Essays and Publications

Sometimes it takes another person or persons to say what you've been meaning to say, or say it better. For years, we've discussed the problems with EQC. Recently, however, an exchange of emails on a popular listserve said everything you need to know about EQC

June 2009
with Sten Westgard, MS

[PLEASE NOTE: AS OF JANUARY 2016, EQC BECAME ILLEGAL WITHIN THE US. LABS USING EQC MUST SWITCH TO USING IQCP.]

Sometimes it takes another person or persons to say what you've been meaning to say, or say it better. For years, we've discussed the problems with EQC. Recently, however, an exchange of emails on a popular listserve said everything you need to know about EQC:

[Question:] "Is it possible to build a levey jennings chart using QC done once a month. We have an [Instrument A] and [Instrument B] and only run controls once a month or when there is a new lot number. We use the ranges set by the manufacturer of the controls. What value would it be in these circumstances. Thanks."

[Answer:] "Both of these POCT analyzers have built-in self-tests and perform various checks with each cartridge, and the [Instrument A] has the electronic simulator as well. Like you do, we run two levels of external QC once a month on our POCT analyzers.

"While possible to construct a QC chart for once-a-month POCT controls, I see no utility for doing so, in the name of process control. The frequency of testing is so low as to render the results statistically meaningless for detecting variations over time, and the possiblity[stet.] of detecting a system failure is included with each cartridge. And many POCT analyzers don't give you a number when running external QC, only a Pass/Fail result.

"Our QC Plan specifies four levels of QC testing (Quantitative, Qualitative, Semi-quantitative, and POCT). Each type has its own system, and Westgard Rules only apply to the first."

There you have it. Once you've reduced the QC frequency for POC devices, the bench-level users view it as practically useless. They won't do anything with once-a-month QC results. So, effectively, doing EQC is the same as eliminating QC.

The Road to EQC...

Let's recall the path that we took to reach the death of QC.

First, it was claimed that use of electronic QC with POC devices was actually the same as performing statistical QC.

Then it was admitted that electronic QC only covers a smaller set of internal steps and does not measure the operator competence. It's the difference between a check engine light and a driver's license test. The check engine light is useful, but it doesn't give you any information about who's driving the car and how well they're doing it.

So then CMS created a set of Validation Protocols to justify the reduction of daily QC to either weekly or monthly QC.

Then it was pointed out that the Evaluation Protocols were scientifically dubious, at best.

So then CMS together with Advamed proposed to task CLSI and expert committees (EP22 and EP23) to develop the scientific justifications for reducing QC. The phrase, "the intelligence and facts were being fixed around the policy" comes to mind.

As the CLSI Evaluation Protocols subcommittees grappled with various scientific rationalizations for reducing QC, years passed and CMS declared victory for the original EQC guidelines. EQC had been implemented and, according to CMS, their inspectors saw no problems in those labs. We wondered what they were looking at since there no longer was any QC data available, but “Trust us”, we were told. No problems were detected by the inspectors, therefore, EQC must work. The comment “the absence of evidence is not evidence of absence” comes to mind here. With No QC data it is impossible to know whether or not there are any QC problems.

Now CMS has a new draft of the State Operations Manual in limited circulation. There appears to be little if any change to the EQC guidelines and little if any reference to the new CLSI guidelines (EP22 and EP23) that are scheduled to be completed this year. Thus, the original EQC guidelines, together with their invalid validation protocols, have "stood the test of time." Or, more correctly, enough laboratories are now using EQC and are happy not to have to worry so much about QC that the practice of EQC has been accepted by the laboratory community, our professional organizations, and our inspection and accreditation bodies.

This should not be misconstrued as progress! As we have seen in our financial institutions, man highly-paid and highly-placed professionals have lost the ability to know right from wrong, and quality is no longer as important as cost and profit. And we can anticipate that we will learn our lesson sometime in the future and at a tremendous cost to the patients who depend on getting the right test done right.

We've said this before, too. This process of chipping away at QC, this death by inches, recalls the parable of the boiled frog. In the parable, if you drop a frog in a pot of boiling water, it will immediately jump out of the water and save itself. On the other hand, if you put the frog in a pot of room-temperature water, then slowly, slowly raise the temperature to boiling, it won't realize the danger. And in the end, the frog will be boiled to death.

That's what's happening in QC. If any standards institute or regulatory body came out with the statement, "We are going to eliminate quality control", it would be met with opposition and probably some outrage. No laboratory would stand for it. But this gradualism, the slow-motion decline of standards and practices, is going to reach the same end. We will not be doing quality control anymore..