Background: A recent SEER study identified significant variations in the care of women with DCIS, yet several potential confounding variables were not included. We report a patterns of care study of women with DCIS to better understand the gap between evidence-based knowledge and the management of DCIS.
Methods: We studied all cases of DCIS diagnosed through the Ontario Breast Screening Program from 1991 to 2000. Data was obtained by database linkage and chart abstraction. A logistic regression model using generalized estimating equations to adjust for clustering was used.
Results: About 320,236 women were screened and 727 individuals were diagnosed with DCIS. The rate of mastectomy was 30% and was associated with multifocality (OR: 3.5 [1.7, 7.1], P = 0.0005), tumor size (OR: >2 cm vs. <or=1 cm: 2.7 [1.3, 5.9], P = 0.01), high nuclear grade (OR: 2.4 [1.1, 5.2], P = 0.03) and surgeon's practice pattern. The rate of axillary dissection (AND) decreased from 36% in 1991-1993 to 20% in 1998-2000. AND was associated with mastectomy, year of surgery and hospital volume (OR: 2.7 [1.3, 5.6], P = 0.01). The use of radiation (XRT) following breast-conserving surgery increased from 39% in 1991 to 51% in 2000. XRT was associated with age<70 years, high nuclear grade (OR: 2.7 [1.2, 6.3], P = 0.02) and tumor size > 1 cm (OR: 2.4 [1.3, 4.4], P = 0.006). Half of cases with margins <1 cm did not receive XRT.
Conclusions: Our study corroborates previous reports on the persistent rates of mastectomy and axillary nodal dissection and the limited use of XRT in the treatment of DCIS.