PERIODONTAL REGENERATION OF HUMAN INFRABONY DEFECTS .3. DIAGNOSTIC STRATEGIES TO DETECT BONE GAIN

ASSESSMENT OF ACTUAL TREATMENT OUTCOME by simple and available diagnostic tests represents an important aspect in the evaluation of periodontal regenerative procedures. This report focuses on the diagnostic accuracy of different strategies to detect the original extent of bone loss and bone gain in deep vertical defects following guided tissue regeneration. Characteristics of probing attachment level and radiographic bone linear measurements were compared to a gold standard obtained as intrasurgical clinical measurements at baseline and at the 1 year re-entry operation. Radiographic linear measurements underestimated bone loss (CEJ-BD) in 55% of cases while the estimate was within 1 mm in 25% of cases. The goodness of agreement could be significantly described by a general linear model incorporating measures of defect morphology. The original level of bone loss could be best detected by an approach which consisted of adding 1.5 mm to the measured attachment level loss. Such an approach correctly estimated bone loss within 1 mm in 85% of sites. In terms of ability to detect bone gain, radiographic measurements underestimated bone gain in 55% and overestimated it in 15% of sites. Diagnostic accuracy was assessed at different threshold values of true gain. The sensitivity and specificity plots intersected at a threshold value of 2.7 mm bone gain. Conversely, measures of probing attachment level gain agreed with the gold standard within 1 mm 45% of times. Intersection of the specificity and sensitivity plots occurred at 3 mm. Classification tables are presented for both diagnostic tests at different threshold values. The best results to detect true bone gain were obtained employing a strategy consisting of the agreement of the radiographic and attachment level measurements. When such agreement was observed (70 to 85% of the observations) the overall diagnostic accuracy approached 100%. The significance of the findings and their clinical applicability in the assessment of actual treatment outcome following GTR procedures in deep vertical defects are discussed.