Three-Step Quality Network Transformation: A Case Study

Feb. 21, 2012
Large-scale benchmarking and local intervention allowed one manufacturer to diagnose network performance in three months.

Pressure on pharmaceutical quality functions is increasing, from all sides. Regulators have bigger budgets and a more aggressive approach to inspection and enforcement. The direct financial impact of quality issues can run into the tens or hundreds of millions of dollars, but pharma companies can’t afford to tackle theses issues by throwing huge amounts of money at problems. Faced with a shortage of high-value new products, spiraling R&D expenditure and aggressive competition from low-cost market entrants, they must learn to do much more with much less, calling for a step-change in efficiency from quality assurance functions.

But even unearthing the key issues and opportunities for performance improvement can be difficult across a global network of plants.

When the FDA demanded that it shut down one plant and issued warning letters to several others, managers at one global pharmaceutical manufacturer knew they had to define innovative ways to uncover the root cause of their quality issues. High-level benchmarking could give an overview of the performance of different plants, but would do little to indicate where the company should take action to improve performance. While individual quality processes are relatively simple, the overall effectiveness of a quality system relies on complex interactions between hundreds of separate actions, and a considerable amount of human judgment. The failure in a single process or a single interface between processes can have profound consequences.
Detailed investigations of every individual plant, on the other hand, would identify action points but would take too long. Instead, the company adopted a hybrid approach, combining the breadth of global benchmarking and the application of modern risk management tools with the depth of focused diagnostic assessments in selected sites.

First Step: Taking the Global High-Level View

The first step was building a high-level picture of the fitness of the company’s quality function, using benchmarks to compare its current performance to that of industry peers, and to identify the variation of performance between its own plants.

Out of 13 key quality activities investigated, the company measured the effectiveness of quality activities by examining the recurrence rates of issues, indicating a failure to get to the root cause of issues. It studied efficiency by looking at the number of worker-days devoted to each quality issue.

Some plants had such inefficient processes that a single quality deviation would absorb considerable resources, while others devoted very little to each problem, suggesting a tendency to go for fast fixes instead of attempting root cause analyses.

Beyond the specific benchmarks, this effort also revealed some important basic issues. Another important aspect of the high-level view is the risk evaluation—understanding the relationship between current performance and potential future quality risks.

Second Step: Focused Assessment of Selected Sites

For deeper analysis, the managers selected three sites which stood out from the others, either due to a particularly high value at risk, strategic importance, or because they were recognized as having exceptionally good quality performance. It dispatched an evaluation team to investigate those sites. The 15-strong team included senior quality managers from all of the company sites worldwide, together with specialists from the local and central quality functions to provide assistance collecting and analyzing data.

Because the members of the evaluation team had the results of the benchmarking and risk analyses, they could focus their time and attention on the processes in the plants that gave the most cause for concern, or on those that seemed to perform much better than those elsewhere (Figure).

At each plant, the team worked with staff on the ground to map the processes, identify the root causes of issues, streamline and standardize. Critically, because the team included quality personnel from many other plants, they could share their own experience of tackling similar problems elsewhere, and when the team discovered something new, they could see immediately if the issue was specific to the plant in question or more generally applicable.In many cases, the teams identified situations where current procedures both increased costs and risk and reduced quality. For example, as many as four different people could be responsible for checking the settings on a particular machine after tooling changeovers. Not only did waiting for four people to complete the checks add cost and delay, but no individual felt truly responsible for the state of the machines, so incorrect settings were relatively common.Where an established solution to an issue was already in use elsewhere, the team would encourage the local plant to adopt that approach. Where no such solution existed, the team brainstormed with local staff to find one. They then established pilot projects to evaluate the effectiveness of their proposals.In the changeover case above, for example, best performing plants relied on a single operator to complete machine checks, ensuring quality with tool designs that made changes more error proof, and a plant culture that encouraged operators to take responsibility for the output of their cells, instead of relying on external process quality checks.The team didn’t just focus on problems however. It also looked at areas of the plants that benchmarking suggested were performing particularly well. Again, team members were able to see very quickly whether local good practice might have wider application. One plant, for example, had managed to embed an impressive continuous improvement culture: shop floor teams were constantly identifying opportunities to improve quality and productivity in their cells, and used a formal process to ensure ideas were tested and implemented.Third Step: Sustaining the ChangeThe evaluation team’s work had put in place the foundations of a fundamental improvement in the company’s quality performance. Comprehensive quality scorecards and metrics were developed from top-to-bottom across the organization, including supply chain and third parties. The company also put in place a two-year quality transformation roadmap.When the company repeated its benchmarking and risk heat mapping exercise for a second time 12 months later, it found overall risk levels had dropped by 20 percent and efficiency improved by about 25 percent. Beyond this initial impact, however, the benchmarking and diagnostic effort had a profound effect on the culture of the company. Teams understood the value of experience sharing through workshops and shop floor visits. They recognized the power of finding and replicating solutions to quality problems developed elsewhere in the network, and the importance of mechanisms that would allow them to do so. Quality costs and metrics are more transparent, included in each scorecard and discussed during all performance reviews across the organization. Finally, quality teams placed a new emphasis on operator mindsets, training and coaching: while formal procedures provided an important safeguard,As this manufacturer learned, building a quality culture and enhancing problem solving skills not only improves quality but it can also have a huge impact on productivity by reducing downtime and rework.

About the Authors
Nicolas Esmaïl ([email protected]), Lorenzo Positano ([email protected] ) and Vanya Telpis ([email protected]) are leaders in McKinsey & Company’s Pharmaceutical Operations practice.

About the Author

Nicolas Esmaïl | Lorenzo Positano and Vanya Telpis