Evaluation of the agreement by examiners according to classifications of third molars

Abstract

Objectives: This study recorded and evaluated the intra- and inter-group agreement degree by different examiners for the classification of lower third molars according to both the Winter's and Pell & Gregory's systems.

Study design: An observational and cross-sectional study was realized with forty lower third molars analyzed from twenty digital panoramic radiographs. Four examiner groups (undergraduates, maxillofacial surgeons, oral radiologists and clinical dentists) from Aracaju, Sergipe, Brazil, classified them in relation to angulation, class and position. The variance test (ANOVA) was applied in the examiner findings with significance level of p<0.05 and confidence intervals of 95%.

Results: Intra- and inter-group agreement was observed in Winter's classification system among all examiners. Pell & Gregory's classification system showed an average intra-group agreement and a statistical significant difference to position variable in inter-group analysis with greater disagreement to the clinical dentists group (p<0.05).

Conclusions: High reproducibility was associated to Winter's classification, whereas the system proposed by Pell & Gregory did not demonstrate appropriate levels of reliability.

Figure 1

Intra-group disagreement analysis.

Figure 2

Boxplot to inter-group agreement for position variable. The top and bottom of the Box are the 25th and 75th percentiles. The line drawn through the middle of the Box in the median (the 50th percentile). The length of the Box is the interquartile range (IQR). The Box represents the middle 50% of the data. Median = 0.77, 0.70, 0.80, 0.70 for undergraduates. OMFS, radiologists and clinics, respectively. The interquartile range for clinics Box shows an enlargement with less frequent responses for the agreement.

Figure 3

Scatter plot to dispersion of professional inter-group agreement for position variable. Scatter plot showed a greater dispersion (since 85% frequency to 20%) for the clinics group and in a global evaluated the others professional groups were mixed in the same area.