The Antibiotic Arms Race

Bacteria will win unless we nudge drug companies on R&D.

Enlarged to show texture.

Photographer: Getty Images

Antibiotics are probably the most important medical advance of the 20th century. Medical professionals at the time chronicled the transformation: You had patients who were choking to death with pneumonia or wasting away with tuberculosis, and then suddenly, a few days later, they were cured. If you're a fan of the James Herriot books about working as a veterinarian in 1930s Yorkshire, you'll already have a sense of what I mean. This scientific advancement, in practice, looked like magic.

Unfortunately, our magic powers are eroding. Antibiotics face a worthy, endlessly adaptive enemy: bacteria. The more we use each drug, the less effective it becomes. That means that we need a constant flow of new antibiotics. And we're not getting them, in part because the market for developing these vital drugs is broken.

Here's the basic process for a pharmaceutical company to develop and profit from a drug. First, you spend a bunch of money looking for compounds that might do what you want to do. Then you file a patent application, and register with the FDA and regulators in other countries. This allows you to start spending a boatload of money on clinical testing that will find out whether your new drug actually does something you want it to do, without killing or maiming the patient. At some point during the clinical trial process, your patent will be awarded. Then you finish clinical trials, and if the drug looks good to everyone, you can start selling it. You have from now until the patent runs out to recover the vast sums you have spent on developing this drug (plus some extra for your shareholders, and to cover the costs of all the R&D that did not lead to blockbuster drugs). After that, generic competition will enter the market and compete your margins down to almost nothing. That generally gives you a little over 10 years to make all of the money on a new drug. For example, Opdivo, the exciting and fantastically expensive new cancer treatment that was approved late last year, will probably be eligible for generic production in 2027.

Do you see why this is such a problem for antibiotic development? If a company develops an exciting new cancer treatment that does a great job, oncologists will start pumping the stuff into patients as fast as they can get permission from insurers. If a company develops an exciting new antibiotic, infectious disease specialists will hold onto it as long as possible, using older generics whenever possible and bringing out your new drug only when everything else has failed. By the time they're ready to use it more broadly, the drug will be off patent, and the company that developed it won't make any money off of it. 

As health policy, this is absolutely the right thing to do; it stretches out the time we get to use an effective antibiotic, and saves costs to boot. But as drug development policy, it's disastrous. Companies are less and less interested in creating new antibiotics, because it's too hard to make money there. 

A commission studying this problem for the British government since 2014 has just released its third report, tackling the misaligned financial incentives that plague the sector. Here, in a nutshell, are the suggestions:

  1. Create a global entity to pay drug companies a flat sum for developing an antibiotic that meets specified benchmarks. Instead of getting paid per pill sold, they'd get paid for adding weapons to our arsenal. Even if we rarely use them.
  2. Set up a global fund to support "blue sky" research into new antimicrobials.
  3. Reduce regulatory barriers that add cost and time to drug development.

All three of these are excellent suggestions. The question is whether they can survive the massive political barriers that exist to this sort of thing. Global deals are hard to make. Legislators and regulators might fear this would be seen as compromising public safety to goose profits for large corporations. Global budgets are strained. 

However, the barriers are not nearly so large as they are to something like coordinated action on global warming. The sums involved are relatively small, and many fewer entrenched interest groups stand the way. This program is eminently possible. And it is urgent. If we continue to allow this market to fail, our 20th century miracles will fail too.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

    To contact the author on this story:
    Megan McArdle at

    To contact the editor on this story:
    Philip Gray at

    Before it's here, it's on the Bloomberg Terminal.