Fee For Service: The Problem (Part One of Three)

Commentators ranging from leaders of provider organizations, to academics to health plan executives  to government administrators have focused on fee for service payment methodology as a major cause of health care inflation.   Fee for service really does encourage too much, and I’d like to share a metaphor.


Imagine if we paid Toyota not for a finished car, but by the bolt.  You can envision an automobile brimming with bolts just like minivans are brimming with cupholders.   The bolts would add cost, add weight, and they might be ugly.  But worse still, each bolt is a potential point of failure.  You might catch your clothing on a bolt – or a bolt might loosen and cause noise or even mechanical damage.  The fee for service system that encourages us to do too many uncoordinated procedures encourages us to deliver extra visits and extra diagnostic tests.  Each of these is costly and takes away from our patients’ productive time.  Worse still, each extra step we put into a patient’s care adds handoffs, decreases reliability, and increases the chance of medical error.   Just like those extra bolts in the automobile, unnecessary uncoordinated care puts costs too much, and also puts our patients at risk.


Some examples of fee for service payment at work:

1)      1. Disease Management:  There is good evidence that provider-based disease management interventions can be quite cost effective.  However, hospitals across the country have made initial investments in disease management during good times, and have pulled back when it became clear that preventing hospitalizations might be a social good, but was costly to hospitals unless they faced a shortage of beds for high-margin elective care.

a.       University of Pennsylvania canceled its dozens of disease management programs in the late 1990s when it became clear that the money they were saving was lost revenue for the system.  (Linkto this is Harvard only, and the article is Burling S. Ounce of prevention proves a costly cure. Philadelphia Inquirer. 1999 Nov 29. I cannot find this article in publicly-accessible web, but email me if you'd like a copy)

b.      Beth Israel in New York cut out its diabetes disease management programs abruptly when it made the same discovery. 

2)     2. Imaging is increasing at a much higher rate than all other medical services.   The fee for service system, coupled with very high margins on a per test basis, encourage excess investment in imaging technology.  Once the investment is made, utilization is likely to increase.

3)      3. There is little coordination among providers in most medical communities, leading to repeated laboratory and other diagnostic tests.  Regional health information exchanges can help prevent this, although most RHIOs are having a hard time finding funding (since those duplicating the tests are generally being paid handsomely for this) 

4)      4. Many providers insist on “decoupling” visits. For instance, a patient might be asked to return for a follow-up visit to discuss a laboratory or imaging finding, or a patient might be instructed to make a follow-up appointment for a procedure (such as ear wax removal or minor dermatologic procedure) to allow for higher billing than if this procedure was performed at the time of the initial visit.   “Max-packing,” or dealing with all of a patient’s issues at today’s appointment, has been shown to improve access and adherence, and can save valuable patient time and effort.

a5    5 . There is also evidence that the current fee for service structure does not reflect what we actually value from health care.  Distortions in market prices persist, such as the national Medicare CPT code fee schedule.  Once a price for a service is set, it is rarely adjusted to reflect the evolving market for competing services.  Newer, cheaper-to-produce innovations in delivery or technology do not lead to lower prices;  prices are not changed if a procedure is found to be ineffective.   This engrains current practice patterns and disengages payment from what evidence suggests should become the standard of care. As the private market tends to follow Medicare's lead, distortions in fee schedules are propagated throughout the system. 


         Fee for service has its problems, but it also has its adherents.  Small, independent practices have a hard time accepting any type of bundled payments – so large swaths of the current delivery system are not optimal for capitation, episode payment, or other methods of bundled payment.  As long as it appears the fee for service system is sustainable, it will be very difficult to move to a more integrated payment system. 


There is an emerging consensus that the current system is not sustainable – which opens the door to payment reform that moves us away from fee for service.    I have two upcoming posts on this issue – the next will detail some of the requirements to make a transition operationally and politically feasible, and the last will review some of the current efforts to transition from fee for service.

 (Thanks to Cecilia Gerard, graduating from Harvard School of Public Health this spring,  for valuable comments and adding to the text of this post)