From the Editor: When Voluntary Becomes Mandatory

April 7, 2009
As the period of introspection within FDA comes to an end, so does the time for playing “wait and see” with PAT, statistical process control, process capability and risk triage

This magazine came into being during an interesting time in the history of drug manufacturing. FDA had been studying the way it approached compliance, why there were so many consent decrees and compliance problems, why manufacturing had become so inefficient and why companies weren’t availing themselves of the latest IT and other tools. The fruits of this study were GMP’s for the 21st Century, the Process Analytical Technologies (PAT) guidance, Quality Systems guidance and Quality by Design.

The works of Deming, Shingo and Ohno became required reading for many industry professionals. It was a period of introspection at FDA. Officials at CDER may have criticized the industry’s poor compliance record and continued reliance on 19th century approaches, but they admitted FDA’s role in creating its intense aversion to risk. 

They also designated these improvement programs as “voluntary’ rather than prescriptive, expounding repeatedly on how all these concepts would improve the status quo for regulators, regulated, and the public at large.

FDA professionals received training in Lean Six Sigma, and in manufacturing, to better understand, on a holistic level, the broad industry issues at hand. Then some were sent out to preach a new gospel—first, with PAT, and then with QbD. Slowly, more companies responded.

This approach was unprecedented. After all, you never see IRS officials go on the road to stump for the benefits of smarter recordkeeping and accounting approaches. Unfortunately, FDA’s efforts may have been misunderstood in some circles, particularly political ones.

But somehow even all this FDA outreach, and the hopes that enlightened self interest would take over, were not quite enough to get stubborn companies to change how they handle compliance. Talk about PAT, continuous manufacturing and quality improvement would often go right back to questions about the three validation batches. What was missing is a document tying these concepts directly to process and product validation.
That guidance, on process validation, which first appeared this January, may represent the final document in the early 21st century flowering of FDA thought.

It will likely be finalized next year, within a whole new FDA.

It’s too soon to guess what the Agency will look like, but it is safe to say that the period of introspection is over. President Obama has already nominated a deputy commissioner who has been openly critical of the industry.

Guidances developed over the past decade will now be driving what promises to be a much tougher enforcement agenda.

As consultant Jim Agalloco pointed out recently during a webcast on FDA’s Process Validation Guidance, companies with well-founded validation and development programs will welcome the flexibility of the draft guidance, while the rest will complain, as they have all along with these voluntary guidances, that they aren’t explicit enough and don’t spell out exactly what to do.

The time is over for waiting for more explicit guidance from FDA, or waiting to see how other companies implement these approaches. If anyone doubts that, they should heed the words of Joe Famulare, deputy director of compliance at CDER, who ended a very upbeat presentation during the webcast by dissecting the language in a hypothetical consent decree letter. Mentioned specifically were process control and control of variability.

Drug companies that do not get smarter about addressing variability and its root causes at their facilities, about employing statistical methods and Quality by Design concepts, may pay very dearly for this failure. These concepts are a reality at FDA, and, although some inspectors may not be singing the tune, we see the language in an increasing number of 483’s.

Whether your organization is large or small, name-brand or generic, mastery of the principles outlined in the process validation draft guidance will be critical: process understanding, control of variability, root cause analysis, statistical tools. For our part, we promise to bring as many best practices as we can, to help in this effort. To watch the webcast, please visit our web site.

The U.S. auto industry has become a negative role model for pharma in many ways.  Never mind their current predicament, car makers were dragged, kicking and screaming, to consider better process efficiency and design improvements. After Ralph Nader hit the scene in the 1970’s, consumer awareness and legislative attention made car safety a requirement. Seat belts became mandatory.

In many ways, smart validation methods are the “seat belts” for your processes. Why not use them before they’re explicitly mandated and the costs of compliance are likely to be much, much higher?

About the Author

Agnes Shanley | Editor in Chief