Going Digital

Pharmaceutical manufacturing processes continue to evolve, periodically embracing technological developments such as disposable components and regulatory initiatives such as process analytical technology (PAT). These changes have improved many operations, yet the top two causes of deviations—human error and equipment failure—remain the same. These factors create anywhere from 50% to two-thirds of manufacturers' headaches. How could companies remove them from the equation? One possible solution is automation.

(PHOTO: COMSTOCK IMAGES, GETTY IMAGES)

Drugmakers use corrective and preventive action (CAPA) programs to address deviations, modify procedures to prevent their recurrence, and measure the effectiveness of these modifications. Automating CAPA processes with software can help manufacturers find the best ways to resolve production problems and, ideally, eliminate deviations altogether.

Knowledge is power

Personnel must know as much as possible about a deviation to resolve it effectively through a CAPA program. Most drug companies have a software application system dedicated to tracking production problems and helping personnel figure out how to avoid them in the future. To enable thorough investigations, this system must receive accurate information about the manufacturing process. But if personnel enter data into the system manually, the opportunity for human error increases, and it becomes difficult to categorize deviations correctly and find their root causes. These difficulties can delay a product s release for shipping.

Automation that captures data and integrated software that facilitates the tracking and trending of actions and commitments are important tools for supporting a successful CAPA program. "With the move to at-line or on-line data-capture tools, the capture of events and excursions during processes is achieved more readily through automation," says William (Al) Kentrup, vice-president of manufacturing and supplier quality assessments at Pfizer (New York). Integrating the related event logs into appropriate CAPA software allows a company to manage the manufacturing process, including product disposition and resource allocations, effectively.

Employees prepare for cell-culture transfer from a seed bioreactor into a perfusion bioreactor. (IMAGE IS COURTESY OF PFIZER)

On the other hand, sometimes data-capture tools are too effective for their own good. "In today's environment, the ability to capture data and events can exceed the ability to assess them and react appropriately," says Kentrup. But automation or software tools can help manufacturers convert data into knowledge easily, especially in highly automated or complex processes such as packaging operations.

In addition to PAT tools that enable on-line monitoring, an automated, centralized quality-control system also helps address the challenge of getting consistent, granular data, says Tim Mohn, industry principal at Sparta Systems (Holmdel, NJ). Integrating the CAPA program with an overall quality-management solution can provide a comprehensive view of the process and facilitate root-cause analysis and trending.

Close the loop

Trending of data, a crucial part of the CAPA process, allows a manufacturer to assess its performance in comparison with established standards. Trending deserves more of drugmakers' attention because it allows them to determine how effective their corrective actions are in preventing the recurrence of deviations. Over time, pharmaceutical manufacturers have become skilled at identifying root causes and developing preventive actions for deviations. But companies often do not follow up by performing the appropriate tracking, trending, and effectiveness checks, says Susan Schniepp, vice-president of OSO BioPharmaceuticals Manufacturing (Albuquerque, NM) and a member of Pharmaceutical Technology's editorial advisory board. This lack of continuity can allow a deviation to recur if the preventive action did not address the problem directly. Automation can make it easier for companies to track manufacturing operations and perform effectiveness checks, which are crucial steps for closing the loop of the CAPA process.

Each system plays a part

Drugmakers often use dedicated software solutions to perform tasks such as maintenance, manufacturing, and analytical testing. Each of these systems can play a part in reducing the deviations that might otherwise trigger a CAPA process. For example, operations-management software uses automated workflows to walk personnel through standard operating procedures step by step, thus ensuring that manufacturing steps are carried out correctly. The software also verifies that equipment has been cleaned or sterilized and that personnel have the training required for a given procedure. By checking this information before performing work and making it available to workers in real time, an operations-management application can help to prevent deviations from happening, thus ensuring the manufacture of acceptable products and saving time that might have been spent on investigations.

Software from companies such as EtQ (New York) can prompt manufacturers to take corrective actions in response to deviations. (IMAGE IS COURTESY OF ETQ)

A quality-management system (QMS) can confirm the identity, strength, and purity of raw materials and finished products. QMS helps manufacturers maintain accountability by allowing them to assign employees the responsibility for investigating the cause of a given deviation. The software also allows managers to establish a timeline for resolving deviations and to develop a structured escalation process for overdue actions.

Companies have been successful in establishing QMS at a high level, but they often are cited for not implementing it or for not having a uniform process for mitigating deviations, says Michael Sieve, life-sciences solutions engineer at Seapine Software. With an effective CAPA procedure, aided by QMS, organizations can prevent recurring manufacturing failures, save time and resources, and help generate revenue.

Unite and conquer

More and more, pharmaceutical companies are realizing that the information in their various data systems can improve CAPA processes appreciably. During the past few years, manufacturers have begun to flow information from software such as QMS and maintenance systems into their operations-management system. And links between companies' CAPA and operations-management systems are becoming increasingly common, to the point where they are expected for new facilities.

"You're starting to see much more of an integration of information flows between what had been various silos," says Bob Lenich, director of Emerson Process Management's Syncade product platform. If these data systems are linked, an excursion during production automatically triggers a deviation process within a CAPA system. This link helps resolve the deviation faster and allows products to ship sooner as part of the normal release process. Integrating information from laboratory, maintenance, and operations systems is crucial, says Lenich, because it can help the CAPA process find the root cause of production problems early so that they can be avoided.

The case for full conversion

Much of the pharmaceutical industry has automated various steps of the CAPA process such as requirement gathering, event tracking, and testing. But full automation in CAPA is currently not widespread. Manufacturers ultimately need a single system that provides full traceability of requirements through testing, defects, corrective action, and closure so that teams in successive segments of a product's life cycle are informed about previous deviations and understand why particular events were resolved in certain ways. "Being able to automatically inform people about the current status of a product can be invaluable in the long-term life of the product," says Sieve.

Many companies believe that automating CAPA processes is expensive, time-consuming, and does not provide a return on investment. But this belief is mistaken, says Sieve, because some vendors offer an all-encompassing automated CAPA system that is much easier to set up than an enterprise resource planning system, which can require extensive customization. Commercial off-the-shelf software can offer a wealth of configuration options.

Removing bottlenecks

On-line monitoring, linked to an infrastructure that integrates a company's various data systems, can provide personnel with excellent visibility of the manufacturing process. But the wealth of information that this infrastructure provides, coupled with memories of production problems that have brought regulatory scrutiny, can prompt pharmaceutical manufacturers to be overly cautious with their CAPA programs. "Not all deviations will be CAPAs. I think a lot of people tend to funnel all their deviations into the CAPA system, and that's not necessarily appropriate," says Schniepp. A deviation should lead to a CAPA only when the result of that deviation would affect the product's quality and safety.

A common problem in CAPA programs today is many companies' failure to distinguish crucial from noncrucial events. "Events in a quality system often always become a CAPA, until the company has more CAPAs than it can manage effectively," says Tim Lozier, manager of strategic development at EtQ (New York). As a result, manufacturers often focus on addressing the most overdue CAPAs rather than completing those that represent the most crucial events. The excess of CAPAs can create a bottleneck in an otherwise healthy quality system.

Risk-assessment tools can help companies judge whether or not an event is crucial. First, a company develops risk formulas based on its past events. When the software system records a deviation, the risk-assessment tools evaluate the threat that it poses to factors such as compliance, safety, and quality. The company also assesses the deviation's severity and frequency, and these calculations result in a risk ranking for each deviation. The ranking helps the company decide whether the event is crucial or not. "This strategy allows for a more focused approach to corrective actions and allows companies to focus their attention on the events that matter most to the business," says Lozier.

A cosmetic error such as a smudged logo on a finished product may not be deemed crucial, especially if it occurs in a single lot, and can be immediately corrected. Less crucial events are handled in the event record. A mislabeled finished product, however, can have a higher risk ranking and be deemed crucial. The definition of a crucial event can vary from company to company, depending on each manufacturer s history.

The industry is beginning to accept this risk-assessment approach to CAPA as the best practice. For some pharmaceutical manufacturers, crucial process-parameter deviations automatically trigger a CAPA process. The number of noncrucial deviations that triggers a CAPA process depends on the culture of the particular company, says Lenich.

Automation for all?

A fully automated CAPA process helps pharmaceutical manufacturers in different ways, and to different degrees, depending on their size. Although small, medium, and large drug companies face similar regulatory scrutiny and compliance challenges, the resources available to these companies differ, which may explain why pharmaceutical manufacturers have embraced automation to different degrees.

Automation is not necessarily cost effective for companies with 100 employees or fewer, says Schniepp. Setting up automated CAPA systems can be time consuming and might not provide many advantages to firms that manufacture low volumes of a small portfolio of products. Instead, small drugmakers are likely to rely on their contract manufacturing organizations (CMOs) to track CAPAs.

The extent to which small companies are automating their CAPA processes is unclear. Small companies do stand to benefit from automation, says Lenich, but they often view automation as a cost rather than as a value. This perception is a salient distinction between small drugmakers and mid- to large-sized manufacturers.

Even if they cannot automate their CAPA systems fully, smaller companies can find automation for specific parts of the CAPA process, which may help them address performance concerns and focus on continuous improvement, says Kentrup. "The challenge is to determine a prioritization scheme for automating parts of the process that is consistent with expected business outcomes," he adds. Considerations that would promote the automation of a CAPA system include minimizing the initial financial investment, monitoring process performance, and containing capital-investment costs.

It can be difficult for mid-sized pharmaceutical companies to determine how cost effective it is for them to implement an automated CAPA system, partly because they are uncertain about whether their products will reach the market. A mid-sized company with a mixture of small- and large-molecule products will likely want to track the active-ingredient manufacturer, final-product manufacturer, and packager for each late-stage product. If the company hires 12 CMOs to provide four products, it might consider automating its CAPA process to keep track of its contractors' operations, says Schniepp.

"My impression is that most mid to large companies view automation as being required for CAPA," says Lenich. During a US Food and Drug Administration audit, manufacturers are expected to show that they have a formal CAPA process, that they enforce it, and that they record its history and metrics.

"An automated CAPA process, or at least a partially automated process, is vital for any mid-sized to large pharmaceutical company," says Pfizer s Kentrup. Improved compliance, operational effectiveness, and continuous improvement are major reasons that these companies invest in automation. Automated CAPA systems make a lot of sense for large facilities that manufacture the same products consistently because they provide a standard infrastructure and enable manufacturing, investigations, remediations, and trending to be carried out quickly and effectively.

An automated CAPA process is virtually a necessity for large organizations with multiple production facilities and many products, says Lozier. To reduce time to market, meet customer demand, and maintain a high level of quality, Big Pharma uses automation as much as possible to streamline its complex operations. Large companies must not only automate their processes, but also seek ways to integrate manufacturing software with their other business systems, says Lozier.

CMOs can derive the same benefits from automation as pharmaceutical companies do, says Lozier. Some brand owners are enabling CMOs to participate in their CAPA systems. "Using robust flexibility and security, contract manufacturers can participate in the CAPA process without compromising the brand owners' system," says Lozier. This arrangement helps promote collaboration between CMOs and brand owners while tracking quality and compliance throughout the supply chain.

Several large CMOs have invested in automated CAPA systems to help them keep track of multiple contracts and customers. An automated CAPA process enables a CMO to improve its transparency and show clients that it is operating in a state of control. Software also helps reduce the amount of time needed for routine oversight and quality audits. Another benefit for CMOs is that a potential sponsor company might be more confident in a contractor that uses an automated CAPA system similar to the one the sponsor uses in house, says Kentrup.

Automation advances onward

CAPA processes are becoming increasingly automated in every industry, but regulatory scrutiny makes automation all the more important to the life-science industry. "The difference between an automated versus a manual CAPA process can mean a great deal, especially if a company is being audited by the US Food and Drug Administration," says Lozier. Automated CAPA systems help provide traceability and visibility, which are important to a successful audit. The more data a company can provide about adverse events and corrective actions, the better it can demonstrate regulatory compliance.

Although automation can make a wealth of information available and greatly enhance process control, drug-makers should keep in mind that it is an extension of a company's existing CAPA process, which begins with its personnel. "The key is to keep the automated system simple—use it as a tool to drive the CAPA process, not as a solution to the process," says Lozier.

Effective automated solutions should be able to adapt to companies' existing CAPA processes and provide a way to streamline them, thus making them easier for users, says Lozier. Software also should be flexible enough to adapt to changing processes without requiring a burdensome amount of customization.

Software systems surely will become more common in drugmakers' CAPA systems, just as they have become integral to other operations. Automation "is one of the most sought after features in any quality system, since it is the heart of the process and a catalyst for change," says Lozier.