Methods used to evaluate growth modification in Class II malocclusion

Abstract

The methods used to study growth modification in orthodontic patients can have considerable impact on the conclusions that may be drawn. Because of the large "between-patient" variation and small mean changes usually observed, apparent differences in response may sometimes be more attributable to study design than to treatment effectiveness. A systematic review of four major orthodontic journals (1980 to 1987) identified 50 studies reporting treatment of young patients with Class II malocclusion. Variables defined to classify the studies included appliance systems, patient selection, comparison groups, research design, data collection, analysis, and reporting. The appliance systems most frequently investigated were the function regulator and the activator, used with and without headgear. Only 11 (22%) studies were prospective, and random assignment to alternative treatments was never used in this sample. Comparison groups used in 76% of the studies were untreated Class II patients (n = 18) and/or patients with alternate appliance systems (n = 17). In only 24% of the reports were groups tested for pretreatment equivalence. Few studies reported fully how patients had been selected, how decisions had been made to discontinue or change treatment, or whether patients had been lost to study. While most studies reported "p values," in only four were alpha levels adjusted for the number of tests (type I error), and no study included a post beta estimate (type II error). Age, sex, maturation, and duration of treatment were usually reported but seldom adjusted for in the analyses. Given the multiple indices of treatment effect, the generally small sample sizes, weak research designs, and incomplete reporting of important data, we cannot yet conclude whether orthodontic treatment influences the growth of Class II patients.