It’s been a while since I made the time for a posting, but I’ve seen a trend of companies overusing the “5 Why” technique so I just felt the need to share this. Enjoy one of my favorite songs too. If you could only hear me sing it in the shower too.
If an auditor finds a nonconformity, the process owner is assigned to a CAPA. The first step of the CAPA process is to investigate the cause or causes. Some auditors recommend using a technique called the “5 Why” process to identify the root cause. PLEASE TELL THEM TO STOP!
The purpose of investigating the causes of a nonconformity is to determine how big the problem is (i.e. – breadth) and what the underlying causes are (i.e. – depth). The 5 Why process only addresses depth. So why do auditors recommend using this tool? Because the concept is simple to understand.
Unfortunately, not every problem is deep. Other problems have more than one cause. There is no magic behind asking why five times. The principle behind this overly-used technique is to encourage process owners to look deeper into the underlying issues instead of stopping at the obvious specific causes. The intent behind implementing a mandatory “5 Why” process is admirable—but wrong.
There are other techniques. Please try using them.
Cause and Effect Diagrams, also known as “Fish Bone Diagrams” or “Ishikawa Diagrams”, are popular tools for investigating problems in the automotive and aerospace industries. This method involves systematically investigating six categories of potential causes. These six categories are often referred to as the six “M’s”: 1) manpower, 2) mother-nature, 3) machine, 4) material, 5) method, and 6) measurement. This technique is particularly effective, because it is a systematic technique that investigates breadth of problems. Is/Is Not Analysis is another tool used to help process owners narrow down the source of a problem when they really don’t know which category of potential causes matter. For example, an investigator might ask the following questions in conjunction with the Cause and Effect Diagrams: 1) “Is this problem observed with any other employees?”, 2) “Did this problem only occur during hot and humid summer months?”, 3) “Did this problem occur in cavity #4 of parts or all cavities?”, 4) “Is this problem observed in cobalt chrome parts only or titanium implants as well?”, 5) “Do we observe failed crystals for only circuit boards that are reworked by hand or for circuit boards that are built entirely by automated assembly as well?”, and 6) “Were all the defective parts inspected using the same thread gauge?”
The above questions are sophisticated questions that may require a cross-functional team to properly answer. It may be necessary to review lots of records from previous lots. It may be necessary to interview operators to gather information that is not included in the records. It might be necessary to design experiments to verify theories. Often it is necessary to use multiple techniques together. The point is that a thorough investigation is not simple.
If a finding is a simple administrative oversight, there are four common causes: 1) insufficient training, 2) insufficient resources, 3) insufficient management oversight, and 4) insufficiently detailed procedures. If multiple people believe that one of these four causes is the problem, look to other areas to see if a similar problem occurs elsewhere—this is investigation of breadth. If the problem is observed in other places, there is a systemic problem that requires asking “Why?” at least one more time. If the problem is isolated, to the area where it was found, then there is probably no need for asking “Why?” again and again.
If your company’s CAPA procedure requires that you use the “5 Why” technique, please consider rewriting the procedure to encourage process owners to “investigate the cause with an appropriate root cause investigation technique such as one of the following…”. This flexible approach requires more thorough training, resources, and management oversight. However, an overly prescriptive CAPA procedure is often ineffective at preventing recurrence of problems.
13485cert
QC is Dead 