CAPA and Complaints: Ascertaining Root Cause

By Bob Mehta

This past August, FDA issued a warning letter to Soleetech Corp., a Taipei, Taiwan-based manufacturer of airway connectors. The agency was not impressed with this organization’s level of compliance. FDA’s issuance of a warning letter is not an earth-shattering event, but two violations laid out in the Soleetech warning letter—regarding corrective and preventive actions (CAPA) and complaints—really stood out:

and

Let’s examine what Soleetech did to warrant the letter and dive into one of the underlying concepts associated with effective CAPA and complaint management: ascertaining root cause.

Soleetech’s Mistake

Consider these two excerpts from the warning letter:

and

Any medical device establishment located inside or outside the United States has placed itself in a precarious position when it informs FDA that it has no intention of complying with the quality system regulation (QSR). Making such bold statements will most assuredly result in the removal of this establishment’s product from the U.S. marketplace.

CAPA & Complaints: Root Cause

Having compliant procedures for CAPA and complaint management is a must for any firm in the medical device industry, regardless of marketplace. The QSR, Ministerial Ordinance 169 in Japan, and EN ISO 13485:2012 in Europe all require device manufacturers to effectively manage CAPA and complaints. Most medical device manufacturers manage to establish reasonable procedures and are able to resolve issues placed into their CAPA systems, including customer complaints.

However, identifying root cause continues to be challenging.

According to BRC Global Standards:

“Root cause analysis is a problem solving process for conducting an investigation into an identified incident, problem, concern[,] or nonconformity. Root cause analysis is a completely separate process to incident management and immediate corrective action, although they are often completed in close proximity.”

Additionally, the individual(s) tasked with ascertaining the underlying root cause must look beyond the obvious and make a serious attempt to pinpoint root cause. The good news is that there are tools available for investigators to facilitate their quest for ascertaining root cause.

Understanding Root Cause

To determine root cause, it’s essential to first understand what the term means. The best way to explain root cause analysis is to use the example of a weed. Weeds can be difficult to remove once they start to grow and spread. On the surface, the weed is easy to see; however, the underlying cause of the weed, its root, lies below the surface and is not so obvious.

Conversely, the word root in root-cause analysis refers to all underlying causes and not just one. That is why it is imperative to be open-minded and objective when performing root-cause analysis. Beginning an analysis with a preconceived idea of what appears to be an obvious root cause could result in the incorrect root cause being identified and the wrong correction being implemented.

Tools for Ascertaining Root Cause

There are a plethora of tools available for assisting in the identification of root cause. The underlying goal is to achieve an accurate root cause, so the appropriate corrective actions can be pursued to prevent recurrence. If the incorrect root cause is identified, it is inevitable that the incorrect solution will be implemented. In the medical device industry such errors can compromise device safety and efficacy. Some of the tools available for quality professionals to employ in ascertaining root cause include the following:

The Five Whys Model. The five whys model is a root-cause analysis tool originally created by Japanese inventor and industrialist Sakichi Toyoda. The effectiveness of the model became apparent in the Japanese automotive market in the 1960s and ‘70s. Toyota became a big proponent of the five whys model, which ultimately became a critical component of the company’s problem-solving training and the foundation for its scientific approach to performing root-cause analysis. Today, the five whys model is being effectively employed in the medical device industry, with evidence of the model’s use within Kaizen, lean manufacturing, and Six Sigma.

Fishbone Diagram. The fishbone diagram, made famous by Kaoru Ishikawa, is similar to the five whys model in that it captures the cause-and-effect relationship of problems. The fishbone diagram is prevalently used as a tool to identify defects associated with design, development, and product realization activities. The underlying premise is that defects are typically driven by process variation. Sources of variation are placed into six categories to facilitate the root-cause analysis process: people, methods, machines, material, measurements, and environment.

Pareto Analysis. The Pareto analysis is better known as the “80/20 Rule.” The fundamental concept of Pareto analysis is the identification of the most likely sources of variation that are resulting in product defects and QMS nonconformances. As part of the root-cause investigative process, the investigator and/or investigative team identify a number of potential sources causing defects and nonconformances to occur. The sources of the most prevalent causes become the focus of the investigative process. However, this approach can also be problematic, as minor sources driving defects and nonconformances may be excluded from the initial investigation. Conversely, Pareto analysis is an excellent tool for supporting risk management activities because of the need to focus on big-picture product issues.

Fault Tree Analysis. Fault tree analysis is a deductive investigative process in which an undesired state of a system is analyzed using Boolean logic to combine a series of lower-level events. This analytical method is employed as a tool for ascertaining system failures and identifying risk removal and risk mitigation activities. For example, in system engineering the fundamental goal is assess and address all “undesired states.” As high-level events associated with fault tree analysis, each failure condition is categorized premised on the severity of its effect. Simply stated, the more severe a condition, the more extensive the fault tree analysis.

Typical applications of a fault tree analysis include the following:

FMEA

The FMEA has been a longtime device industry staple. Originally designed to support complex aerospace and defense systems, there is significant value today in the design, development, and manufacture of medical devices that are safe and effective in their intended use. The FMEA can be categorized as a qualitative analysis tool used to assess components and processes, and their cause and effect on finished medical devices. An effective FMEA can be used by a device manufacturer to identify potential failure modes based on experience with product performance, the performance of similar competitive devices, raw materials employed in the manufacturing process, manufacturing processes, and unexpected field failures.

The medical device industry routinely employs three types of FMEAs:

Effective Root Cause Analysis

There are multiple reasons why CAPA and complaints directly relating to warning letters have remained at the top of FDA's list for several years. Some of the underlying reasons driving warning letters include the following:

Conclusion

It is difficult to fathom the logic behind telling FDA that a device manufacturer has no intention of complying with any aspect of the QSR. Industries outside the medical device industry have robust requirements for pursuing corrective action and the need for addressing customer complaints. Regardless of the industry, it is imperative that accurate root cause be ascertained. There are a plethora of tools available to support root-cause analysis. If proper training is not provided to employees, accurate root causes are not determined and the chances increase that device manufacturers may implement the incorrect solution. Implementing the wrong solution may potentially impact device safety and efficacy, so it is imperative that great care and attention to detail be employed as part of the root-cause investigative process.

References 

1. Code of Federal Regulations. 21 CFR 820.

2. Determine the Root Cause: 5 Whys, [online] (Ridgefield, CT: iSixSigma, 2013 [cited 27 August 2013]); available from Internet: http://www.isixsigma.com/tools-templates/cause-effect/determine-root-cause-5-whys/.

3. Warning Letter: Soleetech Corp 8/13/13, [online] (Silver Spring, MD: FDA, 2013 [cited 26 August 2013]); available from Internet: http://www.fda.gov/ICECI/EnforcementActions/WarningLetters/2013/ucm365317.htm.

4. D. Gano, “Comparison of Root Cause Analysis Tools and Methods,” in Apollo Root Cause Analysis— A New Way of Thinking 3rd Ed., Dean L. Gano [online] (HVACR & Mechanical Conference, 2007 [cited 27 August 2013]); available from Internet: http://www.instructorworkshop.org/App_Content/instructorsworkshop/files/PRESENTATIONS/2013Presentations/Reality%20Charting_ARCA_Appendix.pdf

5. Understanding Root Cause Analysis, [online] (London, UK: BRC Global Standards, 2012 [cited 26 August 2013]); available from Internet: http://www.brcglobalstandards.com/Portals/0/media/files/Certification/BRC026%20-%20Understanding%20Root%20Cause%20Analysis.pdf 

Bob Mehta is the principal consultant and recruiter at GMP ISO Expert Services, where he provides consulting service in pharma, biotech, medical device, API, and food/dietary supplement industries. Bob has more than 23 years of experience, including as a principal consultant, in the quality systems, training, and regulatory compliance areas.