Validity, reliability and reproducibility of three methods used to measure tooth widths for bolton analyses

Vernier calipers have traditionally been used as the 'gold-standard' for tooth width measurements. New digital methods may prove to be as valid, reliable and reproducible as caliper measurements.

To determine the validity, reliability and reproducibility of mesiodistal crown measurements made with calipers, using the DigiModel method and on digital photographs, and to determine the validity, reliability and reproducibility when the measurements are employed in Bolton tooth size analyses.

Twenty-five consecutive study models were used. The maximum mesiodistal crown widths of 12 teeth in each arch (first molar to first molar) were measured with digital calipers, using the DigiModel software (OrthoProof, Caulfield South, Victoria, Australia) and on standardised digital photographs of the models by eight examiners. One examiner measured all models three times using the three measurement methods, two examiners measured all models once using the three measurement methods and five examiners measured all models once using the calipers only. Validity for the widths of individual teeth, the mean tooth width and the Bolton ratios were assessed using a general linear model two-way analysis of variance and pair-wise comparisons between the two digital methods and the calipers. Reliability was assessed with Cronbach's alpha and reproducibility with the intra-class correlation coefficient.

There were statistically significant differences between the tooth widths measured with DigiModel and the calipers (p < 0.001) and with the calipers and digital photographs (p < 0.001). However these discrepancies were judged not to be clinically significant. The Cronbach's alpha scores for all methods were classified as 'excellent'. The intra-class correlation coefficient values exceeded 0.75 for the digital calipers, 0.79 for DigiModel and 0.54 for the digital photographs.

The validities of DigiModel and digital photographs for tooth width measurements and Bolton analyses are clinically acceptable. The reliability of both methods is also excellent. However, DigiModel software provided more accurate Bolton's ratios and demonstrated greater reproducibility than the digital photographs.