Friday, February 17, 2006

When Good Insurance Isn't Good Enough

by Tom Bozzo

There was a lot going on on the high cost health care front this week.

An NYT article on the stupendous prices being set for certain cancer drugs made an excellent case for Dean Baker's modest proposal to nationalize pharmaceutical research. Genentech is planning to charge about $100,000/year for new applications to lung and breast cancer of its blockbuster angiogenesis inhibitor Avastin, "citing the inherent value of life-sustaining therapies." Which is a polite way of saying that Genentech would like to capture more of the value of sustaining life as profit. The catch is that the price is so high that it induces sticker shock even for patients with relatively good insurance (i.e., low co-insurance rates, as in the 5% coinsurance for "catastrophic" costs in Medicare Part D). Genentech also appears to be entering something of a game of chicken with insurers, which (as Paul Krugman repeatedly points out) maintain large bureaucracies for what amounts to the purpose of denying expensive coverage.

In the battle of the Medicare Part D pathologies, the worse policy error here is preventing Medicare from using its purchasing power to negotiate the price of the drug. Since the marginal cost of a patient-year of Avastin for is reportedly very low, a buyer with market power could readily foil Genentech's pricing plans, make the drug available to more patients (demand curves slope downwards!) and transfer some of the surplus from Genentech and parent Roche executives' wildly overpaid hedge fund managers to very sick people and their families.


Joan McCarville, an area dairy farmer, 44 and a mother of two, made the front page of the Wisconsin State Journal when it appeared that the state's Health Insurance Risk Sharing Plan would not cover a lung transplant — which would be her third (*) — under a one-transplant-per-organ policy. In the absence of coverage, UW Health had insisted upon an advance payment or guarantee of $330,000 for the procedure. As she wrenchingly put it, "It's cheaper for me to die than it is for me to live. That's what it has come down to." HIRSP relented, in part since the previous transplants were performed prior to the establishment of the current policy, though the cost McCarville's ongoing treatment means that the pending procedure will likely exceed the HIRSP lifetime cap in this case.

Along similar lines, the Phantom Scribblersphere went into action to help in the case of Annika Tiede (whose mother blogs here). Annika is five and has had two liver transplants to treat a rare bile duct disease; the harrowing story is here. A recent hospitalization caused Annika's $1 million/year cap on insurance payments to be exceeded, making a financial disaster most of us are really fortunate never to have to face. This page:
suggests several ways (including tax-deductible donation methods) to help. Please consider helping.

It's important to keep in mind that these disasters are happening despite insurance — even relatively good insurance.


A couple observations on the economics. As with drugs subject to patent protection, it's important to keep in mind the difference between price and cost for hospital services. The difference between the two, whether to cover drug R&D (and marketing) costs or the 'fixed' costs of the hospital, leads to a 'static' inefficiency — the price induces people to forego treatment whose benefits exceed the marginal cost of provision — that is hoped to be offset by 'dynamic' efficiencies. Drugs get developed, hospitals can stay open. That doesn't automatically imply that it makes sense to expose individual consumers to non-cost-based prices, which is a gist of the Health Savings Account nonsense. Indeed, for lots of consumers, a tax deduction for routine health expenses is worth less than an insurer's power to obtain discounts from the list prices of services. See also Alan Schussman, who has an excellent post on the absurdities of trying to turn individuals into health-care bargain hunters via HSAs.

Also, the "stop loss" provisions for the insurers — the annual and lifetime caps that Annika and McCarville face — seem higher to the relatively (or currently) healthy than they really are. For example, our policy has a $2 million lifetime benefit limit. That seems to be an effectively infinite pot when it's just $150 doctor visits chipping away at it, but when you have a combination of the prices charged for expensive treatments far exceeding their economic costs, and an outcome when caps of typical magnitudes of premature death, disability, and/or bankruptcy of the patient's family, it's conceivable that the caps are much too low. The value of a "statistical life" is typically measured to be several million dollars (the amount may vary over the life cycle), which may be lost because market failures prevent patients from obtaining treatments that cost, say, only a few hundred thousand dollars. This is to say nothing of the value of the actual lives facing the dilemma.

One of the underlying problems is inadequate risk pooling, even at the level of state pools such as the Wisconsin HIRSP. This is why proposed reforms such as Kerry's national reinsurance plan would constitute significant improvements to the U.S. health insurance system, even if they're less bold than might be warranted.


(*) The first having failed due to a poor match with the donated organs, i.e., through no fault of McCarville's.
If pharmaceutical research was nationalized, drugs like Avastin would likely not be available, nor would most other important drugs.

Avastin isn't cheap in the first place. It is not a small-molecule drug, it's a monoclonal antibody. These are very expensive to produce in the quantities necessary for humans. The cost of development of these types of therapies is gigantic too. That's not to say that $100,000/yr doesn't provide Genentech with a very large profit, but the drugs based on antibodies that I know of are sometimes $5,000 per dose. $100,000/yr for life-saving treatment with monoclonal antibodies for doesn't seem that outrageous.

I don't understand why you are on Genentech's case. Without them, Avastin would not be available at all. If it can help cure certain types of cancers, then people with those types of cancers are better off with it than without it, right? Some people might not be able to afford it, but if Avastin weren't available in the first place, no one could benefit from it at all.
Bryan, while I think you should note the presence of the words "modest proposal," I think you are stepping out on a limb in suggesting that publicly funded research couldn't replace the present pharma R&D system in principle. There isn't much uncertainty as to what diseases are out there, or which stand to benefit from improved treatment; plus a lot of the basic research is publicly funded anyway. I suppose there's also a legitimate question as to whether corporations are the appropriate entities to bear drug development risk, and there's definitely a question as to the value of some copycat drugs.

In the case of Avastin, the NYT reports that the higher doses for the new uses don't materially increase production costs, so the pricing is nearly pure profit for Genentech/Roche. On the economics, you'd need to be careful that the $5K represented the marginal cost of the dose of the drug in production, and not a price that makes a contribution to R&D costs, corporate overheads and profits, etc.

While I would argue that their pricing model is legal but unethical, it's unquestionable that it results in inefficiencies given the existence of the drug. (As you might have noted, I did not deny the need to cover costs.)

Dean Baker's point is that if the differences between pharma prices and costs were an explicit tax (rather than an implicit one, via the grants of patent monopolies), everyone would be hopping up and down as if on fire -- the implicit monopolist tax is that stupendous. The upshot is that you university-based researchers could be wildly inefficient with your research dollars and still potentially be cheaper than the current system.
$100,000/year is actually a good value compared to previous surgical and radiological therapies (especially transplants) that had lower efficacy and higher risk.

Nationalizing research is not going to get drugs to market, only to the lab. Just because a compound works in the lab (and even in early trials), doesn't mean it can be efficiently manufactured and brought to market.

We need a system that rewards discovery and translation without the anti-competitive effects of patent protection that drive prices through the roof on an inelastic demand curve.

One way to do this is to separate research firms from manufacturing and distribution firms, and make compulsory licensing a condition of issuing the patent to the research firm. In the cases of products with incredible social benefit, the government could buy the license from the R&D firm (either on the market or by emminent domain) and then re-license it for manufacture at a lower price. Or, it could use its incredible buying power to negotiate lower prices -- as you said, the lack of purchasing power is a major failure of Part D, which replaced many state and private plans that did negotiate.
Paul: Mitigating the 'static' inefficiencies of the patent monopolies is really what I'm getting at with the reference to Dean Baker. Off the cuff, I'd certainly leave manufacturing to the private sector even with a maximalist government R&D role. ("Nationalizing" is something of a misnomer, since you couldn't really prevent firms from doing their own research.)

I think there's merit the compulsory licensing idea and, of course, allowing all buyers to use such purchasing power as they have.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?