The Superbugs are Coming. Data Science Can Help.

As concerns mount about the growing risk of drug-resistant microbes, scientists and health professionals are scrambling for solutions. Data science can play a central role in the fight against the looming global threat.
Since the dawn of our species, we co-evolved with bacteria, viruses, parasites, and other microbes that inhabit our bodies and environs. Many of these microbes are harmless; some are symbiotic and even essential for our health. Others are dangerous, but thanks to miraculous advances in public health and medical science over the past century, we can prevent and treat many common microbial infections.
Yet some in the health industry fear that may be changing. We misuse and overuse antimicrobial drugs on a massive scale, and the bad bugs are beginning to evolve new resistance mechanisms. As the world flattens and gets more connected, drug-resistant pathogens can spread faster than ever. Meanwhile, we are not developing new medicines and upgrading global health systems fast enough to keep up with our microbial foes. The result could be new strains of “superbugs” with devastating impacts.
Scientists have long warned about these risks, but a recent study by the Review on Antiomicrobial Resistance (AMR), a UK-based initiative to provide research and policy guidance on AMR, lends a new urgency to the issue. Commissioned by the UK Government and the Wellcome Trust, the study makes some frightening predictions about the scale of the threat. They estimate that drug-resistant strains of tuberculosis, malaria, HIV/AIDS, and other pathogens already kill around 700,000 people per year. By 2050, the annual toll could reach 10 million.
The report advocates for an array of economic, political, and social interventions to tackle the problem. It also acknowledges the importance of data science in supporting those interventions. In fact, the report devotes an entire section to ideas for improving AMR surveillance in humans and animals. Fortunately, many organizations are already working to improve the availability and quality of AMR data.
The World Healthcare Organization leads one major effort. In 2015, it launched a Global Antimicrobial Resistance Surveillance System that aims to create standards and processes for data collection and sharing. It will initially focus on clinical monitoring of eight high-risk bacterial strains, but may eventually morph into a platform for sharing AMR data of many types.
Advances in bioinformatics could supercharge these efforts. Andrew McArthur, a researcher who holds the Cisco Research Chair in Bioinformatics at McMaster University, imagines a future in which clinicians can instantly sequence the DNA of a disease-causing microbe and use that data to determine the best course of treatment for the patient. That data could then be shared with global AMR surveillance networks and analyzed in real time.
“We can already sequence bacterial DNA with remarkable speed and affordability,” says McArthur. “The big challenge is creating software and protocols to make all that data useful for AMR surveillance.”
Since overuse of antibiotics in humans is one of the biggest AMR risk factors, tracking their consumption is also critically important. Almost everywhere in the world, physicians routinely prescribe antibiotics when they are not needed. In some places, including many emerging markets where the burden of infectious disease is particularly high, antibiotics are sold as over-the-counter medicines and used with virtually no discretion.
My own work at mClinica, a Singapore-based health data and technology firm, involves building health data collection and patient engagement platforms in Asia’s emerging markets. Antibiotics are among the most overprescribed medicines in every country we cover, and our data shows that sub-therapeutic dosing is common. While this may be discouraging, we can also use this data to identify hotspots of antibiotic misuse and plan targeted interventions.
Pharmaceutical companies, whose R&D pipelines for antibiotics have all dried up in recent years, recognize that they must be part of the solution. At the World Economic Forum in January 2016, many of the world’s top pharmaceutical, biotech, and diagnostics companies signed a declaration that called for efforts to encourage prudent stewardship of currently-available therapies, hasten drug discovery, and undertake other measures to prevent us from slipping into a post-antibiotic era.
Achieving these goals will be all but impossible without smart use of lots of data. Enter the data scientists. Armed with an array of technical skills, from statistical methods to data visualization techniques, they can help bring together complex datasets from varying domains, find patterns in the noise, and develop actionable strategies for fighting AMR.
Data science can provide insights that help us develop new antimicrobial medicines, prevent overuse of the ones we already have, and reshape health systems to reduce their need in the first place. It is not a silver bullet for the AMR problem, but it can turbocharge efforts to find and implement solutions.
Will Greene is a writer and strategy consultant focused on Asia’s emerging R&D ecosystems. You can find him on LinkedIn.