When a U.S. physician sees a patient - either for a routine visit or to administer some sort of treatment - there's a good chance she'll be paid a different amount for her work than another doctor doing the same thing one state, or perhaps even one county, away. Variation in the amounts that private insurers pay physicians is a known phenomenon, but extensive research on the practice - and the factors that account for such variation - has been lacking.
To better understand these differences in payments, Stanford health-policy experts Laurence Baker, PhD, and M. Kate Bundorf, MBA, MPH, PhD, teamed up with an Indiana University–Purdue University researcher to comb through more than 41 million insurance claims for four kinds of services: office visits with established patients, office visits with new patients, office consultations, and preventive visits with established patients. What they found was that physicians at the high-end of payments received were generally paid more than twice than what physicians at the low end were paid for the same service. They also found that the variation couldn't be explained by patients’ and physicians’ characteristics - things like the age and sex of the patient, the physician's specialty, and whether the doctor was a "network provider" - but that about one third of the variation was associated with the geographic area of the practice.
To find out more about the study, which was published online yesterday in Health Affairs, I contacted Baker. He answers my questions below.
Were you surprised by what you found?
Sort of. Some news reports have highlighted variations in health-care bills, so we were ready for some variations. But since we were looking at services that are quite common and pretty consistent from place to place, we weren't expecting to find very big variations, which is what we got.
The other thing that's interesting is the amount of variation that isn't explainable by the things we looked at. I had expected a lot of it would be explainable, but most of it isn't. This is another indication of the complexity of the health care system and the lack of understanding we have of the factors that determine prices.
Did you expect geography to be more of or less of a factor?
I had expected more. Geography is a proxy for many things - such as the costs in different areas, the competitiveness of areas, the preferences of the population and doctors. These could all influence prices. I had thought these would play a bigger role than what we found. But there's a chance that these things still do, but in ways that are specific to individual doctors or groups, so that we need to do more work to fully measure them.
Why is a better understanding of price variation important?
Price variations could signal important problems with the functioning of health-care markets. Large price variations for similar services normally only exist where someone in a market has a lot of power to dictate prices, which is often a problem for consumers. Price variations can also exist for reasons we'd be less concerned about - for example, if some providers are much higher in quality than others. But knowing about the existence and patterns of price variations can guide us to examine areas that we may need to work on to improve the system.
Informing patients about price variations can also be important. Some patients - for example, the uninsured - can end up paying widely different prices for the same services. If they have more information, they'll perhaps be better able to manage their health-care experiences and bills.
Your analysis was done on claims from 2007, well before health reforms took effect. Do you think your findings would be different if you examined claims from this (or next) calendar year?
I'm not sure health reform is likely to change price patterns quickly. There are some aspects of health reform that might encourage more transparent pricing, which would probably make prices less variable. But the main thrust of health reform is to get more people insured. That might influence pricing in the longer run, but probably not a lot.
What are your next steps in regards to this line of work?
We're continuing to study prices and the causes of variations. One of our current projects focuses on measuring the effects of large physician practices on prices, and another is looking at what happens when hospitals acquire physician practices. We hope to have more to say about these things in coming months.
Photo by ASSOCIATED PRESS