STEN WESTGARD: Six Sigma helps separate out effort, good intentions and marketing from actual performance. When you do a Sigma metric, you're stripping out all the hype around a process. You're getting to the actual performance. The unvarnished truth. It can be frightening to learn that despite all of our good intentions, and despite all of our hard work, we're still not producing the right level of quality for our patients. But that also becomes the fuel for real change. Take an honest look at your processes and learn where you’re doing great so you don't have to work so hard; and where are you doing poorly so you know where you need to improve.

JAMES WESTGARD: The amount of quality control (QC) that the laboratory needs to run is minimized when you're at Six Sigma; the laboratory must maximize the QC implemented as the process declines to Three Sigma. If you’re below Three Sigma, you simply can’t do enough control to ensure correct results.

STEN WESTGARD: Before you add an instrument, you're supposed to assess it and think about whether it fits in your laboratory. So, while a Three Sigma test may be less expensive initially, in the long run it may generate too many unreliable results. It's going to cause laboratories to do more recalibration, run controls and troubleshooting.

JAMES WESTGARD: There is a way that the industry looks at cost, which I call quality cost. It's the cost to prevent things from occurring, the cost to monitor and detect, the cost of failures. Those are internal failures that affect a laboratory and increase costs. The most expensive cost, however, is an external failure where you report wrong results, and those wrong results are acted on and the patient is misdiagnosed and mistreated. You only need to imagine one of those to realize that all the money you're spending in the lab to try to assure quality is probably less than that one cost to a patient who’s mistreated.

When we first started looking at this industrial model, we could easily demonstrate that the costs in the laboratory to do better quality control could be saved by reducing the false rejections, the repeat costs. Doing better quality control, managing the system more carefully, pays for itself. There's nothing mysterious about it. But it can be difficult to come up with the numbers because you can’t estimate the external failure costs, which are the biggest part of the cost. But by achieving strong Sigma performance laboratories can control costs by ensuring quality and, in turn, fewer or no instances of results leading to misdiagnoses.

STEN WESTGARD: One is Winchester Medical Center, part of the Valley Health System in Virginia. Over the nine-year lifespan of one of its sets of chemistry instruments, it calculated that it saved over $160,000 in hard costs and over 3,200 hours in soft costs. Essentially, it was saving an hour a day because the technicians did not have to stand in front of the instrument repeating a control, recalibrating, troubleshooting or calling up technical support to complain, all because it was able to right-size its quality control operations based on the Sigma metric analysis. That meant a whole set of outliers that caused extra work, like false rejections, simply disappeared. There were fewer errors, so its technicians had fewer things to take care of. It consumed less reagent and QC material.

Winchester was running QC three times a day, and it reduced it to twice a day because they were increasingly confident in their performance. Many laboratories run QC once a day, so their main savings will be realized by reducing troubleshooting and recalibrations, not through a reduction in frequency.

STEN WESTGARD: There are three components. The first is the instrument itself, because if you buy a Three Sigma instrument, no matter how good you are in the lab and how much effort you put into it, you can't improve what it's been engineered to do. You must have a good diagnostic partner.

Second, the people in the laboratory must understand how to do QC. It would be useful if the staff understood how Sigma gets calculated. The staff at every lab in our Sigma Verification of Performance program has to get trained and prove their competence in managing quality.

The last component is technology: informatics enable you to easily figure out the Sigma metric and act on it. Sigma metrics are increasingly integrated into the informatics environment. This means we no longer have to train the person to extract the data, build the spreadsheet, understand the outcome, then make a decision. Now a laboratorian asks, can I flip this switch on? It's being done for them.

JAMES WESTGARD: When the Sigma metric was introduced to the laboratory 20 years ago, everything was manual. Now it’s integrated into more and more software, so it's more available to laboratories. This means you can implement it with less effort. It's never going to be completely turnkey; you still need professional judgment to monitor how the Sigma metric is being chosen or defined with the allowable total error and how the results are being handled.

 They say anytime you have an innovation it takes around 20 years to become part of the industry you're trying to change. If anything, we're just reaching the maturation of Sigma in implementation.