Customer Story

A leader in pain relief overcomes its own growing pains

At Johnson Matthey, JMP^® is the ‘active ingredient’ in process optimization

With an estimated one in five adults around the world now living with chronic or acute pain, the development of better long-term treatment regimens has become a true public health priority. Medical practitioners view integrated pain management as a central part of a patient’s overall recovery plan, with research linking shorter recovery times and improved quality of life outcomes with better managed pain.

Johnson Matthey, a global leader in science that enables a cleaner and healthier world, is now one of the world’s leading manufacturers of opiate alkaloids, a family of compounds that serve as the active ingredient in most pain medication formulations. With a 200-year history in controlled substance manufacturing, Johnson Matthey was among the first to develop the natural product extraction and chemical synthesis methods for opiates that remain at the forefront of the industry today.

Johnson Matthey’s active ingredients are now found in a wide range of pain therapies, from the tramadol or codeine prescribed to patients with rheumatoid arthritis to the hydromorphone that might be used on cancer wards or in palliative care. As scientists learn more about the benefits of managing pain – and how the risks associated with controlled substances can be safely minimized – research partners like Johnson Matthey are advancing science by innovating not only pharmaceutical ingredients but the processes by which they are made.

Resolving production challenges threatening a 'strategic growth area'

With global demand for pain medication on the rise, Johnson Matthey decided to grow its extractive operations. But the rapid scaling-up also threatened to introduce new challenges. Once the company’s newly expanded Edinburgh facility was up and running at a higher capacity, engineers observed that frequent processing issues were thwarting the increase in production that the project was designed to support.

“It was a key strategic growth area of the business,” explains Johnson Matthey’s Site Director David Payne. Failure to resolve these production problems “would have put in jeopardy a huge amount of work that had already been done.”

Even the best efforts of a dedicated and experienced workforce couldn’t ferret out the cause of issues that resulted in process variation and manufacturing chokes. “Every day or two the plant would have to be shut down and started up again,” Payne recalls. The effects were both financial and psychological. “When there are chokes or the equipment goes down, it can really begin to damage morale and cause frustration. Issues materialized in different places depending on how you were running the plant and the types of straw you were processing.”

Though Johnson Matthey’s raw material extraction processes have been running for many years, the company has not stopped working to improve operations. In fact, the drive to advance both laboratory and manufacturing science over its long history has become a core part of Johnson Matthey’s identity, with continuous improvement enabling the company to grow its already expansive portfolio of active pharmaceutical ingredients and controlled substances.

So when processing problems were getting in the way of yield in Edinburgh, Payne teamed up with Andrew Ruddick, co-founder of the UK consultancy Process Insight, a firm that helps companies like Johnson Matthey to improve operational efficiencies using Lean Six Sigma. “David brought along this project because it was a critical problem for the business,” says Ruddick. “He recognized it needed a fairly heavy dose of data analysis to figure out what was going on.”

Process improvements at the point of extraction help scale up efficiencies

Deriving opiate alkaloids from raw materials is a multistep process. Plant matter must be milled and formed into pellets before the resulting straw can be extracted in a solvent and subjected to a purification process that captures morphine.

The first step in improving efficiency at the extraction facility was to determine the optimal rate for feeding straw into the extractor. When Payne and Ruddick examined parameters they had assumed were responsible for the feed rate – weight changes in big siloes that stored the straw – the results were surprising. “It became apparent very quickly that those weight changes weren’t measuring what [we] thought they were measuring,” says Ruddick. “We would never have known that without JMP. It was some of the modeling and visualization tools in JMP that got [technicians at Johnson Matthey] to see that.”

Payne and Ruddick used JMP Profiler and other modeling features in the software to identify a more reliable set of variables, including the speed of the belt that feeds straw into the extractor and the height of the gate that controls the depth of raw materials on the belt.

JMP also helped adjust for the “almost infinite” variations among several types of plant varietals that fed the extractor, says Payne. “Because you’re dealing with a natural product, that product varies,” Payne explains. “And we didn’t really have the understanding of how to set the plant up to deal with different types of straw.”

Detective work helps optimize delivery systems

Payne’s team built a JMP database that included thousands of bits of information – growing and harvesting practices, extraction procedures, equipment performance and more. “It was easy to bring the data together for analysis in JMP using simple scripting,” said Ruddick. Then JMP modeling, profiling and what-if simulation tools did the detective work needed to determine a formula that optimized the delivery of straw to the extractor.

Next, Payne and Ruddick focused on the second step in the extraction process: separating the morphine from the solvent that washes it out of the straw. Too concentrated, and there is an increased risk of choking the equipment, causing a shutdown. Too diluted, and the equipment is underutilized and yield slumps. Again, JMP helped refine the process to reduce the variability of flow and stabilize the morphine concentration. “What we’re seeing on-site is the use of JMP expanding more and more to solve all kinds of problems. JMP is our chosen statistical analysis tool,” says Payne.

Throughout the analysis process, JMP captured results in interactive graphs and charts that let managers and operators alike see trends and patterns in the data and observe how changes to even a single variable affected the overall process. “It engaged the whole organization in problem solving by opening data analysis to everyone, not just statisticians,” Ruddick explains.

“You can show them a profile and they understand that. It was important within the organization to illustrate both what was going on and demonstrate how you could bring this back into control. It gave them a confidence that hadn’t been there before.”

Putting JMP^® to work on more complex problems

After seeing how optimization efforts led to real results, Payne has since deployed JMP to many of Johnson Matthey’s processing plants where predictive modeling helps plant operators to adhere to specific production parameters. One example where JMP has really made a difference, Payne says, is in helping to determine target specifications that allow plant operators to mill powdered products to a precise particle size.

“For a long time, we did some blind experimenting and a bit of hit-and-hope. Previously, we got away with it because you can re-mill particles that are too large. But you can’t make particles milled too small bigger again without reprocessing – and that can be very costly.”

Since the company first implemented Lean Six Sigma with JMP across the organization, Payne’s team now gathers a huge amount of data on different mill settings, screen sizes, speeds, hammer types and products. “With that data, we’re building up a picture – a model – for any starting particle size we might need,” he says. “We’re now able to set the mill up to hit the target size we’re looking for. Once you’ve built that model, almost anyone can plug the data in… In fact, the success of this [particle milling] project, and specifically using JMP within it, elevated the importance of JMP at this site. More and more projects will be using JMP to do this sort of analysis going forward.”

Among other things, Payne is now exploring ways to build out dashboards that will further this proactive approach. “We’ve undergone a slight restructuring in how we support our plants,” he says. “We now have process technical specialists for each product – Lean Six Sigma Black Belts, that is – who have helped us move from a reactive use of JMP to a proactive one. Dashboards will help us to monitor how processes are performing before complications arise.

“As we’ve grown, we’ve got people who are more comfortable with JMP, and because of that, we’re finding more complex problems and building more and more advanced analyses (to tackle those problems). JMP changes the way you collect data, clean data, how you organize the basic analyses and how you put it into a format that tells a story.”

Through science – and the science of statistics in particular – Johnson Matthey has overcome its processing pains and is continuing to evolve and improve with each new challenge. That means better pain medications and a better quality of life for millions of people around the world.