The relationship between quality, spending and outcomes among women with breast cancer.

BACKGROUND Substantial shortcomings in the quality of breast cancer (BC) care have been identified. While breast cancer is responsible for the largest share of cancer care spending, little is known about the value of care provided to US women with BC. METHODS For women aged 65 to 70 years diagnosed July 1997 through December 2005 with stage 0-III BC who were continuously enrolled in fee-for-service Medicare, we evaluated performance relative to 20 measures recommending for proven therapies and seven measures recommending against unnecessary therapies. Using health care service area as the unit of analysis, we characterized quality for recommended and unnecessary therapies, median per-patient cost in the year after diagnosis for Medicare parts A and B, and five-year overall survival. We analyzed the relationships between quality, cost, and survival. All statistical tests were two-sided. RESULTS We assessed the care provided to 15357 women and compiled quality, cost, and outcomes data for 99 regions. The median number of patients/region was 85 (interquartile range = 47-158); the five-year overall survival was 87.5%. Part B expenditure correlated positively with measures of recommended therapy (P = .027) and negatively with measures of unnecessary therapy (P = .004). Survival did not correlate with quality or cost. Regions demonstrating lower quality for both recommended and unnecessary therapy measures exhibited higher part A expenditure. CONCLUSIONS BC patients experience wide variation in quality, cost, and survival. Process measures that assess recommended and unnecessary therapies did not correlate with survival, but did shed light on spending for related and unrelated services. Additional efforts to identify factors that impact the relationships between quality, cost, and outcomes are needed.