We Can Predict Injuries, Just Not With 100% Accuracy


By Phil La Duke

The first snow of any significance of the 2015 winter hit the Detroit yesterday dumping 5 or so inches of snow on an otherwise grey and rainy November day, marking the beginning of yet another Ingmar-Bergmanesque bleak and depressing Michigan winter.  There are those here who love winter.  We call them idiots and we don’t let them use a table knife and fork unattended.  But there was something enlightening about this snow storm that I think is a nice analogy for what’s going on in the world of Safety.

This storm was predicted as most weather is.  Now you can joke about a weatherman being the only person who can get things wrong half the time and keep his job, but with all the Crank Coxes out there, I wonder if that’s not equally, or even more true of safety practitioners.  What’s more, I just read a wonderful book that had nothing to do with Safety (and believe me, I scour everything I read for some kernel of wisdom that can be used to make the workplace safer), Guns, Germs, and Steel.  In this book, the author points out the importances of agriculture in the development of dominate civilizations, and cites a study that showed that weather forecasts are surprisingly accurate—for example, in cases where the meteorologist predicted a 70% chance of rain, that it did in fact rain in about 70% of those cases.

Weather prediction is an apt parallel to safety.  When the weatherman (or lady) is wrong many people openly deride him or her as “not being able to predict anything”.Just as there are many who believe that some injuries are just an act of God.  But in both cases the believe that something is IMPOSSIBLE simply because one failed to do it is asinine; everything was impossible until someone figured out how to do it.

And this brings me, fairly verbosely, to my point.  Zero Injury goals have been vilified by many because they aren’t possible and setting that as our goal demoralizes the organization.  To some extent I have made that argument, saying that to achieve zero injuries we have to be able to predict every conceivable hazard and possible injury.  But is this so different from predicting the weather?  One of the most difficult thing about predicting the weather is that there are so many variables acting in an extremely complex system that one could throw up one’s hands and say that its essentially just a guess. Except it’s not a guess, it’s science.

Science Versus Luck

This article should not be seen as an endorsement of zero injury goals, I have enough whack jobs in my life without stirring up the fervor of the Zero Harm zealots.  The point is, as many have made it more artfully before me, that many Zero Injury workplaces are just the result of dumb luck.  It’s like me predicting that it will snow again tomorrow (by the way the forecast doesn’t call for snow) base on nothing but the pain in my knee.  If it snows my prediction was accurate.  Should I be then given a job as a weather forecaster? What if I made correct predictions for a month?  Unless we know WHY we are correct our predictions can’t be trusted.  And we cannot ever prevent hazards that we didn’t anticipate, all we can do is argue amongst ourselves who bears the blame for failing to predict  the injuries.

Why Are Predictions Are So Poor

For starters, we shouldn’t be predicting injuries, we should be anticipating the possibility of an injury.  People who accurately predict injuries should be jailed (Sorry Mrs. Kelsey, we knew this would happen but we let it kill your husband anyway, what can I tell you? He’ll be missed.”) But assuming for the sake of this conversation that we use the word “predict” to anticipate the consequences of one interacting with a hazard, why then do they consistently fail us?

  1. We don’t try.  I’ve been in enough workplaces where the frontline supervisors believe (and are indifferent toward) the inevitability of injuries.  People get hurt, it just part of the job.  Well if you believe that, then there is absolutely no point in trying to predict how someone might be injured and do anything to mitigate the risk.
  2. We believe an accurate prediction is impossible.  “Yeah right, what are the odds of that happening?” Too many of us believe that it is impossible to predict that an injury will happen.  People are understandably skeptical. Unlike the weather, however, we don’t need to know anywhere near as many variables as a meteorologist.  We aren’t trying to predict the precise moment and injury will occur, or the specific type of injury, heck we don’t even need to predict how severe an injury will be, only that an injury is likely to occur unless there is some sort of intervention.
  3. We look at too many variables. Safety has become alchemy; a blend of science, superstition, snake oil, and guess-work.  We don’t need a good share of the information and tools that we fiddle with to anticipate that a workplace rife with hazards, lax enforcement of safety rules and a culture with a high risk tolerance and a knowing or unknowing contempt for worker safety to prognosticate a high risk of injuries.

Mark Twain reputedly said that everyone talks about the weather but nobody does anything about it, sometimes I feel the way about safety, we all talk about it, but we do so little of substance about it.