:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/33706811/1321427.0.jpg)
The best-intentioned health care policies don't always lead to the best outcomes for patients.
This is one problem at the core of the Veterans Affairs' scandal: administrators received financial bonuses keeping wait times short. The aim of those bonuses was to incentivize good behavior: If there's money to be made in getting patients seen quickly, the reasoning goes, then hospitals will focus on reducing wait times.
Instead, those extra bonuses appear to have incentivized bad behavior. They gave administrators an incentive to do something easier than improving health care: falsify records to make it look like wait times were shorter, and still net bonus payments.
This is not how the bonus program, obviously, was supposed to work. And it underlines a challenge that's actually much bigger than the Veterans' Affairs scandal: its really difficult, in any hospital, to design payment programs that incentivize doctors to do the right thing.
This is something that American health care system has struggled with for ages — and something that Obamacare hopes to address, too.
"There certainly can be unintended consequences," says Jerry Cromwell, a senior fellow in health economics at RTI International, which helps Health and Human Services evaluate the effectiveness of health incentive programs.
The goal: Pay doctors for providing great health care
Veterans' Affairs was trying to do something that is becoming increasingly common in health care: paying providers more money when they deliver better health care. This is known, in the health world, as "pay-for-performance" or "value-based care."
This is different from the way that insurers and the government usually pay for health care services. For decades now, the health care system has relied on a fee-for-service system, where doctors are paid a flat fee for every service they provide. The financial incentives of a fee-for-service system are pretty terrible. They encourage doctors to provide as much care as possible, regardless of whether it actually helps a patient get any better.
In response, the health care system is slowly shifting towards incentive-based reimbursement that tethers financial rewards to better health care.
The health care law's value-based purchasing program, for example, takes a small portion of hospitals' reimbursements and puts them in a bonus pool. Hospitals that score high on metrics — how satisfied patients are, for example, and if they get appropriate care — get money out of the bonus pool.
The hospitals that don't perform don't get bonus payments. And the hope is this will lead to improvements in care. That doesn't, however, always happen.
There are lots of ways for incentive programs to go wrong
In the case of the VA, the bonus payments appear to have incentivized fraud rather than excellent care. The Daily Beast has reported hospital administrators across the country received between $10,000 and $15,000 in bonus payments for submitting falsified records of wait times.
"Yes, it is gaming the system a bit," one administrator wrote in an email obtained by the Daily Beast. "But you have to know the rules of the game you are playing, and when we exceed the 14-day measure, the front office gets very upset, which doesn't help us."
Officials could have, of course, reported the wait times and raised a red flag about the challenge. But instead, with financial incentives in play, reports suggest they decided to lie about the wait.
This isn't how other incentive programs go wrong — but there are less nefarious reasons that trying to reward good behavior in health care can create perverse insurance.
"There are potentially some concerns about these measures," Cromwell says. "Some have to do with reporting, and the accuracy. But others are a concern of teaching to the test. If we identify, for example, 10 major areas to reward and penalize, you could see a lot of effort being put into hitting those metrics."
The value-based purchasing program, for example, partially bases bonuses on how satisfied patients are with their care. That seems to make sense: hospitals should care about whether their patients are treated well.
But some research shows that patient satisfaction actually correlates with worse care. The most satisfied patients, one study found, were most likely to die, possibly because they have doctors who prioritize the care patients want over the care they actually need.
"Without additional measures," a team of researchers at University of California—Davis warned in 2002, "an overemphasis on patient satisfaction could have unintended adverse effects on health care utilization, expenditures, and outcomes."
Teaching to the test
Take, for example, one quality measure that hospitals regularly use: whether cardiology patients die within 30 days of their procedure. Hospitals will look at that metric as a way to figure out if their cardiologists are doing a good job, sometimes using it as a way to assign financial bonuses.
The measure, on the face, seems sound: Doctors should be rewarded when their patients don't die. But a new analysis last month showed something bizarre: there was an inexplicable uptick in cardiac patients' deaths one day after the 30-day penalty period had ended.
"At least no one I've talked to has come up with a hypothesis about why you would see that increase," says lead study author Bryan Maxwell, a cardiologist at Johns Hopkins University. "If you have a person whose job it is to do three things, and pay them just for the first thing, naturally they'll focus on doing the first thing really well."
Maxwell says its difficult to come up with great quality metrics that aren't too burdensome to implement. Hospitals could, theoretically, track patients for a longer time and across different facilities to get rid of the weird incentives around hitting the 30-day mark. That would require more resources though on their part.
"It's not benign to say let's just measure five more things," Maxwell says. "Ther'es a balance to be struck between saying we should measure more, and the burden of taking those measurements."
