Objective: To assess interobserver and intraobserver reproducibility of the cervical vertebrae maturation method (CVMM) among three panels of judges with different levels of orthodontic experience (OE).
Materials and methods: Fifty individual lateral cephalograms of good quality with complete visualization of cervical vertebrae 1 to 4 were selected. Thirty clinicians, divided according to their OE into three groups (junior group, JU, OE ≤ 1 year; postgraduate group, PG, 2 ≤ OE ≤ 4 years; specialist group, SP, OE ≥ 7 years), evaluated the cephalograms in two sessions (T1 and T2) at 3 weeks apart. Kendall's W and weighted Cohen's kappa (κ) coefficients were performed to assess interobserver and intraobserver agreement. The level of significance was set as P < .05. For both the interobserver and the intraobserver datasets, the percentage of perfect agreement (PPA) and the number of stages apart for each disagreement were calculated.
Results: Kendall's W at T1 was SP = 0.61, PG = 0.70, and JU = 0.87; at T2 it was SP = 0.78, PG = 0.85, and JU = 0.86. The percentage of total interobserver perfect agreement (Inter-PPA) was 42.3% at T1 and 46.3% at T2. The JU group had the highest Cohen's κ coefficient at 0.78, while the PG and SP had coefficients of 0.64 each. The percentage of total intraobserver perfect agreement (Intra-PPA) was 54.2%.
Conclusions: The reproducibility of the method was not improved by the level of orthodontic experience. The group with the lowest level of orthodontic experience had the best performance.
Keywords: Cervical vertebrae maturation method; Reliability; Reproducibility; Skeletal maturation.