Calibração de examinadores para exame epidemiologico da carie dentaria : uma avaliação mediante diferentes limiares de diagnostico / Calibration of examiners for the epidemiological examination of dental caries, an evaluation under different diagnostic thresholds

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

This thesis, comprising 4 studies, aimed al evaluating the examiners calibration for the dental caries epidemiological survey, under different diagnostic thresholds. Methodology of the 4 studies: Eleven examiners, having experience in epidemiological surveys, were trained for the dental caries calibration. The inclusion criteria were based on those recommended by WHO (WHQ, 1997), including the initial lesions (IL). The examinations were performed in 6-7 year olds with and without history of dental caries. Two caries thresholds were used for data analysis: WHO and WHO+IL. The aim of the 1st study was to analyze the intra-and inter-examiner reproducibility (reliability) and to verify the accuracy (benchmark validity) though sensitivity. (S) specificity (SP), positive (PPV) and negative predictive values (NPV) of this calibration trial. The aim of the 2nd study was to analyze the reproducibility of calibration trial in a 12-month evaluation. lntra -and inter-examiner Kappa showed excellent mean values for both diagnostic thresholds during the calibration phases. However, the most relevant errors were related to the decayed component and to the IL diagnosis. When the results of the benchmark examiner were compared to those of the examiners in the first phase of the evaluation (study 1), moderate to high validity values were observed (0.71 1.00), with some loss mainly for S and PPV, when IL was included. In conclusion, the methodology used to examine the mixed dentition described in the two studies above could be used in epidemiological surveys, although new strategies for a better training and calibration of IL are necessary. The aim of the 3rd study was to develop an adequate approach to measure inter-examiner s agreement during the calibration process under different diagnostic thresholds of dental caries. Kappa statistics (K) was used to analyze the teeth in four stages: all together; the superior, arch; the inferior arch; the sextants; and each tooth individually. The results showed high mean values for inter-examiner reproducibility, concerning both diagnostic thresholds, except for the teeth analyzed individually, which showed a great decrease in Kappa values. The aim of the 4th study was to compare three measuring tools, Kappa (k); General Agreement Percentage (GAP); and Dice Index (D), which are used to determine the reproducibility of examiner?s in surveys under two criteria of dental caries diagnosis. It was concluded that Kappa could be considered the best statistical test for evaluating examiners? agreement for dental caries calibration because it assesses the proportion of agreement that is achievable beyond chance. However when it is not possible to be used the authors recommend GAP associated with D to improve the examiners ability to diagnose the carious lesions.

ASSUNTO(S)

caries dentarias diagnosis diagnostico epidemiology dental caries epidemiologia

ACESSO AO ARTIGO

Documentos Relacionados