Root-cause analysis (RCA) is an invaluable tool used in our research. The goal of RCA is to find the true cause of an observed
defect, failure, or problem and use that information to correct it. RCA is a problem-solving tool widely used in a variety
of situations and industries. In the pharmaceutical industry, it is a prelude to determining the appropriate corrective and
preventive action (CAPA) to be taken for an unexpected occurrence.
RCA starts by having a good understanding of the problem, coupled with available information (both quantitative and qualitative),
data, and facts. The process for determining root cause is systematic and consistent, as the root cause must be supported
by evidence. Additionally, a defect, failure, or other problem may occur due to a combination of factors. It is very possible
to identify more than one root cause or probable root cause for a given situation. Of course, careful documentation is essential
at a company (and, by extension, FDA) to enable later "forensic" identification of the cause.
Note also that there can be multiple levels of causation. For example, in industry, the root cause of many recalls is the
failure of management to implement and maintain a robust quality system (also known as a "quality-management system"). The
test of an industrial quality system is its capacity and effectiveness in preventing manufacturing inconsistencies and quality
defects. This central objective is met through sound design, control, monitoring, detection, and adaptive improvement practices
throughout the product lifecycle. Both elusive and easily identified manufacturing problems can cause substandard product
to enter the market, and it is essential for a manufacturer to institute robust performance monitoring and CAPA programs to
prevent such occurrences (4, 5). Thus, RCA in this project attempts to drill down to identify the mechanistic manufacturing
cause(s) that should have been detected at the design stage (i.e., during experimentation or scale-up) and/or by the quality
system2 prior to distribution. In accord with the RCA concept, the cause(s) should be defined with enough precision and certainty
to suggest a clear course of action. Our research project realized the importance of this objective, even as we observed that
the root causes posited by the company at times significantly changed in the course of the iterative investigations we reviewed
and that the inspection findings occasionally revealed other likely causes. All available relevant information was considered
in our root cause determinations.
The general categories of root causes are:
- Human error
- Materials and components
- Machines and equipment
- Physical facilities and utilities
- Environment and nature
- Process and product design
- Measurement methods, standards and specifications.
An ideal RCA should:
- Give a clear and specific cause
- Be substantial and nontrivial
- Be supported by available information and data
- Be supported by experienced and knowledgeable experts
- Provide for correction and/or prevention
- Have management authority and resources to correct it
- Provide correction in an acceptable length of time
- Track and trend the effectiveness of the correction over time
- Lead to fewer or no events of that type again
- Lead to learning that can be applied to other areas.