Flush drowning research - correlation is not causation

 My commentary on recent research: whitewater paddlers and flush drownings. Published in Paddling Magazine issue 62 (link at bottom to full article).

I’ve had my share of swims, as I’m sure you have too. Especially in the early days when my kayaking universe was expanding outwards and I was compelled (as perhaps you were too) to climb the difficulty ladder to somehow prove my abilities (prove to whom?). As a raft guide in the east, flipping and swimming is part of the game, but those tend to be pretty predictable and ‘controlled’, if such a thing can be said. I’ve had long swims, cold swims, stuck in holes and bashing rocks swims. One where I was left stranded clinging to a rock face and had to await my buddy’s lowered rope kind of swims. One that left a bruise on my thigh that took 5 months to go away kind of swims. But I’ve never had a desperate swim.

New research into the causes of whitewater drowning is attempting to understand flush drownings. In their research article “Flush drownings as a cause of whitewater deaths”, two Colorado medical doctors looked at whitewater fatality data from the American Whitewater Association and attempted to parse out the variables that may be significant (link here). Flush drownings could be construed as mysterious for these do not seem to have a direct cause. Us whitewater folks would colloquially call a flush drowning a situation where a long swim results in death, with no apparent complications like head impact or strainers. The research authors focused on water temperature and geography, as their data set compared western river fatalities to those in the east. The authors concluded cold water plays a role in the likelihood of flush drowning.

An aside: as an academic researcher myself, there are rigorous protocols in examining data and making claims of absolute truth. For our niche whitewater activity, we have largely escaped academic interest, and as such any new findings are going to seem pretty self evident. The research literature, however, has to build itself one block at a time, taking pains to prove what is already self evident to everyone else.

So, the research authors conclude there is a correlation between cold water and flush drownings, primarily due to the data showing far fewer flush drownings in the east. Perhaps this is because of water temperature (which is warmer in the east), but (I would argue) this is more likely due to the continuous nature of western rivers. Any swim is a long swim in the west. Every second in the water increases the chance of drowning. Longer swims equals more exposure to the risk of drowning, regardless of temperature. By contrast, the eastern rivers data showed a far greater proportion of entrapment fatalities. By these authors’ logic then, entrapment fatalities are correlated with warm water? If one fact is proposed, its opposite has to be true. The primary research trap drilled into Ph.D. students is that ‘correlation is not causation’. Cold water is correlated to flush drowning; warm water is correlated with entrapment. That says very little about what caused what.

Desperate swims. In my guide role I’ve had my share of chasing down long swims; the Tutshi near Whitehorse is one notable memory, where the paddler swam for 2 kilometres before finally getting him into an eddy, exhausted, puking up water – desperate. Another was the Yampa in Colorado, early summer flood levels and swimmers in the water for a mile. But I’ve also hauled up desperate swimmers who were in the water for just seconds. One was a kid, no pfd, ten years ago at a favourite play spot. He was playing along the shore with his buddies and fell in. After 15 long seconds he resurfaced, where I hauled him onto my bow and got him to shore, near drowned, dazed and terrified.

Flush drowning is drowning. We don’t need an overriding reason to drown i.e. we don’t need to be hit on the head or stuffed under a rock. We only need to suck water into our lungs. It can happen to any of us, which is why we wear pfd’s but that is no guarantee. Every second spent swimming in whitewater ups the odds of sucking in water, no doubt complicated by cold water slowing us down, high gradient and water speed keeping us from getting to shore, a crappy old pfd that does not really float anymore – all correlated, but the cause is sucking in water. We need to wrap our heads around the opposite of our collective experience: swimming in whitewater exposes us to drowning, even though it does not happen very often. 

link to original publication:

https://dashboard.mazsystems.com/webreader/69048?page=60