Reducing Analytical Cycle Time Through Six Sigma Manufacturing Theory Implementation

"We Have to Ship Tomorrow!  Why Does It Take So Long to Test?"

These are just a few of the dreaded words that every lab manager will hear in the course of his or her day-today operations. In the past, my response was either to enjoy the bliss of feigned ignorance or to shift responsibility to the company management sectors that oversaw manufacturing scheduling operations. However, in the spirit of teamwork and professional progress, I began to consider the possibility of improving the cycle times in my lab while maintaining or perhaps even raising the already high level of testing quality. So I decided to cautiously go along with a joint educational improvement project proposed by our top management and a business professor from a local university. I thought that inviting the professor and a group of his students to learn and work in our lab environment would not only be a source of intrinsically motivated, no-cost labor, but would enhance company relations with the local academic community.

The first step in our improvement model to identify sources of waste was to identify the process. Each step in the testing process was analyzed to determine loading versus capacity and to identify opportunities to reduce variation and waste within the testing process.

What has happened, however, has not only radically restructured our traditional laboratory testing environment, but has dramatically reduced our lab release cycle time from an average of 17 days to less than 5 days, resulting in significant monetary savings, happier customers, and reduced overtime for lab personnel.

Additionally, testing quality first pass yields have increased from 97.8 to 99.99 percent. This article will briefly elaborate on the conceptual management paradigm shifts required of both the leadership and personnel team as well as outline some of the procedural changes that helped streamline the process while still maintaining a high level of GMP compliance.

Although I had agreed to support this local educational adventure, I was initially fairly skeptical about its outcome. The very suggestion that, after 25 years of lab management, I needed outside input from university students and a business professor to solve the issue of slow cycle times in our pharmaceutical testing facility was difficult to accept. It seemed improbable that a group of outsiders, who did not have the technical chemistry experience or knowledge of the lab procedures, would be able to provide a realistic solution. As a lab scientist, however, I realized it was best to allow the data to speak for itself before making a final decision. Thus, I decided to be open to learning some new concepts outside of my realm of expertise, and conveyed to upper-level management my somewhat contrived enthusiasm for the project. Convincing the rest of the lab team, however, was going to take some time and a plethora of supportive data. Furthermore, I realized that the lines of communication and management initiatives coming from my office needed to produce greater team unity without threatening the roles and importance of any of our analysts.

Japanese Toyota executive Taiichi Ohno originally introduced the term “muda” to convey removal of waste. Muda can be defined as types of defective work, not just products. It includes wasted time, motion, and materials (Pyzdek 705). The first step in our improvement model to identify sources of waste was to identify the process. We proceeded by beginning to map our current lab processes. The students, having been classroom trained in process mapping, shadowed our analysts to determine the amount of time that specific lab tasks took and to obtain critical cycle time data. Each step in the testing process was analyzed to determine loading versus capacity and to identify opportunities to reduce variation and waste within the testing process. Our team of students and lab personnel was able to identify the easiest, most productive changes and targeted those for immediate improvement.

Analysis of the data revealed a bottleneck within our instrument lab analysis area.

Prior to work cell, dance chart reveals excessive analyst movement . Analysts were walking ~780 feet per sample.

Initially our staff felt that this was the data we needed to support the allocation of additional personnel resources, which would quickly solve the problem. The obvious solution: We needed more analysts! The request was denied. Site management believed cycle time improvements could be achieved through process/procedural changes that did not require additional staffing. What initially appeared to be difficult policy, however, actually opened doors for a major paradigm shift.

During a brainstorming session with the university students and Professor DeLoach, we came upon the idea of developing a position that we would call the flex analyst. The flex analyst would be a qualified lab analyst trained to “flex” into the instrument lab, set up basic chromatography analysis on HPLCs, and then return to the main lab. In other words, the task of the individual analyst would alter, or “flex,” based on the load in what was called the work cell—another manufacturing concept. We realized that what we primarily needed was an individual geared for specific tasks, and not necessarily someone experienced and knowledgeable about the full spectrum of HPLC theory. As a lab manager and chemist, the idea made me feel a bit uncomfortable, as anything less than a fully trained instrument chemist would increase the opportunity for errors and rework. Yet it was successful. Training our current lab analysts to perform basic process HPLC analysis allowed us to begin balancing our lab workload and enabled samples to move more smoothly through our processes.

The flex analyst concept was not a quick-fix scheme and was never intended to solve all aspects of our analytical procedures. The success of the flex analyst innovation was contingent upon and integrated with our willingness to adopt a broad array of Six Sigma manufacturing concepts and apply them to the entire lab environment, not just to one area. One of these concepts is called the 5S. This, coupled with other efficiency improvement concepts, has greatly reduced many forms of wasted motion and effort by our analysts.

HPLC column storage prior to implementing the 5S process

The 5S concept simply refers to five basic manufacturing fundamentals that also have their origin in Japanese manufacturing theory, and which begin with the letter “S.” The terms are as follows: sort seiri, set in order seiton, shine seiso, standardize seiketsu, and sustain shitsuke. Briefly explained, “sort” refers to determining what is necessary to do the job. “Setting in order” is just what it says—putting items in their proper place for easy retrieval and use. “Shine” refers to keeping a workspace clean and free of clutter. “Standardize” is the procedure for keeping the process in the order that has been established. And finally, “sustain” means creating maintenance habits for the work environment (Pyzdek 715). The concepts themselves were certainly not “rocket science” for our lab analysts, yet their revolutionary simplicity has a proven track record in manufacturing. Proper implementation, however, was made possible only by utilizing our cross-functional team and being open to frank and honest input from the consultant and his team.

It was in “sorting” and “setting in order” that design of the work cell, as mentioned earlier, came into play. Instead of every analyst taking each individual product from “cradle to grave,” the lab was rearranged into work cells so as to promote continuous flow and improve FIFO (first in, first out). Routine tasks (e.g., HPLC mobile phases and standard preparations) were designated as suppliers to the main work cell configuration. Orchestrated tasks and testing moved from process to process with improved transitions and more clearly defined areas of responsibility instead of each analyst performing all testing on a sample. Then the lab was “set in order”: We placed all equipment and machines strategically within the flow of the work cell—not according to the personal preferences of the various analysts.

All equipment and supplies were strategically placed within the flow of the work cell. Then the lab was “set in order” and the work cell implemented. Analysts now walk ~ 240 feet per sample saving over 1500 miles of walking per year.

The processes of “sorting” and “setting in order” went fairly smoothly. “Standardizing” was a different kind of challenge. Initial rebellion showed up in the work cell prototype arrangement. Pieces of equipment that had been designated to specific places within the work cell prototype began to mysteriously disappear and reappear in their former locations around the lab. It was clear that the entire group had not “bought into” the work cell concept. So, taking a step back, I encouraged the analysts to decide within the work cell parameters where they wanted some of the equipment placed. I realized how important it was that they not only feel empowered, but actually be empowered, as they participated in the restructuring process. Soon, equipment items were being placed back in the most strategic and appropriate areas according to the work cell design.

New HPLC column storage system allows for column selection within seconds.

Another issue in the standardization phase was making sure that work flow was not interrupted between lab team shifts. Since our factory operations work three shifts, six days a week (with a brief pause on the weekends), there was a need to make sure that the various lab teams entering and exiting were exchanging relevant information. With this in mind, we established a shift changeover meeting in order to establish priorities and specific assignments so that we could effectively transition from one shift to another. Thus, communication between the shifts skyrocketed to new levels of collaboration. Additionally, in order to maintain dialogue and focus, we integrated reinforcement of work cell concepts into the content of our weekly lab team meetings so that the laboratory staff might continue to increase their knowledge and comfort level of implementing the various process improvement ideas. We discussed “what’s working” and “what’s not,” allowing the analysts to express their ideas and find solutions to the “what’s not working” list. Although not the main agenda of those meetings, the regular repetition of these concepts strengthened the overall paradigmatic shifts in the lab.

The initial lab team response revealed a resistance to change. At this point, my lab manager role needed to expand from that of technical specialist to team leader in order to rally the crew behind the project. The idea of the work cell had produced the most resistance. Some analysts felt that this mirrored an assembly line. Other members of the team were incredulous—just as I had been at first—that a team of students, first shift lab personnel, and a business professor without any actual knowledge of chemical and analytical processes could tell other shifts and the rest of the lab how to do their job better and more efficiently. The older, longertenured staff members were most reluctant to embrace the concepts. It was the comment of a younger lab analyst, however, that finally encouraged the team to try the new methods: She pointed out that we were all paid the same regardless of the lab process being used. Additionally, it might even free the staff up from working excessive overtime.

The most difficult leg of the journey to the Efficient Lab has not, however, been the conceptual and procedural shifts. It has been in learning a fundamental lesson about the nature of leadership itself. I have realized that there are times when the lab manager must stand alone, maintain a commitment to a chosen path of action, and be patient while colleagues develop their own understanding and conviction about new processes. Lab management must move beyond overseeing procedures and work toward building a true team approach to innovation and implementation. Overall, the results of the project have been extremely gratifying; not only in the tremendous savings yielded from decreased cycle time, but also in the feeling of success created when a group of hardworking individuals, each gifted and talented in his or her own scientific field, unites to accomplish a challenging goal together.

References

Pyzdek, Thomas. The Six Sigma Handbook. Rev. ed. New York: McGraw-Hill, 2003.

More information regarding Lean manufacturing concepts in the pharmaceutical laboratory environment can be obtained by contacting Mr. Guy DeLoach, Lean Advantage Consulting Group, at gdeloach@skydancercorp.com or gdeloach@leeuniversity.edu.