C4.3 Prevention and detection

ASQ FMEA [8]

The third component requires that every control be tagged as prevention, acting on the cause of a failure mode, or detection, acting on the mode once it has begun, and that both the occurrence and the detection ratings of the mode be revisited after any action with the previous values kept. The source is the failure-mode tradition [8], which holds the two kinds in two separate columns and refuses to let one score stand for both. The reason is practical: a process with excellent detection and no prevention catches every failure after the cost has been incurred, and a process with excellent prevention and no detection does not know when the prevention has stopped working.

This reconciles the financial matrix, which has one control column and one effectiveness score, with the quality tradition, which has two columns and re-rates after action. It also gives barrier analysis in C5.3 its two questions in advance: when an event occurs, the investigation asks which prevention control should have stopped the cause and which detection control should have caught the mode, and it can only ask them if the controls were tagged before the event. The rule that the previous ratings are kept is what lets a register show whether an action worked, rather than merely that it was taken.