desk piled with papers

A number of years ago, a family doctor friend of mine took me on a tour of his small group practice. He proudly showed me the exam rooms, his medical equipment, and other parts of the facility that related to patient care. Then, we came to a large room with a bunch of desks piled high with paper. He explained, bitterly, that this part of his office was for the people he had to keep on the payroll to do nothing but deal with insurers.

This administrative expense was cutting his margins to the bone and did not help him take better care of his patients. He eventually left practice to pursue a second career as a physician executive—a job that was, for him, more remunerative and more satisfying.

Part of the problem is that physicians in the U.S. have to deal with multiple health plans—each with its own set of managed care rules, formularies (or list of approved drugs), requirements for prior authorization, rules for billing, submission of claims, and adjudication. Until recently, almost all of this administrative work was done by phone or fax. Picture this: Rooms full of practice-based nurses talking to insurance company nurses about the details of a case that may or may not lead to payment for medical care.


Just how much is our multi-payer system costing us?

Researchers at the University of Toronto, Weill Cornell Medical College, and the Medical Group Management Association collaborated to help shed light on this important question. They decided to compare how much time (and money) is being spent interacting with payers (private and public plans) in the U.S. vs Canada, a country with a single-payer system. The results were published in an article in the August 2011 issue of Health Affairs.

The researchers designed survey instruments for physicians and business managers to assess the time spent in minutes on such issues as formularies, billing, credentialing, and prior authorizations (the latter two issues are relevant only to the U.S. practices as the Ontario, Canada single-payer system examined in this study does not credential physicians or require prior authorizations.)


The results?

Here is what they found:

  • Physicians in the U.S. spent 3.4 hours per week interacting with multiple payers compared to Ontario physicians who spent only 2.2 hours per week interacting with the single-payer system. Most of the difference was related to time spent obtaining prior authorizations, an agreement by the insurance company that the service meets criteria for payment.
  • U.S. nursing staff (including medical assistants) spent 20.6 hours per physician in the practice per week interacting with payers—almost ten times the amount of time spent by Ontario nurses (2.5 hours). Again, a lot of it (13.1 hours per physician per week) was related to prior authorizations.
  • Clerical staff in the U.S. spent 53.1 hours per physician per week compared to 15.9 hours in Ontario. Most of the difference was related to billing issues and obtaining prior authorizations.

When this time is translated into U.S. dollars (adjusting for salary rates and specialty mix), Ontario practices spend $22,205 per year per physician compared to—are you ready for this??—$82,975 in the U.S. That is a difference of $60,770 per physician. The authors calculate that if U.S. physicians had administrative costs similar to Ontario physicians the total savings would be ~$27.6 billion per year! You could insure a lot of people and/or pay for a lot of healthcare with that type of dough.

The authors are careful to point out that we really don’t know the value of the benefits that may be reaped by these insurance company interactions. For example, how much inappropriate care is avoided by prior authorizations and how much innovation is stimulated by competition between the various payers? I have raised the former issue with health plans I have worked with, however, to date, there really aren’t good studies that take into account the costs of doing the prior authorizations, the savings related to denial of inappropriate care, and the value of any “sentinel effect” related to just having the “watchdog” process in place. It would clearly be helpful if our friends in the health services research community could help answer some of these important questions.


The bottom line…I think?

Meanwhile, I do think this study should stimulate a vigorous discussion, perhaps during the Presidential debates, about how much we are willing to pay for “choice” of insurer… perhaps, we might find that $27.6 billion a year is way too much.


  1. Thanks Al and Dan,

    There are no silver bullets when it comes to the health care cost problem, rather what we need are a thousand golden BBs!


  2. Interesting metrics – and I do understand Al’s point ($27B = 1% of $2.7T), but here’s the thing. There are NO $500B healthcare problems to go solve. What we do have are 100’s of these smaller ones – and it’s so fragmented – we successfully discount and trivialize all of them. I say turn the statistic around. It’s “only” 1% of $2.7T, but if we applied that $27B to say health insurance using a rough estimate of say $8,000/yr/person (which itself is *very* high), there would be almost 4 million more people with healthcare insurance.

    This may all be back of the envelope math – but it’s in the ballpark – and it serves to highlight the larger point. It’s not that $27B is “only” 1% of $2.7T. It’s that $2.7T is insanely high – and we need to find dozens of ways – like this one – to scale it back.

    FYI – I wrote a list of about 13 others worth considering here:

  3. You’ve kinda proven the opposite. $27-billion is about 1% of healthspend. I’d happily pay an extra 1% for a choice of suppliers in just about anything — cars, supermarkets, and certainly health plans.

    It’s also possible the Health Affairs article missed the mark. The “cost” of choice might be that the buyers lack the market power to force fees down. That could be studied by looking at the cost of insurance in oligopsonistic markets vs. others.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.