The cost of medicines

An interesting article (probably not free full text) in the October 26th JAMA about why medicines cost so much. It is a good introduction for those who haven't thought much about the problem including my colleagues who drive around in Mercedes thinking the pharmaceutical companies should just lower their prices.

It does a good job pointing out many of the problems with the current system:

Six major problems with the patent system are (1) recovery of research costs by patent monopoly reduces access to drugs; (2) market demand rather than health needs determines research priorities; (3) resources between research and marketing are misallocated; (4) the market for drugs has inherent market failures; (5) overall investment in drug research and development is too low, compared with profits; and (6) the existing system discriminates against US patients. Potential solutions fall into 3 categories: change in drug pricing through either price controls or tiered pricing; change in drug industry structure through a "buy-out" pricing system or with the public sector acting as exclusive research funder; and change in development incentives through a disease burden incentive system, orphan drug approaches, or requiring new drugs to demonstrate improvement over existing products prior to US Food and Drug Administration approval

but finds many suggested solutions are equally problematic.

For example, one commonly suggested remedy is too require new drugs to demonstrate improved efficacy compared to currently available medicines not just to placebo. The idea being that many new meds are very similar to existing ones (think beta-blockers or statins). The authors point out that having several medicines in the same class may decrease costs by allowing insurance companies to bargain for good prices. In addition, there may be significant medical differences among medicines in the same class.

They have some good suggestions, including requiring post-approval comparison studies of the new medicine vs. established treatment, public funding for study of rare illnesses and something called patent "buy-out" in which the government pays off an inventor for the value of his invention (in this case a medicine) for the amount he/she could have expected to make in monopoly profits and then puts it in the public domain. In this case, the government would pay a pharmaceutical company what it would make from selling a given medicine and then allow generic manufacturers to make it and sell it near the marginal cost of production. This would have the benefit of making the medicine cheaper but still encouraging innovation. Of course how to determine the profit the company would make would be hard and the government would spend a lot of money (not to mention problems of international access).

Nonetheless, there could be significant benefits. For example Omalizumab (Xolair), an anti-IgE antibody for the treatment of asthma, is very effective but very expensive. Only a few asthmatics qualify for it. On the other hand the companies profits are limited because althought very expensive only a few patients use it. If the government paid Genentech/Novartis their expected profits from charging $15-20,000/year for 5% of asthmatics, they'd get the same amount of money. But if the cost came way down, a much larger number of patients could benefit. It seems like a win-win situation. The article notes downsides including difficulty in establish which drugs deserve buyout and calculating a fair price and the difficulty generating enthusiasm to pay a huge amount upfront. In addition, GlaxoSmithKline would be justifiably upset having to Xolair priced cheaper than Advair thanks to a huge government subsidy.

Apparently, the only time this has actually been done is in the case of the dagguerotype patent in France in the 1800s. This may be an example of something that makes sense to economists but is hard to put into practice.

Read the whole article if you are interested, it does a good job addressing the issue of drug costs in an even-handed way.

There were also a couple of interesting statistics in the article. In 2002 drugs accounted for just over 10% of health care spending, up from 5.8% in 1992 but similar to the proportion in the middle 1960s. Also, the average cost of brining a new medication to market is $800 million.

My own sense is that there needs to be effective price competition among medicines. Either consumers (maybe with health savings accounts) have to pay for a lot of the cost themselves so they make tradeoffs like "is $20 more a month worth it for once daily dosing?" or HMOs/insurers have to negotiate prices for specific meds in a class and charge more for patients who use other ones (I know they do this to some degree and it is pretty effective, since almost none of my patients is willing to pay a higher copay than necessary).