Siegfried Schmitt, principal consultant, PAREXEL International, discusses how to find the root cause of the problem.
Siegfried Schmitt, principal consultant, PAREXEL International, discusses how to find the root cause of the problem.
Q:Deviations in clinical trials happen, and as such, FDA requires deviations to be investigated and documented. Our company has an established deviation investigation process. What advice can you give to ensure that deviation investigations provide the most optimal corrective actions and preventive actions (CAPA)?
A:Standard operating procedures (SOPs) for deviation investigations are generally similar from company to company, detailing a logical process flow with execution instructions, and a section on roles and responsibilities. Following a risk-based approach, deviations are classified as minor or major, requiring investigations that differ in resource requirements and detail.
Finding the root causes of the deviation, rather than merely causal factors, is absolutely essential to follow through with CAPA, a concept that focuses on the root causes of identified problems in an attempt to ultimately prevent their recurrence.
Root-cause investigations can be performed using a plethora of methodologies (see Table I), providing the deviation investigators with all the tools they need. These tools require familiarity with the methodology and regular practice to understand which method(s) will provide the answers you are looking for. It is almost always necessary to apply two or more methodologies in order to find the root causes of a deviation.
Table I: Selected root-cause analysis methodologies.
8D methodology
Affinity diagrams
Awareness and problem definition
Barrier analysis
Brain storming
Cause and effect diagrams
Define, measure, analyse, improve, control (DMAIC)
Failure modes and effects analysis (FMEA)
Five whys
Flow diagrams
Hazard analysis and critical control points (HACCP)
Impact assessment
Ishikawa diagrams
Kepner-Tregoe
Mind mapping
Modeling
Pareto charts
Pivot tables
Risk tree
Run charts
Statistical methods and sampling
Statistical process capability analysis
Statistical process control
The interview
Trend analysis
Trending
One factor that can derail the investigation is if too many operators and managers are involved in root-cause investigations. Unless the individuals regularly participate in root-cause investigations, they will not be fully familiar with either process or methodologies. Instead, it is best practice to have core teams of experienced subject matter experts (SMEs) fully support and execute all deviation investigations. This practice delivers consistency, quality, and assurance that the real root causes are found.
Another factor is that SMEs might be tempted to mistake root causes for causal factors, since the latter would require less investment or resources to fix. This is not good practice-almost always it will be possible to find at least partial solutions that can be implemented.
To give an example, an operator picked up the wrong tool to fasten a pipe. An investigation revealed the causal factor as ‘human error’ and the actual root causes as ‘tools not clearly and unambiguously labeled’ and ‘insufficient lighting to read the labels.’ Thus, instructing the operator in the procedure would not have resolved the problem. The solutions needed are improved labeling and lighting.
This example addresses why deviation investigations are so often found to be lacking when reviewed by agency inspectors. These issues can be grouped like this:
The source of the problem with deviation investigations is not the procedure or methodology; it typically is how root-cause analyses are being performed. Improving on that step and following up with logical CAPAs will almost certainly guarantee compliance and satisfied regulatory inspectors.
Article Details
Pharmaceutical Technology
Vol. 40, No. 9
September 2016
Pages: 98
Citation
When referring to this article, please cite it as S. Schmitt, "Root Cause Analysis: Finding the Root of the Problem," Pharmaceutical Technology 40 (9) 2016.