Method comparison studies: an introduction to acceptability criteria

In this third article in his series on method comparisons, Stephen MacDonald moves on from experimental design and analysis, the sources of samples and the number required, to focus this month on what differences are expected and acceptable, and what other factors need further investigation before implementation of an assay.

When comparing methods our goal is to assure ourselves of the consistency of results between methods so that methodologies can be used interchangeably, or one can replace another without adversely affecting patient outcomes. This is an interesting concept for a number of reasons. It makes sense that we would not want to introduce an assay that is less clinically useful than what we currently have.

So, should our benchmark be to

perform as well as the current method? Should we be aiming for better? Ideally, our goal is to achieve performance to specifications derived from clinical outcome studies. What if these studies were themselves performed on assays that are less sensitive or specific than what we have at our disposal now? Does that potentially change patient outcome? What specifications could we use?

Performance specifications At its simplest, the assay must measure an association between the test and the condition or disease of interest. This is such a fundamental principle that it is decreed in EU regulation to determine the scientific validity of any analyte. Furthermore, the regulation ([EU] 2017/746) defines requirements for assays providing continuous (numerical)

and dichotomous (positive/negative, diseased/non-diseased, detected/not- detected) measures.

The choice of what performance specifications we use has been an evolving process over the past 50 years. The latest framework has built upon the Stockholm and Milan conferences of 1999 and 2015, previously referenced in both the measurement uncertainty and statistical quality control series of articles.

Initially, five models of quality specifications were proposed in 1999. Four of them were further subdivided. This has been condensed to a more manageable three models covering a more conceptual approach. Although the framework is not meant to be viewed as a hierarchy of standards, there is an apparent order to the levels of evidence used in each of the three models. The minimum standard is based on the current state of the art. Sources for such information may come from performance in proficiency testing schemes or other schemes where a large assessment of performance data are available. Peer-reviewed publications referencing methodology development and performance characterisation are included as state-of-the-art knowledge. Biological variation data are common to all historical and current frameworks, and are a very useful measure to guide performance standards, being linked

Some assays are established as screening tests, while others are more specifically used for diagnosis. WWW.PATHOLOGYINPRACTICE.COM DECEMBER 2020 15

Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60  |  Page 61  |  Page 62