By Evan Falchuk

At the Wall Street Journal’s Health Blog, Jacob Goldstein writes that programs that pay more for better quality care are a “tricky proposition:”

Even if you can figure out what to measure, and how to account for differences between patient populations, you still need to have a decent sample size; anybody can have a bad (or good) day, so you need to measure a large enough number of events to minimize the role of random variation.

But 65% of primary-care docs work in practices that are too small to draw meaningful conclusions about the quality of care they provide (at least if you rely on Medicare data), according to a study published this week in JAMA.

Fair enough.  But it begs the question:  why would we need to pay doctors more for delivering quality service (whatever that term might mean)?

The hidden assumption in these kinds of programs is that quality problems in health care are caused by bad financial incentives. Since doctors are paid the same regardless of the quality of their work, they don’t care enough about whether they do a good job or not.  Pay them extra if they do well, and you’ll see improvements as they try to earn that extra pay.

It might be true if the practice of medicine were like an assembly line.  In that kind of work, the goal is to deliver large numbers of standardized products through a series of repetitive, simple tasks.  People are good at this kind of work, and can be incentivized to deliver different kinds of outcomes, depending on how you pay them.  Want more cars?  Pay extra for hitting some production target.   Want higher quality manufacturing?  Pay extra for lower levels of defects.  Robots do this kind of work even better, and you don’t even need to pay them anything at all.

But doctors aren’t robots, and this isn’t what medicine is all about.  Of course, this hasn’t stopped the exact opposite from becoming the conventional wisdom about health care.

Costs are too high?  Increase throughput and reduce unit costs.  Or, in English, make the doctor see more patients and pay him less for each one he sees.  Still too high?  Get nurses to do some of the work so throughput can be increased even more.  Oh, and let’s cut the pay, too.  Meanwhile, let’s buy some new computer systems that will make work more efficient.  Computer systems always do that.  Let’s also come up with some quality measures, like maybe, 1,500 or so, and pay doctors a little extra if they meet some of them.

If it weren’t for all the politics surrounding health care, I think many people would find it amusing that anyone thought this was a very good idea.  But it’s the state of the art in reform.  And it’s characterized by a fixation on metrics that too often misses the bigger, more important, picture.

So what’s wrong?  In health care, we do not have a consensus on what “quality” means.  Some say quality should be measured based on outcomes, but even the very best doctor can’t ensure a good outcome.  Sometimes the news is bad, and the outcome will be what it will be.

Instead, why aren’t we measuring quality based on what people really want from their doctors – the right diagnosis and treatment?  As a patient, I would gladly pay more to go to a doctor who I knew was better at this than her colleagues.

As I’ve noted many times before, the fundamental mistake is prioritizing money over medicine.   By focusing on ever more clever ways to design economic incentives, our system undervalues the very things that make it possible for doctors to deliver the right diagnosis and treatment. Things like time with the patient, thinking about his or her problems, consulting with colleagues, and coming up with sound advice.

If we want to remake our health care system, we need to start with the idea that the right diagnosis and treatment is the fundamental goal.  Everything else we create should be based on whether it helps serve this goal, above all others.