In my blog relating the Norwegian experiences of fatal casualties in recreational marine incidents, Investigators highlighted an association between drowning and ‘inebriation’. On average, across more than 70 studies, alcohol was involved in 28% of fatal and 35% of non fatal leisure drownings.
Recreational boating and alcohol use are culturally linked in most of the world. Casualty statistics from many organisations suggest that alcohol use is factor in recreational drowning.While you don’t have to be an expert to suggest why inebriation might predispose people to drowning in leisure settings formal research studies into this linkage produce widely varying results. There’s enough data to suggest that mixing alcohol and water (except in a glass) is a very bad idea: at least in certain circumstances; but not enough to be precise about exactly why and how.
A recent systematic review 1 (pre-print here) which included 48 studies of incidents where alcohol was involved in drowning, gave involvement rates of less than 5% to more than 70%, with an average of about one third of incidents (specifically 4.46% – 72.22% mean, 28.10%, SD 17.03%, 95%CI [3.15%, 53.06%]. The studies reviewed looked at all incidents involving alcohol from a population of all deaths in leisure boating incidents, in a given time period, which is to say they were measures of ‘prevalence’.
Human science and behaviour studies show that consistent clear messaging is important for behaviour change. For this reason, trying to understand the reasons behind the apparent variability in the effect of alcohol on fatal leisure marine accidents is important because it will allow better guidance to be formulated by training providers, associations and others.
The authors of this systematic review recognise that the studies they review were mostly less good quality that we would hope for, and in many cases had sources of implicit bias. The astonishing variety in methods used to determine the effect of alcohol on drowning in the prevalence must be a significant contributor to the variability in the reported results. Of the 48 prevalence studies only 26 used blood alcohol content as the decision variable. But even within these there was a lack of coherence in decision criteria for an alcohol related incident.
It is possible to cherry pick findings from individual studies in this group, or small groups of studies – but this is a risky strategy prone to confirming bias in the person doing the picking. This review retrieved 9049 possibly relevant studies from academic search engines. As is usual in a review of this type these pass through a filtering process. This review specifically excluded non accidental drownings, drownings occurring during effective or attempted rescues, studies in languages other than English, and studies that hadn’t been peer reviewed.
Systematic reviews are massively time consuming andcan only do the magic of revealing the state of a field where all studies are reasonably compared. In this study it might have beenn better and much more revealing if the filtering had been more aggressive. For example it feels unreasonable to try to mix findings from prevalence of drowning and near drowning where the decision criteria were quantitative with those which were qualitative.
- Hamilton, K., Keech, J.J., Peden, A.E. and Hagger, M.S. (2018), Alcohol use, aquatic injury, and unintentional drowning: A systematic literature review. Drug Alcohol Rev., 37: 752-773. https://doi.org/10.1111/dar.12817 ↩︎