A "statistically valid" physician rating site?

Another physician rating site has joined the din of online patient forums, but this one is taking a different approach.

The creator, Consumers’ Checkbook, says this site brings a more balanced assessment of each doc. Read more

Physician rating sites abound, and many of them just have a few entries from patients, making it easy for a couple rants or raves to skew the picture. Many docs say these sites are a disservice to patients – and a headache for the physicians – because the sample size isn’t statistically significant and can encourage patients to pick doctors based on factors other than quality of care. (Read more here about the sites – and how to defend yourself against this public forum.)

Consumers’ Checkbook (which AMNews notes here is the same group that in 2006 unsuccessfully sued the government for access to raw Medicare claims data for individual physicians) has completed pilot programs in Denver, Kansas City, and Memphis. The nonprofit consumer organization says their efforts show the approach can be scaled up nationwide.

Consumer Checkbook’s site relies on “scientifically valid surveys” of patients on information such as how well their doctors communicated, made themselves available, and arranged for a courteous staff - information the group says is "medically important."

They take a statistically significant number of surveys per doctor – such as an average of 58 completed patient surveys per doctor for more than 700 physicians in the Kansas City area.

“The rigorous survey design has enabled us confidently to identify real differences among doctors,” President Robert Krughoff said in a release on the group’s Web site. The group plans to do the survey in New York City this fall.

Does this sound like a more balanced approach to physician rating sites? Ever run into problems with rating sites, or do you have a solution to dealing with them?