The Dec 2010 issue of The Atlantic contains an investigative report by Robin Fields on dialysis quality and costs in the U.S. and how lessons learned can apply to national health care policy. The story also appears on ProPublica, which includes photos. It is a very good story, though it is unbalanced. (It’s the nature of investigative journalism.) In addition to reading the article, I encourage you to read the comments at the bottom of both versions of the article. I’ll address two points from the article below.

The dialysis business is big and concentrated but there are outliers

As reported in the Economist Apr 2010, the market for dialysis treatment is large. There are about 350,000 patients in the U.S. receiving dialysis therapy. The average cost of dialysis treatment is about $70,000 per patient per year. Further, it is highly concentrated, with one big buyer (called a monopsony) and two big suppliers (called an oligopoly).

In the U.S., the cost of dialysis care is covered mostly by Medicare, and the federal government spends about $24 billion per year, or about 85% of the total cost, which represents about 6% of Medicare’s total budget. End-stage renal disease is the only medical condition that Medicare covers regardless of age. Dialysis reimbursement is the single biggest Medicare cost category and is growing faster than overall medical costs.

There are two major manufacturers of dialysis equipment in the world, Fresenius of Germany and Gambro of Sweden. These manufacturers sell to dialysis service providers that tend to buy all their equipment from one provider or the other. In fact, there are now two major dialysis service providers in the U.S., both of which are for-profit enterprises. The largest is Fresenius Medical Care, a subsidiary of Fresenius and uses only Fresenius equipment. Next is DaVita, an independent company that acquired all of Gambro’s clinics and uses Gambro equipment. Between them, they operate about two-thirds of the 5,000 dialysis clinics in the U.S.

There are also many smaller regional clinic chains, many of which are not-for-profit. Not-for-profit organizations need not be any better run than for-profit ones. In fact, their clinics are often inefficient, poorly maintained, and less likely to used advanced technology. But some of them, like Northwest Kidney Centers in Washington are well run and have clean, comfortable, safe facilities. NKC operates 14 dialysis centers around Seattle, making it the biggest provider in the Puget Sound region. [Disclosure: I am a volunteer for Northwest Kidney Centers and have contributed to it.]

Reimbursement policy affects how dialysis treatment centers are run

Currently, Medicare reimburses dialysis providers a flat amount per patient plus pharmaceutical costs. This has led to several behaviors by the dialysis providers that are unintended, but should have been expected.

1) They use high blood flow rates through the dialysis machines. This reduces the time each patient is in the clinic, which means more patients can be handled per day and less technician labor is needed per patient. However, high blood flow rate causes low blood pressure in the patient during dialysis. This is correlated with higher rates of cardiac events and death.

2) They limit each patient to three visits per week and discourage home hemodialysis (which allows the patients to treat themselves more frequently). However, fewer visits are also correlated with higher rates of cardiac events and death.

3) They prescribe higher doses of drugs such as heparin and especially expensive ones like erythropoietin than dialysis centers in other countries. They also use more injectable drugs, which are more expensive than oral ones. Medication now accounts for one-quarter of the total cost of dialysis treatment.

Medicare will soon switch to bundled reimbursement of $230 per session (indexed for inflation) and institute a 2% bonus system under its Quality Incentive Program. This has been a very controversial change.

First, the change itself will reduce revenues for nearly all dialysis centers, putting even more pressure to reduce costs. In the U.S. there has been considerable pressure to consolidate and cut costs as Medicare keeps reimbursement rates low.

For examples of discretionary costs, consider the Northwest Kidney Centers. It runs kidney fairs to encourage the public to get tested for hypertension and diabetes to avoid end-stage kidney disease. It also works with nephrologists to encourage its most healthy patients to consider getting a kidney transplant. And it partners with the Univ. Washington Medical Center to support the Kidney Research Institute. All of these actions are good social policy, but are expensive, have little benefit to existing patients, and hopefully reduce the total number of patients needing its services in the future. Thus, they are all bad for the bottom line.

Second, even if improved quality reduces costs, the benefit may not reach the dialysis centers. Medicare may pay a small bonus to the dialysis center, but may keep most of the cost savings  for itself. For more on the perverse impact of improved quality on care provider profit, see this article in the Nov 2002 amednews.

Finally, Medicare plans to measure quality via blood tests rather than base it on medical outcomes. Blood tests have the advantage of being fast, cheap, and less prone to measurement errors than other tests. However, blood tests are not a reliable indicator of quality. Some better measures would be mortality rates and morbidity rates. But both are hard to measure and expensive to track. Blood tests may be easy to adjust to meet the QIP goal and may not be closely correlated with desired medical outcomes.

Bill Peckham, who is quoted in the article, is an expert on how bundled reimbursement will affect dialysis centers and their patients. Mr. Peckham is a patient and an advocate for Northwest Kidney Centers. You can read more at his blogs, Dialysis from the Sharp End of the Needle and Fix Dialysis.