While we like to think of ourselves as rational creatures, there’s no doubt that human beings are actually quite awful at assessing risk. So I can understand why Ethan Linck thought to contextualize the risk of drinking from backcountry streams with data. “Life is triage, a constant series of negotiations between risks of varying severity,” he wrote. “And how we talk about those risks matters.”
Yes, it does—which is exactly why his piece in Slate last week was so damaging. It was anything but a careful, scientific evaluation of the risks. Wes Siler over at Outside Magazine already pointed out a myriad of issues with the article, but I want to zero in on the actual data, because Linck claimed to be looking at the matter scientifically. Instead, he cherry-picked sources to argue against doing one of the simplest things you can do to protect yourself from some truly awful diseases when you’re backpacking: treating your water.
— Outside Magazine (@outsidemagazine) February 7, 2018
Simply put: when you drink water straight from a stream, river, or lake, you have no idea what’s in it. And that’s bad, because, as epidemiologist Tara Smith, PhD, explained for SELF last month, it can be contaminated with all sorts of nasty things. “These include Giardia, a parasite found in streams and rivers that causes “beaver fever” in campers and hikers, and bacteria like Shigella and Campylobacter that can cause bloody diarrhea,” she wrote.
These organisms exist in waters because they exist in our digestive tracts and those of other animals. So anywhere that there’s poop near water, that water could contain pathogenic strains of Escherichia coli, Salmonella, Campylobacter, Aeromonas, Yersenia enterocolitica, Leptospirosis, Listeria, or Vibrio, in addition to a suite of viruses and protozoan parasites like Giardia and Cryptosporidium. Some of these bugs only cause short-term, if severe, gastrointestinal distress. Others can cause issues that last for weeks, months, or even years.
“Even water that appears pure or clear can be contaminated by people or animals,” explains Jonathan Yoder, MPH, who is deputy chief of CDC’s Waterborne Disease Prevention Branch. I asked Yoder and his colleague, epidemiologist Kathy Benedict, PhD, what their thoughts were on the bold claim that science doesn’t support backcountry water treatment (that “the scientific evidence shows that this mandate [to filter water] rests on a shaky foundation”). Needless to say, they disagreed.
Because appropriately assessing risk requires a clear understanding of what’s at stake, it’s important to point out that with many of these pathogens, we’re not just talking inconvenient cases of the runs. The gastrointestinal symptoms that can occur—intense diarrhea, vomiting—aren’t easily managed by hikers on long trips in the middle of nowhere. “If they have an acute situation, it can actually be quite scary,” says Benedict, “especially if they’re out there by themselves, because they can get themselves into a lot of trouble very quickly.” And people, even people with rapid access to medical care, do sometimes die.
To his credit, Linck doesn’t exactly downplay the dangers of Giardia or other potential pathogens. But, he argues, that their dangers are not important. It’s totally ok to drink any water you might come across while backpacking (or as the title says, “You Don’t Need to Filter Your Stream Water”) because: “The idea that most wilderness water sources are inherently unsafe is baseless dogma, unsupported by any epidemiological evidence”.
It’s a lovely little straw man that he immediately sets to tearing apart. To paraphrase his argument: everyone says most water is chock full of terrible things. So as long as most water sources are safe, then you’re good to drink from any stream. And look! Here’s a study that says feces contamination was only found in a minority of the sites tested! And one by… a magazine… that says the same thing!
But let’s take a quick look at those sources. The first study examined lakes and streams in Kings Canyon, Sequoia, and Yosemite National Parks looking for feces-associated bacteria (fecal coliform), because where there’s feces, there could be something dangerous in the water. And they found it at 22 of the 55 chosen sites. So yes, that’s a minority, but it’s 40%—and of those, ~16.3% (9) had “higher levels”. And that magazine one? Well, Linck didn’t directly link to it, but Google is a wonderful thing. They surveyed seven locations three times throughout the year for parasites, five of which had at least one test come up positive for either Giardia or Cryptosporidium—just over 71%. But, they claimed, only one had close to dangerous levels. That’s still 1/7—or a little over 14%. It’s a shame that Linck didn’t include the percentages in his article, because for all his talk about appropriately assessing risk, he doesn’t give the information needed to actually do that.
And those studies—if you can really call the magazine investigation that—were conducted in the early 2000s, each testing a miniscule fraction of the waters that are accessed by U.S. hikers every year on a handful of occasions. There is other research he could have cited—like this 2009 study from Georgia, which found 79% of water samples from rivers and streams in southern Georgia over a year tested positive for Salmonella. Or this 2011 one which found “Salmonella, Campylobacter, Staphylococcus aureus, Vibrio vulnificus, and V. parahaemolyticus were widespread—12 of 22 O’ahu streams had all five pathogens.” And even if he didn’t want to count Hawai’i (though it is a U.S. state), then he could have cited this 1987 study in Washington instead, which found Campylobacter at ~36% of sites tested (5 of 14), including a mountain stream. According to the authors, “Campylobacter spp. are widely distributed in central Washington and are present in a variety of aquatic habitats including ponds, lakes, and small mountain streams, which ranged in elevation from 1,460 to 5,400 feet above sea level.” Or this 1982 study, which detected fecal coliform bacteria in ~86% of streams in Great Smoky Mountains National Park, with ~5% testing extremely high (>1 organism per milliliter). They concluded that “it is clear that water in natural streams in the Great Smoky Mountains does not meet standards for safe drinking water.”
Even if he just looked at other papers by the same authors as his study, he’d have found this 2004 study, which found that out of 31 backcountry sites, 45% tested positive for fecal coliforms, and just over 25% had high levels. Or this 2006 study which found fecal coliform at 1/15 sites used by backpackers (~6.7%). Or this one from 2008, which found coliform bacteria in 18% of human day use areas, and 14% of backpacker sites. But really, they all tell the same story: one-in-five to one-in-ten sites test positive, which means they might get you sick if you drink their water untreated.
Of course, even counting every sample in every study I just mentioned, only a tiny fraction of potential hiker drinking sources have been tested, so a lot remains unknown. And because these pathogens are associated with human and animal activities, their presence can be as transient as the wildlife. None of these studies really explains just how variable the risks can be, which is actually something that was noted in that Backpacker article:
“Risk in this area is very hard to quantify,” explains Tod Schimelpfenig, curriculum director for the Wilderness Medicine Institute, of the Wyoming-based National Outdoor Leadership School. “Sample a creek at one point in time and you could have a flush of organisms from an animal that just defecated upstream. Sample it 20 feet upstream 2 hours later and you could find nothing. The risk of drinking untreated water in the wilderness depends entirely on when and where.”
If any conclusions can be drawn, it’s that not all areas carry the same risks, and that even remote streams can harbor dangers. Even if you could assume something like 10%-20% of water sources test positive for fecal contamination (and thus may get you sick)—then, yeah, technically, only a minority of sites are ‘unsafe’ (not that anyone was arguing that most of them were). But just imagine going to a buffet and seeing a sign that said “Nine out of ten of our menu items tested negative for fecal bacteria!”—would you want to eat?
Yoder wouldn’t. Of course, he’s had Giardia before, so he knows exactly what he’d be risking. “I think after you have a one of the more severe diarrheal infections—and I speak from my personal experience—I think you understand that even though you know ninety-nine times out of a hundred, if you drank from that water source, you’re not going to get infected, it’s worth it to protect yourself for that one percent chance.”
While Linck’s prevalence analysis is a straw man at best, he goes on to claim something that’s patently false: that “research to date has failed to demonstrate any significant link between wilderness water consumption and infection with these threats.”
Again, he turns to decades-old data, citing a 1993 study where only 5.7% of backcountry travelers in California’s Sierra Nevada had Giardia, and none of them felt sick. Mind you, he failed to note that more than half of the travelers did purify their water, so they’d have been protected from the parasite. And he didn’t mention that 16.7% of them did return with gastrointestinal illnesses—they just weren’t that one. Then he cites a survey of health departments and a meta-analysis from the same researcher looking at Giardiasis in the US, both of which found that the majority of cases came from non-wilderness sources. But again, that’s a straw man—no one was arguing that backcountry streams were the main source of Giardia infections. That’s like saying most house fires aren’t started by deep-frying a frozen turkey, so by all means, it’s totally safe! And Giardia seems like a very specific hill to die on. What about the myriad of other possible disease agents? Nothing about Campylobacter, Leptosporosis, Shigella, Norovirus, E. coli—and the list goes on.
Heck, if we’re going back decades for our data, why not mention this 1977 study of 256 known Giardia cases in Colorado, where they found a significantly higher proportion of cases among people who “visited Colorado mountains (69% vs. 47%), camped out overnight (38% vs. 18%), and drank untreated mountain water (50% vs. 17%), p less than .001” [emphasis mine]. They also found that when they tested 16 Rocky Mountain streams in uninhabited areas, they found unsafe concentrations of fecal coliforms in all of them. Or this 1984 study that found people who drank untreated water taken from a stream, river, or lake were about ten times as likely to have Campylobacter jejuni infections than their neighbors who didn’t? Or this 1983 one, which found both Campylobacter jejuni (23%) and Giardia lamblia (8%) in people who came back from Grand Teton National Park with diarrheal disease. They also found that Campylobacter occurred “most frequently in young adults who had been hiking in wilderness areas and was significantly associated with drinking untreated surface water in the week before illness” [emphasis mine]. They even isolated Campylobacter from one of the mountain streams suspected as a source.
Or, Linck could have looked at more recent data—like this 2016 study of Giardia outbreaks in the US from 1971-2011, which found six linked to rivers or streams. Or this 2017 study looking at waterborne illness outbreaks in the US in 2013 and 2014, which found another six outbreaks of Giardiasis affecting 91 people which were “caused by ingestion of water from a river, stream, or spring.” And these are just outbreaks where multiple people got infected and sought treatment—they don’t look at the countless individual cases, many of which are not reported to authorities because we don’t go to the doctor for diarrhea unless it’s really bad (diarrheal diseases are notoriously underreported). And it can take a week or more for signs to show in some cases, so you might not even connect your bowel troubles to your recent hiking experience.
Linck does mention a camping related outbreak of Giardia from 1976 which was thought to be waterborne. But he claims, citing a 2004 editorial, that the analysis was wrong. Instead, “the afflicted campers failed to properly wash their hands after using the bathroom.” (Yet, none of the campers had the parasite in their stool before the trip, so even if poor hand-washing spread it faster, where did the disease come from?)
“I don’t disagree with the author that, in addition to water treatment, there are other very important things that people can do to stay safe in the backcountry,” says Yoder, “and those include the proper disposal of waste and using good hand-washing hand hygiene, particularly after using the bathroom and before eating before preparing food.”
By pointing to hand-washing, Linck is basically making the turkey argument again; the fact that poor hand-washing also causes outbreaks, maybe even more of them or worse ones, doesn’t mean that drinking untreated water never causes any.
And there is a study that compared the two—a paper examining the outcomes of 280 people planning to hike the Appalachian Trail in 1997 for a week or more found that both bad hygiene and drinking untreated water increased the backpacker’s odds of cancelling their trip early. Those that practiced good hand-washing were about 50% less likely to have gotten sick, while those that drank untreated water were almost 8 times as likely to have had to pack it in because of diarrhea. They concluded: “Hikers should purify water routinely, avoiding using untreated surface water. The risk of gastrointestinal illness can also be reduced by maintaining personal hygiene practices and cleaning cookware.”
In fact, the idea that every case reported from backpackers comes from poor hygiene instead of streams is not just hard to believe—it’s unsupported by the science (excellent summaries of the many papers that show this can be found here). “We have data that there is a risk from backcountry water,” says Yoder. “There certainly are waterborne disease outbreaks—more than twenty that have been reported to CDC—where consumption of water in the backcountry has been linked to illness.”
The simple fact is, if you drink untreated water, you’re taking on a non-negligible amount of risk. The good news is that risk—however large or small it may be—can be mitigated. And no, not just with fancy filters.
Linck is quick to chastise the outdoors industry for “claiming the average hiker or camper needs a $99.95 microfilter pump to avoid illness and death.” Maybe expensive devices are talked up by their makers (are we really blaming companies for wanting to sell their products?), but luckily, the people that study waterborne diseases and pretty much everyone who has had any kind of training in wilderness survival will tell you that decent water treatment can be done a myriad of ways, many of which are dirt cheap. “Boiling water is one of the most effective ways to inactivate parasites, viruses, and bacteria,” Yoder notes. Or you could get chlorine dioxide tablets that treat a liter of water for about $0.50 each.
Ultimately, the choice of what you drink is yours and yours alone. The point of this isn’t to shame you if you choose to drink untreated stream water—it’s to provide you with what is known from scientific research, rather than hyperbolic rhetoric. Now that you have the information, you can decide to trust in your ability to pick safe water sources, or to take precautions.
As for me, I’ll go with the instincts of the guy who studies these things. Yoder has analyzed the data, and his choice is simple: “I think that having that extra level of protection and making your water safer is worth it, because you’re trading that for more enjoyment out of the backcountry and not having to experience an illness that, at least temporarily, is pretty debilitating.”
It seems like the overwhelming majority of Slate’s twitter followers agree. (Ah, the ratio…)
— Slate (@Slate) February 1, 2018
UPDATED 2/8/18 to add additional papers I was alerted to after publication.