Radiographs in clinical periodontal trials

Fifty-eight patients were selected from an ongoing study of periodontal therapy at The University of Michigan. Pre-and post-treatment series of full mouth radiographs obtained by conventional paralleling long cone technique were available. Mesial and distal radiographic bone height was scored using the technique developed by Björn et al. A total of 1416 teeth were scored from the itial radiographs, and subsequently at one, two, three, and four years after the treatment. Radiographic bone height scores were compared with level of attachment and pocket depth scores for the same teeth at the same time. Statistical significance and correlation coefficients were derived using computer analysis of the data. The data were analyzed using three different data groupings: individual teeth, patient means, and half-mouth treatment methods. High positive correlations were found between initial measurements of radiographic bone height and attachment level as well as pocket depth. The correlations between changes in measurements of radiographic bone height and attachment level after treatment were markedly lower but are statistically significant. A highly significant correlation between radiographic bone heights and measurements of attachment level also appeared in follow-up data one to four years after treatment. The method of data grouping resulted in different correlation coefficients. Highest correlations were found if the data were pooled for patients. Lowest correlations were found if the data for individual teeth were analyzed. When the data were combined to produce patient scores, sample variation is reduced and correlations increase in magnitude. The generally high correlations between radiographic bone height and attachment level scores before and after treatment tend to confirm the fact that radiographic assessment of alveolar bone height using the method of Björn et al. can provide fairly accurate assessment of interproximal periodontal support.