This is the next chapter in my Wednesday is Student Day series. Rocky, a medical student, asks below: "What is your take on pay for performance, and will it be integrated into BIDMC?"
My economics professors in college set forth a series of theories and formulas that described the functioning of the free market. We all knew that this formulation was unrealistic, in that most markets are imperfect. There is often "friction" between parties in a marketplace that result in imbalances between supply and demand, that result in uneven knowledge between and among buyers and sellers, or that otherwise gets in the way of an economically efficient equilibrium condition.
But it was not until I joined the health care world that I discovered the extent to which an economic system could be so convoluted that there is virtually no relationship between the value of services provided and the compensation for those services. In health care, there are not only intermediaries between the procurer of a service (i.e., the patient) and the supplier (i.e., the doctor or the hospital), but the actual pricing of specific services is often based on the wrong premises.
I think that most people would like to think that a doctor or hospital would be paid based on the quality of the service provided, but that is not so. Most recently, insurance companies have introduced surrogates for real measures of quality. They attempt to reward providers with "extra" payments for certain accomplishments -- administrative or clinical -- that are deemed to be of value to the insurance company in the plans it offers to its subscribers.
This is a crude system in several respects. For one, the measures chosen do not always add quantifiable value. For another, even when they do add value, the amount of the performance bonus is not related to the value. For another, the bonus does not necessarily pass through to the specific providers who deliver specific services to patients. For another, the bonus is often not really a "bonus" that provides extra revenue to the provider. Rather, it is often in the form of a withholding of a portion of the fair compensation to which the provider is entitled even if the chosen metrics are not accomplished.
It is my hope that, over time, insurance companies will actually base payments on accurate and measurable levels of service quality. It is also my hope that the current imbalance in payments between "cognitive" specialists like primary care doctors, neurologists, and nephrologists and "procedural" specialists like surgeons and interventional cardiologists will some day be set aright. In the meantime, places like BIDMC live with the rules that are decided by the insurance companies. There really is no other choice for us, for, in the parlance of my economics professors, we are price takers. (By the way, the same is true regarding our payments from Medicare and Medicaid.)
Pay for Performance is one of the "trendy" topics across the country. Here in Seattle, there is a coalition of insurance companies, providers, self-insured companies that are working to develop quality measures that will be linked to Pay for Performance but there is actually very little evidence to show that it is effective.
ReplyDeleteWhat people often forget is that we currently have a system that already "pays for performance." Although we all agree that their are inequities in the non-free market allocation of rewards to health care providers(if the market was free wouldn't more Docs go into high-paying practices and over time the market would work it out) The more creative and effective solutions are to pay providers for episodes of care and reimburse for preventive care.
Over 70% of all health care costs are the result of 4 major chronic conditions but the current "P4P" models rewards the provision of services and not the health outcomes of the patient.
In one Robert Wood Johnson Foundation study in Whatcom County, WA (shared care plan) that put patients at the center of their care (with patient controlled Personal Health Records combined with Case Managers) they cut costs $3,000 on average per patient. Despite saving millions of dollars one of the largest local clinics with 40 providers opted out of the system as it would end up costing their specialists money (lost income) to participate.
The current system doesn't reward providers who answer email, provide educational instruction or keep their patients out of the hospital.
People also often forget that only a small portion of care is actually evidence based with clear indications (ie a hip fracture). The vast majority of care is preference based (lump vs mastectomy) or even supply side driven (why for examples does the Dartmouth Atlas study show that patients in New York hospitals are seen 3 times as often as in Seattle with no difference in outcomes)?
In England when they implemented P4P they found that the high quality providers simply received more compensation without any increase in quality or patient outcomes. In fact some would argue that it can institutionalize poor care since patients don't really have the ability to see the high quality providers as their practices are often already full.
Maggie Mahar discusses P4P in her book "Money-Driven Medicine," which I'm reading right now and loving. Also heard a number of great points at a regional SGIM meeting:
ReplyDelete* Who should we be rewarding with P4P? The people who make huge improvements in their quality, or the people who are already doing great with quality?
* What are we measuring, and are we sure it's the right thing to measure? (How many patients are getting multiple influenza vaccines because we don't have documentation to show they've already received one?)
* Are the incentives there? It costs money to collect all the data and make institutional changes, is the amount Medicare is dangling enough to encourage people to change?
* Shouldn't we be focusing on quality for quality's sake? Should we really be tying quality to payment--shouldn't we expect that our physicians and hospitals are providing quality care, no matter what they're getting paid?
* Take a deeper look at quality, and one starts to wonder, as Mahar asks, "Sure, we can show that our mortality rates after CABG are better, but what about figuring out which CABGs weren't necessary in the first place?"
The Missouri Healthcare Executives Group of which I am secretary had this morning a panel discussion on this very topic, and we hard approximately the same points raised at your SGIM meeting.
ReplyDelete> It costs money to collect all
> the data and make institutional
> changes
> shouldn't we expect that our
> physicians and hospitals are
> providing quality care, no
> matter what they're getting paid?
These two seemingly contradictory points are I think the nub of the disagreements over P4P.
Buyers disappointed with practice variation do expect consistent quality for the price that is being paid right now. Sellers say they're being asked to provide an improved, more consistent product for free, as it were.
The only thing I have to say to this is "Look at the example of GM and Toyota." Whether anyone likes it or not, buyers are in control.
The question of the cost of collecting data for quality improvement is related to the "shouldn't we simply expect quality" question. I personally think the incentive payments are quite modest and cannot possibly pay for organizational transformation. Therefore the organizations best able to "grab" the incentive payments they way the programs are currently structured are probably the ones that need them least.
Which brings us back to the first question in your list. And then it brings up Maggie's question of what the healthcare "market" really is anyway — a market for discrete services delivered to anyone willing to lie still, or a market for (public) health outcomes.
Its interesting stuff.
t
I think the medical specialty societies should take the lead in developing quality metrics that each thinks are relevant, appropriate and measurable. I suspect this would be much more difficult for primary care physicians than for surgeons. It would also be necessary to adjust the metrics for the riskiness of the served patient population, which, in turn, means we need a decent individual risk scoring system. I know Medicare already does risk scoring to adjust Medicare Advantage payments that insurers receive. I don't know how good the system is or how easily it could be improved. Presumably, health risk assessments and prescription drug history would be good starting points for individual scoring.
ReplyDeleteP4P may well be more trouble than it's worth for primary care doctors. I think what might be more useful, at least from an insurer perspective, is to track utilization by both the providing and the referring doctor. Ideally, all insurers, including Medicare and Medicaid should be able to pool their claims data to do this thoroughly and then risk adjust the utilization data for each provider's patient population.
UnitedHealth Group assesses doctors based on quality (I don't know the metrics they use) and cost-effectiveness. Those who achieve certain targets are awarded a premium designation. While doctors who achieve the premium designation are not paid extra, patients will soon pay a lower co-pay if they use a doctor from this group. Obviously, this is similar in concept to tiered drug co-pays. As a practical matter, insurers have found it extremely difficult to remove underperforming doctors from their network, so lower patient co-pays might be a good approach to reward the higher performing docs with more patients and steering business away from the underperformers.
Paying for performance in manufacturing or service related business is sensible, rewards lead to higher levels of performance, and performance leads to quality. Healthcare financing weaves a complicated web of bonuses and quality measures that don’t add up, at least just yet.
ReplyDeleteMost of the private insurer's pay for performance plans are predicated on the reduction of costs, and not always intended to improve quality of service to the patient. In fact, some insurers have delivered decreased costs and higher quality with P4P programs. Aligning payment with quality goals, as so many have said is the right direction for payments. I am questioning if the results are sustainable and beneficial in the long term.