Objective assessment of aesthetic outcome after breast conserving therapy: Subjective third party panel rating and objective BCCT.core software evaluation

Abstract

We analysed intra- and inter-rater agreement of subjective third party assessment and agreement with a semi-automated objective software evaluation tool (BCCT.core). We presented standardized photographs of 50 patients, taken shortly and one year after surgery to a panel of five breast surgeons, six breast nurses, seven members of a breast cancer support group, five medical and seven non-medical students. In two turns they rated aesthetic outcome on a four point scale. Moreover the same photographs were evaluated by the BCCT.core software. Intra-rater agreement in the panel members was moderate to substantial (k = 0.4-0.5; wk = 0.6-0.7; according to different subgroups and times of assessment). In contrast inter-rater agreement was only slight to fair (mk = 0.1-0.3). Agreement between the panel participants and the software was fair (wk = 0.24-0.45). Subjective third party assessment only fairly agree with objective BCCT.core evaluation just as third party participants do not agree well among each other.

Aesthetic outcome following breast-conserving surgery assessed by three evaluation modalities in relation to health-related quality of life

“…In other studies, the agreement between panel evaluation and BCCT.core has been measured using κ statistics. Haloua and colleagues and Heil et al . reported a level of agreement of 0·64 and 0·27 respectively.…”

“…In other studies, the agreement between panel evaluation and BCCT.core has been measured using κ statistics. Haloua and colleagues 22 and Heil et al 23 reported a level of agreement of 0⋅64 and 0⋅27 respectively. Merie and co-workers 24 noted a κ value of 0⋅57 for comparison of dichotomized BCCT.core results (excellent/good versus fair/poor) with those of a panel of five assessors.…”

“…The interobserver agreement in the present study, measured using weighted κ, ranged from 0·43 to 0·60; this can be considered to show moderate agreement. Heil and co‐workers reported a lower value of 0·37, whereas Haloua et al . documented a higher value of 0·66.…”

“…Vrieling and colleagues reported a weighted κ value of 0·70 for intraobserver agreement in a panel of five assessors. Heil et al . constructed a panel of 29 assessors and showed an intraobserver agreement of 0·32 after the same pictures were shown in a different order following a break of 60 min.…”

Cosmetic outcome and quality of life are inextricably linked in breast‐conserving therapy

“…Furthermore, two retrospective studies reported no correlation between objectively assessed cosmetic outcomes (Breast Analyzing Tool and Cosmesis Index) and quality of life . Computerized measurements are known to result in a significantly worse cosmetic outcome compared to patient and panel evaluation, with a fair to moderate agreement between BCCT.core and panel evaluation . Patient and panel evaluations are based on observer evaluation of the breast, taking into account specific factors, whereas some computerized measurements fail to include aspects such as scar appearance and color differences between breasts .…”