Can Big Data Tame MRSA Superbugs?
Last month, in a PBS Frontline interview, Dr. Arjun Srinivasan, associate director with the Centers for Disease Control and Prevention (CDS), told the interviewer that the age of antibiotics was coming to an end. Can big data bridge this problem?
“For a long time, there have been newspaper stories and covers of magazines that talked about ‘The end of antibiotics, question mark,’” Srinivasan told the Frontline interviewer. “Well, now I would say you can change the title to ‘The end of antibiotics, period.’”
He’s not alone. This past weekend, new alarm bells were sounded by European health officials through The Lancet health journal, as doctors warned that humanity is under attack, and current coordination efforts are insufficient.
Worse, the problem is spreading. “[The MRSA Superbug] were initially found in hospitals and infecting severely ill people that were treated by a lot of antibiotics,” said Dr. Ruth Massey, a researcher with the Department of Biology & Biochemistry with the University of Bath in a video last week. “Unfortunately, in subsequent years, they’ve really broken out of the hospitals and they’re now causing big problems in the community with healthy people with no hospital exposure.”
In fact, recently three members of the NFL’s Tampa Bay Buccaneers were diagnosed with MRSA, with one player, an otherwise healthy All-Pro, having to undergo surgery due to an infection that had gotten into the bone in his left foot.
“Staph Aureus is very commonly associated with people,” explained Dr. Tim Reed, an associate professor at Emory University in Atlanta, Georgia, who is working with Dr. Massey on a big data approach to the MRSA issue. “It’s estimated that naturally, 50 percent of everybody on earth carries Staph Aureus – it’s very common. When Staph Aureus acquires an antibody resistance, especially MRSA, it becomes very hard to treat in a hospital setting.”
Across the ocean, Massey and Reed are working together to try and find a solution to the problem through the utilization (and mobilization) of data. In a crowd source fundraising video last week, the researchers explained their approach and hopes for finding a solution.
“In my lab here in Bath, we are very interested in phenotyping, so looking at how the bacteria… cause severe disease,” said Massey. “The main way they do that is through the secretion of toxins which are proteins that go and digest the human cells that surround it… We’ve been looking at the genome sequences and trying to identify the genetic signatures in these bacteria, that would lend one bacteria to be very toxic and one to not be so.”
While Massey identifies the phenotype signatures, Reed indicates that it becomes a big data challenge for his group at Emory. “What we’re trying to do is take the signatures that Ruth has identified and apply that to the sort of worldwide collection of MRSA and see if we can find some fundamental genetic signals that would predict toxicity and potentially antibiotic resistance.”
Together, they are trying to build a system that would give hospitals resources to use data to target antibiotics towards specific superbug strains. Reed has built a database, called “Staphopia,” which catalogues more than 6,000 MRSA bacterial genomes, and aims to be the data heart of this issue.
“You can sequence the genome of a bacteria in a matter of hours,” explained Massey, “and we envisage that in time this will become the primary means of diagnosis in clinics. So the end goal will be when genome sequencing is common in a clinic, they will then go to Tim’s website (or an equivalent) and upload the sequence to it, and [that information will be used to] tell the clinician wherever they are in the world that the strain is either highly toxic or not very toxic, and they can treat the patient accordingly.”
The duo, who say they are short on staff to do the research, launched a crowdfunding initiative on indiegogo last week. “What we need is bodies to come and put these things together,” said Massey. “A student would end up with an amazing multidisciplinary set of skill that they could apply beyond this project, as well as producing some amazing data that we have great promise, great reason to believe is going to work very well.”
While not speaking to this project specifically, Srinivasan addressed the needs for this type of effort to become more prevalent, at least in the United States, and indicated the CDC has been moving in that direction. “There is a definite need to bring the surveillance of both antimicrobial resistance and antimicrobial use into the modern era, and that means we need to do this electronically. We can’t rely on the paper-based reports and those kinds of paper-based systems that we have historically been using a lot of to do this type of surveillance.”
According to Srinivasan, the CDC has built a real time national system to track infections in hospitals called the National Healthcare Safety Network [NHSN], which recently branched to contain a new Antimicrobial Use [AU] Module. “[It] is a system that allows for completely electronic capture of the antibiotics that are being used in a hospital. …It sends it to us at CDC, but most importantly, it gives it back to the facility in real time.”
Srinivasan says that the CDC is aiming to build a fully electronic system that, like Reed’s Staphopia, will extract electronic information on resistance from hospitals and summarize that information and provide it simultaneously both to the health-care facility where the information comes from and to CDC.
“That module, which will be launched, we hope in the early part of next year, really will be a giant step forward,” he says. “It will be transformative in our efforts to monitor resistance. It will take us out of the old model, which was doing this manually, trying to collect all this information from a variety of different laboratory systems, and take us to a position where we can do that electronically.”
Will it work? We can only hope at this point. But it goes to show that this big data stuff isn’t all advertising and stuffing online shopping carts. There are life-saving implications as the technologies to manage and utilize huge amounts of data are being built.
The clock is ticking. Maybe big data is the answer.