What is the cause of excess costs in US healthcare? Take three - signs of reform

We've already extensively discussed why it costs twice as much for the US to provide healthcare for it's citizens all the while failing to cover health care for all. Most recently, we discussed the hidden tax of the uninsured and the perverse incentive structure of US healthcare which encourage costlier care, more utilization, and more procedures.


To summarize, the US spends more on healthcare compared to other industrialized nations because

  1. We deliver it inefficiently
  2. Without universality problems present when critical and in the ER
  3. Fee-for-service incentives in the form of excessive reimbursement for procedures and hospitals ramp up costs by encouraging doctors to overuse expensive tests and perform more procedures
  4. Direct-to-consumer advertising (we are one of two countries that allow advertisement of prescription drugs) and medicare part D encourage overuse of pharmaceuticals while tying providers hands when it comes to bargaining for lower drug prices
  5. Defensive medicine
  6. Poor management of end-of-life decisions and excessive and futile overuse of resources at the end of life
  7. Absence of a universal electronic medical record (or record format) to prevent redundancy and waste.

Now, what about the new Affordable Care Act? Are there going to be measures to address these sources of excess cost while creating universal coverage? The WaPo has an article outlining reforms addressing many of these specific problems.

First off, fee-for-service is going to be discouraged with increased use of "bundling" of costs:

Zucker is the chief development officer at Baptist Health System, a five-hospital network in San Antonio. For 37 common surgeries, such as hip replacements and pacemaker implants, it would soon collect "bundled" Medicare payments. Traditionally, hospitals and doctors had collected separate fees for each step of such procedures; now they would get a lump sum for treating everything related to the patient's condition.

If a hospital delivered care for less than the bundled rate, while hitting certain quality metrics, it would keep the difference as profit. But if costs were high and quality was too low, Baptist would lose money. For the first time in their careers, the doctors' paychecks depended on the quality of the care they provided.

Four surgeons quit in protest.

While I agree that bundling is good because it disrupts the perverse incentives to use healthcare resources excessively, I wouldn't characterize this as linking the doctors' paychecks to the "quality" of care, as much as it creates an incentive for physicians to use resources more judiciously. We already have a financial incentive to provide high quality care, if you're a crappy doctor you cause more complications, make more mistakes, you get sued, the hospital gets sued, docs stop referring to you etc. Eventually people complain to the state medical board, they review your records, and if you are indeed below the standard, you lose your license. Using general quality indicators can be very problematic, because they might actually harm patients that are poorer, sicker, or older.

If your pay is tied to outcomes, you will bias your results by trying to cherry-pick the healthier, wealthier, younger patients. This effect has been demonstrated in studies of an incentive scheme tied to outcomes by the NHS. If, say, an obese diabetic patient who smokes presents for a hernia repair, a doctor might have an incentive to refer him to another surgeon rather than do it himself, because that patient is more likely to have complications, wound infections and readmissions. The surgeon who doesn't discriminate and takes all comers pays a penalty as, even if she's a better surgeon, she'll still have more complications and readmissions. Tying medical reimbursement to outcomes has to be done very carefully to prevent incentives against treating more difficult patient populations.

Metrics of morbidity and mortality can actually be very misleading. Private hospitals may often tout a lower complication rate and lower mortality in their hospital, but what is missed is that they also treat healthier, insured patients. And when those patients have serious complications what do they do? They transfer them to the nearest academic medical center. The academic medical centers take the sickest, poorest, oldest, most challenging patients, and I think the physicians at such centers are better doctors for it. However, they're rewarded for taking on the tough cases with higher morbidity and mortality statistics.

Luckily, the metrics so far described are rewarding appropriate physician behavior, rather than outcomes:

The program launched in June 2009 with a checklist of quality metrics. To earn a bonus, surgeons would, among other things, need to ensure that antibiotics were administered an hour before surgery and halted 24 hours after, reducing the chances of costly complications.

Only three doctors hit the metrics that first month, but their bonuses caught the attention of others. "There was a lot of, 'Why are those doctors getting more, and I'm not?" Zucker says. Eight doctors got bonus payments in July; two dozen got them in August. Compliance with certain quality metrics steadily climbed from 89 percent to 98 percent in three months.

Two-and-a-half years later, Baptists' surgeons have earned more than $950,000 in bonuses. Medicare, meanwhile, has netted savings: Its bundled rate is about 5 percent lower than all the fees it used to pay out for the same services. "It wasn't a home-run," says Zucker, noting the start-up costs in administering the program -- not to mention a handful of lost employees. "But I'd call it a solid triple."

This is a very positive sign, as rewarding physicians for both good and cost-effective care will help discourage the perverse incentives of fee-for-service, which is a major target of reforms:

In Washington, the Obama administration was facing a different problem: Medicare was eating up a growing share of the federal budget. The program's costs had more than doubled in a decade, from $212 billion in 1999 to $499 billion in 2009. Since it started in 1965, the program has paid providers based on volume. Most private insurance works this way, too: In 2008, 78 percent of health plans paid for coverage on a fee-for-service basis. That system, economists argued, was driving up costs: It pushed doctors to provide as much care as possible, regardless of whether it was effective.

Across the country, however, a few health-care systems had made high-profile moves in another direction. Places like Kaiser Permanente in California and the Mayo Clinic in Minnesota were setting strict budgets for their patients and demonstrating, in study after study, that they could deliver higher quality outcomes at a lower cost than other hospitals and doctors.


What these systems had in common was a model called integrated care, where doctors, hospitals and insurers work together to deliver the most cost-effective treatments. In integrated care systems, doctors are often paid a flat salary, rather than charging for each procedure they perform. They often receive incentive payments for hitting certain quality metrics.

I don't agree with all of these quality metrics though, for instance, penalizing hospitals for readmissions for "preventable" complications.

But the remaining changes will be mandatory. Beginning in October, hospitals stand to lose 1 percent of their Medicare revenue if they can't hit key metrics on "preventable readmissions" -- patients who turn up at the hospital with a complication from an earlier procedure. That's a big change from the current, volume-based system in which those readmissions generate additional revenue for a hospital.

The problem is, we also know who is more likely to suffer from complications and "preventable readmissions", namely poor, sick, old, obese, and poorly-educated patients and those without good home or out-of-hospital support. I hope to see that these quality metrics take care not to discriminate against doctors and hospitals that take care of the neediest patients, and don't perversely reward the doctors and hospitals that pick and chose the easy patients to care for.

Overall though this is encouraging news. Dismantling the incentives for overuse is critical. Now all we have to do is allow collective bargaining for drugs, ban direct-to-consumer advertising, fight back against the "death panels" nonsense that's preventing a mature discussion about end of life care, institute an effective, universally-compatible electronic medical record, and have some reasonable tort reforms to prevent the worst kinds of defensive medicine.


More like this

We've discussed it before, why are costs so much higher in US healthcare compared to other countries? The Washington Post has a pointless article which seems to answer with the tautology costs are high because healthcare in America costs more. How much more? Well, we spend nearly twice as much…
Monday was the start date for an Affordable Care Act provision aimed at reducing high rates of hospital readmission among Medicare patients. This year, hospitals determined to have excess readmissions for patients with acute myocardial infarctions, heart failure, and pneumonia can lose up to one…
A New York Times article by Jane Gross highlights a costly healthcare problem: avoidable hospital readmissions, which affect one in five patients and account for $17.4 billion of the current $102.6 billion Medicare budget. (When people talk about readmissions, they're generally referring to an…
Crazy ranting about impending socialism/fascism aside, there are legitimate critiques to be made of Obamacare. One policy in particular that raises my ire is penalizing hospitals over performance metrics and penalizing readmissions in particular. The way it works is, patients are admitted to the…

From a quality assurance perspective I agree with you that rewarding or punishing end results is a bad idea. At the individual level the SNR (signal to noise ratio) is too low to be effective in behavior modification. As Deming taught, the trick is to improve the process, and you get what you measure.

Deming was a huge pioneer in getting the world to recognize the "cost of nonquality," and there's nowhere that nonquality is more expensive than in medicine.

So modern SQA (statistical quality assurance) tools have gotten very good at identifying process factors (e.g. gloves and other infection control steps) which influence key outcomes. Work on identifying those factors (which is only feasible at large scale) and then improve compliance with best practices, and you'll get your best quality improvement results.

Even if practitioners aren't comfortable thinking of themselves as components in a process, it's a very useful model for improvement. The "artisan" model is another way of, in effect, denying that it's possible to do any better because you can't change the key element: the individual artisan.

dcs, an old fogie artisan (engineer subspecies) who is forcing himself to adopt new methods to improve his own quality. Because they work.

By D. C. Sessions (not verified) on 23 Mar 2012 #permalink

Note another factor helping as new buildings replace old ones the single room has taken over from semi-private ones, partly for hippa reasons, but I suspect we will find fewer cross infections as well. I know my local hospital went all single rooms when it rebuilt a couple of years ago.

A guy like Cheney can afford this heart transplant. I bet he had insurance, on our dime, to cover it. As i is now, I can not even afford to have my father live OR die. Check this out http://igg.me/p/78164?a=494862

By Chrometuna (not verified) on 24 Mar 2012 #permalink

A very large source of increased utilization comes from physicians who own their own imaging equipment. Many studies have found that patients of these physicians undergo much more testing than those who see physicians who do not own their own imaging equipment. This self referral system seems like a prime source of potential cost and efficiency savings yet there seems to be very little discussion publicly about it.

Kilani RK, Paxton BE, Stinnett SS, et al. Self-referral in medical imaging: A meta-analysis of the literature. J Am Coll Radiol. 2011. doi:10.1016/j.jacr.2011.01.016