Evaluations of aesthetic results in breast reconstruction: an analysis of reliability

Abstract

This study evaluated the reliability of three commonly used measures of aesthetic outcomes of breast surgery: a four-point ordinal scale of overall aesthetics, five four-point subscales, and a visual analogue scale. Fifty patients were randomly selected from women who underwent breast reconstruction surgery at University of Michigan hospitals between July 1989 and May 1993. Postoperative photographs of these patients were provided to three plastic surgeons, who were asked to rate the photographs using the three methods. The same process was repeated 4 weeks later. Intrarater and interrater reliability ranged from poor to good for the three methods, with the subscales showing the highest reliability. The lowest reliability occurred for those scales with the least-explicit rating criteria. Without explicit criteria, raters must develop and use their own criteria, which are likely to differ for each rater. Separating the various components of the aesthetic results of breast surgery into different subscales helps make the rating criteria more explicit. Scales with demonstrated reliability are critical for ensuring comparability of results across studies.