Blended learning: an approach to operator training

Abstract

In this white paper we establish what a blended learning approach is and the advantages it has over traditional biomanufacturing operator training methods, where reading and acknowledging procedures can easily become a repetitious but necessary tick box. We aim to discover how blended learning increases operator engagement, contributes to job satisfaction, and can reduce operator error in drug manufacturing processes. The byproducts of choosing blended learning techniques as an approach to training should be a reduction in batch loss, maintenance of batch quality and yield, and retention of skilled operators.

How big a problem is operator error in drug manufacture?

The twentieth annual BioPlan ‘Report and Survey of Biopharmaceutical Manufacturing Capacity and Production’, is an extensive source of trends and analysis in the biopharma industry. This collation of survey responses from more than 300 biopharmaceutical manufacturers reported that in 2023, annual GMP batch failure due to operator error, was equivalent to an average of 2.9% for commercial manufacturing and 2.8% at clinical scale manufacturing. Batch failure due to equipment failure was reported at 3.2% (commercial) and 3.1% (clinical), and batch failure due to contamination reported at 3.1% and 2.9% respectively (1).

Our recent study of Cytiva equipment uptime surveyed 60 large and medium biotech drug manufacturers and CMO/CDMOs. This study confirmed that contamination is the number one cause behind batch loss, closely followed by equipment failure and human error. In this survey respondents were asked, ‘If you experience a production batch loss what is the typical financial consequence? Please estimate the total financial impact.’ And, ‘What other consequences does your organization suffer as a result of a production batch loss?’ They equated a single batch loss to an average of nearly $ 2.7 million with disrupting impacts rippling beyond the financial (2) (Table 1).

Table 1. Financial impact of batch loss (trimmed average), N

$ 2.696.852
By organization type
N, $	All	Medium biopharma biotech	Large biopharma	CDMO/ CMO
< />	10	0	6	< />
1m to 5M	27	6	8	< />
5M to 10M	8	1	6	< />
>10M	3	1	1	< />
>NA	12	1	4	< />

The same study explored the consequences of lower-than expected quality as well as the consequences of batch production rescheduling (Tables 2 and 3).

Table 2. Financial impact of low batch quality (trimmed average), N

$340, 413
By organization type
N, $	All	Medium biopharma biotech	Large biopharma	CDMO/ CMO
<100k$340, 413	10	0	7	3
100k - 500K	19	0	10	9
>500K	13	4	4	5
NA	18	5	4	9

Table 3. Financial impact of batch rescheduling (trimmed average), N

By organization type
N, $	All	Medium biopharma biotech	Large biopharma	CDMO/ CMO
<100k	16	0	8	8
100k - 500K	18	3	8	7
>500K	5	2	1	2
NA	21	4	8	9

Non-financial impact survey responses

An analysis of open-ended survey responses revealed the non-financial impact of batch loss, low batch quality and batch rescheduling.

Consequences of batch loss

• Supply chain disruptions — delays in production batches lead to supply chain disruptions, affecting the timely delivery of materials and impacting the overall production schedule.
• Impact on subsequent batches — a common consequence is the slippage in plans for subsequent batches, leading to the need for unexpected work hours, rework of planning activities, and potential delays in subsequent steps.
• Quality control and deviations — longer production times can result in in-process control issues, triggering deviations and investigations. This can lead to a loss of product quality, especially for unstable products.
• Operational disorganization — overlapping with another batch can lead to a prolonged culture seed train, shifting subsequent steps, and causing disorganization in production shifts and staff schedules.
• Reputational damage and regulatory impact — delays in production can impact reputation, cause delays in quality assurance release, and affect clinical start timelines. Regulatory issues may also arise, impacting the overall planning of projects and resource availability.

Consequences of low batch quality

• Not meeting target production output — lower than expected quality of drug substance leads to issues in meeting target production output, impacting overall productivity and requiring additional batches.
• Patient safety and clinical trials delays — quality issues may lead to delays in clinical trials, regulatory filing, market application approval, and commercialization. There are concerns about patient safety.
• Reputation and client trust — lower quality affects company reputation, client trust, and may result in client complaints. Unhappy clients and delays in closing projects contribute to reputational damage.
• Regulatory impact — regulatory concerns arise during filing. The possibility of not being able to release the batch if out of specification is mentioned.
• Supply chain disruptions and resource impact — lower than expected quality causes disruptions in the supply chain, impacting prioritization, supply chain timelines, and clinical trial timelines. It also results in the need for additional resources and has organizational, resource, and supply chain impacts.

Consequences of batch rescheduling

• Scheduling issues for operators — rescheduling production batches leads to scheduling challenges for operators, affecting workforce planning and potentially causing stress and errors.
• Clinical trial delays — requiring reorganization of schedules for staff, rooms, and materials. There are also delays in respective clinical trials and potential impacts on commercial lot supply chain issues.
• Supply chain disruptions — rescheduling has consequences for product supply and supply chain planning, leading to disruptions and challenges in meeting customer commitments.
• Impact on timelines and capacity utilization — there are delays on timelines for other activities, under-utilization of facilities, and the need for additional resources to manage changeovers in facilities. Rescheduling also affects other planned batches, reducing overall capacity utilization.
• Reputation and customer confidence — rescheduling has a significant impact on reputation, client trust, and customer confidence. It may lead to client complaints, reputational damage, and potential project cancellations.

Re-training operators as a CAPA

A standard approach to handling deviations occurring during batch manufacture would be to document for quality purposes and undergo investigative root cause analysis. Corrective and preventive actions (CAPA) would then be instigated. If a thorough root cause analysis has been drilled down to ‘operator error’, the CAPA is likely to be ‘re-training of operator(s)’. This is a good time to ask the questions:

• Is the current operator training sufficient?
• Is the current operator training efficient?
• Do the same errors keep occurring?

Clearly, drug manufacturing processes can be highly complex, using multiple pieces of equipment, flow paths and testing protocols. It could be argued that ‘operator error’ as a root cause may be skewed by other factors that could be improved, such as (3):

• Complicated or unclear written instructions
• Complex equipment and software
• Not enough experienced operators or system users assigned to a task or process

The beauty of life is that although most problems can be costly, they do bring opportunity. Human error and ‘re-training operators’ can open new doors if we ask ourselves the questions:

• How do individuals learn best?
• Is there a better way?

The bottom line is that effective training will always be a key factor in reducing operator error. With this in context, the better way could be investment in more experiential and diverse training techniques as an alternative to traditional read and acknowledge methods.

Criteria for successful operator training programs

Inadequate training/non-current GMP training (poorly trained employees and contractors) was listed in a 2022 ‘Synopsis for Most Common FDA Audit Findings’ reported in recent years (4). This clearly highlights the need for successful operator training programs. The design of efficient training programs can be accomplished by applying the best learning methodologies while understanding the requirements specific to process or technologies. It should be suitable for employees with limited or no experience in biopharmaceutical manufacturing, largely self-directed, typically include instructions on the rules and procedures for maintaining aseptic conditions, as well as taking into consideration that access to cleanroom suites may be limited. Effective training goes beyond meeting compliance requirements; it ensures that operators can perform processing tasks competently under real-life conditions.

Learning styles and the fundamental aspects of training

So, it is clear that the goal of any training program is to ensure that everyone who receives training executes the same tasks in the same manner in the same timeframe, leading to the same output every time. To achieve this goal in a drug manufacturing process is to protect product quality and patient safety, and it is essential to identify training needs and address them holistically. In essence, a good training foundation must support transfer of knowledge, as well as skills, and behaviors required for consistent high-level performance (Fig 1). Adult learning styles must be considered, and it has been identified that there are generally four ways to learn.

• Listening (auditory)
• Viewing (visual)
• Reading/writing
• Doing (hands-on)

Fig 1. Key components of effective training programs.

 

These learning styles can be addressed with either passive or participatory teaching methods. The former includes lectures, reading, audiovisual tools, and demonstrations, while the latter comprise group discussions, practice, and teaching others. Adult learners are focused on learning what they need to be successful in their current role and meet their assigned goals and objectives. It is only natural that increased engagement leads to increased job satisfaction. As such, some of the best practices for adult learning, include face-to-face interactions and participatory programs that require problem-solving, are topic-centered, and involve real-life scenarios linked to actual daily tasks. In this way, interest and engagement remain high. Several studies have shown that most adults prefer to learn by doing and retain more knowledge when actively participating in the learning process (5).

So, what is blended learning?

Blended learning is best described as an approach that uses a blend of digital and experiential learning from face-to face training to virtual reality. This is necessary because we all learn differently. We see this more clearly with children at school, but it remains true with adults.

• Face-to-face training — utilizing subject matter expert (SME) trainers and specialists, training can occur in the classroom and on the manufacturing floor.
• Hands-on training — experiential, whether in-house or at the equipment manufacturer, this training allows operators to use the actual equipment.
• Online eLearning — interactive and online, this type of training engages the learner and helps build application and technology knowledge at appropriate levels from foundation to expert.
• Virtual classroom training — gives training flexibility through remote access, removing travel pressures and operator downtime.
• Virtual reality training — a fully immersive virtual ‘hands-on’ experience that can be utilized on-demand to allow repeatability of process and grow muscle memory and confidence.
• Mixed reality training — adds interactive digital elements to a real-world environment for efficient on-the-job guidance.

Each of these different learning methods has unique benefits. The advantage of a blended learning approach is to use more than one type of learning to meet the needs of individuals and provide a holistic training path. According to the National Training Laboratories, retention of learning content ranges from 5% to 30% for passive methods and 50% to 90% for participatory methods (5). Adults often learn as much from committing errors (trying and failing), because they must think about what went wrong and why, which leads to greater understanding of the process and prevents reoccurrence.

Virtual reality (VR) for more efficient training

VR is quickly becoming a go-to solution as part of a blended learning approach. Forward thinking users in the biotechnology space have been taking advantage of its on-demand nature and the positive statistics reported when training personnel in other critical industries.

For instance, Ho et al, proposed a VR approach in their report, ‘Virtual Reality Training for Assembly of Hybrid Medical Devices’, finding that training time reduced by 26.7% and trainee assessment scores were on average 62.0% better, when compared with traditional training methods (6). For the assembly of hybrid medical devices operator skill level must remain high to ensure correct and efficient assembly where potential contamination is a risk to patient safety. Their proposed VR approach used a range of human computer interaction (HCI) techniques that allowed for the gradual step-by-step learning to gain knowledge over time, with objectives to be completed successfully before moving to the next stage. The HCI were developed with game-based themes to increase trainee engagement with practical applications, tutorials, and assessments. The final stage used motion-based HCI to allow for virtual ‘hands-on’ assembly. Not only did this VR approach reduce training time by an average of 26.7%, but each tutorial session was also reduced by an average of 31.2%. The latter result was partly equated to the nature of VR which allows trainees who learn faster to move straight to the next stage, whereby traditional training methods bring tutorials and trainees back to the classroom.

Results in this study also highlighted the benefit of using VR in step-by-step stages as they state to, ‘enhance training through various ‘game’ levels of familiarity building processes (6).’ A further efficacy study by PwC asked the question, ‘How does VR perform as a training tool for soft skills?’ (7). They found that people being trained by VR were:

• Trained four times faster than in a classroom setting
• 75% more confident to apply skills learned post-training

A 2 h classroom training session in this study, took 45 min with eLearning and just 29 min with VR.

Time to complete training

Fig 3. Time to learn study.

Industry 4.0 and the BioPhorum digital technology roadmap

The use of digital tools such as online learning and VR is part of the vision of Industry 4.0 and the ‘BioPhorum Digital Technology Roadmap’ (8). This industry group paper looks at the complexity and risk of digital transformation, asking the questions:

• What will the biopharma industry look like in 2030+?
• What digital technologies will enable this vision of our industry’s digital future?

The paper outlines how we are heading towards ‘digital first’ skill sets highlighting that, ‘the workforce will be strongly supported by digital technology to enable learning, guidance or help with prediction.’ Digital technologies will positively enable:

• Ease of access
• Interactivity
• User experience
• 24×7 availability

This vision of our industry’s digital future is of augmented and virtual reality (AR and VR) manufacturing execution systems (MES) that refine manual processes where human error has the potential to occur. It visualizes operator guidance via standard operating procedures displayed in headgear and the application of video-based verification of manually executed steps.

Advantages of a blended learning approach

VR and online eLearning are just two digital parts of a blended learning approach, which taken as whole can provide an effective training strategy whose advantages are:

• Reduced costs — industry studies have shown that VR training costs less and reduces training time when compared to traditional methods.
• Increased effectiveness and reduced time to competency — when training operators new to a role, industry studies have shown that VR training provides faster and more effective onboarding.
• Consistency of output — ensures batch quality and yield by streamlining training and re-training making people and process more robust (increasing right first time).
• Flexibly on demand — digital training offerings add flexibility especially when access to commercial scale equipment, particularly equipment located in a cleanroom setting, is severely limited for training purposes (needed in production). Conditions within a cleanroom may also be limited with respect to training needs (number of people allowed/ability to bring in training materials).
• Engagement and job satisfaction — multiple ways to train operators increase engagement and job satisfaction which is equated with staff retention and an experienced base of operators.
• Improved sustainability — the greatest reduction in environmental impact must be equated to highly trained operators making fewer errors and minimizing the need to repeat manufacturing processes. However, reducing the amount of processing consumables needed when face-to-face learning is the only training tool and reducing travel to facilities by trainers are also positive factors.

Conclusion

We live in an age that recognizes the uniqueness and individuality of people. However, we work in an industry defined by the demands of quality and accuracy, that requires us to consistently run manufacturing processes in the same way, time after time. Human error cannot be entirely removed. However, with our understanding of modern adult learning techniques, cutting edge technology and digital creativity, we are best placed to train our operators in an engaging way that has the added advantage of saving both time and money. For a culture defined by experiential and digital offerings, a blended learning approach offers a solution to reduce operator error in the drug manufacturing process.

References

1. 2023 20th Annual Report and Survey of Biopharmaceutical Manufacturing Capacity and Production, BioPlan Associates, Inc., 2023.
2. Cytiva internal survey report.
3. Eylath A. Is human error the cause or outcome of GMP process deviations. Bioprocess Online. bioprocessonline. com/doc/is-human-error-the-cause-or-the-outcomeof-gmp-deviations-0001. Published 6 October 2023. Accessed 20 March 2024.
4. May C. Top 10 GMP Audit Citations: FDA and TGA Inspections. onlinegmptraining.com/top-10-gmp-auditfailure-reasons-fda-tga/. Published 6 September 2022. Accessed 27 March 2024.
5. Starting Strong: A Guide to Pre-Service Training 1998, MOSAICA, AmeriCorps National Provider.
6. Ho N, Wong PM, Chua M. et al. Virtual reality training for assembly of hybrid medical devices. Multimed Tools Appl, 2018;77;30651-30682. doi.org/10.1007/s11042-018- 6216-x
7. PwC. What does virtual reality and the metaverse mean for training? pwc.com/us/en/tech-effect/emerging-tech/ virtual-reality-study.html Published 15 September 2002. Accessed 20 March 2024.
8. Adkins M, Alford G, Coombs S. et al. BioPhorum Technology Roadmapping: Digital Technology Roadmap. 8 July 2022. doi.org/10.46220/2022TR004