Throwing CAPAs at the Wall
By DR. JAMIE WEISS, Senior Consultant, Kepner-Tregoe Inc.
At Kepner-Tregoe, we do a fair amount of problem-solving facilitation. A client will call us up and say, “We’ve got this problem, and we’ve been working on it for months, and we still don’t have a consensus on the cause. It’s gotten so bad that you can’t even ask a question without getting into a political fight. Could you send someone who can take over the analysis and act as an unbiased analyst, with a cold clear eye?”
And, if the issue matches our process, we often say, “Yes.”
We also do a fair amount of diagnostic work to help clients assess how their problem-solving process is working, where it may be having trouble, and where it may need to be revised. In the course of doing this, we get to read a lot of investigation reports. I mean, a lot of investigation reports.
And we have noticed a short-cut to tell if they have the cause nailed. It’s frighteningly simple, actually. All you have to do is count the corrective and preventive actions (CAPAs).
If the report lists a dozen separate corrective and preventive actions, you can be pretty sure that they don’t know the cause, and are “throwing spaghetti on the wall to see what sticks.” If you look into the CAPAs, you can often see that they have two or three causes they are trying to cover, and not enough evidence to nail down which one it is.
Some of these CAPAs may not actually be corrective or preventive actions. They may be data-gathering steps. They may be actions to determine the cause, not to make it go away. And some may be generic “nice to do” actions, which probably won’t hurt, but might not help. Some folks add these in to make it apparent that they are taking the issue seriously and investing significant time and effort in its pursuit. But in the end, they’re clutter.
On the other hand, if they only have one or two CAPAs, they probably have the right cause, or at least have enough confidence to focus on one possibility. And if there are only one or two corrective actions, it is easy to track the logic back to the data: If this is the fix, how does it address that as the cause? And if that is the cause, how does it explain these as the symptoms? It’s a clear logic chain.
In between two CAPAs and 12 CAPAs, you need to dig into the data to see what is there. One trick we often recommend is to pull three sections from the report — the symptoms, the cause and the CAPAs — and put them together on a single page to see if they line up. If you can’t figure out how the cause causes the symptoms, that’s a problem. If it’s not clear how the CAPAs make the cause stop causing the symptoms, that’s a problem.
Of course, when you get down to it, each problem and each investigation is unique; if it weren’t, you would know the cause because you have seen exactly these symptoms before. So in the final analysis, you have to read and analyze each one on its own. But as a starting point, as a short-cut, just count your CAPAs.