Predictors of late stage breast cancer diagnosis: A multilevel approach

Introduction: Early stage breast cancer diagnosis has a better prognosis and improved survival rate compared to late stage diagnosis; survival rate for localized breast cancer is 98.6% compared to 24.3%, if distant metastasis has occurred. Identifying factors that influence stage at diagnosis could inform strategies to reduce the occurrence of late stage diagnosis.

Methods: Using the Surveillance, Ends Result and Epidemiology data, a retrospective study was conducted to identify the association between stage at breast cancer diagnosis, and individual level, county level and tumor characteristics. Analyses were limited to patients diagnosed with breast cancer between years 2000 and 2011. Descriptive statistics were conducted. A two-level multilevel logistic regression was also conducted; individuals (level 1) were nested within counties (level 2). Dependent variable was stage at diagnosis (0=early; 1=late).

Results:  Descriptive statistics showed that rural residents were more likely to receive a late stage diagnosis compared to urban residents (33.49 vs. 31.4; p-value=<0.0001). On multivariate analyses, rural residents (OR=1.10; 95%=1.04-1.19) and blacks (OR=1.29; 95% CI=1.27-1.33) were more likely to be diagnosed at a late stage compared to urban residents and whites respectively. Those who had a high grade tumor were more likely to have a late stage diagnosis compared to those with a low grade tumor (OR=1.69; 95% CI=1.66-1.71).

Conclusion: Even after controlling for tumor characteristics, minority populations and rural residents were at an increased risk for late stage breast cancer diagnosis. Efforts are needed to reduce these disparities.