Reliability of plaque and periodontal measurements estimated by the internal consistency method

Abstract

Abstract

The purpose of this paper was to assess whether the internal consistency method could be used to estimate reliability of periodontal and plaque measurements. This method is based on the assumption of comparability of data from the left and right halves of the mouth. The material comprised data from 61 subjects (mean age, 43 years). Percentage of tooth surfaces with plaque index scores of at least 1, percentage of bleeding units (GI = 2), and the mean pocket depth were calculated for each half of the mouth. At the group level the mean prevalence of the two sides of the mouth did not differ significantly with regard to any of the variables, and there was a substantial degree of bilateral symmetry for all variables as measured by Pearson's r. Reliability coefficients were computed in accordance with the Spearman-Brown formula. Highest reliability coefficient was observed for pocket depth measurements (0.97), followed by plaque measurements (0.95), and, finally, gingival bleeding measurements (0.90). The internal consistency method seems promising in estimating reliability in epidemiologic surveys, but further studies should be initiated to substantiate this contention.