Radiotherapy for Prostate Cancer: Economic Considerations

Evidence suggests that men with favorable-risk prostate cancer achieve very high disease-control rates regardless of the primary therapy they receive. However, there are differences among the treatment approaches in terms of therapy-related toxicity and cost. For example, intensity-modulated radiotherapy (IMRT) is considerably more expensive than radical prostatectomy or brachytherapy.

During the past decade, some urology practices have developed financial relationships with radiation-oncology facilities, to which some of their patients with prostate cancer have been referred for IMRT. To examine the association between ownership and use of IMRT services, an academic investigator used Medicare claims data from 2005 through 2010 to perform two comparisons. The first was between 35 self-referring private urology groups and 35 matched non–self-referring private urology groups. The second was between non–self-referring urologists at 11 National Comprehensive Cancer Network centers and 11 matched self-referring private urology groups.

The investigator compared the use of IMRT in the periods before and during ownership and used a difference-in-differences analysis to assess IMRT use changes according to self-referral status. Using a clinical diagnosis and management algorithm developed in consultation with a private practice urologist, the investigator used a 6-month window from the date of a prostate cancer diagnosis to evaluate treatment.

The first comparison showed that IMRT use increased among self-referring urologists from 13.1% to 32.3% (P<0.001) and among non–self-referring urologists from 14.3% to 15.6% (P=0.05); difference-in-differences analysis showed that self-referral was associated with an unadjusted 17.9 percentage point increase in IMRT use (P<0.001). The second comparison showed that IMRT use increased among self-referring urologists from 9.0% to 42.0% (P<0.001) and remained stable among non–self-referring urologists at 8.0%; difference-in-differences analysis showed that self-referral was associated with an unadjusted 32.6 percentage point increase in IMRT use (P<0.001).