If this blog’s motto weren’t “Helping idiots who desperately need my assistance by calling them fucking morons since 2004”, it would probably be “WASH YOUR DAMN HANDS!” That being said, I’m somewhat shocked by the negative reaction the recent ‘superbug’ outbreak study has received. That might have something to do with that this paper was the first most people outside of the field had heard of this and with this occurring at an NIH facility (although why people didn’t lose their shit when the same group reported an Acinetobacter outbreak–and for the same reasons–escapes me). Nonetheless, the dumping on genomic epidemiology is, well, stupid and short-sighted.
First, there’s the argument that the sequencing really didn’t inform what to do during the outbreak because it took too long (and wasn’t started until too late). This misses the point. Five years ago, if you want to sequence one isolate, it took weeks, and that involved unleashing the firepower of a major sequencing center. At the end of it all, you wound up with a mediocre genome with a price of high five digits.
Now we can easily crank through a hundred genomes of much higher quality at one-hundredth the cost, and we haven’t begun to realize the cost savings yet (it’s not the cost of the actual sequencing, but everything else surrounding it). If a group (non-profit or profit) wanted to speed up the process and lower costs, it’s not hard–and such efforts are ongoing.
Which brings us to the second point: speed. I don’t anyone who argues that genomic epidemiology is a ‘bedside intervention.’ It does not cure a sick patient. What routine surveillance does is it prevents the next patient from getting sick. As the NIH outbreak illustrates, these are not like food outbreaks–one source where everyone gets sick and rapidly. Hospital outbreaks usually go on for quite a while. As I’ve outlined elsewhere, if you have a targeted, routine genomic surveillance program in place, you can identify outbreaks that you would otherwise never even see–and do so much earlier. Why do I say this? Because this is already done with mid-20th century technology (along with some nineteenth century tech too).
By all means, we need traditional infection control measures. And, while it pains me to say it, we know hand hygiene (and machine and surface hygiene) are important, yet they are not that effective in practice. People just don’t wash their hands like they know they should.
Those measures are far more effective when coupled with epidemiology. Infection control requires many pieces, often of small or no effect by themselves. If we’re smart and focused, genomics can provide significant benefit, especially as we get faster and cheaper.