Posted by Sten Westgard, MS
Recently, a user submitted a question about the mandatory use of the 10x rule on chemistry analytes as mandated by ISO 15189 through NABL:
"Specific criteria of accreditation of NABL India (for ISO 15189) say that 10x should be considered as a violation for clinical chemistry and Immunoassay parameters. However we understand that Westgard rules consider it as a warning only. Please advise."
So, the central question is, does ISO 15189 mandate the use of the 10x control rule?
"5.6.2 Quality control
The laboratory shall design quality control procedures that verify the attainment of the intended quality of results. Special attention should be given to elimination of mistakes in the preexamination and post examination processes.
NOTE In several countries, quality control, as referred to in this section, is also named “internal quality control.”"
That phrase, "shall design quality control procedures that verify the attainment of the intended quality of results" indicates that (1) labs must do this (when ISO says "shall" it really means, "Thou shalt do this!"). However, there is no specific guidance in this clause. Everything here is conceptual.
An additional clause provides a bit more advice:
"5.6.2.2 Quality control data
The laboratory shall have a procedure to minimise the risk of singificantly different or aberrant patient examination results being reported in the event of QC rule failure.
NOTE: When the quality control rules are violated, examination results should normally be rejected and relevant patient samples re-examined after the error condition has been corrected and within-specification performance is verified. The laboratory should also evaluate the results from patient samples that werre examined after the last successful quality control event.
Quality control data shall be reviewed at regular intervals to detect trends in examination performance that may indicate problems in the examination system. When such trends are noted preventive actions shall be taken and recorded.
NOTE: Established statistical techniques such as Shewhart/Levey-Jennings charts and process control rules should be used wherever possible to continuously monitor examination system performance."
Again, while the clause suggests the use of control rules and Levey-Jennings control charts, it is not specifically dictating what rules to use. ISO 15189, as usual, gives conceptual advice, not detailed instructions.
However, when we dig a little deeper, we discover that this is the "Indian" version of ISO 15189, which takes all of the ISO standard and then adds additional details. So here, in the NABL specifications, are the directions which are mandating the use of certain control rules:
"Clinical Biochemistry
The Laboratory must establish and document procedures for monitoring and evaluating analysis of testing processes including procedures for resolving ‘out-of-control’ situations. The laboratory is encouraged to use control material similar to or identical with patient sample matrix. The laboratory shall incorporate in the procedure, the multi-control QC rules used to detect systematic (trends or shifts) and random errors.
The laboratory shall include a minimum of one level QC at least once a day. However, where the number of patient samples analysed for any parameter exceeds 25 per day, the laboratory shall employ 2 levels of QC at least once a day for such parameters. Further, if the number of patient samples analysed for any parameter exceeds 75 per day, the laboratory shall employ 2 levels of QC at least twice a day at appropriate intervals.
The daily QC values shall be documented along with the calculation of %CV from the monthly QC data. The laboratory shall maintain control charts to demonstrate stability of the analytical measuring systems.
The laboratory shall follow the multi control QC rules as described below:
The rules to follow when one level QC material is used:
Reject QC if:
a. it is outside 3 SD (13s)
b. two consecutive values obtained are outside 2 SD on the same side but within 3 SD (22s)
c. ten consecutive values are above or below the mean, but within 2 SD (10x)
The rules to follow when 2 level QC materials are used:
Reject QC if:
-
- either QC values is outside 3 SD (13s)
- both QC values are outside 2 SD on the same side, but within 3 SD (22s)
- difference between both QC values is >4 SD i.e. one level QC is > 2 SD and other level QC is <2SD (R4s).
- ten consecutive values of the same level QC are >/< the mean, but within 2 SD(10x).
- five consecutive values of one level QC and five consecutive values of other level QC are >/< the mean but within 2 SD (10x)"
Here then, is very specific advice on what controls rules to use for QC in clinical biochemistry. The description is basically that of the "classic Westgard Rules." Where ISO 15189 was completely vague, NABL spells it to the letter.
The problem here is, are all these control rules really needed for every clinical biochemistry test? If you look at QC Design and Six Sigma analysis, you'll find that the answer is no. Not every test needs the full "Westgard Rules." Some tests do need that amount of QC effort. But other tests have either (1) large total allowable error and/or (2) very good method performance, and, as such, require only single control limits - simple rules like 13s, 12.5s, etc. To implement a 10x rule on a Six Sigma method, for example, is overkill. It will generate flags that aren't significantlly impacting the performance required of the test for appropriate use.
There are two extremes on display: ISO 15189 is conceptual, without any specifics. The NABL, on the other hand, is so very detailed that it is imposing requirements unncessarily. There's a regulatory dilemma at play: it's hard to inspect a laboratory and determine if they're are properly implementing the vauge clauses of ISO 15189 alone. It's much easier to impose a set of specific rules, like the NABL guidelines, and make them apply to all tests. It's very efficient for the inspector when every test must do exactly the same kind of QC. Unfortunately, it's not very efficient for the laboratory operation. True, there are tests where the full "Westgard Rules" are needed. But there are others where this level of QC is too much - and for the ease of inspection and regulatory compliance, we are wasting the time and effort of the laboratory.
I know it's difficult - and perhaps it's even heretical - to question the correctness of ISO 15189 and its related guidelines. But ISO 15189 isn't divinely inspired - it's the product of a lot of committee work. And the secondary regulations that flesh out those conceptual standards are also the product of a lot of committee work. So we shouldn't be too surprised if there are rare cases where the regulations aren't perfect.
To return to the specific question: the use of the 10x rule should not be mandatory for all tests in all situations in all laboratories. The 10x rule is a tool. Like any tool, there are scenarios where it's useful and scenarios where it's not.
Comments