Tuesday, February 15, 2011

Guideline Experts Fail Again: and this time it costs us

One of the buzz-phrases of health reform and cutting costs has been, "We are going to pay physicians for quality, rather than quality." This sounds good. I would agree with this method.

The question becomes: What is quality health care? For hospital care, one can get pretty good measures of things that happen acutely. Some examples:
-Paying less to a hospital when a patient gets a hospital-acquired illness (e.g. a fall or infection due to lack of sanitary conditions)
-Pay more when a hospital discharges patient with a complete plan for post-hospital care

But for outpatient care, the measures are much harder. Here are some candidates for diabetes care:
1. Pay doctors for how much their patients like them
2. Pay doctors for how long their patients live or what their quality of life is
3. Pay doctors for what their patient's lab values are

Each of these choices has its problem. Option #1 seems like a pretty good option. But you could be a really bad, friendly doctor. (That's not me.) Option #2 seems the best measure, because this is what patients care about. But you can imagine the problems: patients switch doctors, some patients do things that a doctor cannot control, etc.

So we are left with #3, which is how the current proposals urge to pay doctors. The following article reveals how this process has gone terribly wrong:
Sudden Acceleration of Diabetes Quality Measures, February 16, 2011, Pogach and Aron 305 (7): 709 — JAMA

Basically we are using measures that have little research behind them. We are using lab values that do not matter much to patients. Why do we use them? Because the quality experts messed up. They jumped the gun and made quality standards before the research was finished that would give them the answers. I do not think this problem will end. I think we should not rely on paying doctors based on lab values, or intermediate tests that don't have a big effect.

My solution: a mixed payer model. Pay primary care doctors a set fee to take care of patients for a year. Based on a person's age and some other factors, a doctor would get a set amount for the year. The doctor could do whatever they want for that patient. But, if the costs goes above this amount, it's on the doctor's shoulders.

Next: pay doctors like all other professionals, by the hour. Some patient visits take 5 minutes (a cold). Others take an hour (a cancer diagnosis). Fairly simple.

Finally: Pay a group of doctors based on the community's health. So if one city's overall health is increasing more than another's, increase their pay a little bit.

Any of these options is better than paying doctors for lab tests that patients do not care about, and that may not even indicate how good a doctor is.
Post a Comment