Westgard Rules!
Sitting Down With… James Westgard, Emeritus Professor in Pathology and Laboratory Medicine at the University of Wisconsin Medical School and President of Westgard QC, Inc., USA.
You are one of the world’s most recognized experts in laboratory QA/QC; what inspired you to follow this career path?
I majored in chemistry in college, in part because my father had studied it. I was then inspired to pursue a graduate degree by a couple of professors who were outstanding teachers. Teachers have been critical in my personal development and I don’t think they are properly appreciated at all levels of education. The actual decision to work in laboratory medicine was somewhat fortuitous! I finished my graduate studies mid-year while my wife was under contract to teach elementary school for the whole year. I had a couple of colleagues who worked in the medical school and wanted some help to implement new automated analytic systems. Having been trained in analytical chemistry and performed graduate studies on automated continuous flow systems, it was a good fit. The position at the University of Wisconsin turned out to be the job of a lifetime!
And your interest was piqued during your first project evaluating the performance of a new multi-test analyzer?
That’s right. In studying the scientific literature on evaluation of methods, I found that the statistical results and decisions on acceptability of performance didn’t make sense; decisions often depended on correlation statistics, rather than the size of errors occurring and their effect on the use and interpretation of test results. That led us to propose new criteria for judging the acceptability of performance, including the introduction of the concept of Total Analytical Error. That experience helped me understand the importance of employing proper tools and techniques to measure and manage quality.
Westgard Rules are internationally recognized. Why do you feel that you have been so successful?
Westgard Rules have a theoretic rationale, but they were also evaluated for practicality by the medical technologists in our clinical chemistry laboratory, who found them to be a logical and fit with their thoughts about inspecting QC results. It was just that no one had formalized those thoughts as “rules” and justified and approved their use in the laboratory. We did a lot of education and training over the years, documented the practice in the scientific literature and laboratory textbooks, expanded the applications to tests outside of clinical chemistry, and then further to laboratories outside the US.
Analytical error is still unacceptably high. Why, and what needs to change?
Laboratories need to pay attention to the quality of all phases of the Total Testing Process. There has been a tremendous amount of work on preanalytic errors in the last 20 years and those error rates have been significantly reduced. Analytical errors actually have more serious impact on patient care and I worry now that laboratories continue to assume analytical errors are not as important as preanalytic errors. Our studies to characterize the quality of processes using sigma metrics are now showing that preanalytic processes are often as good or better than analytic processes. On a scale of 3 to 6, where 3-sigma represents the minimum acceptable quality for production and 6-sigma world class quality, we observe that preanalytic processes have often improved to better than 4-sigma, whereas there are some critical analytic tests, such as HbA1c, that often operate at 3-sigma or less.
What advice would you give to laboratory professionals who feel that things could be done better?
Quality of service and quality of work life should go hand-in-hand. Everyone has a responsibility to contribute to providing the services they would want to receive for themselves, for their family, for everyone. This often means additional training, but anyone who is committed to quality can find ways to make improvements, starting with their own work processes.
If you could start your career over again, would you do anything differently?
I would probably include some formal study of statistics. On the other hand, I have sometimes found that my lack of it has been an advantage, because I start by focusing on the problem and then work with data to understand what statistical tools are useful. Statisticians often have their favorite tools and techniques and attempt to define the apparent problem so that it fits the statistics, rather than solve the real problem. I’ve been fortunate to work on some interesting problems and develop practical solutions that others have also found useful.
What is your most satisfying achievement?
I would definitely say my success as a teacher. I see myself first as a teacher, second as a researcher, and third as a laboratory analyst or service provider. I’m also incredibly proud of my son Sten, who developed and maintains the www.westgard.com website, publishes our books and training materials, and continues our teaching and training around the world. Together we have been able to help laboratories develop a better understanding of quality and provide tools and techniques to improve quality management.
Is there anything that you would still like to achieve before you can say… “yes, I’m happy now”?
I am actually happy and have been for many years. But, it would be nice to see the term Total Analytic Error and its definition appear in the official international vocabulary of metrology (VIM). That would be heaven!
