We thank Drs Sabour and Dastjerdi for their valuable comments about our article. They pointed out that the diagnosis of an outcome can be made cross-sectionally by using 2 separate prospective cohort data sets. However, our study used retrospective data available in a hospital setting. The data obtained for the study were mixed longitudinal; some patients had follow-up radiographs, but others did not. This drawback was accounted for and adequately explained in the article; it is this nature of the study design that hampers the use of the statistical tests mentioned. Furthermore, the t test provides an exact test for the equality of the means of 2 normal populations with unknown but equal variances. Both groups in the study adequately fit this criterion (since they were derived from the same data set); hence, it was decided to use the t test to test for means.
The diagnostic accuracy and validity of a test can only be ascertained when there is genuine diagnostic uncertainty. In our study, the diagnosis was already established, and only then were the prior radiographs evaluated. Hence, tests such as sensitivity, specificity, and likelihood ratios would be trivial; subsequently, the other dependent tests such as the receiver operating characteristic curve would be rendered inconsequential.
Moreover, the geometric measurement model has been used previously to predict the impaction of maxillary canines. We simply used a modified version of the same model in a previously untested population, which demonstrated varied and interesting etiologic considerations for impacted maxillary canines. Also, simple geometric measurements do not intend to substitute diagnostic tests used to determine impaction of maxillary canines. They merely serve to alert clinicians of the possibility that the maxillary canine will fail to erupt.
Thus, it would be difficult to refute the hard, obstinate facts obtained from observing logs of raw data used in this study. Certainly, reporting clinically discernible differences by using simple, timely, elementary-school geometric measurements obtained from tracing panoramic radiographs (often part of a routine radiographic examination) can provide valuable information for predicting impaction of maxillary canines.