Patient compliance is critical for equivalent clinical outcomes for breastcancer treated by breast-conservation therapy

Objective To determine the compliance with a standard breast-conservation t herapy (BCT) program in a predominantly indigent, minority population of pa tients with early breast cancer (stage I and II) served by a rural state in stitution in the South; to compare the clinical outcomes of this group with those reported in clinical trials; and to examine the socioeconomic factor s that may have contributed to the rate of compliance. Summary Background Data Disease-free survival and overall survival in early breast cancer treated by BCT versus modified radical mastectomy are report ed to be equivalent in prospective randomized trials. However, patients enr olled in clinical trials may not be representative of patients living in th e various diverse communities that make up the United States. The authors' hypothesis is that patients enrolled in clinical trials at the national lev el may not be representative of indigent patients in the rural South and th at clinical trial results may not be directly applicable. Methods A retrospective review of 55 women with early-stage breast cancer t reated from 1990 to 1995 was performed. Clinical data, compliance with trea tment and clinical follow-up, and recurrence rates were examined. Statistic al analysis performed include the Fisher exact test, Kaplan-Meier survival analysis, and log-rank test. Results Full compliance (defined as completion of the entire course of radi ation therapy and clinical follow-up) with the BCT program was observed in only 36% of patients. Fifteen of the 35 noncompliant patients did not compl ete radiation therapy. A significantly higher local failure rate was observ ed: 8 of these 15 patients (53%) have had local failure. In contrast, patie nts who were either in full compliance with the BCT program or were deficie nt only in that they missed part of their clinical follow-up had local fail ure rates of 5% (1/20) and 10% (2/20), respectively. Age, race, stage of ca ncer, economic status (measured by availability of medical insurance), dist ance of patient's residence from the hospital, and education level were eva luated as potential predictors of compliance. None predicted patient compli ance, although a trend toward higher compliance was noted in patients with a higher education level, as determined by literacy testing. Conclusions Compliance with the BCT protocol at the authors' institution wa s worse than reported in clinical trials, and noncompliance translated into a significant increase in the local failure rate. Factors examined suggest that literacy may play a role in predicting compliance. Although BCT shoul d be discussed with all breast cancer patients, the judicious application o f clinical trial data to an institution's local population is warranted.