Annelotte van Bommel

70 DISCUSSION This is the first nation-wide study investigating the variation in the use of IBR after mastectomy for invasive breast cancer and DCIS between hospitals in the Netherlands. A large variation was found; IBR was performed on average in 17% of patients with invasive breast cancer (range 0–64%) and in 42%of patients with DCIS (range 0–83%). Although various patient and tumor characteristics were found to have a significant effect, adjustment for these factors using multivariate analyses did not result in less variation between hospitals. Apparently, there are other yet unidentified factors, such as patient preferences, surgeons’ beliefs, or hospital organizational factors, which probably affect the use of IBR to a larger extent. Previous studies have reported on breast reconstruction rates after mastec- tomy 1,8,11 ; however, the results of these studies cannot be compared with our results because immediate and delayed breast reconstructions and invasive breast cancer and DCIS were combined in other studies. Some studies reported mean postmastectomy IBR rates of 21% in the United Kingdom 1 and 24% in the United States 11 when combining invasive breast cancer and DCIS. In our study, we decided to analyze DCIS and invasive breast cancer separately because certain factors such as hormone receptor status are only available and relevant for patients diagnosed with invasive breast cancer. Moreover, the IBR rate for patients with DCIS was more than two-fold higher than that for patients with invasive breast cancer, which is consistent with literature. 2 Furthermore, previous studies often combined immediate and delayed breast reconstruction. A large meta-analysis (n=159,305 cases, 28 studies) showed an average of 16.9% of patients receiving immediate or delayed breast reconstruction. Comparison of the 10 largest population-based studies with a total of 10,000 mastectomy cases resulted in breast reconstruction rates (immediate and delayed) varying between 4.9% and 30.3%. 8 Combining immediate and delayed reconstruction for analysis is not preferred in our opinion because treatment approaches and patient populations may be different. Most importantly, the exact numerator to calculate the delayed breast reconstruction rate in a given time period is unknown because a delayed reconstruction may be performed many years after the initial surgery.