| Sumario: | Objective To compare the performance of partial-mouth periodontal examination ( PMPE) protocols with different cut-off points to the full-mouth examination ( FME) in the assessment of the prevalence and extent of gingival bleeding in adolescents. Methods A cross-sectional study was conducted involving 12-year-old adolescents. Following a systematic two-stage cluster sampling process, 1134 individuals were evaluated. Different PMPE protocols were compared to the FME with six sites per tooth. Sensitivity, specificity, area under the ROC curve ( AUC), intraclass correlation coefficient ( ICC), relative and absolute biases and the inflation factor were assessed for each PMPE protocol with different cut-off points for the severity of gingival bleeding. Results The highest AUC values were found for the six-site two-diagonal quadrant (2-4) (0.97), six-site random half-mouth (0.95) and Community Periodontal Index (0.95) protocols. The assessment of three sites [mesiobuccal ( MB), buccal (B) and distolingual ( DL)] in two diagonal quadrants and the random half-mouth protocol had higher sensitivity and lower specificity than the same protocols with distobuccal ( DB) sites. However, the use of DB sites led to better specificity and improved the balance between sensitivity and specificity, except for the two-diagonal quadrant (1-3) protocol. The ≥1 cut-off point led to the most discrepant results from the FME. Conclusion Six-site two-diagonal quadrant (2-4) and random half-mouth assessments perform better in the evaluation of gingival bleeding in adolescents. However, when a faster protocol is needed, a two-diagonal quadrant assessment using only MB, B and DL sites can be used with no important loss of information.
|
|---|