Jan 03, 2013
Singapore is to embark on a major review of health-care costs this year. In a new book out next week, David Goldhill analyses what ails America's health financing system, and explains why cost-plus funding leads to higher prices for medical services all round.
IN 1983, the Ronald Reagan administration enacted one of the most significant cost reforms in Medicare's history. The prospective payment system switched inpatient hospital reimbursement from open-ended fee-for-service to fixed fees, paid per diagnosis.
In theory, this would give hospitals the incentive to treat patients as quickly and economically as possible.
The new rules did drive big changes. Since 1983, the total number of days spent by Medicare patients in hospitals has fallen 40 per cent, even as the number of Medicare enrollees has risen 60 per cent. The average inpatient stay is now just over five days, down from 10.
But even an improvement in efficiency of such magnitude failed to slow the cost train. As the number of hospital days declined, the daily charge to Medicare rose to US$1,800 (S$2,200), from US$300.
The prospective payment system is only one obvious example of a long trend.
Most of the major developments in health care - higher doctor productivity, diagnostic scans, new pharmaceuticals, minimally invasive surgery - could be described as increasing health-care productivity. None of these achievements have lowered prices.
Cost fallacy
WHY not? Strange as it seems, cost is only mildly relevant to the price of care. In the world of health care, cost control is based on the fallacy that there is a fixed amount of care we need. Presumably, the more efficiently it's performed, the cheaper it will be. This ignores how providers actually respond to changes in their business. By focusing relentlessly on the cost of care, we actually drive it up.
To understand why this is inescapable, ask yourself: What would you rather pay for the items you buy - whatever price a retailer charges; or a small amount, say 5 per cent, above the retailer's cost?
Take your time; it's a trick question.
The cost-based pricing seems like the better deal. We imagine going into a store, learning that the merchant paid $10 for a sweater, and buying that sweater for only $10.50. But this assumes that, once you opt for cost-based pricing, the costs will stay the same. In reality, the cost-plus-5 per cent system will change the merchant's economic incentives - so the next sweater the store buys will "cost" far more than $10.
Imagine the impact cost-plus would have on the world's simplest business. Your daughter sets up a lemonade stand outside your house and charges a dollar a cup. (That number just seemed right to her.) She sells 50 cups to people passing by each day.
One day the mayor comes along. He's running for re-election, and he wants to buy a cup of lemonade every week for all 1,000 residents of the town. He doesn't want to pay $1,000 a week, though, so he suggests paying your daughter a "fair" profit of 50 per cent. He knows each cup contains about 10 cents worth of lemons and sugar, so he figures he'll be paying 15 cents a serving.
The moment your daughter agrees to this deal, however, she will try to increase her costs, because higher costs mean bigger profits. She is better off with more expensive lemons and sugar, larger cups (maybe even glasses), an assistant to run the stand and a new Lemonada 5000 mixer, which guarantees a perfect mix of sugar and lemon in every glass.
The mayor is no idiot. He sees what is happening, so he renegotiates his deal. From now on, he'll pay her costs plus 5 cents a cup. Unfortunately, this also creates perverse incentives. Your daughter can make more money by reducing the size of each cup. Or she can cut back on customer service, hygiene or speed. Or she can cut side deals with her vendors: The lemon seller can raise his prices - passed on to the mayor - and share the proceeds with your daughter.
The cost of manipulation
BIZARRELY,
a cost-based pricing structure actually adds a new major cost: the effort it takes to track, manipulate and justify costs. In a US$2.5 trillion industry such as health care, these activities are a big reason that administrative costs exceed US$300 billion a year.
An Economist article on dialysis perfectly illustrates the inflationary impact of cost-plus pricing.
Because United States clinics are paid on a cost-plus basis, they prefer expensive drugs to cheaper ones. In fact, many appear to order drugs in units that exceed what a standard dosage requires, because they can charge the government for the waste. The article noted that many clinics preferred an injected drug with a price of US$4,100 a year over the identical drug in oral form, priced at only US$450 a year.
Not surprisingly, the manufacturer of the oral drug responded by increasing its price above that of the injected version, to make it more competitive.
America's entire health-care system suffers from what I call the cost illusion - the idea that a service has a long-term fixed cost.
But every cost is merely someone else's price. And over time, costs themselves are also determined by prices.
What is the cost of orthopaedic surgery? It is the sum of all the costs of the underlying components - the surgeon, anaesthesiologist, nurses, hospital, device, tests and drugs.
But how are these costs determined? Let's look at the orthopaedic surgeon. We may believe there is some objective way to measure the cost of his time - a fair return on his years of education or training, say. In reality, the cost of the surgeon's time depends on the value of orthopaedic surgery to patients. If more patients need it, the surgeon's time becomes more valuable. In a free market, there are two ways the cost of his time could decline: more orthopaedic surgeons fighting for business, or patients benefiting less from orthopaedic surgery.
In an administered market such as health care, on the other hand, our surrogates - insurers and Medicare - substitute their calculation of cost for the workings of supply and demand. This has the strange effect of preventing costs from ever falling.
Let's say Medicare sets the reimbursement rate for a hip replacement at US$15,000. Now say a new drug is invented that makes hip replacements less useful. In a free market, the price would decline. But in an administered system, these prices are viewed as costs, and once set, there is no mechanism to lower them. A hip replacement still takes the same amount of time from surgeons of the same degree of expertise, so the cost must still be $15,000.
In health care, America's system is designed to shield patients from even knowing the prices. Unfortunately, a world without prices is also one that can't achieve the purpose of prices: the allocation of resources to match what consumers want.
Five weeks after my father died from a hospital-borne infection in the intensive care unit of a New York City hospital, my mother received a bill for his treatment - US$635,695.75! The bill was broken down into 17 items. Had I booked dad a room at the most expensive hotel in town for the five weeks of his illness, filled the room with a million dollars' worth of hospital equipment leased for US$15,000 a month, given him round-the-clock nursing care, and paid a physician to spend an hour a day with him (roughly 50 minutes more than at the hospital), it would total roughly US$150,000.
That leaves US$500,000 left over for, say, drugs (billed at US$145,431), oxygen (US$41,695) and blood (US$30,248).
This comparison with actual prices is absurd, of course, because it assumes that the prices on my father's bill were real prices. No one was actually supposed to pay that bill. The prices didn't even bear a relationship to the exchange of funds for dad's treatment. The hospital billed my mother for her share (US$992), which she wisely didn't pay and the hospital wisely didn't try to collect. Medicare paid the hospital according to its concept of the hospital's cost. Of course, there's no question what the competitive price would be for the service of killing my father: zero.
Prescription costs
A STUNTED price system also distorts investment in new treatments. US pharmaceutical companies spent roughly US$67 billion in 2010 on research to develop new drugs. But many of these new drugs target conditions for which perfectly good drugs already exist. It is the lack of consumer prices that explains their me-too approach.
Once a new drug is approved, it enters the marketplace at a high reimbursement rate, compensating the manufacturer for its expensive research. So what's the punishment for entering a crowded market? Very little. Furthermore, even with a promising new entrant, the prices of the existing drugs don't decline; they have already been set to compensate for their "costs". In any normal market, a new entrant would bear not only the risk of being rejected, but also the risk of a price war.
Administered pricing also explains why America's health-care industry has spent far too little on information technology. Your dry cleaner computerised his inventory system because losing a shirt may mean losing a payment or even a customer. But a doctor who invests in state-of-the-art patient data management can't charge higher prices; insurers won't pay. Nor is there a market mechanism to force hospitals that use paper records to accept lower prices - they don't benefit from being more efficient. So the investment is never made.
If we, the consumers, saw and paid prices, we would be looking at a very different industry. My guess is that many of us would pay only for doctors who spend more time talking to us, providers who invest in computerised records, genuinely better treatments rather than me-too drugs for chronic conditions, and hospitals that kill fewer patients.
The writer is president and chief executive of cable TV network GSN. This is an excerpt from his new book, Catastrophic Care: How American Health Care Killed My Father - And How We Can Fix It, to be published on Jan 8 by Alfred A. Knopf.
The ST reprint above was just part 1. The whole series are here:
http://www.bloomberg.com/news/2013-01-03/to-fix-health-care-turn-patients-into-customers.html