Conversion of plaque-area measurements to plaque index scores. An assessment of variation and discriminatory power.
Plaque areas recorded graphically or photographically provide a permanent record of plaque accumulations on teeth at a moment in time. As such, these records could be re-evaluated and converted into other index scores. The purpose of this study was to determine the reproducibility of scoring a plaque index from previously recorded plaque areas and to compare such scores with the original scores of the same index. A randomised blind, crossover study comparing 5 treatments for plaque inhibition scored by plaque area and index was chosen. 2 examiners, the original scorer PRH and another, NC, 2x scored the plaque area tooth charts according to the criteria of the plaque index system used in the original study. Standard deviations of the differences showed intra-examiner repeatability to be high particularly for the original examiner. Inter-examiner reproducibility for the original index scores was considered good but less than for intra-examiner repeatability. Correlation coefficients were complimentary to the differences analysis, being very high within examiners and less high for between examiners and original and rescored index. Separation between distributions of plaque area measurements for consecutive values of the index were particular good for scores 2 versus 3 and 3 versus 4 and less good for 1 versus 2 and 4 versus 5. Reanalysis of the study for treatment differences using rescored data revealed a similar level of significance as using the original data. Rescored index had similar discriminatory power for the study as plaque area and original plaque index when both were derived from the same buccal tooth surfaces. However, discriminatory power was less by comparison with original plaque index derived from the buccal surfaces of all teeth. It is concluded that plaque area provides a permanent record of plaque distribution which can be converted into index data at a later date. Such data collection could make possible comparisons between studies using different indices.