Saturday, June 26, 2010

Writing a Procedure Doesn't Make it Happen


A firm correctly interpreted that the FDA had lost patience with their inability to sustain compliance as evidenced by repeated FDA 483 observations and the inevitable Warning Letter that arrived by certified mail to the corner office.

Rewind. What happened?

The firm had written an eloquent response to the observations that hit all the right notes. Improvements were evidenced by a revised procedure that was attached to the response, which satisfied the FDA for the moment—though the FDA noted they would see for themselves upon re-inspection.

And see for themselves they did—or rather, they didn’t.

The General Manager was stunned to hear the inspector declare at the close out meeting that there was no substantial improvement since the previous inspection.

“How can this be?” The GM’s question reverberated through the building as everyone ducked for cover and the search for the guilty began.

Unfortunately, this is a common problem, and it’s based on the unfounded notion that if you write a procedure, you make it so.

So, I’m here to declare that writing a procedure doesn’t make diddly happen.

In this instance, the theme of the FDA feedback was the ubiquitous “failure to thoroughly review the failure of a batch or any of its components to meet any of it specifications whether or not the batch has already been distributed.” But the specific issue doesn’t really matter. It could have been anything—the underlying problem remains the same.

It boils down to two things: organization capability and governance.

In this case, reasonable commitments were made to the FDA and appropriate changes were made to their deviation management procedure. However, the firm still lacked the resources and technical bench strength to conduct a thorough investigation and inadequate process knowledge to execute an effective corrective and preventive action.

Nothing in their response changed their organizational capability.

Further, there was no process or forum to monitor how well the deviation management system was working—before or after changes to the system were made. There were no management reviews of quality system performance metrics, and there was no attempt to verify whether they had actually fulfilled their commitment to the FDA.

They had to learn about their failure from the FDA.

The net effect was a continuation of inadequate data gathering, misjudged root cause, poorly written investigations, misguided CAPAs—and the same persistent nonconformities. (Need I mention a mountainous backlog of deviation reports and unpredictable product supply?)

So who are the guilty?

I think that the GM and the head of quality assurance equally share the blame.

The GM is certainly responsible for the overall capability of the organization. However, the head of quality assurance is responsible for reporting on the health and welfare of the quality system—and, he/she should possess no less than the skills of the head of marketing to get the right metrics to the right people at the right time to take action on data-driven decisions.

But, alas—most of us in QA are nerds.

 The QA Pharm

Saturday, June 12, 2010

Patients Take the Individual, not the Average


Many decades ago, my high school science club sponsored a field trip to Eli Lilly. Our little group of longhaired and bell-bottomed kids (after all it was the 60’s) was ushered into a little room to watch the introductory video about the history of the company and Colonel Lilly. This was followed by a guided window tour by a Lilly employee who would never know that he started a sequence of events in my life that led me to the pharmaceutical industry.

I have often thought about that day, which gave focus and purpose to my academic studies. But these many years later, I mostly remember the tour guide.
I recall how he pointed to each piece of equipment and explained what it did, how it worked, why it was important to the process, and even who was important as he pointed to each operator in the area and told us their roles.

When we finally arrived at the end of the tour, our guide showed us the final product being filled and packaged and said that everything we had seen up to this point was important for each and every vial that was coming off the line. It was with a reverential awe that our guide said that each vial would be going to a person whose life depended upon it.

That tour guide got it right. Patients take the individual, not the average.

It is manifestly impossible to sample a sufficient quantity of finished product to assure the conformity down to the unit level. However, it is the unit level that is taken by the patient, sometimes daily for their entire life.

So, if we cannot provide adequate assurance down to the unit level by testing alone, and if it is important that we must for the sake of each use by the patient—then what gives? Is this a double bind?

The answer is in the Quality Management System.

The purpose of each element of the Quality Management System is to collectively work together to control the manufacture of the drug product from the suppliers’ materials received at the dock to the final product in the shipping containers coming off planes and barges around the world and onto hospital and pharmacy shelves—through to the end of the expiration date.

Each step along the way from dock to stock is based on science and engineering. The Quality Management System ensures that the science and engineering foundation is vigilantly protected from undesirable and unintentional variation that could negatively impact the product at the patient level. And if it does, the problem and product impact are thoroughly investigated and the underlying cause is engineered out.

This is the field of Quality Assurance. It’s all about the patient and the individual unit, as well as the controlling the way we work to ensure consistency and predictability. Patients deserve no less.