15 Minutes

(Damiano Papini, M.Sc. Chemistry and MBA, Rephine Consultant and Auditor).

For any manufacturer, identifying, classifying and managing deviations are of vital importance to increase the effectiveness of the quality system.

In particular, the implementation of an effective CAPA (Corrective/Preventive Action) system is the direct consequence of having identified the “true” causes of the deviation, commonly known as “root causes” which, in turn, can be found only if a thorough and well-structured investigation has been carried out. From the other side, we can say that a poor investigation will likely lead to an incorrect interpretation of the reasons why the deviation has occurred, thus focusing on merely causal factors (or even worse on casual factors). Consequently, any formulated CAPA will likely be unsuccessful in preventing the reoccurrence of the deviation itself.

The concept of CAPA has changed over time and different pharmaceutical companies may have dissimilar definitions of the terms “corrective” and “preventive”. One simple interpretation is that a corrective action works on the event, a preventive action on the cause.  A preventive action is intended to mitigate the risk that the event may happen again, even on systems or processes which are different from the one originally affected.

ICH Q7 GMP guide for Active Pharmaceutical Ingredients (API) prescribes that any critical deviation should always be investigated and CAPA identified. A deviation is defined as a “departure from an approved instruction or established standard”. A critical deviation is defined as a “departure from established critical parameters or a significant departure to standard operations which may affect the quality of the API or intermediate” and consequently carry the risk of a product that is hazardous to the health of patients.

A deviation is by its nature an unplanned event. APIC “How to do” Interpretation of ICH Q7 Guide, reports examples of common deviations which include incorrect charging of raw materials, temperature, pressure, vacuum parameters outside defined limits, breakdown of process equipment or failure of utilities, equipment out of calibration, in process control limits not achieved, extended drying or distillation times due to faulty equipment, discrepancy in labels reconciliation, alteration of shipping condition and other unplanned events.

Certain activities are always required for dealing with deviations. First of all, deviations should be thoroughly recorded and investigated to identify root causes. The risk arising from the deviation on the affected batch should be assessed. It will also be necessary to establish whether or not other batches may be affected. Suitable CAPA for all affected batches must be specified to prevent recurrence of the deviation. Last but not least, the effectiveness of assigned CAPA should be examined after implementation and a periodic review of the effectiveness of whole system performed.

The management of any deviation process is the task and responsibility of several key manufacturing, engineering, R&D and quality functions which should actively participate in the investigation, analysis of the causes, assessment of the impact and definition of a CAPA plan.  A key figure is, of course, that of the investigator, who is the person responsible for leading the process (typically QA). One factor that can derail the investigation is if too many people are involved in root cause investigations.  Instead, it is a best practice to have core teams of experienced subject matter experts who fully support and execute all deviation investigations. This practice delivers consistency, quality, and assurance that the “real” root causes are found.

The causes of a deviation are not always immediately evident, so it is essential to identify valid tools to set the boundaries of the analysis and simplify the investigation process. Root-cause investigations can be performed using a plethora of methodologies such as brainstorming, interviews, affinity diagrams, DMAIC, FMEA, 5 Why’s, flow diagrams, Ishikawa diagrams, Pareto charts, statistical tools, run charts, process capability analysis. Many of these tools have been consolidated for decades, broadly used in many industrial fields and nowadays recognized by regulatory bodies as valid investigation approaches.

Two tools have been found to be particularly effective in identifying the “hidden” causes: the “5 Why’s” and the “Ishikawa Diagram”. The 5 Why’s tool consists of challenging the interviewee by “digging” behind every answer given and evaluating each time whether the “true” cause has been found and stop or instead not to be satisfied and continue to ask “why” until the “true” cause is identified. Unfortunately, identifying the “true” cause is often like “finding a needle in a haystack”, which means that in the most complex cases the 5 Why’s tool alone is not always resolutive. When there is no easy evidence of the causes, an exclusion method such as the Ishikawa diagram can be very powerful. This method consists in finding the “zone of the haystack where the needle is likely to be” by dissecting the “haystack” into 5 or more smaller portions (the so called “categories” of the diagram) such as manpower, machines, methods, materials, environment etc. In this way the field of research gradually narrows down and what initially may have seemed like a huge challenge, become a possible mission.

As part of the investigation, one key element is to determine the quality impact of the deviation, starting from the question “Is there any impact on the quality attributes of the batch involved?”.  If the answer is yes, other questions are triggered: “Can the product be marketed? Are other batches affected? Which are the causes of the deviation and which, therefore, are the best CAPA to be undertaken?”. These questions often encourage a clear and quick answer.

Ironically, when a deviation comes from an evident failure such as a laboratory OOS, the task for the investigator is much simpler; since the fate of the batch is established, the qualitative impact is evident and the investigation can just proceed to search for the root causes of failure. On the other hand, when the deviation event is not directly related to a product failure, the assessment of the impact can be fairly complex. One example is the temperature variation that can affect a product during shipment, for example at 60°C for 2 hours. As long as the temperature excursion is covered by supporting stability studies, the maintenance of product quality attributes can be demonstrated. When the excursion is not supported by any study, despite all of the statistical predictions that can be made about the known impurities up to the expiration of the drug, the assessor is faced with two tricky questions: “Do we know the nature of impurities that could have been formed at 60° C for 2 hours? Are our analytical methods capable of intercepting and quantifying them?

Special consideration should be given to those situations where the cause of deviation is generically attributed to so-called “human errors”, as many regulatory authorities no longer accept “human error” as a justifiable cause of deviation.  In fact, when quality defects are associated to “human error”, investigations may be poor and superficial. In such instances, the assigned action is just “re-training” and no other sustainable CAPA is defined. The approach to correct by just “re-training” often fails to produce the desired result, as it only takes care of issues related to lack of knowledge, skill or ability. If the error did not occur because of one of these factors, then just “re-training” can be useless. In fact, “human errors” usually indicate potential deficiencies in the process (robustness), in the procedures (complexity), in the effectiveness/frequency of training (methods), in the equipment (ergonomics), in resources (stress, lack of time).  When we talk about “human error”, we should therefore realise that there are always other contributing factors which provoke the “human error”.

In conclusion, the source of the problem with deviation investigations is typically how well root-cause analyses are being performed.  Finding the root causes of the deviations is absolutely essential to follow through with effective CAPA in the attempt to ultimately prevent their recurrence.  Improving on that step and following up with logical CAPAs will almost certainly guarantee compliance and satisfied regulatory inspectors.

About the Author

Damiano Papini, M.Sc. (Chemistry), MBA. Consultant and Auditor.

Damiano has over 30 years’ pharmaceutical industry experience, in a variety of roles with increasing responsibility within pharmaceutical development, API manufacturing and QC/QA sectors.

Prior to becoming a full-time consultant and auditor, he occupied senior management positions in big pharma (Pharmaceutical Development Manager and Director API Analytical Sciences in R&D GlaxoSmithKline) and global CRO (Quality Director, Third Party and Regulatory in Aptuit Evotec).  Damiano has extensive knowledge of  Investigational Medicinal Product  manufacturing and control,   QC laboratory management and analytical technology transfer, as well as direct experience with regulatory authorities across the world, having approved several CMC dossiers from Phase I up to NDA/MAA filing and participated in many regulatory inspections.  

As a full-time consultant he has been carrying out  more than 130  audits  in both Europe and Asia, for various categories of suppliers (API, excipients, primary packaging, sterile and oral dosage forms, biologicals,  manufacturing contractors,  clinical packaging and labelling,  QC Laboratories,  service providers, medical devices,   according to  various standards  such as ICHQ7, cGMP, EU GMP/GDP,  IPEC-PQG,  ISO9001,  ISO15378, ISO17025 and  ISO13485).  Other consultancy activities included quality gap assessments, delivery of GMP training courses  and QA/QC temporary management assignments in Italy and Ireland.

Damiano has a degree in Chemistry from the University of Padova  with specialisation in Protein Chemistry.  After graduation he has been a visiting scientist for 19 months at Kansas University School of Pharmacy, involved in the development of polymer-based polypeptides delivery systems.  He also obtained a Masters in Business Administration in 2011. Damiano is a registered professional chemist and is eligible to act as a QP under the permanent provision of Directive 2001/83/EC.