Understanding variation in primary prostate cancer treatment within the Veterans Health Administration

To examine the variation in prostate cancer treatment in the Veterans Health Administration (VHA)--a national, integrated delivery system. We also compared the care for older men in the VHA with that in fee-for-service Medicare.

We used data from the Veterans Affairs Central Cancer Registry linked with administrative data and Surveillance, Epidemiology, and End Results-Medicare data to identify men with local or regional prostate cancer diagnosed during 2001 to 2004. We used multinomial logistic and hierarchical regression models to examine the patient, tumor, and facility characteristics associated with treatment in the VHA and, among older patients, used propensity score methods to compare primary therapy between the VHA and fee-for-service Medicare.

The rates of radical prostatectomy and radiotherapy varied substantially across VHA facilities. Among the VHA patients, older age, black race/ethnicity, and greater comorbidity were associated with receiving neither radical prostatectomy nor radiotherapy. Facilities with more black patients with prostate cancer had lower rates of radical prostatectomy, and those with less availability of external beam radiotherapy had lower radiotherapy rates. The adjusted rates of radiotherapy (39.7% vs 52.0%) and radical prostatectomy (12.1% vs 15.8%) were lower and the rates of receiving neither treatment greater (48.2% vs 32.2%) in the VHA versus fee-for-service Medicare (P < .001).

In the VHA, the treatment rates varied substantially across facilities, and black men received less aggressive prostate cancer treatment than white men, suggesting factors other than patient preferences influence the treatment decisions. Also, primary prostate cancer therapy for older men is less aggressive in the VHA than in fee-for-service Medicare.