Aetna plans tiers based on cost, quality

The insurer says encouraging patients to choose more accomplished specialists is economical, but some doctors doubt the veracity of its data.

By Robert Kazel, AMNews staff. Oct. 20, 2003.

Doctors who remember physician profiling from the heyday of managed care may see Aetna's new "tiered physician network" as déjà vu with a twist.

The insurance company is taking old medical claims filed by specialists and separating out those doctors who the payer believes shine brightest in clinical quality and cost efficiency. These best-performing doctors are placed in a new, discrete network that's being marketed to employers as a long-term cost-savings opportunity.

If employees choose the doctors deemed most excellent and cost-conscious, they pocket rewards: a reduction in their co-payment or co-insurance, or a reduction in their deductible. The physician gets no additional pay, only the opportunity to attract more patients. Aetna is not the first to create tiered networks based on cost, but few plans take quality measures into account.

But some doctors say they're worried that the Hartford, Conn.-based insurer's data analysis using medical claims isn't nearly advanced enough to justify splitting doctors into camps.

"I've never heard of Aetna's database," said Howard Lewis, MD, a cardiologist and executive director of the Swedish Heart Institute in Seattle. "I don't know how it's gathered. If Aetna's patients are only a small percentage of my practice, does that reflect my entire practice? I don't know the rules they're using. I think [doctors] and consumers need to know the rules."

The tiered networks are "a worrisome approach," according to Lewis Foxhall, MD, chair of the Texas Medical Assn.'s Council on Socioeconomics. "It's very difficult, in our experience, to use administrative claims data to get information on the quality of physicians. What's the patient going to do? They're under financial pressure as it is. [This] is just steering patients away from perfectly good doctors based on data that may be spurious."