3 Reasons Why QbD projects get Stuck & What to Do about it

Posted On

I have seen and experienced many QbD projects get stuck. There are 3 main pillars in Quality-by-Design: Risk assessment; Design Space; Process Analytical Technology.

FIGURE: 3 Pillars of Quality-by-Design (QbD)

Of the three, many companies seem to be focusing and spending much time on the Risk Assessment piece but not enough on Design Space or PAT. As a result, we begin the QbD journey with a good intent but can not complete it. This was certainly the case for Perjeta of Genentech which fell short of a full QbD submission. Why is this happening and what can we do to prevent it?

Here are my thoughts on 3 Why’s and Do’s:

1. Why: Companies run out of time and energy before completing the QbD package. As the input for Design Space and PAT, Risk Assessment is the first essential step. However at the beginning of a new QbD project, most project managers fail to allocate more time and resources for Design Space mapping and PAT upgrades. A 100 page Risk Assessment spreadsheet may seem comprehensive but what happens if we run out of time to gather experimental data to back up the risk priority decisions? Then the spreadsheet remains as a list of unverified assumptions. Better yet, what if we discover lurking attributes or parameters from the experiments? Should we update our signed-off CQA? FDA pointed this out as an incomplete QbD submission in the Perjeta case. When insufficient time and resources are budgeted for Design Space activities, failure is baked into the plan.

Do: Allocate more resources for running experiments to map the Design Space in the QbD project plan. This does not necessarily mean securing a higher budget. Just plan to spend more time on Design Space and PAT activities and less on Risk Assessment. Prioritize and rebalance the existing budget and time. Emphasize this point to the QbD Project Sponsor so that you have enough budget and time secured to complete the cycle of “Risk Assumptions – Verification” through adequate experiments.

2. Why: We forget that Risk Assessment is an iterative process. Without setting a firm deadline on the 1st CQA-FMEA spreadsheet, the project doesn’t proceed to the next step. How many times have we observed meetings where R&D, Quality, Regulatory, Manufacturing, etc. passionately argue over why every attribute and parameter should be “critical” (without data)? As with most organizations, it is difficult to proceed without reaching an agreement. Who or what should be the ultimate decision maker in this case?

Do:  Set a short time limit for the 1st spreadsheet of CQA / FMEA and move on. Plan to revisit it after the experimental data comes in. Risk Assessment should be iterative. In the absence of data, wasn’t the first FMEA mostly a guesswork anyways? In any project, one  will never have enough resources to explore all attributes and parameters. Again, prioritize based on educated assumptions and get data. Then re-prioritize and revise the risk assessment. Rinse and repeat.

3. Why: Conducting FMEA meetings for CQA and CPP are relatively easy and cheap (although reaching agreement as a group on what “critical” means may be a struggle). In comparison, running the correct experiments to map the Design Space and investing in capital equipment for Process Analytical Technology (PAT) require more in-depth knowledge, discipline and resources. Without management’s reinforcement, we will all go through the path of least resistance.

Do: Develop “design space” capabilities through training, hiring an expert, making it a routine work for development.  Set up a disciplined project management system. Instead of the traditional project management style, I suggest the Agile or Scrum approach. Agile is the lean version of project management. It minimizes overhead in documentation and deals effectively with changing information. I will share how I adapted Agile approach in QbD in future articles.

Let’s remind ourselves that financial benefits and regulatory relief comes from a properly characterized Design Space. Risk Assessment is just the beginning. PAT is a control plan for the design space. When QbD is implemented correctly, manufacturing process becomes robust, product waste is reduced and the company benefits from exemption of post-approval regulatory approval requirements within the pre-approved process ranges.

 

I will discuss more on how the Cycle of Risk Assumption & Verification works in the next article.

Please subscribe to join the conversation.