Table 2.

Adjusted ORs and their 95% CIs of treatment associated with neighborhood socioeconomic deprivation and urban/rural locations among women with DCIS of the breast

Surgical treatmentaRadiation therapybEndocrine therapyc
MastectomyNo surgeryDelayedNo radiation
NOR (95% CI)NOR (95% CI)NOR (95% CI)NOR (95% CI)NOR (95% CI)
Socioeconomic deprivation
 Tertile 1 (lowest)1,2371.00 (reference)891.00 (reference)5161.00 (reference)6681.00 (reference)7941.00 (reference)
 Tertile 29341.16 (1.07–1.35)681.18 (0.83–1.67)3221.16 (0.98–1.39)6001.57 (1.35–1.84)5001.03 (0.86–1.23)
 Tertile 3 (highest)7291.44 (1.25–1.66)641.54 (1.02–2.30)2271.26 (1.01–1.57)4371.90 (1.56–2.31)3210.96 (0.77–1.20)
P trend < 0.0001P trend = 0.04P trend = 0.02P trend < 0.0001P trend = 0.75
Urban/rural location
 Urban2,1561.00 (reference)1661.00 (reference)8711.00 (reference)1,1991.00 (reference)1,2871.00 (reference)
 Rural7441.09 (0.96–1.24)551.02 (0.70–1.49)1940.95 (0.77–1.18)5061.29 (1.08–1.53)3281.17 (0.96–1.44)
  • aMastectomy and no surgical treatment were compared with BCS. The ORs were adjusted for age (<50, 50–59, 60–69, ≥70), race (white, black, other), year of diagnosis (1996–1999, 2000–2004, 2005–2009, 2010–2011), health insurance (private, Medicare, Medicaid/uninsured, others/unknown), tumor grade (well differentiated, moderately differentiated, poorly differentiated, unknown), tumor size (<2 cm, ≥2 cm, unknown), histology (comedo, papillary, cribriform, solid, NOS), and estrogen receptor status (positive, negative, and unknown).

  • bThe analysis was restricted to women with BCS and complete data on timing of surgery and first radiotherapy. Radiation delay was defined as an interval of 8 weeks or longer between BCS and the start of radiotherapy. Delayed initiation of radiotherapy and no radiotherapy were compared with timely initiation of radiotherapy. The ORs were adjusted for age, race, surgery, surgical margins, year of diagnosis, health insurance, tumor grade, tumor size, histology, and estrogen receptor status.

  • cThe analysis included patients with estrogen receptor–positive DCIS and patients with BCS only. The OR was adjusted for age, race, year of diagnosis, insurance, tumor grade, tumor size, histology, estrogen receptor status, surgery, surgical margin, and radiotherapy.