% of INC lesions remained INC, while only 19% of ANC lesions remained ANC. For the control group, these numbers were 42% and 28%, respectively.

The factors that cause this difference in susceptibility are as yet unknown. They may, for instance, involve the ‘openness’ of the lesion surface, the age of the lesion, or its progression stage. At this moment, we therefore suggest only that the terminology be changed to unstable vs. stable lesions.