How do data from deepest pocket per quadrant relate to full‐mouth scores?

Abstract

When the subgingival presence of periodontal pathogens is studied in groups of patients or populations, mostly a number of the deepest sites is sampled. The mean clinical parameters of these deep sites are also frequently used as a the descriptor of the clinical situation of these subjects. It can be questioned, whether these 4 deep sites are capable of predicting a full-mouth situation. The purpose of the present retrospective study was to investigate to what extent the experienced progression of periodontitis as measured in the deepest approximal pocket in each quadrant reflects the disease progression at the approximal sites on a full-mouth level. A data set of a 7-year longitudinal study of 158 young subjects (69 male, 89 female, 15-25 years of age at baseline) was used. Clinical assessments included plaque index (PI), pocket depth (PD) and attachment loss (AL) at baseline (1987) and follow-up (1994). Measurements were made at the approximal surfaces of all teeth. The deepest pocket in each quadrant was determined at follow-up. Changes of the clinical parameters between baseline and follow-up were calculated both as full-mouth mean scores as well as for these 4 deepest sites. A regression analysis was used to evaluate the relationship between full-mouth score and the 4 test sites. For disease progression between baseline and follow-up, significant correlation coefficients were observed between the 4-site and full-mouth mean changes (PD: 0.80, AL: 0.70, PI: 0.77). Regression coefficients were 0.51 for PD, 0.35 for AL and 0.55 for PI. The precision of the estimate for the full-mouth mean, as predicted by the 4-site mean, is determined by the residual standard deviation. This was for PD 0.31 mm, for AL 0.31 mm and for PI 0.29. Compared to the between-patient standard deviation of the full-mouth means, the residual standard deviations were high. In the present population, a reasonable to good correlation between full-mouth and 4-sites data was observed. However, the high residual standard deviation in the regression analysis illustrates the inaccuracy for the 4-sites data when used as a descriptive for changes in the periodontal condition on a full-mouth level. Data evaluating progression of periodontitis based on a limited number of diseased sites should be interpreted cautiously.