PURPOSE - This study estimated the economic burden of breast cancer-related lymphedema (BCRL) among working-age women, the incidence of lymphedema, and associated risk factors.
METHODS - We used claims data to study an incident cohort of breast cancer patients for the 2 years after the initiation of cancer treatment. A logistic regression model was used to ascertain factors associated with lymphedema. We compared the medical costs and rate of infections likely associated with lymphedema between a woman with BCRL and a matched control. We performed nonparametric bootstrapping to compare the unadjusted cost differences and estimated the adjusted cost differences in regression analysis.
RESULTS - Approximately 10% of the 1,877 patients had claims indicating treatment of lymphedema. Predictors included treatment with full axillary node dissection (odds ratio [OR] = 6.3, P < .001) and chemotherapy (OR = 1.6, P = .01). A geographic variation was observed; women who resided in the West were more likely to have lymphedema claims than those in the Northeast (OR = 2.05, P = .01). The matched cohort analysis demonstrated that the BCRL group had significantly higher medical costs ($14,877 to $23,167) and was twice as likely to have lymphangitis or cellulitis (OR = 2.02, P = .009). Outpatient care, especially mental health services, diagnostic imaging, and visits with moderate or high complexity, accounted for the majority of the difference.
CONCLUSION - Although the use of claims data may underestimate the true incidence of lymphedema, women with BCRL had a greater risk of infections and incurred higher medical costs. The substantial costs documented here suggest that further efforts should be made to elucidate reduction and prevention strategies for BCRL.