TUESDAY, July 17 (HealthDay News) -- Informal networks among physicians who share patients demonstrate substantial geographic variability, while within networks, physician and patient characteristics are similar, according to a study published in the July 18 issue of the Journal of the American Medical Association.
Bruce E. Landon, M.D., M.B.A., from Harvard University in Boston, and colleagues used methods adapted from social network analysis to examine Medicare administrative data from 2006 for 4,586,044 Medicare beneficiaries seen by 68,288 physicians practicing in 51 hospital referral regions (HRRs). Networks depicting connections between physicians were defined based on shared patients and were constructed for each of the HRRs.
The researchers found that the number of physicians per HRR ranged from 135 in Minot, N.D., to 8,197 in Boston, Mass. Network characteristics varied considerably across HRRs. The mean adjusted degree (number of other physicians each physician was connected to per 100 Medicare beneficiaries) was 27.3 (range, 11.7 to 54.4) across all HRRs. The relative centrality of primary care physicians compared with other physicians in their networks ranged from 0.19 to 1.06, indicating that, in some markets, primary care physicians were more than five times more central than in others. Physicians with connections to each other were in significantly closer geographic proximity and were significantly more likely to be based at the same hospital. Patient panels defined by race or illness burden were more similar for connected physicians than unconnected physicians.
"Network characteristics vary across geographic areas," Landon and colleagues conclude. "Physicians tend to share patients with other physicians with similar physician-level and patient-panel characteristics."
Several authors disclosed financial ties to network analytics companies.
Copyright © 2012 HealthDay. All rights reserved.
|Previous: AAIC: IV Immunoglobulin Stabilizes Alzheimer's Symptoms||Next: Adding Stroke Severity Measure Improves Mortality Risk Models|
Reader comments on this article are listed below. Review our comments policy.