Phil Nethercote, independent pharmaceutical industry consultant and Stephanie Harden, global pharmaceutical marketing manager, Waters, discuss the types of error mitigation tools that ensure the quality of data and products among pharma companies.
Key insights:
- Pharmaceutical companies go to great lengths to assure the quality of the data they produce and any error that arises during execution of an analytical procedure must be thoroughly investigated.
- It's important that as part of any investigation, the lab supervisor works closely with the analysts to truly understand the route cause.
- Adopting a detailed structured investigation, identifying the true root cause, and implementing an appropriate CAPA are key to reducing the frequency of errors due to humans.
The reliability of analytical data in the pharmaceutical industry is critical in providing assurance of product quality. As such, data quality is a topic which receives significant focus from regulatory authorities. Pharmaceutical companies go to great lengths to assure the quality of the data they produce and any error that arises during execution of an analytical procedure must be thoroughly investigated, and appropriate corrective and preventative actions taken (CAPA). Not only can an error in performing an analysis have an impact on product quality but frequent errors can lead to significant waste of resources and lead to delays in product flow through a facility.
Companies will often adopt structured approaches to identify and address sources of errors and frequently tools employed in lean manufacturing are used to help in error investigation. Causes of errors can be categorised using an Ishikawa, or a ‘fishbone’ or a diagram (Figure 1).
Figure 1: Fishbone diagram categorising the different factors that contribute to errors in the lab.
- While errors may be caused by issues associated with any of the “bones”, those associated with machine, man, and method tend to be most frequent, and therefore receive the most focus.
- Significant effort has been invested to identify the best tools and processes to prevent errors arising due to “method”. These efforts have involved applying the principles of quality by design (QbD), that were developed to ensure robust manufacturing processes, to analytical methods.
- Errors due to “machine” are addressed by ensuring laboratory instrumentation is kept up-to-date and protected with service plans that include timely replacement of worn instrument components.
- It is within human error (“man”) that there remain opportunities for significant error reduction. Often, lab investigations conclude that errors are due to a human factor and that the corrective action required is “re-training”. Inspectors raise concerns when reviewing laboratories that have a high portion of such errors and CAPAs as it is clear that re-training was not effective in preventing further errors. One approach that can help significantly is to use tools typically used by the health and safety community to classify human errors into different categories (Figure 2).
Figure 2: Breakdown of the types of human error and violations.
By adopting such an approach, it is possible to ensure that the root cause of the error is identified and the CAPA is effective in preventing reoccurrence. Skill-based errors are best prevented by removing distractions and interruptions and ensuring there is sufficient time available to complete the task. Errors due to lapses of memory can be mitigated by modifying the design of the task to make it easy and intuitive, introducing simple checklists and clear reminders/warnings or alarms or by introducing a second cross-check of critical steps. Re-training is not an effective CAPA for lapses of memory.
Mistakes occur due to lack of knowledge – either on how to perform the task (rule-based) or in how to deal with non-routine situations (knowledge-based). These types of errors are the areas where re-training should be considered. Where this is the case, it is important to consider the reason why the initial training was ineffective in the first place, as well as the impact on others who have been similarly trained. Having plans and/or drills for potential unusual situations should also be considered.
Violations are intentional actions, and these may be where the analyst routinely doesn’t do the task as described in the method, where the task is not done as described because of the situation (e.g. a lack of time) or where the task is deliberately not done as required, in an attempt to rectify a situation (e.g. an analyst overfills a volumetric flask, and then tries to correct this by removing some of the liquid).
Potential CAPAs for intentional actions include:
- Eliminating the reason to cut corners by job redesign
- Removing unnecessary rules
- Removing unrealistic targets
- Simplifying work instructions
- Updating the method to reflect actual practice
- Improving risk perception by promoting an understanding of the criticality of the task
- Embedding warnings in procedures
- Improving supervision.
Clearly when human error is identified, the potential CAPAs are numerous and not simply restricted to re-training. It is therefore important that as part of any investigation, the lab supervisor works closely with the analysts to truly understand the route cause. Typically, the discussion between the analyst and the supervisor should cover:
Table 1: Sample checklist for lab error investigation
Adopting a detailed structured investigation, identifying the true root cause, and implementing an appropriate CAPA are key to reducing the frequency of errors due to humans.
In addition to reducing errors by improving the quality of the investigation and resulting CAPAs, labs should also consider proactive steps that can be taken to prevent human errors in the first place. A detailed walkthrough between supervisor and analyst of the steps involved can identify a surprisingly large number of opportunities for human error. These can then be risk assessed, and proactive steps taken, to avoid the potential errors deemed to be at highest risk of occurrence. For example, such an activity might identify the potential for slips and spills during sample preparation and result in investigating options for automating that step.
Solutions for the pharmaceutical quality assurance and control sector are continuously evolving, not only to keep up with regulatory requirements, but also to help mitigate the risk of analytical failures. Investments in new technology can support analytical needs and compliance demands, while also providing preventive measures that reduce the opportunity for errors.
Additionally, improvements in employee training, better preventive maintenance of laboratory instruments, and the addition of automation and software tools have allowed laboratories to identify and control potential risks effectively, resulting in greater efficiencies. Substantial improvements are now being realised across the industry as organisations strive to implement a culture of quality, set targets around error reduction and update ageing analytical infrastructure.