Many years ago, while working on developing a Quality Management System for a large automotive company, we heard a story about how the company had perfected Corrective Action, Preventive Action (CAPA). The company had a clearly defined, cross-functional process with a very useful quality tool kit containing templates for data collection and root-cause analysis that were extremely effective for brainstorming solutions. The process also routinely included a detailed cost analysis. Executing the CAPA process was costly — to the tune of several hundred thousand dollars — but very effective.
At one point, the company had identified a problem in one of the paint shops. The problem was recurring, and causing significant rework. Consequently, the shop manager opened a CAPA. The shop used the tools, collected the data, performed the root-cause analysis, and fixed the problem. The issue never recurred in that shop. However, the same problem occurred in 24 of the company’s paint shops around the world. Each shop, independently, followed the CAPA process and solved the problem. However, the CAPAs alone cost the company more than $10 million.
If all sites had visibility into the first solution, the company could have saved millions. Despite “perfecting” CAPA, the company still struggled with the high costs associated with the process. There is some irony in seeking to perfect a process that, by definition, points out imperfections; however, taking a well-structured approach can help pharma companies create a more effective CAPA process that pays off in the end.
Is there a problem?
Many companies have recognized problems with volume and cost, as well as ineffectiveness of the CAPA process. The cost of the CAPA process alone, not including remediation, can run into the hundreds of thousands of dollars per CAPA. Companies can find they generate so many CAPAs that CAPA processing backlogs develop. One senior executive visited a plant and found 15 overdue CAPAs. Such backlogs create compliance issues and can end up as a finding in an inspection report.
As part of its Case for Quality program, the U.S. Food and Drug Administration (FDA), in partnership with the Medical Device Innovation Consortium (MDIC), is currently sponsoring a working group to improve the CAPA process. The initial focus of the working group is to develop a reference triage process that will allow companies to address non-conformances and other issues with confidence — generating CAPAs only when patient health and safety may be at risk or other recognized criteria are met.
Even with manageable numbers, problems with implementation and overall effectiveness remain an ongoing challenge. Some companies rush to close CAPAs prematurely to remain in compliance with their Standad Operating Procedures (SOPs). Others fail to include appropriate stakeholders in the investigation and planning process — further exacerbating problems rather than successfully addressing them.
The CAPA process
The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use guidance refers to a structured approach. The specific process will vary from company to company, involving different parties and leveraging different tools, but will include the same main elements (Exhibit 1).
The process starts with any of a number of triggers. The basic criteria are related to product quality, business efficiency and, in life sciences companies, patient health and safety. Initiation refers to capturing and recording information related to the CAPA, including cause, severity, stakeholders, and target outcome, among others.
Investigate: This involves collecting data to understand the problem and root-cause analysis to make sure the solution addresses more than the symptoms of the issue.
Form solution: Correct the immediate issue (corrective action) and prevent recurrence (preventive action).
Implement: This may require initiating a product or process Change Control, which may trigger a regulatory variation filing (contributing to the high cost of CAPAs). SOP revisions will most likely be necessary, requiring communication and training.
Monitor effectiveness: Note that the process does not end with implementation. Companies can close the CAPA in the system. However, CAPA effectiveness must be monitored on an ongoing basis. Only after the corrective and preventive actions have demonstrated effectiveness can the process be considered complete.
When should CAPAs be initiated?
The process starts with occurrence of any number of triggers including a finding by a health authority, a significant process deviation, recurring non-significant deviations, adverse events, complaints, or a Health Hazard Evaluation. The company should have clearly defined criteria for initiating a CAPA, applied uniformly across the enterprise. Given the significant time and cost associated with a CAPA, determining a reasonable threshold is important.
Generally, compliance-related issues are worthy of a CAPA. Minor deviations, however, would not warrant a CAPA. Recurring deviations that result in rework should trigger a CAPA, as should deviations that could have an impact on patient safety. A Form 483 and warning letter findings are definite triggers. In many cases, management may trigger a CAPA coming out of a periodic Quality Management Review to address an issue or as part of continuous improvement.
The right number of CAPAs
Unfortunately, there is no right number. Companies with complex or high-risk products may initiate more CAPAs than companies with simple manufacturing processes. Companies with mature quality systems and processes can expect fewer CAPAs.
As mentioned previously, some companies are overwhelmed with volumes of possibly unnecessary CAPAs. Other companies try to avoid CAPAs, fearing raising red flags with authorities and wanting to avoid the costs associated with the CAPA process. Authorities will expect to see numbers that reflect the level of recurring complaints, inspection findings or other issues.
There are two categories of tools — one is the QMS solution that would be used to manage the CAPA process, where CAPAs would be initiated and tracked. The other tools are the traditional methodologies used to support investigation: histograms and Pareto charts to support issue identification, and Ichikawa or fish-bone diagrams to support root-cause analysis.
Enterprise QMS solution: As the name implies, this is the primary solution used to support the Quality Management System. Companies often try to use one system to manage all quality processes — deviations, inspections and inspection findings, complaints, change control and CAPA. One advantage of a singular system is that the processes can be easily referenced. Data from deviations or other quality issues can be transferred to the CAPA, saving time and allowing for traceability.
While some small companies rely on paper to manage the process, they are likely to hold on to paper much longer than advisable. CAPAs should be entered into a system with sufficient detail to support effective investigation and reporting. An enterprise quality IT system would have potentially allowed the automotive manufacturer in our previous example to gain visibility at the enterprise level while avoiding millions of dollars of wasted effort. CAPA initiations should be reported regularly as part of standard operational monitoring and periodic management reviews. Another reason to leverage technology over paper CAPA management: There are several software-as-a-service (SaaS) quality solutions that are reasonably priced and very effective.
Quality toolkit: A histogram charts occurrences over time. This gives a sense of frequency and impact of an issue. The histogram can also support root-cause analysis by relating peaks with other events occurring at the same time.
Pareto charts can be used to capture types or categories of issues. By graphing the issue counts by category, one or more categories will emerge as the “winners” and allow teams to focus on the issues of greatest concern.
The Ishikawa diagram is a brainstorming tool that provides a structured approach to exploring possible root causes. The diagram resembles a fish skeleton and is also known as the fishbone diagram. Each arm represents a possible root cause. As team members hypothesize causes, they add potential contributing factors to be explored (Exhibit 2).
A corrective action may be taken to provide a stop-gap solution — to allow a process to be run or a batch to be released. Preventive actions are intended to be long-term solutions. These solutions should be implemented only after undertaking a rigorous root-cause analysis, lest the fix not fully prevent recurrence and the CAPA process be repeated.
Quality is a high-stakes business — not only measured in dollars, but more importantly, impacting patients’ lives. Poor quality is measured in seven figures, or more. Label-related recalls alone can cost pharma companies more than $100 million a year. Industry quality executives have estimated an average cost of $10 million for a Form 483 finding. A warning letter can set companies back approximately $200 million, and a consent decree could cost upwards of $1 billion.
CAPAs are expensive as well, but are clearly preferable to the significant costs of noncompliance. As mentioned, the process alone can consume hundreds of hours. Implementation of a solution requires its own impact assessment. The Change Control process may be involved, which can lead to a regulatory impact assessment and variation filings in a number of countries.
All this points to the pursuit of the goal of “Right First Time” and continuous improvement. Assuming CAPAs are initiated prior to a health authority finding, they are executed in the spirit of continuous improvement, rather than under duress of regulatory oversite.
Measurement is foundational to the CAPA process. Measurement of process and process outputs often trigger the CAPA. Recurrence, trending, financial impact and health hazard evaluations are all factors driving initiation.
Once initiated, managers track progress through the CAPA process. The primary focus is on CAPA closure timeliness and the number of overdue CAPAs. More mature companies track and report financial impact and other data such as cause, manufacturing site, and products to provide better insight, improved forecasting of CAPA durations and improvement of the CAPA process itself.
Of course, one of the primary measures should be CAPA effectiveness. Ineffective preventive action can lead to hemorrhaging of cash as well as potential threats to patient safety.
Perhaps there’s no such thing as a “perfect” CAPA. Indeed, perfect would be achieving optimal results the first time through, eliminating the need for CAPAs altogether. With changing regulations, evolving technologies and a mandate for continuous improvement, this day will likely never come — yet pharma must continually strive for better quality management processes.