Issue link: https://www.e-digitaleditions.com/i/1192284
8 Pharmaceutical Technology REGULATORY SOURCEBOOK DECEMBER 2019 P h a r mTe c h . c o m But, as more companies and managers adopt for- mal gemba walk programs, says Chen Ooi, there is a real need to measure their effectiveness. As she notes, it's not a straightforward science. "It's all about behavior and reactions, and you don't want to blame. You need to build trust with peo- ple on the f loor so that they feel comfortable rais- ing issues and provide truthful feedback instead of telling you what they believe you want to hear," she says. Body language and tone of voice are ex- tremely important, she says, and managers need to understand what goes into creating what Har- vard professor Amy Edmondson has described as a "safe" environment where people can be honest (11). It might be beneficial, says Chen Ooi, for execu- tives to have a better understanding of science and engineering concepts in order to communicate more effectively with the people doing the work. But that understanding needs to f low from the bottom up, too. "Leaders in process development, supply chain, and manufacturing also need to be able to translate the very technical aspects of their operations into the more general language of busi- ness," she says. At the most fundamental level, there is a real need to understand the cost of poor quality in order to weigh potential losses (e.g., of a warning letter or consent decree) against the costs of hiring more quality control staff and investing in more modern analytics or information technology (IT). Georgetown University professor Jeffrey Macher has found that most pharmaceutical companies fail to measure or track the cost of quality within their organizations (12). To be fair, says Chen Ooi, it can be difficult to quantify the cost of poor quality in dollar values. "It's really about continuous improvement and pre- venting issues," she says. At many pharmaceutical companies today, basic metrics are tracked, such as successful batch release rate, rate of invalid or out-of-specification (OOS) results, and inspection findings. But these are lagging metrics, she says, and only available after something has gone wrong. Amgen is shifting to a "predict-and-prevent" focus and using artificial intelligence and data vi- sualization to leverage more of the data gathered within the current good manufacturing practice (cGMP) environment that typically remains un- used, says Chen Ooi. Using open-source code, available at very low cost, the company has devel- oped a tool for deviation trending inhouse that can be used to uncover systemic problems, she says. The alternative until now was to have people review hundreds of issues, a painstakingly slow process. Better training needed If the industry is to sharpen its focus on quality metrics and culture, university training should in- clude more industrial engineering-type courses to prepare students to understand and take charge of quality initiatives, says Chen Ooi. "There's a need to educate students on the concept of quality be- yond compliance so that they realize the value and business benefits of quality," she says. "In pharma, every mistake or OOS situation requires a full- blown investigation, which requires significant time and resources," she says. Forward thinking universities are transform- ing traditional curricula. "We try to inculcate experiential learning into educational programs," says Ajaz Hussain, director of the National In- stitute for Pharmaceutical Technology and Edu- cation (NIPTE), a consortium of 17 universities that include Rutgers, Purdue, Duquesne, and the Quality Compliance