A multi-institutional analysis of the socioeconomic determinants of breast reconstruction

Abstract

Introduction: The National Comprehensive Cancer Network Oncology Outcomes Database contains uniquely rich sociodemographic and clinical data on over 9,000 breast cancer patients from 12 NCI-designated comprehensive cancer centers. This prospective, multi-institutional database was used to evaluate determinants of breast reconstruction and whether socioeconomic factors influence who receives reconstruction. Methods: Data were previously collected by chart abstraction and continuous follow-up for stage 0, I, II, III patients undergoing mastectomy from July 1997- June 2002. Rates of reconstruction were determined. Multivariate logistic regression analysis was used to evaluate which socioeconomic characteristics are associated with reconstruction. We controlled for age, body mass index, number of comorbidities, stage and neoadjuvant and adjuvant treatment modality. Results: Of 2,326 patients undergoing mastectomy, 45% received reconstruction. Of the reconstructions, 42% were implants, 21% rotational TRAM flaps, 31% free TRAM flaps, and 6% other. 90% of all reconstructions were immediate. By multivariate analysis, Medicaid (OR = 0.56, p = 0.03) and Medicare (OR = 0.56, p = 0.005) patients were less likely to receive reconstruction than those with managed care; however, there was no significant difference for indemnity versus managed care insurance. Homemakers (OR = 0.68, p = 0.006), unemployed (OR = 0.57, p = 0.02) and retired (OR = 0.55, p = 0.003) patients received significantly fewer reconstructions than those employed outside the home. Patients with a college education received more reconstructions than those who did not attend college (OR = 1.38, p = 0.008). Race and ethnicity were not found to be statistically significant predictors of reconstruction. Conclusion: The reconstruction rate in this study (45%) is markedly higher than previously reported rates. While this may be due to the institutions included, it also likely reflects changes in reimbursement and subsequent practice patterns. Differences in insurance, education, and employment status, but not race or ethnicity, appear to influence receipt of breast reconstruction. Further investigation of these differences is warranted to determine if they represent patient preference or provider bias.