Labeling complaint investigation tool – Internal auditing

Can internal auditing be used as a tool specifically designed for conducting a labeling complaint investigation?

What is a labeling complaint investigation?

A complaint investigation is required when a customer “alleges deficiencies related to the identity, quality, durability, reliability, safety, effectiveness, or performance of a device after it is released for distribution (21 CFR 820.3b).” In the case of a “labeling complaint,” the customer is alleging that there is a deficiency with the label content, the physical label material itself, or both. In the fictional Halloween story I read about the “Label Monster,” there were five different deficiencies identified in a single unit. The customer refused to accept the entire lot and returned it to the contract manufacturer. “Natalie,” the internal auditor and consultant, was responsible for helping the quality manager (i.e., “Lorelei”) and the operations manager (i.e., “Valerie”) identify the root cause for each of the five label deficiencies. The investigation process involved: 1) interviewing personnel; 2) observing the process, equipment, and environment; and 3) inspecting the returned product.

What is different from a routine internal audit and a complaint investigation?

Routine internal audits are scheduled as part of your annual internal audit schedule. You are the audit program manager, and usually you review your draft annual internal audit schedule with the other members of top management during management review meetings. Since a labeling complaint is not something you can plan or schedule, you would have to amend your internal audit schedule to add a special, “for cause” internal audit of the labeling process to perform a labeling complaint investigation as part of an internal audit. Internal audit scheduling also involves assignment of the lead auditor and team members (if applicable). In the case of labeling at the Red Rum Company, the quality manager was also the audit program manager and responsible for review and approval of the label change control process, inspection of labels prior to use, and final release of each lot. Therefore, an independent lead auditor from outside the company is needed (e.g., a Medical Device Academy auditor).

In addition to differences in the planning of the internal audit, a labeling complaint investigation also reduces the effectiveness of certain parts of the process approach to auditing. Specifically, “Natalie” said, “Reviewing procedures seems like a waste of time.” She was right. Whether the labeling procedure is adequate or not, it is important to evaluate how the process was performed–regardless of what the procedure instructs personnel to do or how well people are trained. Changing procedures and improving training could be a potential corrective action, but correcting procedures and training will not correct the deficiencies observed. We also assumed that the work order content matched the customer purchase order requirements–otherwise the label content would have matched the work order content (i.e., lot numbers should match). Therefore, internal audits performed as part of a complaint investigation should focus on the following elements of the process approach to auditing:

1. Process description – The team described the process to the auditor, but the auditor doesn’t truly understand the process until they “go and see at the source (I.e., genchi genbutsu).” Time in the conference room is usually wasted in a labeling complaint investigation.
2. Inputs – As stated above, the customer’s purchase order triggers the creation of a work order for the production lot. If the labeling does not match the batch record, then the problem is not related to process inputs.
3. Outputs – Batch record(s) and physical labels on returned product require detailed inspection and review.
4. What? – What equipment, materials, and software are involved in the process must be carefully reviewed? including maintenance, measurement, and process validation. In the case of this labeling complaint investigation, raw material mix-ups were identified as part of the root cause.
5. Who? – Who is responsible for performing each task in the process?
6. How done? – Instead of focusing on procedures and forms, the auditor should review the methods used for the potential for mix-ups, contamination, and set-up of the equipment.
7. Metrics – Metrics are often overlooked by inexperienced auditors, but the team was able to identify why approximately 50% of the labels were using the wrong template because the printers were incapable of printing 10,000 labels in a shift. Reviewing process metrics, measurement of product, and quality objectives is essential to value-added auditing.

Why didn’t a pFMEA and process validation prevent the labeling errors in our story from happening?

In response to our live-streaming video that was published on Halloween, we received a comment with two questions (copied below).

The story I read was fictional, but in the scenario I was imagining, I envisioned a manufacturing facility that was running extremely well on the two first shift production lines, but nonconformities occurred when production started on second shift for a new product. I envisioned a team consisting of the quality manager (i.e., “Lorelei), operations manager (i.e., “Valerie”), and a possibly a manufacturing engineer and/or quality engineer working together to proactively review the new production line for any potential problems. They started with their pFMEA for the slower first-shift production lines, and modified that pFMEA for the new production line. Unfortunately, the cycle time of the second-shift production line outpaced the capacity of a single labeling machine. Therefore, the team would have three options for addressing this challenge:

1. split production between two lines on second shift for all of the production process or at least label printing
2. purchase a new thermal transfer printer that is capable of two-second print time, instead of three-second print time
3. batch printing of labels for second shift

All three of the above options are viable, but I imagine most teams would choose the third option. Unfortunately, if your process risk analysis for labeling process controls is designed for an on-demand labeling process, you could easily underestimate the potential complexities of batch building. Even if you evaluate the proposed process controls by conducting a pilot run, production would need to exceed the cycle time of the printers to expose the potential hazardous situation.

The second question that was asked in the YouTube comment would not be relevant, because most of the labeling change control process and the labeling process itself are manual (i.e., no process validation is required, because product is 100% verified). You would still validate that the printer correctly prints the label template in your labeling software, but process validation would not be performed for the manual creation of a label template, manual review of the new label, or manual printing and application of batch-printed labels. Usually, labeling process controls for batch printing of labels focus on initial verification of the label content, verifying the number of labels printed, and reconciling the number of labels printed with the number of labels used. Verification of label content was performed, but only for one label printer–not two printers. There were no problems associated with the number of labels printed or the reconciliation of labels printed with the number used.

What is an “IS/IS NOT” analysis?

In last week’s live-streaming video, we mentioned that an IS/IS NOT analysis would be used as one of the techniques for the labeling complaint investigation. This refers to the process of comparing the areas where a problem is found and where a problem is not found. Problems were not found on product made during the first-shift production, but problems were found on product made during the second-shift production. Additionally, finding problems with half of the product labeling is also critical to identifying the root cause of a failure. Therefore, this tool is extremely valuable when performing a complaint investigation.

What is a “Fish Bone Diagram?”

A Fish Bone Diagram is another popular root cause analysis tool, because it is a six-part systematic method for identifying a root cause when you don’t know what the cause is. Other names for this tool include: Ishikawa Diagram and Cause and Effect Diagram. The six parts of the method are sometimes called the 6M’s, because you can select words that begin in the letter “M”:

1. Materials
2. Methods (i.e., Procedures)
3. Measurement (i.e., Calibrated Equipment or Method of Measurement)
4. Mother Nature (i.e., Environment)
5. Men (i.e., People)
6. Machine (i.e., Equipment)

It was a fictional story

The story read in our live-streaming video was fictional. You rarely have returned product to investigate, you rarely have a supportive team that helps you identify a root cause, and the process rarely as obvious as the story implies. The key to success is hard work and a rigorous investigation that might take more than a week. Rob Packard should not have agreed to a penalty if the CAPA plan was written by the end of the day.