To begin the discussions in this issue about structuring successful training programs, these authors describe how to develop in-house resources. Both emphasize the need to assess the impact of previous training approaches. An important factor is to understand that students will bring varying levels of knowledge and relevant experience to their work with your company. So how do you structure a training program that takes into account individual learning styles and operational expertise? How to you measure the successful application of training to performance? As a manager, how can you confirm that reading training materials has truly led to required levels of understanding within your group? And how long can you realistically expect employees to spend watching a training video before their attentions wander?
At the March 2016 BioProcess Theater at Interphex in New York City, Chad Cooper, Biogen’s director of performance development, and David Yarley, director of training and development for Fujifilm Diosynth Biotechnologies USA, offered examples of successful approaches to internal training programs.
Beyond Read and Understand: Moving Training from Compliance to Performance
by Chad Cooper
At the BioProcess Theater, Chad Cooper offered the following case study. In it, he detailed development of a type of performance paradigm, recognizing that a company can struggle administratively when improving on “read and understand” in training expectations. Moving from training for compliance to true performance improvement can become difficult when employees are bogged down by the administrative burden of multiple read-and-understand training requirements.
Biogen is a worldwide company of about 8,000 employees, headquartered in Cambridge, MA. Its primary therapeutic area is multiple sclerosis, with nine commercial products and an active pipeline of other candidates. The company also works on therapies for neurodegenerative disorders. Its three drug-substance sites are located in Cambridge, MA (2,000-L scale); Research Triangle Part (RTP), NC (2,000- and 15,000-L scales); and HillerØd, Denmark (15,000-L). Another drug-substance site is being built in Solothurn, Switzerland, for completion in 2019, and the company has recently acquired a small-molecule and fill–finish facility from Eisai, also in RTP.
Before the company developed its current training approach, training was a label for documentation and was viewed as an exercise for demonstrating compliance. Manufacturing employees typically had 500–1,000 read-and-understand requirements. A manufacturing operator could have 20–30 pages of read-and-understand standard operating procedures (SOPs) with unclear links between training and tasks. An operator learning how to perform column operations, for example, needed to find all the SOPs associated either directly or indirectly with that and be trained on them all — a virtually impossible task. Additionally, SOPs do not contain other training content such as background theories and troubleshooting guidance.
Our goal was to offer training that was much more supportive of how performance actually was to be accomplished. Among the many published studies about human performance, we learned about six primary factors that have an impact on human performance, both external and internal to employees. We based our program design on these six factors: expectations, resources, incentives, knowledge and skills, readiness, and personal motives and preferences.
Expectations and Feedback: What are the requirements from the boss and from the company? What sources of guidance and reinforcement are available? Where is coaching needed, and what types of feedback are most effective? Such elements have a great impact on how humans perform their work.
Tools, Resources, Job-Site Conditions: What do the work environment and work conditions actually look like? How clearly do your employees understand the procedures?
Incentives and Disincentives: What do you reward? What do you sanction, either implicitly or explicitly?
Knowledge and Skills: How much training has each employee had? What does each person know and understand about the technology behind the tasks?
Capacity and Readiness: Is each person physically and mentally capable of performing a given task?
Personal Motives and Preferences: Do employees want to be there? Are they motivated to do their jobs?
Of those six factors, the first four are controlled to some extent by management, and the remaining two are more inherent and innate to individual employees.
We tried to target our training to address the first four factors. We wanted a structure that truly was based on task. For example, if we wanted somebody to run column operations, we wanted training set up so that someone could say: “I want to run a column next week. Here are all the things I need to be trained on to do that.” To begin with, we had few, if any, materials, aside from the SOPs — and we needed content. We wanted knowledge and performance criteria that were available across our network and across the globe.
“Read-and-understand” was our goal for 100% compliance. But how do you know that somebody has read and understood every document? We needed reinforcement. We wanted some qualification standard that our people had to pass to show that they were actually trained. We knew that training approaches vary widely and that the transfer of knowledge and skills is, to a large extent, on-the-job instruction. But along with that expected variability, we needed standardization and consistency.
Competency-Based Training: We began by focusing on task-based competency training. When developing training for large unit operations such as bioreactor and columns operations, for example, we created training specific to those tasks that would not depend on location, site, or equipment size. We set up training for bioreactors, whether for 2,000- or 15,000-L operations.
As already mentioned, we had to develop a great amount of training content that did not yet exist. At the highest level, we created what we call knowledge standards: basically e-learning courses. A unit operation such as running chromatography columns is one knowledge standard. Bioreactor operations is another. We identified 24 major unit operations before we created knowledge standards for each one.
An e-learning course for us is 20–30 minutes long. An employee has to pass a short assessment at the end. Each course provides some theory, some troubleshooting, and some basic background on the minimum details of operation. For chromatography, it would explain how the resin actually works. The material in this course was the highest level of knowledge conveyed.
The next step was to break each major task into subunits or subcomponents, such as preuse activities, process activities, and postuse activities. For each subunit, we created performance qualifying events (PQEs). PQEs are on-the-job assessments that an employee must complete to be qualified to execute a given operation.
Then we organized SOPs below those subunits — such as SOPs that go with column preuse activities. If a person wants to execute column operations, he or she has to be trained on the knowledge standard and all three PQEs rather than just note that procedures have been read and understood.
That was the brunt of the work to create this system. Along with creating content for 24 new knowledge standards, we had to create more than 140 performance-qualifying events that had not existed before. This was a year-long effort, at least. But the savings were tremendous. As part of this, we converted our SOPs to work instructions. A work instruction is defined as step-by-step instruction that an operator needs to perform an operation. Although some instructions still were included in SOPs, this enabled drastic reduction of read-and-understand and revision training.
We didn’t change the format of the SOPs, but we changed their headers to read Work Instruction.
Current Status, Future Improvement
At our RTP site for cell culture personnel, the read-and-understand requirements went from 285 down to 21, and that was about the same for purification employees as well. Training is logically grouped by task. All major unit operations now have training content. PQEs are standardized. We have seen >80% reduction in the transactions that we have to perform in the learning management system. I can’t quantify how much time was spent before this confirming read-and-understand and logging into the system to simply acknowledge that. This improvement is hard to measure, but it’s a substantial win.
Future improvements are based on further understanding of modern learners. We have determined that modern learners may be overwhelmed, distracted, and impatient. On average they spend 1% of their time on training, if that much. If you think about it, our training should become more “micro.” Most learners are not going to watch videos that are longer than four minutes, whereas most e-learnings are 20–30 minutes long.
Mobile and Game-Based Learning: People check their phones at least nine times per hour. Game-based learning has shown to be very effective, and we think we can actually use that in our future program designs.
Benchmarking with Other Industries: We benchmark with other biotechnology and pharmaceutical companies, but companies in some other industries are doing some pretty amazing things.
Our management goal across the board is to dedicate more time to training: more like 5% rather than <1%. We know that moving from training for the sake of regulatory compliance (read and understand) to true training for performance can greatly improve the effectiveness of our staff and the overall reliability of our organization.
Chad Cooper is the director of manufacturing at Biogen; firstname.lastname@example.org.
Effective Theory-Based Training for Operators
by David Yarley
Effective theory-based training begins with getting back to basics: to training operators on foundational elements. Those could include the practical science behind each operation, practical math that needs to be used and understood, equipment operations, and the theory behind the equipment. Phases of operation and specific things about them need to be understood, as do troubleshooting approaches. Also important to understand is how one area or one unit operation affects other unit operations, whether downstream or specific to financial aspects of your business.
At the BioProcess Theater at Interphex, March 2016, in New York City, David Yarley addressed these issues using examples gained from years of teaching about bioprocessing to students with varying levels of knowledge, experience, and professional expectations. Fujifilm Diosynth Biotechnologies has 1,100 employees worldwide at three sites: Billingham, UK; College Station, TX; and RTP, NC. The company has six licensed commercial manufacturing products and over 35 years of experience working in contract development and manufacturing on more than 270 molecules. Sixty of those are monoclonal antibodies (MAbs) or MAb–like molecules, 30 are vaccines, and 175 are nonantibody recombinant proteins.
Much existing production documentation lacks theory, but theory is foundational. It is often shared, as well. So once that foundation is in place, you can often apply its concepts and mechanisms in other areas. For example, in the case of salting out a protein, the concept of hydrophobic interactions there also can be applied to hydrophobicinteraction and reversed-phase HPLC.
Operators are quick to understand new concepts when they are provided with theoretical bases, and such training encourages more technical communication between operators and engineers and also with scientists. Understanding theory will encourage operators to seek additional lifelong learning.
I’ve had 23 years of experience working the pharmaceutical industry. During one of the phases of my career, I managed production for which I helped develop a training program. I saw firsthand how teaching theory improved overall operations, resulting in fewer deviations and other problems. From there I moved on to the BioNetwork Capstone Center (Raleigh, NC), which was a community college initiative. I was there for 10 years and taught thousands of students about bioprocessing and related aspects. Because it was a community college, many students did not have much biotechnology experience at all. They wanted to learn to get a job. I had to figure out where it started with them. Now I’m working for Fujifilm Diosynth Biotechnologies, and I’ve brought quite a lot of theory-based training into our company.
Certification: Training at Fujifilm consists of a certification program similar to the program that Chad described. We have five functional and four core areas. The functional areas are fermentation, cell culture, filtration, centrifugation, and chromatography. The core technology areas are cell disruption, glass washing and autoclaving, weighing and dispensing, and solution preparation. The certification program in each area comprises a pretest, a standard operating manual, testing of manual contents, and test qualification.
For task qualifications, we determine which specific tasks employees must be qualified to perform. An employee who bundles all those tasks together is certified within that area. Opportunities for becoming a subject matter expert (SME) are also available, although it is a much more stringent pathway.
Concurrent with that, we also introduced a theory-based training in each of these areas so that we don’t depend on “read-and-understand.” Instead, we have open discussions to confirm and ensure comprehension.
One of the helpful nuggets that I came across along the way is the work of Mortimer Adler, editor of the Great Books of Western Civilization series. He was a brilliant mind and a brilliant educator. During the 1980s he developed the three-part Paideia Proposal, an educational approach. It has three parts. The first part is lecture-based learning, which if taken alone is least effective. The next part involves coaching in skills, defining skills as habits not memories. The final part involves Socratic moments that consist of open-ended questions and discussions.
Socratic moments are important. Adler believed that discussions had to include asking the open-ended questions in such a way as to help students (in our case, employees) know why they know what they know. Theory-based training as practiced at my company is built on the following four topics:
- Know your audience
- Prepare your content
- Develop your materials
- Deliver your presentation.
Know Your Audience: If you miss this mark, you miss the whole goal. You need to know where to start. Some good suggestions are to know what previous training your employees have had, maybe in the companies where they previously worked, and what type of training was done there. If you know the demographics of your area, you may also design your training to apply to knowledge your people already have. For example, if your audience includes auto mechanics, you can provide many examples to explain your equipment by using basic auto mechanic theory.
Knowing your employees’ education level is also essential to a good training program. What is their science experience? How much chemistry? How much engineering have they actually been exposed to? Part of this is the importance of knowing what “tribal knowledge” exists in your company — especially legacy information for which the theory not quite correct. That needs to be addressed, reinforced, and changed when you are embarking on theory-based training.
Here is where SMEs can bring their experiences to help you determine, from problems they’ve addressed, what knowledge seems to be missing from the process and where it is needed. So you need to make sure such experts are included in your training development.
Prepare Your Content: There is practical theory, and there is just theorizing. You want to make sure that you create what you need, and that takes some work. Again, multiple SMEs can help you realize that your people may have different visions of where problems exist or where knowledge may be missing.
This process takes time. Plan for one or two hours maximum within your training sessions. Use as many facility pictures as you can because you want to make training as realistic as possible to the operators. And map out a logical approach. When we are teaching something such as filtration that may be present in multiple areas or scattered throughout a plant, I like to use what I call the sandwich approach. Think of it as a loaf of bread on which the top piece is an overview of the overall process, with detailed “slices” offered below it.
Once you have gone through the content, finish the lecture with another overview. Then you start asking probing questions based on your presentation content. Where do you use hydrophilic filters? Or do we use hydrophobic filters? If you are using an isolated unit operation, make sure your students understand where it is located. Is it upstream? Is it downstream? Is it used in recovery?
A Chromatography Example: Start with a definition of the history of chromatography and explain it very simply. Show a column with a mobile phase and a stationary phase (resin). Then go directly into the chemistry of proteins, starting with amino acids. But during the course on amino acids, you can point out a few that are specific to your processes, such as how cysteine is responsible for disulfide bonding or how tyrosine and tryptophan are responsible for your A280 absorption.
Then in our program we expand into looking at the proteins. Here we talk about chemical bonding: ionic, covalent, hydrophobic interactions, and hydrogen bonding. We explain the importance of maintaining the folds within these proteins.
Finally, we go into resin properties, explaining also how the resin is actually made, the importance of crosslinking, and the difference between gel and macroporous resins. In different types of chromatography, what mechanisms are used and how do they operate? I use some animation to help deliver this information, and you can find quite a lot of such resources online.
What are the operations? We talk about the conditioning steps, the loading steps, the washing steps, the elution steps. How are they run? Why are they run that way? And as for performance testing: Here we do explain the math, but we don’t try to delve into the theoretical plates discussion. We simply explain that it is a calculated number and describe how to calculate it. And we go into asymmetry, as well.
We look next at the specific columns that we use, the related systems, the unique features in each column, and the packing method, of course.
Developing Your Materials: I still like to use Microsoft PowerPoint software. But I suggest that when you change concepts, change the backgrounds. And between the presentations we include assessments using iClicker technology that allows us to poll our audience. I like to present answers that are very similar to one another because that automatically prompts group discussion.
Always be prepared to go deeper. Many times we have used process training for both our process development scientists and our maintenance personnel. To serve them, we need to build on that knowledge by asking more probing questions during the lectures.
Delivering Your Presentation: Be sure to keep your presentations to no more than two hours, or you start losing people’s attention. Where and when training occurs is determined by the operators. The operators at Fujifilm Diosynth like me to gown and return to the cleanroom to deliver training because then they don’t have to regown.
Make your presentation enjoyable by using real-life examples, presenting multiple technologies, and encouraging team interactions and discussions. And always listen well. When you are training people who aren’t really understanding as well as they should, then you will want to create times for additional one-on-one coaching.
When applying Adler’s proposal to theory-based training, you want to make sure that your lectures are well-balanced and that you start at the right point. You want to organize your material logically so that the sequence of topics and examples makes sense to your audience. You need to use practical examples and implant pictures. Ask probing questions to make sure they understand the fundamentals, and if possible, include hands-on exercises.
For your Socratic moments, you can prepare group exercises: “We had a recent problem. Why do you think it happened?” Open things up for discussion. In many cases there is no right answer, but there are many answers. Probing into an example can lead to good discussions.
Finally, make that sure you include business and financial aspects and how those are affected by individual operations. For example, chromatography operations can include many cycles, for which a company has to depreciate the expense of the resin. Employees should understand that financial impact. How does a problem occurring upstream affect what happens (and costs incurred) downstream? Such questions can prompt further discussion.
David Yarley serves as director of training and development for Fujifilm Diosynth Biotechnologies USA, Inc. (Research Triangle Park, NC); email@example.com.