By Phil La Duke
Last week’s post that asked why Heinrich’s Pyramid was so popular across all industry segments despite being largely discredited by many in the safety industry angered up some of your blood. I don’t mind, if I didn’t get some of you cantankerous old coots’ blood moving some of you would be declared legally dead. I’m not complaining; anytime anyone challenges any of the cherished charms and totems of BBS one is likely to get blowback. So great is the backlash from the zealous and paranoid torch-and-pitchfork crowd that I know of at least three safety journalists who won’t touch the subject. But in the froth and fury to spew forth on on-line threads one persistent argument kept coming up: are injuries causes by multiple causes or by a single root cause. The answer is “yes”.
Before joining the glamorous and sexy world of safety I spent half my career in performance improvement—both human performance improvement and process improvement—and in the course of my duties I taught problem solving. Problems, you see, can’t be neatly wrapped up in one neat little box.
Categories of Problems
I don’t see problems (any difficult situation to be settled or resolved, a question for discussion or solution, or a discrepancy between fact and observation) as any different from injury causes. Both problems and injuries are unexpected outcomes of a process and both require the organization to find out what happened and why. Problems come in four categories: Broad, specific, decision, and planning.
Broad problems are difficult to get your arms around because typically they have multiple causes and effects, grew over time, and have visible, known causes. Think of problems like world hunger; the causes are known and visible, but they are just too enormous to easily fix. Broad problems are typically the “system errors” that so many safety professionals argue are the causes for injuries. In many cases they are right, but not all injuries are caused by broad problems. Ergonomic injuries are good examples of broad problems that should be attacked using tools like Situation Analysis, fish bone diagrams, etc.
Specific problems pertain to a specific object and a specific defect. In the case of safety the specific object could either be the person injured or the means by which the injury was caused while the specific defect is the kind of injury that was caused. Specific problems have a sudden occurrence—things are going along just fine until something happens and someone gets hurt. In these kinds of problems/injuries there is generally a single root cause and the cause is typically unknown. Slip trip and falls are good examples of specific problems that cause injuries.
Decision problems are those issues that arise because of poor decision making practices. In safety decision analysis should be more widely used to generate an understanding of why people make poor decisions that end in injury and to teach workers to make better decisions. Failure to lockout or to tie off while working at heights are good examples of injuries resulting from poorly solved decision problems.
Planning problems are those situations that are so complex that a failure to plan introduces process variation and risk and too often results in serious injuries. A good example of injuries caused by planning problems are those cases of workers injured doing nonstandard work. The lack of a robust plan often results in deadly improvisations.
Structure of Problems
Understanding the category of problem is only half the battle. Next we need to understand the structure of the problem. Problems can be any of many structures but the most common are:
- Problems with a gradual structure begin with performance at the desired state (or at a minimum within the process control limits) and gradually deteriorate, or drift, away from the standard until a failure threshold has been reached and a failure (injury or near miss) happens. Sydney Dekker explores this phenomenon in the book Drift Into Failure. Essentially we let things get out of hand until failure is all but a matter of time. Think of workplaces where little hazards abound and where any one of these hazards taken on its own, is no big deal, but when working with other hazards can cause a chain reaction of deadly events. Consider, for example, the factory fires where emergency equipment is in disrepair, the alarms aren’t working, emergency exits are blocked, and escape doors are locked. None of these things in and of themselves will injure or kill a worker, but each makes it more likely that should a fire breakout lives will be lost.
- A sudden structure of a problem manifests as everything operating at the desired state until something sudden and unexpected plunges the operation into failure. Think of a flat tire. You are driving along just fine, hit a pot hole and blowout your tire, one minute your cruising up the boulevard and the next you are on the side of the road cursing your teenage son for making off with the tire jack.
- Start-Up. Whenever we start a new operation we generally have a period where we struggle to get to, and remain at, the desired performance standard. This is not a license for us to hurt workers, but it should be an incentive to better protect workers by focusing on mitigating severity in addition to trying to predict start up issues. Too many companies misunderstand start up issues and will dismiss any concerns as a need for “work hardening.” Work hardening is the practice of having employees build muscles and generally get used to back breaking work that causes excruciating pain and usually ends in ergonomic injuries.
- Problems with a recurring structure should be of paramount interest to safety practitioners because, most often, a recurring structure is indicative of a misdiagnosed cause. When you treat the symptoms instead of the cause you frequently see an initial improvement only to see the problem gradually return, sometimes with deadly results.
- Some problems don’t seem like problems at all. Problems with a positive structure are those situations where the outcome is actually better than expected. But because the situation is better than expected it must be researched so that the positive results can be replicated. Think in terms of a major cause of injuries suddenly falling dramatically. Unless you know WHY you saw the improvement you can never be sure that you won’t degrade back to your old ways.
Think of all the good that we could be doing instead of arguing about whether injuries are caused by systems or behaviors, multiple causes or a single root cause, or whether a hair-brained pyramid has “at least some useful parts” and concentrated on using a tool-box approach to injury reduction? As the great Peter Drucker said, “the most common source of mistakes in management decisions is the emphasis on finding the right answer rather than the right question.”