Rationalization of quantitative tooth surface loss data for epidemiological research

Recent UK National Surveys have revealed a high prevalence of tooth surface loss (TSL) because of erosion in Children and Adolescents. Although digital surface mapping and surface matching techniques may be used to quantify its progression with time, reporting TSL of individuals as a function of either mean depth loss or volume loss, this can be inadequate when transferred for epidemiological analysis. For example, a tooth displaying multiple regions of depth loss may deserve to be distinguished from a tooth with a single localised area of erosion, although the total volume change may be the same. A potential solution, explored here, is to use both the quantitative wear data and colour coded surface representation plots generated by such methods to arrive at a single categorical wear score [1 (≤5% of surface exhibits TSL) to 5 (≥51% of surface exhibits TSL)]. Two examiners independently categorized, on two separate occasions, the TSL of 53 maxillary incisors (26 subjects observed at baseline and 9 months). Their performance was assessed for intra- and inter-examiner agreement by; (i) calculating the percentage of agreement, (ii) a Wilcoxon matched-pairs signed rank test. Thereafter, in cases of disagreement a consensus score was allocated and a regression analysis of this versus the numerically derived percentage of the surface that had undergone change was carried out. Intra- and inter-examiner agreement was 100 and 88·7% respectively and for the inter-examiner comparison P = 0·0456. The categorical scores and the quantitative wear data correlated linearly (R = 0·82). It is concluded that the method used to rationalize the quantitative TSL data is both reproducible and reflects the quantitative data. Its use should be helpful in epidemiological TSL studies but continued vigilance is required in cases of examiner disagreement.