Little Bundles of Joy
Why do insurers ignore the most promising way of cutting health costs?
Now that a national health reform bill passed, it's time to get to work on the reforms themselves. To illustrate the big challenges, and also a practical way forward, let's consider asthma in children. A leading cause of misery, the disease accounts for one in six pediatric emergency room visits and is the most common cause of inpatient hospitalizations in many urban areas. And yet, most experts agree, asthma is eminently treatable and most of these hospitalizations are preventable.
Treating asthma requires a sophisticated algorithm. Frequently, parents must try three or more combinations of medications before finding the one that works best. Guidelines to prevent subsequent attacks also advise that schools receive written asthma "action plans" for the child; allergists evaluate and treat specific asthma triggers; homes be treated for insect pests.
Last November, researchers from Children's Hospital Boston reported interim results from a community-based asthma program that used case managers, home health aides, and outreach to coordinate the proper steps for children with severe asthma. Within six months, emergency-room visits dropped by 60 percent. Hospitalizations fell by 80 percent and stayed down for a year. And yet, according to a legislative liaison from the hospital, "in the current traditional health care system, these kinds of workers"—such as home-visiting nurses and case managers—"are not providers that have been able to bill or get any payment for their services." In two years, the philanthropic funds underwriting the pilot program will run out.
Why don't doctors do a better job of matching the right patients with the right procedures for treatable problems like asthma? Observers tend to blame this mess on our "fee-for-service" payment system. The more doctors do, the more they are paid; rather than rewarding quality, insurers pay for quantity. If a hospital's doctors do a terrible job, necessitating longer and more frequent hospitalizations for a child, they get rewarded with more money. That's why many reformers believe the solution is to "bundle" payments: Insurers would pay a fixed, up-front cost for each particular health problem—like asthma—and let the hospital and caregivers determine the best way to use the money to deliver quality care. Bundling could save money, improve care, and encourage innovation by health providers.
Suppose that you hire a contractor to renovate your kitchen. If you paid on an hourly basis, the contractor would almost certainly drag his feet to bill more hours. If he hangs the cabinets all wrong, he bills extra time to realign them. But if you paid a fixed fee for the whole job (a bundled payment), the contractor has an incentive to get it done in a timely manner.
This isn't a new concept. In the 1990s, for example, Medicare bundled payment for heart bypass surgery in seven hospitals as part of a demonstration project, which resulted in shorter hospital stays, better survival, and lower costs. Last year, RAND researchers evaluated the potential savings from a dozen cost-control measures (like computerizing health records or creating in-pharmacy clinics) and concluded that bundling payments, which was projected to shave 5 percent from national health spending, was the most promising.
Why, then, hasn't the idea caught on widely? Elizabeth McGlynn, an associate director at RAND who writes widely on bundled payments, tells me that historically, insurers have tinkered with the delivery of health care—like managed care with its focus on "gate keeping" primary-care doctors—rather than tackling payment reform.
The first objection is that no one knows how to divvy up the money well. Suppose Massachusetts insurers paid a fixed fee to cover child's asthma care at Children's Hospital for a year. If the child goes out of town and has an attack requiring a visit to the ER, who pays? And what about diseases like diabetes, where a patient's care is divided among a primary-care office, physical therapy unit, outpatient dietician, and so on?
The second drawback is that hospitals might randomly get a patient who blows the budget. A single asthmatic who requires, say, a temporary heart-lung bypass machine could cost more than the combined bundled payments of dozens of others. Many small hospitals are unwilling to take that insurance risk.
To overcome these drawbacks, hospitals and clinics merge into accountable care organizations, or ACOs, which are massive, vertically integrated health collectives that can perform or subcontract all necessary services for tens of thousands of patients. But this strategy has its own risk for overall health spending. ACOs can become so huge that they transform into health care monopolies. In Boston, for example, the dominant Partners HealthCare (which includes Massachusetts General Hospital and Brigham and Women's Hospital) has used its market dominance power to drive up prices for everything from routine X-rays to cardiac surgery.
Enter the Prometheus model, which was developed in conjunction with the Robert Wood Johnson Foundation in 2006 and represents the most promising way of structuring bundled payments. At first glance, it seems to continue traditional fee-for-service payments: Hospitals still bill for each procedure or consultation. But once the patient leaves the hospital, the insurer tracks additional costs for complications like infections or rehospitalizations, which account for up to 80 percent of costs for conditions like heart failure. If the patient's total bill care comes in under the "evidence-informed case rate" (essentially, the budget set by the insurer), the hospital keeps the difference. If it goes over, the insurer still pays.
The key to Prometheus is that the initial caregiver is strongly, and positively, incentivized to avoid complications. Francois de Brantes, a former General Electric executive who oversees the Prometheus model, explains, "You start with payment to create organizational change, not the other way around."
Here's how Prometheus could reduce the cost of cardiac bypass surgery, which might be complicated by poorly controlled diabetes requiring extra intensive-care days and a wound infection necessitating readmission. Under the prior fee-for-service, a hospital might get $47,500 for the bypass and the surgeon $15,000. The avoidable diabetes complications cost an extra $14,000, and the avoidable wound infection $25,000. In the end, the insurers cough up about $101,000.
Under Prometheus, the insurer pays the hospital $61,000 and pays the physicians a further $13,000. Now, according to Prometheus data, the average patient might be expected to consume another $15,300 for complications (called the "severity induced allowance"). If no complications result, the hospital and surgeon pocket the $15,300 allowance. The insurer also wins, since the evidence-informed case rate is $89,300, which is $12,800 less than the fee-for-service cost. (If the complications occur anyway, the insurer pays for the care and nobody is worse off than under the previous system.)
Now imagine how this might help asthmatic children in Boston. Suppose the typical child with severe asthma cost $500 in outpatient visits and $3,000 in emergency-room visits and hospitalizations. The total cost per patient is $3,500.
With a Prometheus-style bundled payment, outpatient care for severe asthma might cost $1,500 (which would pay for the innovative Community Asthma Initiative program), and the severity induced allowance for emergency or inpatient care might be $1,000. If 80 percent of emergency visits and hospitalizations are avoided, the outpatient caregivers earn a large windfall for their high-quality care—nearly $1,000 per patient. And the insurer also wins, since the total costs per child is $2,500 instead of $3,500.
Now that health insurance reform has passed, the wonky work of designing the right payment plans has begun. And the Prometheus model is a good start.
Darshak Sanghavi is Slate's health care columnist. He is chief of pediatric cardiology and associate professor of pediatrics at the University of Massachusetts Medical School as well as the author of A Map of the Child: A Pediatrician's Tour of the Body. Follow him on Twitter.