Physician Social Networks and Variation in Prostate Cancer Treatment in Three Cities

In 2010, some 170,000 U.S. men were diagnosed with prostate cancer. The treatment they received varied widely across practice settings, geographic areas, and patient race/ethnicity and socioeconomic status. The variation cannot be explained by patient preferences but rather may be related to physicians’ relationships with other physicians.

Researchers used Medicare claims data to examine patterns of prostate cancer treatment in 4,520 elderly patients in three cities and mapped the relationships of their doctors. They included the following doctors in their network construction: diagnosing urologist, majority urologist (billed most claims in the nine months after diagnosis), primary care provider, plurality provider (other specialist) and radiation oncologist. The researchers hypothesized that practice style and treatment modalities would be similar among doctors in subgroups that share patients.

The cities varied in overall rates of prostatectomy from 8.6 percent to 25.3 percent. Subgroups varied in prostatectomy rates from 14.1 percent to 47.1 percent in one city. In each city, the vast majority of urologists were solo practitioners.

The researchers concluded that variations in care for men with prostate cancer may be related to the doctor that a man sees and how that doctor is connected to other doctors, both formally and informally.