The clinical significance of error measurement in the interpretation of treatment results.

The aims of this investigation were to determine the errors involved in cephalometric landmark identification and to link these to the interpretation of treatment results. Fifty cephalograms were randomly selected from patient files and the following were determined. (i) Accuracy of the digitizer--single tracing digitized on five occasions on each of 10 different positions on the digitizer by one observer. (ii) Intra- and inter-observer digitizing error--35 landmarks on the same tracing (on a fixed position) digitized on five occasions by each of four judges. (iii) Intra- and inter-observer tracing error--five separate tracings of 10 different cephalograms by four judges. There were no significant differences in the variances of the co-ordinates for each landmark between the different positions on the digitizer (mean variance x-axis 0.07 mm and y-axis 0.08 mm). (ii) One-way ANOVA showed no significant intra- or inter-observer differences in digitation. (iii) Levene's test for homogeneity of variance showed significant differences in the co-ordinates of different landmarks and between the same landmarks on different cephalograms. Two-way ANOVA showed significant differences between observers for the same landmark that were greater than the intra-observer differences. The results indicate that tracing accuracy is a limiting factor in cephalometry. The variance of each landmark is dependent on the quality of the cephalogram. Inter-observer differences were greater than intra-observer effects and these were random, rather than systematic errors. Minimal error estimation calculations enable discrimination between treatment results and measurement errors.