Tools, Technologies and Training for Healthcare Laboratories

Westgard Web: Why We Write

December 2006
with Sten Westgard, MS

After ten years of Westgard Web and more than 100 essays written, Dr. Westgard reflects on the reasons for writing for the Laboratory Professional audience on the Internet.

At my recent retirement celebration, my son Sten pointed out something to me that, up until then, I hadn't noticed: I have written more than 100 essays for “Westgard Web.”

Sten did more than count the number of essays. According to him, I’ve written nearly two hundred essays, lessons and articles for Westgard Web, which works out to approximately 170,000 words, or over 1,000 pages. This, in addition, to nine books printed on paper, which represent another 2,000 pages. All of this, mind you, is my published works just for Westgard QC, Inc., not the scientific publications, papers, and books that have preceded and paralleled the company work.

All of this points to the fact that the Westgard website is now ten years old. In internet years, that’s probably equivalent to fifty years. Back in 1996, I didn’t know what a website was, but when we were looking for a way to communicate with our customers and colleagues, Sten assured us this was the wave of the future. If Sten had told me that I would end up writing 170,000 words for this new technological marvel, I probably would have turned him down.

Today, Westgard Web gets more than 12,000 visitors a week. It’s not Google, MySpace, or YouTube, but if you consider that this is a website about statistical quality control for hospital laboratories, which is a limited audience if ever there was one, this is a huge number.

So after writing thousands of pages and hundreds of thousands of words, the question arises, “Why do you write?” It’s a question that I am often asked, and a question that both Sten and I sometimes ask ourselves.

Quality is an opportunity to advance laboratory science and scientists

Today, there are special problems for laboratory scientists because laboratories are often the target of cost cutting. Doing more and more with less and less is a way of life in laboratories today. In spite of our tremendous efforts to improve productivity, we are still left with low pay and little professional respect in the healthcare community. This has led to a serious shortage of laboratory scientists. Many already in the field are looking for more rewarding and better-paying jobs in business and industry. Recruitment of new students is difficult because they are drawn to more attractive, lucrative fields.

All of this causes productivity to be paramount. Quality becomes the neglected step-child.

Yet it is in the area of quality that laboratory scientists could make the biggest impact. Healthcare quality programs are almost always the province of nurses and hospital clerics, even though laboratory scientists have the best understanding of variation, data, and data analysis. Rather than having the present quality programs - which operate with the goal of accumulating stacks of paper for documentation - healthcare organizations could develop true quality management programs under the guidance of skilled laboratory scientists.

Therefore, we write to support you - the professional laboratory community - in your own commitments and efforts to improve quality in your laboratory, organization, and institution. Hopefully, we provide practical advice, tools that are useful for your own applications, lessons that will help you teach others about quality management, and occasionally some inspiration to continue with the hard work. You are important in improving quality in healthcare today!

Quality is still misunderstood in diagnostic testing

We write because too often, despite the Internet-fueled explosion of information available to the laboratory professional, despite the increasing emphasis on quality management in healthcare, quality is still not well understood and not well managed. Quality control is even less understood. Furthermore, QC is often considered too mundane to be of interest in our professional organizations and scientific journals. Despite the STARD initiative of years ago, (whose aim was to improve the quality of scientific studies on the evaluation of tests and technology on healthcare outcomes), quality control is still missing in many of the outcome studies that provide the basis for national testing and treatment guidelines.

Trace the recent history of hs-CRP and you see the marketing hype for this new test, even in supposedly scientific articles in highly regarded professional and academic journals. Part of the problem in this case was the lack of objective assessment of the quality required for the intended application and the performance specifications, particularly the precision and accuracy that are allowable for the measurement procedure and the QC that is needed to detect unstable method performance when the test is applied to the next individual patient. Studies may demonstrate a difference between the averages observed on different groups of patients, but the routine application of the test to the next individual patient is a different scientific problem, one not adequately addressed in many clinical studies.

Nevertheless, such clinical studies of new diagnostic tests create a demand for access to those tests in routine laboratory services. Manufacturers provide new methods and technology on automated systems to make those tests available, market the new technology to laboratories, and soon the tests are widely available and in use, regardless whether the quality of the testing methodology can be adequately controlled for the intended medical application under routine service conditions. In fact, the trend seems to be to marginalize the need for quality control when new methods are developed and new technology is introduced.

So we write to educate, to explain the power and the limitation of current testing. We write to remind you that quality must be considered a factor in every method, even the new, better, more advanced, faster, leaner tests. When new tests or studies promise new heights, we make sure they remain grounded in reality.

Quality is often assumed, not assured!

There is a belief that technology is the solution to all our problems in healthcare, as well as the corollary belief that technology will also solve any problems with quality. The reality is that many healthcare organizations today assume quality is adequate, rather than adopt practices to assure the necessary quality is actually being produced.

In this age of evidence-based medicine, it is surprising to find little objective data to support these assumptions about quality. The main defense used by the "quality is just fine" crowd is to cite the lack of complaints by physicians and patients. But, as our former Secretary of Defense once noted, “the absence of evidence is not evidence of absence.” As scientists, we should recognize that the absence of data cannot prove the presence of quality. The absence of complaints might be explained by the functioning, or lack thereof, of the complaint system itself, or by the fact that it is difficult for physicians to recognize poor quality results unless extreme. Complaint systems in healthcare organizations may be considered proprietary and confidential in order to limit legal liability; existing complaint data may not even be available to people who could utilize that information to identify and solve problems with quality.

So we write to dispel complacency. We write to insist that quality, even though it's been talked and talked about for decades, is still something that requires attention, investment, and commitment.

Quality has gone missing in today’s Money-Driven Medicine

It’s in almost no one’s financial interest to talk about quality in a plain and understandable way. Even large diagnostic companies with great instruments hide behind statistics that customers are unlikely to understand. Just take a look at a manufacturer’s claims for any new instrument system! They usually give you statistical results (such as SDs, CVs, regression equations, etc.), but they don’t tell you whether the test performance provides the quality necessary for patient care. In fact, on this website, you will find assessments of manufacturer’s own performance claims that indicate the quality achieved in terms of sigma-metrics is unacceptable for use by laboratories. It is likely that neither the manufacturer nor the customer understands the quality of test results, otherwise such systems would never make it to the marketplace.

Professional organizations are also guided by their own financial interests, and in turn are often dependent on the diagnostic industry for financial support. They publish their academic journals to satisfy their responsibilities to their professional customers, but they cater to the diagnostic industry to support many of their activities, particularly those annual national meetings.

Many other organizations are decidedly “neutral” when it comes to quality. They agree that quality should be important, but neglect to take any steps that back that up. And with much of their support increasingly dependent on contributions or revenues from diagnostic companies, harsh demands on quality potentially eat into their bottom line.

Westgard QC faces the same pressures as other organizations and individuals, but we have more freedom and incentive to be honest. With revenues less dependent on corporate influence, we worry less about offending the major players and concentrate on the facts. And frankly, by being a small company, we have a lot less to lose by telling the truth.

So we write because when others won't or can't write about quality, we can.

New quality programs often aren’t

We can write the truth about what’s happening in the world of quality and challenge the assumptions of emerging practices and programs.

For example, Lean manufacturing or production techniques are beginning to be applied in laboratories to improve workflow, efficiency, and productivity. Great cost savings are often claimed, usually due to installation of new automated systems that reduce the number of laboratory analysts. The main evidence of improvement is usually faster turnaround times for tests, with little consideration of the analytical quality of those test results. Again, analytical quality is often being assumed, not assured. Some advocates of Lean go so far as to say that QC is a waste that should be eliminated. Others talk about Lean-Six-Sigma, often implying that addition of Six Sigma to the name somehow assures that quality is being measured and managed.

Pay-for-performance is another "new" quality program that assumes quality can be improved by financial incentives. Right now, laboratories get paid for a billable test no matter the quality of the result. Pay-for-performance will supposedly change that, even though we lack objective measures of quality to support such a strategy. Worse still, all existing pay-for-performance metrics focus on factors other than analytical quality. When incentives favor faster over better, the marketplace drives quality down. Pay-for-performance will likely have the same effect, but you’ll never hear that from its advocates.

Quality indicators are gaining prominence because of the need for objective measures of quality in laboratory medicine and healthcare. However, most of the proposed laboratory measures are focused on pre-analytic and post-analytic characteristics of quality, even though the most basic measure required for a test is analytic quality. Again, analytical quality is being assumed rather than assessed and assured.

Risk management is now being presented as the new approach for laboratory QC, particularly as the new “option 4” that is to rescue CMS’s flawed “equivalent quality testing” or “equivalent QC procedures.” Industry favors this approach because it allows them to identify the potential sources of problems, mitigate the risks of the problems as best they can, then leave it up to the laboratory to manage the remaining residual risks, which they can’t without implementation of effective statistical QC procedures. So we have an effort to eliminate statistical QC (because it is too complicated) that ends up requiring the laboratory to design and implement optimal statistical QC procedures (which requires a higher knowledge and greater skill) to monitor the residual risks.

When the new emperor has no clothes, we write about nudity. When the latest management fad promises magic solutions, we voice our skepticism.

There is a need for Truth in today’s world of “Truthiness”

Given the present situation where quality is not well understood and quality is often mis-represented and mis-sold, we write about quality because the laboratory industry needs an honest broker. In a sea of slick marketing and overblown promises for healthcare, truth has been replaced by “truthiness” (something which sounds and feels like truth but really isn’t).

To better understand the present attitudes and values, the year 2006 has been officially recognized as the year of “truthiness.”

December 9, 2006, SPRINGFIELD, Massachusetts (AP) -- After 12 months of naked partisanship on Capitol Hill, on cable TV and in the blogosphere, the word of the year for 2006 is ... "truthiness." The word -- if one can call it that -- best summed up 2006, according to an online survey by dictionary publisher Merriam-Webster. "Truthiness" was credited to Comedy Central satirist Stephen Colbert, who defined it as "truth that comes from the gut, not books." "We're at a point where what constitutes truth is a question on a lot of people's minds, and truth has become up for grabs," said Merriam-Webster president John Morse. "'Truthiness' is a playful way for us to think about a very important issue."

Unfortunately, in healthcare, “truthiness” is not very funny because it leads to serious consequences for our patients.

Quality in healthcare is a serious issue and must be scientifically managed with objective data and metrics, rather than “from the gut”. We have often used the analogy that quality is like truth – a laboratory test should tell the truth, the whole truth, and nothing but the truth. These three dimensions of truth should help us understand the complexity of quality and also provide guidance for the strategy to achieve quality. Truth must include how good the test must be for medical applications, the whole truth must consider the precision and accuracy of the measurement procedure, and the “nothing but the truth” must consider efforts such as QC to eliminate misleading results.

We write to separate truth from the "truthiniess." We write to point out the differences between the two.

Why not “publish” instead of post?

With all that we want to say, and feel compelled to say, another question arises - why not publish in the traditional journals and trade publications? In fact, I still do publish, but most of the material is now going straight to the website. Why is that?

Admittedly, one of the reasons is that what I say is not “new” in the sense of scientific research. It’s not ground-breaking to suggest that there are problems in the laboratory. Pointing out the obvious flaws in current regulations is also a hard sell for a scientific journal – these are issues of a transitory nature, not of scientific discovery. Articles on what we’re not doing are at a disadvantage to articles that discuss novel methods and discoveries.

This is an age-old problem. The true adoption of an important technique often lags its discovery by decades. It takes time for the innovation to disseminate and build critical mass as well as the infrastructure support. The new ideas my colleagues and I published in the 1980s and 1990s are still percolating through the industry; and writing about them here helps them along.

Another problem with the scientific publication system is that the time between submission, acceptance, and publication can often be months, even a year or more. The world moves too fast for that now. If I had published an objection to the “EQC options” in a journal, it would have come out too late. By the time those objections had been published, the faulty regulations would have been in the field too long, and the momentum would have been difficult to halt. By speaking out online, early and often, I believe that we helped bring attention to serious limitations in the EQC recommendations and stimulate a more critical assessment of the proposed practices.

The final question posed by publishing in scientific journals is the audience. While the scientific journal has the imprimatur of acceptance, it has a limited audience. Any article published in a journal is only read by a small percentage of the subscribers, which tend to be Western Europeans and Americans. In contrast, distribution over the web makes this available to everyone, everywhere with an Internet connection, and (because we manitain these articles in a public archive) for a long period of time. By making access easy and broad, Westgard Web can reach more people. This year the website had visitors from nearly every country, with the exception of North Korea and a few other similarly afflicted countries.

I’ve also reached a point in my career where continued publications in the literature are nice but not necessary. Retired from the University, I no longer need a raft of publications to keep my tenure afloat. It’s more important to me to get the message out than how the message is packaged.

All good reasons to write, but not THE reason.

Quality matters!

Deep down, what compels me to write is the message. The purpose of Westgard Web, and of my life’s work, is to deliver this message in as many ways possible:

Quality matters. Quality is important. We are called not just professionally, not just ethically, but deep down, we are even called spiritually not to make things just faster or cheaper, but to make things better.

In the speech Sten gave at my retirement, he said that my departure from the University of Wisconsin was not the close of my career, but the opening of a new horizon. He has confidence that in the years ahead, there are more things we can discuss and discover.

I look forward to working with him, and with you, on the quality of the future.


James O. Westgard, PhD, is a professor of pathology and laboratory medicine at the University of Wisconsin Medical School, Madison. He also is president of Westgard QC, Inc., (Madison, Wis.) which provides tools, technology, and training for laboratory quality management.