Thursday, June 2, 2016

The Collective Action Problem of Resistance to Antibiotics

"We estimate that by 2050, 10 million lives a year and a cumulative 100 trillion USD of economic output are at risk due to the rise of drug-resistant infections if we do not find proactive solutions now to slow down the rise of drug resistance. Even today, 700,000 people die of resistant infections every year. Antibiotics are a special category of antimicrobial drugs that underpin modern medicine as we know it: if they lose their effectiveness, key medical procedures (such as gut surgery, caesarean sections, joint replacements, and treatments that depress the immune system, such as chemotherapy for cancer) could become too dangerous to perform. Most of the direct and much of the indirect impact of AMR [anti-microbial resistance] will fall on low and middle‑income countries. It does not have to be this way. ..  The economic impact is also already material. In the US alone, more than two million infections a year are caused by bacteria that are resistant to at least first-line antibiotic treatments, costing the US health system 20 billion USD in excess costs each year." 
This is from the report "Tackling Drug-Resistant Infections Globally: Final Report and Recommendations," from the Review on Antimicrobial Resistance that was set up by the UK government, funded by the Wellcome Trust and the UK Department of Health, and chaired by Jim O’Neill (who was chief economist at Goldman Sachs for many years and is known as the originator of the acronym "BRICs" to refer to the emerging economies of Brazil, Russia, India, and China.) Background reports are supporting documentation for the report are available here.

For economists, antibiotic resistance falls into the analytical category of collective action problems, which are situations where economic actors in pursuit of private gain have no incentive to take a social cost into account. Problems of air and water pollution can fall into this this category. In the case of antibiotics, they clearly help many sick people and can help livestock gain weight, too. But those using antibiotics for private gain have no incentive to take into account that when they are commonly used, resistance to them evolves in a way that can make them less effective, or just plain ineffective. The report(on p. 16)  includes a discussion of the issue of antibiotic resistance in the terms economists prefer to use: externalities, imperfect information, and public goods.

The policies to address this issue are conceptually straightforward. Over the longer-term, provide incentives for companies to do research and development that can lead to new antibiotics (as well as other methods of fighting bacterial infections). Given that many existing antibiotics are off-patent and available in cheap generic versions, and also given that doctors might prefer to hold fancy new antibiotics in reserve unless or until the current versions don't work, trying to create new antibiotics may not look like a very encouraging market to pursue without some additional policy steps. But in the shorter-term, the policies need to be about reducing the overly casual use of antibiotics. If antibiotics are used only when really needed, then the problem of antibiotic resistance can be mitigated. Here are a few of the steps along these lines that jumped out at me from the report.

1) In the past, many doctors have prescribe antibiotics on the "it can't hurt" philosophy, and while antibiotics are unlikely to hurt that particular patient, the broader social problem of antibiotic resistance can indeed hurt. Thus, one set of policies would encourage doctors to prescribe antibiotics only when really needed. "One study showed that in Belgium, campaigns to reduce antibiotic use during the winter flu season, resulted in a 36 percent reduction in prescriptions. Over 16 years, the cumulative savings in drug costs alone amounted to around 130 Euros (150 USD) per Euro spent on the campaign."  The key point here is that antibiotics only work against bacterial infections, and do nothing at all against viruses. The report points out that diarrhoeal illness kills about 1.1 million people per year in low and middle-income countries. Howver, about "70 percent of episodes of diarrhoeal illness are caused by viral, rather than bacterial infections, against which antibiotics are ineffective – and yet antibiotics will frequently be used as a treatment."

Perhaps the most important development here would be rapid diagnostic tools, so that doctors could tell more or less in real time--or perhaps within a few hours--if an infection is bacterial and which specific bacteria is involved. This technology would mean that antibiotics could be used much less and targeted much better. As the report notes, that is not what happens now.
"When doctors and other medical professionals decide whether to prescribe an antibiotic, they usually use so-called ‘empirical’ diagnosis: they will use their expertise, intuition and professional judgement to ‘guess’ whether an infection is present and what is likely to be causing it, and thus the most appropriate treatment. In some instances, diagnostic tools are used later to confirm or change that prescription. This process has remained basically unchanged in decades: most of these tests are lab-based, and would look familiar to a doctor trained in the 1950s, using processes that originated in the 1860s. Bacteria must be cultured for 36 hours or more to confirm the type of infection and the drugs to which it is susceptible. An acutely ill patient cannot wait this long for treatment, and even when the health risks are not that high, most doctors’ surgeries and pharmacies are under time, patient and financial pressure, and must address patients’ needs much faster."

2) Take public health measures to avoid people getting sick in the first place, so that antibiotics are less-needed for that reason. Especially in developing countries, major steps to reduce disease include better sanitation and clean water, along with vaccination campaigns. In developed countries, a main focus should be to reduce infections that arise in health care settings: "Across developed countries, between seven and 10 percent of all hospital inpatients will contract some form of healthcare‑associated infection (HCAI), a figure that rises to one patient in three in intensive care units (ICUs). These levels of incidence are even higher in low and middle-income settings,
where healthcare facilities can face extreme constraints, sometimes as fundamental as access to running water for cleaning and handwashing."

3) Dramatically reduce the use of antibiotics in agriculture, where they are often used not just to treat animals who are sick, but as a sort of all-purpose aid to keep animals from falling sick and to help them gain weight.  These antibiotics often work into the environment--say, through disposal of animal waste products--and thus spur bacteria to become resistanc. "The quantity of antibiotics used in livestock is vast, and often includes those medicines that are important for humans. In the US, for example, of the antibiotics defined as medically important for humans by the FDA, over 70 percent of the total volume used (by weight) are sold for use in animals. Many other countries are also likely to use more antibiotics in agriculture than in humans but they do not even hold or publish the information."

Moving ahead with these kinds of policy steps should be an urgent priority. A family of bacteria resistant to the antibiotics usually saved for a last resort has recently been found in a US patient. (The scientific article on this discovery in the journal Antimicrobial Agents and Chemotherapy is available here.)

For two previous posts on antibiotic resistance, see:

Homage: Like many others, I suspect, I ran across this particular report on antibiotic resistance  because of a cover story in the Economist magazine of May 21, 2016: the Economist leader is here; the more detailed article here.