Are outcomes worse in low volume and non-teaching hospitals for female breast cancer patients?

Introduction: Disparities in survival associated with receiving treatment at low volume and/or non-teaching hospitals have been found in some cancers. We explored whether survival in women with breast cancer was related to the treating hospital's characteristics. Methods: Florida Cancer Data System Registry (1996-2007), Agency for Health Care Administration, and US Census data were linked (N=127,754). High volume facilities (HVF), as compared to low volume (LVF), were defined as facilities treating ≥1% of the sample; and teaching (TF) vs. non-teaching facilities (NTF) were identified by the 2005 Association of American Medical Colleges. We calculated descriptive statistics and ran Cox regression models exploring the association of hospital type and survival. Results: Overall, 41% of women were treated in HVF and 10% in TF. The distribution of race and ethnicity differed between NTF (Whites 91%, Blacks 8%; non-Hispanic 91%, Hispanic 9%) and TF (Whites 84%, Blacks 15%; non-Hispanic 82%, Hispanic 18%); as did the distribution of socioeconomic status (SES) ¨C from lowest to highest quartiles in NTF (10, 29, 39, 22%) and TF (20, 32, 30, 19%). No differences in demographics between HVF and LVF. Cox regression adjusting for sociodemographic, clinical, and treatment characteristics showed that NTF were significantly associated with worse outcomes, (Hazard ratio [HR] 1.13; [95%Confidence Interval 1.03-1.24]), p=0.013; but LVF were not, (HR 1.0; [0.90-1.10]). Conclusion: TF are caring for more minority and lower SES patients. However, controlling for comorbidities, survival is better for patients treated at TF.