Managing Costs by Keeping Patients Healthy

HMOs–a bad word in health reform circles–flourished and then declined 20 years ago.  But were they entirely worthless?  Often they were bullies that took no patient or provider input into their decisions, and their cost control strategies could be brutal.  Their payment schemes to providers could cause unscrupulous doctors to skimp on care, but they did bend the cost curve for a few years.

Their best feature was that their payment schemes did address overutilization.  This had the potential to save lives and reduce morbidity.  Nothing good can come to a patient who is subjected to a test or procedure that he doesn’t really need, and a great deal of risk comes with every medical intervention.  I see countless instances of overutilization every day: repair of fractures that don’t really need it, shoulder and knee scopes for vague or non-existent indications, MRIs ordered for pain without any physical findings.  Our current reimbursement schemes reward all of this.

If we knew then what we know now regarding patient satisfaction, physician strategies to improve patient outcomes, the importance of good communication and physician empathy, would the HMO payment model have had more success?  Could we have kept patients satisfied, feeling well cared for, and been given good care while eliminating the financial incentives for dangerous overutilization?  What if HMOs had understood the precepts of patient centered medicine as we currently understand it.  Could primary care doctors have told their patients “no” without causing a backlash?

Now, I’m not suggesting that we return to HMOs, but alternative payment strategies that deliver just the care that patients really need can succeed so long as they are coupled with doctors practicing patient centered care.  Office visits, especially to primary care docs, will need to be better reimbursed to make this model attractive to general practitioners, but HMOs floundered primarily due to poor patient satisfaction rates, and we know enough now to manage costs better and keep our patients healthy and happy with their care.


from Malcolm Maclure, Prof & BC Chair in Patient Safety, University of British Columbia

I have one suggestion based on experience in British Columbia: when restrictions on coverage must be implemented, use design delays (staggered implementation.)

Ideally these would be randomized control groups, but controlled time series designs, especially with matching, are almost as rigorous as RCTs in my experience, and institutionally they seem to be more feasible.

I was inspired by my interview with Tom Chalmers almost 20 years ago, to advocate within the British Columbia Ministry of Health for control groups to be used when the public drug plan (‘PharmaCare’) introduced restrictions on drug coverage. I call this approach ‘designed delays’ because delays are the norm, and ‘designed’ feels like adding orderliness to the process, whereas ‘randomized’ feels like adding chaos to a program that is already complex.

Chalmers told me he had urged the Veterans Health Administration to pair-match hospitals and randomly select one of each pair to be first with a new service while the other hospital was delayed as a control. Despite his close connections with the VHA, he thrice failed to convince the VHA to use designed delays — for Coronary Care Units, for CAT scans and for MRIs. He was very frustrated by the experience and stymied on how to routinize randomization.

In British Columbia, we now routinely implement prescribing education initiatives with randomized delays but our success in using designed delays for changes in drug coverage has been much more limited. Our main success in using designed delays was in 1999 when we partially restricted coverage of nebulized medications. It went so smoothly that the method of designed delays was endorsed for 2 other drug restriction policies (but the restrictions were never introduced for other reasons.)

In contrast, designed delays when ADDING coverage of a new medication seems to be much more difficult. The hope (or hype) that the new drug will yield better outcomes, makes delays unacceptable. On the other hand, SUBTRACTING a medication or service that is believed by many (but not everyone) to be of little value, is made easier by promising rigorous evaluation by use of delays controls.

The most vehement critic and the strongest proponent of the policy AGREED that randomized delay was a good idea. They were both convinced that rigorous evaluation would prove they were right. Commitment to rigorous evaluation thus instantly refuted the criticism that the government was only interested in cost control, not patient outcomes.

A university ethics committee had no difficulty with randomized delays of the restriction because it eventually would apply to everyone, unless the evaluation showed the restriction was harmful (in which case the drug plan director was committed to reassessing the policy.) In contrast, another ethics committee was critical of our proposal for delayed coverage of a NEW medication.

My hypothesis is that when a health organization is forced to restrict services (e.g. the UK National Health Service must cut 20 billion pounds in the next four years), there will be many opportunities to evaluate the balance of benefits versus harms of the restrictions by using delayed controls.

Although Tom Chalmers’s advice mostly fell on deaf ears during times of plenty, I predict his advice will become more appealing to decision makers as cost pressures increase.

(These comments are my own opinions and observations in my academic capacity as Chair in Patient Safety at UBC. That appointment includes my continuation half-time as Co-Director of Research and Evidence Development, an advisory role in the Pharmaceutical Services Division of the BC Ministry of Health. I wish to stress that my opinions are not to be interpreted as representing the Ministry’s positions on these matters.)

Excessive Spine Surgery

Heading the class in “Avoidable” is multi-level spinal fusion as a primary treatment for low back pain in patients with degenerative disc disease.  Despite the many disclosures over the past few years in the press and medical journals regarding this important issue the surgeons participating in such excessive surgery have not yet been held accountable for their action, which thereby continues unchecked.  Given the fact that poor behavior does not change unless there are significant consequences for it I would be most appreciative if this conference would spend some time discussing the issue of developing appropriate consequences and even more appropriately, effective behavior modifiers which would also make sense. 

C. Burton

2019 © Lown Cardiovascular Research Foundation | 21 Longwood Avenue Brookline, MA 02446 USA | P: 617-732-1318 | F: 617-277-0347 | Contact Us