199 questions from the Beckman Coulter Webinar.
Another common inquiry I have gotten over the last few months concerns the advanced technique of Six Sigma and its applicability in not-so-advanced laboratories. In other words, what can resource-challenged laboratories do with Six Sigma metrics? This question applies not only to hematology, but to any type of testing.
This is a tricky challenge, and one that will depend on just how challenging the resources are for the specific laboratory.
Let's highlight the benefits of using the analytical Sigma metric: device and method selection, rule and level optimization, run/frequency optimization.
Device and Medical Selection
Knowing the analytical Sigma metrics of an instrument or method can help you decide whether or not you want that box in your laboratory. Regardless of whether or not you can support Sigma metrics inside your laboratory, knowing if the instrument under consideration has more 6 Sigma assays or 2 Sigma assays will have a major impact on the laboratory. Choose wisely, and you'll have less work to do, fewer outliers to troubleshoot, and happier clinicians and patients. Choose unwisely, and you'll see any savings on the "cheap" instrument evaporate under a tidal wave of outliers, troubleshooting, repeating, recalibrating, delaying of results, and frustration for all involved.
Unfortunately, there are challenges even getting reliable trustworthy data about methods and instruments. Like any statistic, the analytical Sigma metric can be manipulated, and as the importance of the Sigma metric has grown, so have the manipulations of the diagnostic manufacturers. I have seen increasingly that the bids and the demands from laboratories include assessments of Sigma performance. But that's like asking the fox to describe how he would perform inside the chicken coop; he'll find the very best way to convince you. I now take the perspective that a laboratory investigating the performance of a method should consider getting access to the "raw" performance data of a customer site using that instrument in question, and calculate the metrics themselves, thus avoiding any kind of manipulation that the diagnostic manufacturer might perform. That, or find really, really trustworthy sources of performance data.
The point here is to avoid a poor method or instrument, which will cause too much expense and produce too many defective results. The additional factor, of course, is that in resource-challenged areas, the best performing methods and instruments are often out of financial reach, and the only equipment available may be cheap, and also of poor quality. Avoiding the worst is still an accomplishment.
QC optimization, rules and levels
If the laboratory has the ability to calculate analytical Sigma metrics on their current or just-acquired methods, they can identify opportunities to do less QC (4 Sigma and higher performance), as well as the assays where more effort is needed (3 Sigma and lower). When a laboratory can't find a method that's got an acceptable Sigma metric, and there's no alternative, consider adding the context around that measurement using the Reference Change Value (RCV), sometimes known as the critical difference. If you have some test results that are less trustworthy than others, giving the clinicians the ability to understand noise from signal is important.
The lab doesn't have to calculate analytical Sigma metrics every day, probably the most frequent should be once a month, at the same time a QC review is performed. Other labs perform it quarterly, or even every 6 months. Thus, it may be possible to even do manual calculations, given the infrequency of calculating the Sigma metrics. And remember, just because a Sigma metric changes, that doesn't mean your QC will change. Any value above 6 Sigma, be it 8, 10, 20 Sigma (these are possible), still receives the same QC procedures (1:3s with 2 or 3 controls). So the QC procedure may be stable for a long period of time.
QC frequency optimization
The most advanced application of the analytical Sigma metric is a change to the frequency of running QC. Instead of clock-based QC, every 24 hours or every 8 hours, the ideal optimization would be based on the number of patient samples being run. Run a QC for every 1000 patients, or 500 patients, etc. While this optimization is mathematically possible, it's practically impossible to implement with today's controls, without some compromises. Given that controls are getting more and more consolidated, the frequency of running QC is fixed for a number of analytes at the same time. Say you have 40 analytes in a single QC tube. You probably won't have the same Sigma metric for all 40, which presents a conundrum: do you run QC at the frequency for the lowest Sigma metric? Or the highest? Or somewhere in the middle? That question has made most laboratories – even in resource-rich parts of the world – balk at changing their QC frequency. QC remains clock-based, not patient-based. This level of optimization is out of reach for nearly every laboratory, so a lab in a resource-challenged area shouldn't feel too bad about not being able to implement it.
To conclude, the main benefits of analytical Sigma metrics to a resource-challenged laboratory will be
- 1.Method/Instrument qualification. Avoidance of the worst performing tests and instruments.
- 2.Abandonment of waste-plagued QC practices like 1:2s rule for warning and repeating
- 3.Optimization of QC based on Westgard Sigma Rules, reducing QC to the bare minimum rules based on performance