Here are some worksheets that describe the "classic" set of multirule procedures that came to be popularly known as "Westgard Rules." But also included are two multirule combinations that may be considered more modern and appropriate to today's different testing capabilities and needs.
Years ago, one of our regular visitors constructed a worksheet to the traditional multirule procedure. This extensive description of the multirule QC procedure was presented in a convenient worksheet format. After years of making this available to the public, we realized that in addition to the "classic Westgard Rules", there were two other combinations that we frequently recommend.
What's the major difference?
Given the advances in instrumentation, software, and automation, the traditional use of the 12s 'warning' control rule is no longer recommended. This is not a new recommendation from Westgard QC, indeed it is prominently mentioned in nearly all discussions of the rules. However, we've just never made up a graphic worksheet demonstrating the more appropriate multirule combinations - until now.
Why are there two new "Westgard Rules" worksheets?
Again, let's stress that these are not "new" multirules. We have simply made graphic logic diagrams of the recommendations we've been making for many years. The the two main recommendations are tailored to the typical number of controls in a single run. For chemistry instruments, controls are usually run in pairs (i.e. 2 controls, 4, etc.). The "traditional" rules were developed for that practice. However, for hematology and other instruments, controls are usually run in triplicate (i.e. 3 controls per run, high, low, normal, etc.). A different set of rules, substituting a 2of32s control rule for a 22s control rule, is more appropriate.
Do I need to implement all of these new rules?
One more time. These multirule combinations are not new, nor are they cast in stone. Use as many of the rules as needed to provide the necessary error detection. For example, you may find that a particular method only needs to use a 13s/22s/R4s combination, without the 41s/10x rules. You may also determine that another method needs a 12x instead of a 9x as the final part of a multirule combination. By performing a QC Design process, you can determine exactly which rules you need to implement, as well as which rules are unnecessary to your QC procedures.
The one thing that you can't do when you build the rules is "mix and match" in a random fashion. They multirules are built in a rigorous order. You can't rely on just a 10x rule for QC, for instance, or build a 22s/8x multirule combination. The worksheets make clear how the rules fit together.
