The screening and categorization of patients: a case for natural selection?

A preliminary study was performed on the feasibility and validity of objective categorization of patients into basic 'treatment types'. The Reflex Metrograph was used for a three-dimensional cast analysis, and an H.P. 9874 A digitizer was used for a two-dimensional radiograph analysis. In both measurement techniques the method error was found to be acceptable. Using the Ricketts 10 Factor Analysis, it was found that by means of 'Clinical deviators' radiographs could be used to distinguish orthognathic cases from the other treatment groups. The three-dimensional study cast analysis showed promise, distinguishing a number of statistically significant factors within the lower arch and between simple and fixed appliance cases. A number of variables which might have been expected to be significant were confounded by their range of values within the groups. Further investigation is required with an improved technique of initial categorization of patients, perhaps by means of a panel of clinicians, and more sophisticated statistical techniques.