A conversation with John Wilkes (AstraZeneca), Clifford Berry (Takeda), Amy D. Wilson, Ph.D. (Biogen), and Jim Morris (NSF Health Sciences)
This article is the first part of a two-part roundtable Q&A focused on human performance in pharmaceutical operations. Part 1 discusses key drivers for human performance improvement, compares lean manufacturing and human performance programs, and provides perspectives on human performance in the context of the rapid scale-up and production of COVID-19 therapeutics and vaccines. Part 2 reviews human performance in the context of company investigation and CAPA programs. The four participants in the Q&A have significant combined experience in manufacturing operations and human and organizational performance:
- John Wilkes is the human performance lead for biologics at AstraZeneca. He has more than 25 years of experience in industry, with experience in manufacturing operations, operational excellence, quality systems, and quality control.
- Clifford Berry is the head of business excellence for Takeda at its Massachusetts Biologics Operations site. He has been a human and organizational performance practitioner since 1999, with experiences in commercial nuclear electrical generation, electric transmission and distribution, and biopharma.
- Amy D. Wilson, Ph.D., is the global human performance lead for Biogen. She has more than 20 years of experience in biopharma manufacturing, with focuses on human and organizational performance, operational excellence, risk management, and technical training.
- Jim Morris, executive director at NSF Health Sciences, with over 30 years' pharmaceutical operations experience in quality and manufacturing, is often leading consulting and training projects in investigation and CAPA management.
Although the responses to each of the questions were provided independently, common themes stand out — offering insight into how forward-thinking organizations view human performance and treat "human error" in the workplace.
What underpins human performance improvement in pharmaceutical operations?
John Wilkes, AstraZeneca: Underpinning human performance improvement in pharmaceutical operations begins with the organization's mindset toward error. Science has shown that error is a normal performance attribute of human behavior, occurring not only in work with failed outcomes but also, and most importantly, in work with successful outcomes. Shifting the organization's mindset from "error" as the exceptional cause of failure to a mindset that it is normal and not causal is challenging. However, the effect of this shift is profound. It moves the focus of improvement from fixing the "errors" of workers toward improving the effectiveness of the management of performance risk and operational learning.
Jim Morris, NSF Health Sciences: What underpins human performance in an organization is, in my view, open communication across levels and between functions. In other words, an open reporting environment where employees are encouraged to report error traps and are empowered to take action will lead to long-term improvement. Co-worker relationships based on trust will encourage reporting and a collaborative approach to problem solving.
Amy D. Wilson, Ph.D., Biogen: Human performance at Biogen is primarily characterized by a recognition of people as necessary drivers of success rather than as a variable that creates failure through error. This perspective drives us to ask different questions when we seek to understand both how failure occurs and how success is achieved, which ensures our improvement efforts are focused in areas that will actually improve overall system performance. The underpinnings of human performance in pharmaceutical operations are principles that include a recognition that people are fallible but come to work to do a good job, that systems and context drive behaviors, that systems are complex and imperfect, that our response to failure matters, that learning is vital and blame fixes nothing, and that reliability is the presence of controls and not an absence of accidents or deviations.
Clifford Berry, Takeda: From my perspective, the underpinnings always begin with the principles – those desired shared basic underlying assumptions that shape espoused values and beliefs and ultimately guide how we design, execute, and learn from work. Those principles could be a version of Todd Conklin's five principles, the principles that originated with Tony Muschara at the Institute for Nuclear Power Operations, or Takeda's seven human and organizational performance principles. You need some variant that works for your organization that includes the concepts that people have good intentions, error is normal, systems are complex and fail, how we respond to failure matters, and blame fixes nothing.
Takeaways For Improvement:
- Shift the mindset from "fixing" errors to reducing performance risk and increasing operational learning.
- Create a safe environment for reporting near misses and possible error traps.
- Shift the mindset to recognizing that systems and context drive behaviors and our response to failure matters.
- Establish shared principles to guide how we design, execute, and learn from work.
What changes do you feel bring about the greatest improvement in human performance on the plant floor?
Berry: Make time for conversations about risk before work begins. This dialogue could be informal or more formal, like a pre-job brief. Discuss the human actions that could trigger immediate, irreversible, and intolerable failure. Talk about where controls to prevent harm to people or product may be insufficient and how that risk will be managed. One way to create positive control during work where there is risk of failure is by staying in role for the natural duration of the task – avoid swapping performer and verifier roles. Lastly, when the task ends, talk about the work – how it went well, surprises and difficulties encountered, and share the learning with leaders.
Wilkes: Operational learning – improving human performance – means strengthening the understanding of performance risk, improving the risk competence of workers, and establishing defenses that effectively manage the consequence of human error where it matters most. The key action is to improve the organization's ability to receive feedback from operations. This feedback is the fuel for improvement. Acting on this feedback consistently and with urgency defines learning and continuous improvement.
Morris: Pre- and post-action reviews of activity at the unit level will bring about a significant improvement in human performance. Those organizations that institutionalize practices that prepare people to execute work and embed after-action reviews will see a significant improvement in human performance. Subject matter experts must be involved in these reviews. And, organizations that prepare SMEs as trainers and mentors will indirectly ensure these reviews are effective. Institutionalizing these practices is hard work, requiring discipline and leadership support.
Wilson: First, I think it's important to equate human performance with overall performance. Human performance is not separate. If one makes it easier to "get it right" and harder to "get it wrong," you will see overall improvement in success rates and operational metric performance. In terms of what drives significant improvement, we have seen the greatest impact from: (1) implementing formal risk management practices on the plant floor, such as pre-job briefs and three-part communication, (2) improving the way we learn from deviations, and (3) improving the way we learn in general from operations through learning teams, after-action reviews, post-job reviews, and work observations.
Takeaways For Improvement:
- Make time for conversations about risk before work begins.
- Improve the organization's ability to receive and act on feedback from operations.
- Develop subject matter experts to serve as effective trainers / mentors.
- Focus on operational learning, considering pre-job briefs, after-action reviews, learning teams.
The themes within human performance, such as work observation and speak-up culture, seem similar to those of typical operational excellence or lean manufacturing programs. Is human performance essentially the same thing?
Wilson: Human performance is absolutely not the same as typical operational excellence or lean manufacturing programs. One difference is in how they are focused. Many operational excellence or lean programs emphasize cost savings and efficiency gains. This is often in contrast with a human performance focus, which takes an integrated systems approach to strengthen risk management and enable resilience. It should be recognized that this difference in focus stems from fundamental differences in their basis in principle. Lean and operational excellence efforts that I have experienced are typically based on a position that standardizing work and controlling through management will yield better results. Often, results are defined in lagging ways, such as measures of timeliness. Human performance, however, integrates views from social and safety sciences and places emphasis instead on understanding more about where adaptation is required and where variability becomes difficult to manage. Generally, it will be the case that operational excellence and lean programs will co-exist and need to be complementary to human performance. To do this well, recognize where their associated methodologies or tools should be applied. If you have a need to stabilize and define a work process, eliminate waiting between steps in a process, or look for waste reduction opportunities across a supply chain, then operational excellence and lean methodologies will likely yield great results. If, however, you think you have a "right-first-time," success rate, or "human error" problem, you will have greater success applying approaches based on human performance principles.
Wilkes: Certainly, they are not the same thing. I would begin by reframing "operational excellence." In my opinion, operational excellence should be viewed as being inclusive of any number of integrated operating philosophies applied to deliver the goals and objectives of the organization. It is important to recognize that the value propositions of lean manufacturing programs and human performance are different. At its core, lean manufacturing is about maximizing value for the customer through efficiency gains and waste elimination at a process level. Human performance is focused on building capacity and resilience at a systems level with the intention of increasing the frequency of successful performance outcomes. This does not make them incompatible, and any attempt to choose between them is a false choice. Rather, the goal of the organization should be to integrate these approaches in the spirit of reframed operational excellence – using lean approaches to eliminate waste where the system can benefit from efficiency and using human performance approaches where the system requires capacity and essential controls to effectively manage performance risk. The key consideration will be who makes the decision about what is considered "waste." Waste in the eyes of a lean practitioner may be a practice to build capacity or an essential control in the eyes of a human performance practitioner.
Berry: Lean manufacturing, or operational excellence if you like, has an emphasis on process issues that relate to production quality control and efficiency. Human performance, or human factors and system safety, considers overall system performance in order to improve the likelihood of success and includes humans as part of the system. A potentially unintended result of lean manufacturing removing what is perceived to be waste in processes is systems becoming more brittle, where human performance recognizes slack, margin, and capacity are essential to system resilience. An example of lean creating brittleness on a global scale was the breakdown of supply chains during the COVID-19 pandemic. Just as the emphasis of each are similar but different, so too are the methods used. For example, while the use of 5-Whys in lean manufacturing may be satisfactory to whittle away at solving a production quality issue through repeated iterations of PDCA (plan, do, check, act), the application of system mental models and related methods, such as event and causal factor (E&CF) charting, by human performance is more likely to lead to learning about multiple systemic contributors, interactions between contributors, and emergent system properties that can be adjusted to prevent event recurrence on the first pass. On the topic of speak-up culture, pharma has had good intentions and limited success due to being overly reliant on telling people to speak up to report problems, rather than taking the time to learn about the downward pressure that may inhibit reporting, as well as changing how leaders respond to failure in order to create an environment where people feel safe to be candid about risks and work adaptations.
Morris: There is clearly overlap; however, human performance is about reliability of work performed by people, whereas lean management is focused on efficiency, waste reduction, and eliminating non-value-added activity. I would recommend training the operations excellence and lean management experts on the principles of human performance in order to put greater weight on the interface of people in the context of work and more focus on what it takes to set people up for successful work outcomes.
Takeaways For Improvement:
- Human performance recognizes slack, margin, and capacity are essential to system resilience.
- Operational excellence and lean programs will co-exist and need to be complementary to human performance.
- Organizations should reframe operational excellence by integrating lean manufacturing and human performance.
- The methods/tools are different and warrant training to ensure practitioners are effective in both arenas.
COVID-19 related production demands are unique, and some companies are facing unprecedented scale-up challenges under extremely tight timelines. What are the likely human performance challenges that such companies will be facing and what advice can you offer to proactively confront these challenges?
Morris: Manufacturers will be managing multi-shift operations and, in some cases, running around the clock 24x7. They will most likely hire temporary workers and transfer workers from other sites. The products they are scaling up are new and technically difficult to manufacture. To manage human performance in the "perfect storm" of challenges requires fast tracking a lot of activity and a tremendous level of risk-based thinking. At NSF, we would advise recognizing the unique role of every employee and especially preparing the subject matter experts on each shift/each function. Every day must be considered "game day" for the entire staff, where the routine of PEC – prepare, execute, check – is in the mindset of every employee. This requires management support not only in words but in actions – establishing expectation, opportunity, and a process for PEC. Distributing experts across the organization empowered to make in the moment decisions will help tremendously. Maintaining the discipline of "normal work" by continuing pre- and post-action reviews is recommended. These become even more important during scale-up activity and will help track error traps or key learnings that can be embedded in other operations beyond the COVID-19-related production. Finally, stay positive, since errors will occur and the response to an error often has greater impact than the error itself.
Berry: Scaling up under tight timelines will result in delivering equipment and processes that will set people up for failure if managers and engineers neglect to engage end users as partners during work design. This collaboration will improve the ergonomics related to the location of new equipment, enhance the helpfulness of automation prompts and feedback, and ensure human-centered design principles are leveraged during the creation of standard operating procedures (SOPs) and master batch records (MBRs). The people on the floor always help to complete the design and create safety when they execute work, so it makes sense to involve them earlier in the design of work. Involving end users in the design of work will reduce rework and workarounds and result in less of a difference between work as imagined and work as done. This will help to create more operational success and support the mission of delivering life-changing therapies to patients.
Wilson: There are many human performance challenges in our current COVID-19 environment. To increase the likelihood of success, I would suggest continuing to provide focus on critical controls in operations. This is the time to ensure you have a good understanding of where you have risk and what controls need to be reinforced to mitigate that risk. Many production facilities are maintaining only essential personnel physically on-site. I would suggest augmenting that with lead/supervisory presence during the most critical operations. This helps especially in those areas where one relies solely on administrative (human based) controls to prevent undesired outcomes. How do you know where these exist? You can look for situations where the only thing preventing a product loss is that a person performs step 4.2 in a procedure correctly, or, situations where if an automation prompt is answered out of sequence, a sterile boundary might be breached, or, operations where the only thing preventing chemical exposure injury is proper PPE. An even easier way to find where risk is greatest is to ask your workers. I would also recommend doubling down on operational learning. Do more after-action reviews and get people together to ensure learning both when things go wrong and when things go well. Especially in our current unprecedented, dynamic environment, it is through these learning discussions that people can determine where to focus efforts and energy toward success with production demands and scale-up.
Wilkes: It is easy to foresee that the gap between work as imagined and work as done being significant during these times. Operational learning is the key for closing this gap. Now is the time to emphasize the practices of operational learning and rapid improvement. Proactive operational learning processes, such as after-action reviews, post-job discussion, and work observations, create opportunities for feedback from the frontline workers to develop an understanding of the challenges of work as done. Utilize this feedback to improve the effectiveness of critical controls where performance error is consequential. Be intentional about operational learning. Create time and space for it – workers know the performance challenges and they know where they must be at their best to get successful work outcomes – that is the goal of human performance.
Takeaways For Improvement:
- Errors will occur and it is the response to error that matters more than the error itself.
- Engage end users as partners in work design to reduce rework and workarounds.
- Augment operations with lead/supervisory presence during the most critical operations.
- Create time and space for operational learning and opportunities for feedback from frontline workers.
The responses provided by these subject matter experts offer points to consider as you reflect on what your company is doing in the area of human performance improvement and, in particular, at this time where COVID-19 adds stress to the work environment. Developing mechanisms to facilitate operational learning is a key theme and opportunity area for most unit operations. In Part 2 of this Q&A, the experts look at human performance as it relates to conducting investigations and identifying meaningful corrective and preventive actions.