Examples of number processing tools include:

Run charts
Control charts
Check sheets, etc.

Idea processing tools include such things as:

5 why’s
Brainstorming
Fishbone diagrams
A combination of several tools

In an investigation, the challenge is to know what hard data you need, where to get it, and how to use it.

Idea Processing Tools

These tools are ways to find root cause quicker and to generate, collect and organize facts. They help teams to think faster, wider and deeper. Done well, they lead us to consider possibilities we might miss, what data are we missing, where should we go with our investigation and who to talk to.

The FDA is looking for investigations that are thorough.

The strategy and tools we select are driven by the degree of risk our problems present. These tools help us to expand our thinking to match that risk, and to see more possibilities - and then process them.

Brainstorming

This is a common tool, often misused in many organizations. Most people are familiar with the rules of brainstorming:

All team members must contribute
No criticisms or other comments during the process
Ideas may be piggy-backed
All ideas recorded exactly as they are expressed
Rapidly generate ideas
Clarify and combine ideas afterwards

The intent is to use this tool early in the investigation to generate a list of all possible causes for a problem. It allows teams to uncover all kinds of possibilities. Admittedly some of them will be impossible. However, this process will get the team to consider possibilities they would have failed to see and then never investigate.

The Issue

Where many teams fail is after brainstorming. They don’t process the information any further or they will ‘cherry pick” ideas off the list and ignore the rest. Not only does that leave potential ideas or possibilities on the table, but discourages team members’ involvement.

Our Recommendation

Done well, leaders will organize the data into categories that can be understood, and then acted upon. We recommend using a fishbone chart. The fishbone then uses the 4M’s (or 5 or 6) to manage the data: Man, Methods, Materials, Methods, Mother nature, and Management.

Using these tools ensure that all information has been organized and is in a form that can be processed and is actionable. Now we have groups of data we can use that can guide us into an thorough investigation.

Stay tuned for next month’s session where we’ll share more helpful investigation tools!

Do you have any additional advice to share? If you have had other methods of success in conducting investigations, we’d love to hear about them.

The Last Step: Test

Assuming that each of those first five steps have been completed accurately and thoroughly, the test will bring the investigation home, so to speak. Remember how our first three steps were built on facts - just the facts. Clues and likely cause were a combination of expertise and inference. We “connected the dots” and reached logical conclusions.

The Process of Testing

The test forces us to test against the facts - to verify that our intuitions were correct - or not. We ask “if this is the likely root cause, does it explain the Is (investigate) facts and not the Is not (comparison) fact? Yes or No. Then we must justify our answer. If we determine it is yes in each of the four “W’s”, we have probably found our root cause. If we get more than one “no”, we should probably eliminate it as a likely cause.

Challenges

Not conducting a thorough test is often the reason why teams fail at this last step. At this point, teams have done a lot of hard work. You don’t want to quit until you check all your facts. You need to test all likely root causes against all facts.

The Need for Process

Also, often times teams will arrive at this point thinking that one of their likely root causes HAS to be the one. That’s not always the case. You may realize that none of your likely causes fits the facts. That doesn’t mean you failed. As is often the case, your process may highlight where you need more data, or need to go explore more. That’s the beauty of following this process. It can highlight those areas and help you do a more thorough investigation.

Do you have any additional advice to share? If you have had other methods of success in conducting investigations, we’d love to hear about them.

Which aspect of your quality department needs the most improvement?

Vote at http://www.surveymonkey.com/s/GXV8VCD

A CAPA System must be compliant, effective and efficient.

So what is required of a CAPA System to make it compliant, effective and efficient? Let’s take a look:

Compliant CAPA System

The applicable regulations and standards are dependent upon where your organization plans on selling its products or services.

United States
• 21CFR820.100 Corrective and Preventive Action

Europe
• ISO 13485 8.5.2 Corrective Action
• ISO 13485 8.5.3 Preventive Action

• MDD 93/42/EEC Annex II 3.1

Canada
• CMDR SOR/98-282 Section 57
• MDD 93/42/EEC Annex II 3.1

Japan
• JPAL Ministerial No. 169 Article 63
• JPAL Ministerial No. 169 Article 64

Once it is determined what regulations or standards are applicable, it is important to ensure those specific requirements are addressed procedurally and executed accordingly in order to be fully compliant. Being fully compliant, however, is not enough to be successful in CAPA. The regulations and standards define requirements that specify “what” to do but they do not specify “how”. That is up to you, and appropriately so. Your success with CAPA is dependent on the “how” of CAPA in order to truly be effective in improving your organization.

Effective CAPA System

In order to positively impact an organization, CAPA projects must address the most important issues and must do so in a way that desired results are obtained. How is this accomplished?
Various sub-system elements of a Quality System (QS), business and manufacturing processes produce non-conformances. All relevant QS sub-systems must be designed to identify and correct their non-conformances. These QS sub-systems are also expected to operate within a state of “control.” These QS sub-systems are unique and separate entities from the CAPA System, which is also a QS sub-system. There is, however, a hierarchical relationship between these QS sub-systems. QS sub-systems that produce non-conformances must have the ability to analyze and appropriately escalate non-conformances that are identified as either “high risk” or are the result of “out of control” process conditions.

What does “appropriately escalate” mean? Wouldn’t all identified non-conformances be escalated into CAPA? The answer is no. Doing so could cause the organization to become overwhelmed with the trivial many without properly addressing the significant few. It is important to understand that each QS sub-system non-conformance will be corrected. However, only those identified as “high risk” or are the result of “out of control” process conditions should be required to be escalated and formally addressed through a CAPA System. This system is designed to uncover the “root cause” of the issue and take corrective action in order to prevent the recurrence of the issue.

Upon escalation of a non-conformance issue to CAPA, effectively addressing the problem is of paramount importance. The “infrastructure” of the process used to execute a CAPA project, the “how”, will ultimately determine if the organizations improvement efforts are successful. Consider the various phases an effective CAPA process might go through in order for that work to be executed in an orderly manner:

Initiation – Phase where the identified issue is scoped and documented and becomes a CAPA Project.
Investigation – Phase where the issue is investigated and the root cause is determined.
Action Plan – Phase where the results of the investigation are turned into a plan of action for correction and/or prevention.
Implementation – Phase where the execution of the action plan occurs.
Verification of Effectiveness – Phase where the corrective or preventive actions are verified to be effective.
Closure – Phase where the documented activity necessary to gain closure of the CAPA occurs.

These are all uniquely different phases in terms of the type of work conducted in each phase. It stands to reason that keeping a separate focus on these distinct phases of a CAPA project would enable the work output to be successful within each individual phase and for the CAPA project in aggregate.

It is clear that the instructions for executing CAPA project work should be carefully built. The separate work components should be individually defined, yet housed within a system of documents that work together to accomplish the effective execution of CAPA projects.
A CAPA System, however, consists of more than the project phases defined above. A properly designed CAPA System should consist of the following elements:

CAPA System Management – The requirements to manage the CAPA System as a business process.
CAPA Projects Review – The requirements for review and management of CAPA projects.
CAPA System Review – The process of analyzing the CAPA System Metrics and escalation to Management Review.
CAPA Escalation Decision – The process of analyzing the severity of a non-conformance issue and documenting the outcome. This is the linkage between a QS sub-system that produces non-conformances and the CAPA System.

CAPA Process – The requirements to execute and document the six phases of a CAPA project, as shown above.
NC and CAPA Trending – The requirements to analyze and trend nonconformance quality data and CAPA project effectiveness.

The instructions for executing CAPA project work and the various additional elements of a CAPA System are unique and separate. The separate system components should be individually defined within a system of documents that work together to accomplish the effective execution of all CAPA activity within the organization. In their entirety, these documents then comprise the organization’s CAPA System.

Efficient CAPA System

A properly constructed CAPA System should consist of streamlined work processes that provide an organization with some significant benefits:

• In using appropriate escalation rules, the organization is less likely to become overwhelmed with the trivial many and can properly address the significant few non-conformance issues that arise from the QS sub-systems.
• By enabling a separate focus on the execution of the distinct phases of a CAPA project, the work output is capable of being efficiently completed within each individual phase resulting in a more successful CAPA project in aggregate.
• By enabling a distinct phases of a CAPA project work, with different required skill sets required for execution, CAPA project work is capable of being assigned by phase and in accordance with individual skills.
• Appropriate trending process will identify the “common cause” variation for non-conformance reduction improvement initiatives that can provide significant return in the investment of a CAPA project.

Key Advice

Be specific and and make sure the cause statement consists of two clear and separate parts- Cause and Effect. Because of this condition - because “this” happened, we got this effect - or result. A good cause statement defines WHY something happened as well as WHAT happened.

Example: Imagine you are investigating that customer XYZ got the wrong parts. A typical cause statement we often see is “the operator selected the wrong label.” The likely answer here is to do retraining. However, consider this as a cause statement: “Orders are not kept separate. Parts and shipping labels were mixed together in the shipping area, causing the wrong parts to be shipped to customer XYZ. Looks more like a process issue doesn’t it?

It’s interesting to see how we arrived at a totally different conclusion with the better cause statement.

Do you have any additional advice to share? Please share in the comments below. If you have had other methods of success in conducting investigations, we’d love to hear about them.

What percentage of your quality system investigations are repeats?

Vote at http://www.surveymonkey.com/s/89X2LM3

This month, we are going to connect the dots by actually listening to what our data is telling us. That is going to lead us to the clues that will uncover the root cause of our problem. Turns out, it’s really not any different from what your favorite investigator does on TV.

They are asking what’s different, changed or unique about our comparison data vs. our investigation facts.

Specifically, we are going to apply our four M’s:

Manpower
Material
Methods
Machinery

Something has likely changed in one of those ares that gave us our unfavorable results. As investigators we need to explore each of these in the areas of our 4W’s that we spoke of: the What, Where, When and Weight.

We need to ask what has changed in the area of manpower in terms of the “what”. Did we add new workers, did someone miss critical training because they were on vacation? We ask the same type of questions with each of the M’s within each of the W’s. Following this discipline will help us to uncover clues.

The Biggest Challenge for Teams

Oftentimes, teams want to give up after they find one or two clues. Or sometimes they will quit when they decide they found “their” likely cause. A team leader has to be persistent.

The biggest mistake that teams make is that they get confused about how they document their clues. This is critical. The information you record must always be true of the “IS” data - your investigation facts. Not of the “IS NOT” data - your comparison data. To ignore this will confuse your investigation, frustrate your team, and often keep you from finding the root cause.

Stay tuned for next month’s session, we’re getting closer to finding our true root cause!

Last month, we talked about the importance of the problem statement and investigation steps. Just to review - the problem statement is about identifying the object and defect, and the investigation step is about gathering facts about the problem.

The next step in the investigation process is the comparison step. You might think of it as the companion to the investigation step. These two parts of the process are often referred to as IS (investigation) and IS NOT (comparison). We start the search by looking for things that don’t share the problem.

Again, we use the 4 W’s to organize our search:

What
Where
When
Weight

Sounds easy enough, but like the other steps, we often see some common pitfalls here, too.

The most common problem is using comparisons that share the problem. When you identify things that have the same problem or defect you are investigating, it cannot serve as a comparison. Using them will complicate, or derail, your investigation.

For example: Let’s say the grass in your front yard is dying and you want to find out why. You might compare it to your back yard provided the grass was healthy there. Or your neighbors yard if their grass was ok. But if the grass is dieing in either of those cases, you can’t use it because it shares the problem. You need to find comparisons that will provide differences. Without those, you won’t be able to find any clues later.

Remember, good comparisons are essential to finding good clues. Poor or inappropriate comparisons will mean that it is very unlikely you’ll find the true root cause.

Do you have any additional advice to share? Please post on our Blog or LinkedIn Group. If you have had other methods of success in conducting investigations, we’d love to hear about them.

Yours for training that makes a difference,

Herb Miller
Product Development and Training Manager
PathWise, Inc.

This month, we’ll discuss the importance of fact gathering and common issues that get in the way of good investigations. Specifically:

I see many investigators who don’t want to appear like they mistrust their witnesses. So they don’t challenge the meaning of words, and they often accept vague or general statements. They attach a meaning to the words and move on.

My advice: Don’t be afraid to ask the “stupid questions.” For example, a witness may say “The truck was late.” I can either write that down and move on or I can ask “How late” or “What do you mean by late?” or “Whose truck?” By challenging, I get more concrete information. Remember, challenging helps us to be thorough.
Also, make sure you’ve talked to all the witnesses. Gather all information using the necessary quality tools: checksheets, run charts, process maps, etc. I see too many investigations that are shallow because the investigator quits too early and in doing so, never identifies the root cause.

Fact gathering means all the facts- leaving no stone unturned.

Have you ever experienced this scenario while training a group? Please share your experience with us.

In your opinion, what is the most common tool your investigators use when conducting a root cause analysis?

Vote at http://www.surveymonkey.com/s/R6LCRK9