Intra- and Inter-Examiner Reliability and Inter-Method Comparison in Physical Anthropometry and Photogrammetry
Date of Award
Master of Science (MS)
Bosio, Jose A.
Objective: Orthodontic treatment can often effect changes in the facial complex. In order to assess what those changes are, precise and reliable methods for measuring facial structures need to be used. While the techniques used for measuring have become increasingly more sophisticated they have also become more expensive and cumbersome for daily use. This study investigated the reliability of two methods, physical anthropometry and photogrammetry, that were inexpensive and relatively easy to set up.
Materials and Methods: Ten examiners measured a sample comprised of 20 dental students (10 male, 10 female) twice over three weeks. Eighteen measurements were acquired directly using a digital caliper. The 18 measurements were comprised of 20 facial landmarks previously defined by Farkas (1981). In addition, standardized facial photographs were made of the 20 participants, and the examiners were asked to identify the same points. The images were then calibrated and the same facial measurements computed. The intra-class correlation coefficient was used to determine the intra- and inter-examiner reliabilities. The Bland-Altman method was used to compare the two methods.
Results: Anthropometric intra-examiner reliability was very high for all measurements, while inter-examiner reliability exhibited a wide range of values, Overall the reliabilities were higher for easily identifiable landmarks, such as landmarks around the mouth, eyes, the nose, while bony landmarks covered by soft-tissue produced less reliable measurements. With few exceptions, photogrammetric reliability was high for both inter- and intra-examiner reliabilities. The least reliable measurement was the interlabial gap probably due to placement error of the points which was compounded by the small distances measured. The Altman-Bland plots showed large variability around the average difference.
Conclusions: The study found that examiners demonstrate high intra-rater reliability regardless of which method was used. Inter-examiner reliability showed larger variability dependent on the method. When using a caliper the examiner was not as consistent as on a photograph when selecting the facial landmarks. While both methods allow for accurate intra-examiner measurements, this study found that the photogrammetric method had greater inter-examiner agreement. However, large variability was found when comparing the two methods.