Interpreting Differences in BREAST-Q Scores following Reduction Mammaplasty: Minimal Important Difference

The BREAST-Q is a multiscale, condition-specific, patient-reported outcome measure that was developed and validated by Pusic and colleagues to assess important aspects of health-related quality of life and patient satisfaction following various types of cosmetic, functional, and reconstructive breast surgical procedures.1 The reduction module of the BREAST-Q is one of the most widely used patient-reported outcome instruments worldwide to measure surgical outcomes specific to women with breast hypertrophy undergoing reduction mammaplasty. While it is useful to obtain quantitative scores to measure and compare health-related quality of life before and after surgery, the value is somewhat limited if there is no defined threshold as to what constitutes a clinically meaningful change. The interpretation of BREAST-Q scores based solely on statistical significance may be misleading, because small differences in scores can be statistically significant even in the absence of the change being clinically meaningful. A minimal important difference has become a standard approach for interpreting the clinical relevance of changes in health-related quality-of-life scores and can be defined as the smallest difference in health-related quality-of-life score that is perceived as clinically important by a patient or clinician. However, there is no universal measure of minimal important difference, and this threshold may vary depending on the context of disease, on disease severity, by population characteristics, and by type of outcome instruments, highlighting the importance of estimating minimal important difference based on the specific clinical setting. While minimal important difference estimates have been described for the augmentation module2 and reconstruction module3 of the BREAST-Q, estimates for the reduction module remain undetermined. Therefore, to further provide a meaningful interpretation of BREAST-Q scores following reduction mammaplasty, a minimal important difference estimate for each scale of the BREAST-Q reduction module was established using distribution-based methods. In this study, we used frequently reported distribution-based criterion for determination of the minimal important difference, including one-half of the baseline standard deviation and effect sizes of 0.52,4 and the variation of change scores (standardized response mean) value of 0.3 for each of the BREAST-Q subscales. Minimal important difference estimates were derived from data collected from a prospective study of 132 women who were undergoing reduction mammaplasty.5 Women in this study completed the BREAST-Q before surgery and again at 12 months after surgery. Summary statistics for the BREAST-Q scores for surgical participants are presented in Table 1. Table 1 also presents the range of estimated minimal important difference values based on a 0.5 standard deviation, 0.5 effect size, and 0.3 standardized response mean for each scale of the BREAST-Q. On the transformed 0 to 100 scale, the mean minimal important difference estimates for each scale were as follows: satisfaction with breasts, 6 (range, 6 to 7); psychosocial well-being, 7 (range, 6 to 8); physical well-being, 7 (range, 5 to 8); and sexual well-being, 9 (range, 8 to 10). Table 1. - BREAST-Q Scale Mean Scores, Standard Deviations, Effect Sizes, and Minimal Important Difference Estimates Scale Before Surgery 12 Months after Surgery Change from before Surgery to 12 Months after Surgery MID Estimate Mean SD 0.5 SD Mean SD Mean SD d 0.5 d 0.3 SRM Based on 0.5 SD Based on 0.5 d Based on 0.3 SRM Mean MID Satisfaction with breasts 21.9 10.9 5.5 73.3 17.3 51.4 20.2 3.6 7.1 6.1 6 7 6 6 Psychosocial well-being 32.4 12.8 6.4 69.3 19.8 36.9 21.7 2.2 8.4 6.5 6 8 7 7 Physical well-being 44.0 15.5 7.8 76.6 12.1 32.7 17.9 2.3 7.1 5.4 8 7 5 7 Sexual well-being 30.4 17.6 8.8 61.9 22.1 31.4 25.0 1.6 9.8 7.5 9 10 8 9 MID, minimal important difference; SD, standard deviation; d, effect size; SRM, standardized response mean. This study describes minimal important difference estimates for the BREAST-Q reduction module, enhancing the interpretability of scores for both patients and clinicians as to what represents a meaningful change in patients undergoing reduction mammaplasty. A future direction would include using a combination of anchor-based and distribution-based methods to establish the most accurate estimates of minimal important difference. DISCLOSURE The authors have no financial interest to declare in relation to the content of this article. No funding was received for this article.