Effect of prior knowledge about treatment on cephalometric measurements

Objective: We hypothesised that prior knowledge of details for a growth modification treatment influences cephalometric measurements and results in a detectable bias. Design: Observational study. Setting: University teaching facility. Methods: Six orthodontic residents assessed 48 lateral cephalograms taken before and after functional appliance treatment from 24 patients. The residents assessed six cephalometric measurements, (Cd-Pog, Cd-Me, Ar-Pog, Ar-Me, Go-Me, SNB) over three separate sessions, in either a random concealed order or as matched pairs with information about treatment and time disclosed. Results: When information was disclosed, five out of the six cephalometric measurements were significantly higher that the corresponding cephalometric measurements taken randomly with undisclosed information. The bias was in the range of 1.6-3.2 mm for linear measurements and was 1.1° for SNB. Conclusion: Disclosing treatment information does introduce systematic errors in cephalometric measurements. Cephalometric analysis in orthodontic clinical research should be carried out by assessors who are blinded to treatment details, to minimise risk of bias.