"A big data revolution is under way in health care," said a recent McKinsey report, pointing to pharmaceutical companies' aggregation of years of data into medical databases. Meanwhile, a Harvard microbiologist is trying to figure out how new technology that can now digest huge volumes of data quickly can be married to her laboratory's newfound ability to generate reams of data on the deadly bacteria that causes tuberculosis.
But what role will big data play in an evolving tug-of-war to better understand the enormous use of antibiotics on farms, and their role in helping drive a global crisis in superbug infections?
We know that antibiotic use generally drives the development and spread of multidrug resistant "superbugs". But until Congress amended the Animal Drug User Fee Act in 2008, the FDA never publicly reported data on total antibiotic use in agriculture. That changed late in 2010. The Agency issued the first public figures on antibiotics sales of products intented for livestock use, and from that summary data we now know around 70% of all antibiotics important to humans are sold with the intent of giving them to pigs, cows, chickens and turkeys. Summary data like these, unfortunately, tell us nothing about how antibiotics actually are being used on farms including which drugs are given to which animals and for what purpose. Without the latter, experts trying to better understand exactly how livestock antibiotics worsen the crisis in antibiotic resistance -- and which policy measures to reduce antibiotics overuse will be most effective -- are mostly thwarted.
For years NRDC and its allies have pushed FDA to use its authority to collect better data. To its credit, the FDA in May finally proposed modest new rules under which pharmaceutical companies will need to provide not only overall sales figures for their animal antibiotics, but also estimates of the percentage of product sales by animal species. Predictably, the pharmaceutical and feed industries are fighting back, saying FDA's is overreaching its authority.
We laud FDA's quest for better data, but the proposed rule doesn't go far enough. As pharmaceutival companies acknowledge, they may have limited information on how their products are used on farms. To better meet the FDA's public health goals, data also should be collected on the actual use and distribution of antibiotics -- the closer to the farm, the more useful such data are likely to be.
That's why NRDC filed comments this week asking that FDA reach even higher. It can and should (but currently doesn't) ask medicated feed mills to provide information on how much antibiotics they use, what products they are put into and what animal species are the intended recipients. (Feed mills are FDA-licensed operations that fill the "prescriptions" for antibiotics put into animal feed, and thus have far better information on the species or which antibiotic are being used, and how much are being used). Feed mills already collect such information, and hang onto it for up to two years; they just have never been required to report it to the FDA, or the public.
Stay tuned, because the fight for better data on livestock antibiotics is far from over. FDA just announced a public meeting in Washington, DC on the topic on September 30th. We'll be there.