Annelotte van Bommel

54 The rate of BCS has been promulgated as a quality indicator. 30 When performing primary BCS, a delicate balance exists between the esthetic and oncological aims of the surgery: a wider excision may lead to a worse esthetic result, while a too narrow excision may leave residual tumor tissue. Striving for a high BCS rate may unintentionally lead to the perverse incentive of aiming for the lowest possible positivemargin rates by resecting larger amounts of breast tissue. BCPP serves the aim of measuring esthetic outcome more appropriately, as it appreciates at least the combined efforts and different treatment strategies to maintain the shape of the breast, which is in itself a desirable esthetic outcome. While BCPP more or less annihilated conventional age-specific BCS rates, no such effect was observed for institutional differences. Despite an apparent interplay between the various strategies used to preserve the breast contour (illustrated by the observed inverse association between the rate of BCS and the proportion of patients who underwent BCPP), the net effect of the hospital variation in BCS after NAC andmastectomywith IBR resulted in an observedwider range of the proportion of BCPP than the hospital variation in BCS rates. Previous studies using data fromthe NBCA studied the variation of NAC rates 25 and the proportion of patients undergoing mastectomy combined with IBR. 21,31 Patient and tumor characteristics and hospital factors did account for institutional variation, but the number of treated patients per hospital was not a factor associatedwith higher rates of NAC or IBR. In another study, we also observed that surgeons’ and plastic surgeons’ preferences had an impact on the institutional IBR rate. 32 Much of the observed institutional variation remains unexplained. Several hospitals in the present study never applied NAC or provided IBR, which might explain the wider range of BCPP rates. As these hospitals had no means other than primary BCS to enhance their BCPP rate, these institutions fell behind as others were improving their BCPP rate. Obviously, this hypothesis urges the need for additional in-depth analysis of the observed institutional variation. Having a national multidisciplinary audit for breast cancer care enabled us to analyze questions with large numbers of patients. This is a strength of the present study, and the population-based data are also suitable to study time trends. The absence of information regarding important patient characteristics such as smoking status and