Facial disfigurement: is it in the eye of the beholder? A study in patients with Graves' ophthalmopathy

OBJECTIVES The importance of facial disfigurement in many diseases necessitates a reliable and valid measure of disfigurement severity for clinical studies. The hypothesis is that a universal concept of disfigurement exists and can be measured in a reliable way. The objectives of this study were to investigate if persons, in particular patients and physicians, can agree on facial disfigurement severity; and to determine the relative contribution of predefined clinical characteristics of patients with Graves' ophthalmopathy (GO) to the overall rating of facial disfigurement severity. DESIGN A panel study was carried out in four different panels, each consisting of four members. PATIENTS We randomly selected 100 slide pairs of GO patients from four available study populations, involving mild, moderate and severe GO patients (mean age 49 years, 76% female) who were treated with either radiotherapy, sham-irradiation, prednisone or orbital decompression. MEASUREMENTS All panel members individually scored the disfigurement severity of 100 GO patients shown on standardized slides on a Visual Analog Scale. In total, 1600 ratings were collected. We calculated within- and between-panel agreement of disfigurement severity and identified determinants of disfigurement. RESULTS Agreement within a panel varied from 0.65 to 0.79 and was highest within the panel of ophthalmologists. Between-panel agreement was 0.67 and was highest between ophthalmologists and laypersons. Compared with the global average, patients overrated and endocrinologists underrated disfigurement severity. Female panellists rated the patients, on average, more disfigured than male panellists. Important determinants of disfigurement were eyelid retraction, severe eyelid swelling and proptosis. Their relative importance was consistent across panels and in contrast to current measures of GO severity. CONCLUSION Facial disfigurement severity can be measured in a reliable way using panels of panellists. Except for some systematic differences between panellists, facial disfigurement does not seem to be in the eye of the beholder.