A comparative assessment of cephalometric errors.

Measuring a cephalometric radiograph is an imprecise task, and the errors associated with this should be quantified and understood. The potential sources of error are outlined and the literature relating to the assessment of measurement errors is reviewed. Three techniques are examined: the error attributable to a single recording; its associated confidence limits; and the coefficient of reliability were the most commonly applied. Data from two error studies were pooled and used to compare the accuracy of 12 skeletal, dental, and soft tissue measurements, using various assessments of each of the three techniques. Within each group, the use of different mathematical criteria led to variation in the results; one method was totally unsuitable. As with all statistical procedures, a knowledge of the technique facilitates interpretation of the results. Some suggestions as to the most satisfactory approach to the estimation of measurement error in cephalometry are given.