Recently, I received the new Clinical and Laboratory Standards Institute’s “Evaluation of Total Analytical Error for Quantitative Medical Laboratory Measurement Procedures” (CLSI, EP21 2nd Ed. July 2016), which is used by healthcare professionals to assess the performance of diagnostic tests.
Reading about the newest standards made me reflect on the amazing progress we’ve seen in diagnostic performance during the past two decades, including a new point-of-care assay that Alere recently submitted for regulatory approval that had a total analytical error (TAE) of less than 5% deviation from true value.
Thanks to performance improvements, diagnostic tests have become increasingly valuable in enabling healthcare professionals to quickly and accurately screen for dangerous illnesses and administer the best care possible. Putting this progress in context requires defining a few terms. TAE is a standard and useful metric for assessing the quality of an assay. It takes into account the two “halves” of diagnostic performance: accuracy and precision. Accuracy refers to how close a measured value is to the true value; precision refers to the reproducibility of a result.
Imagining a dart board is a good way to visualize the difference. Shooting ten bull’s eyes in a row is both accurate and precise, but ten shots that hit the board in the same spot outside the bull’s eye is precise, but not accurate. Similarly, a test can be precise yet not accurate if it consistently gives the same wrong results. Conversely, a test can give accurate results sometimes, but not consistently so (precisely).
Thanks to performance improvements, diagnostic tests have become increasingly valuable in enabling healthcare professionals to quickly and accurately screen for dangerous illnesses and administer the best care possible.
Although manufacturers today do not consistently report TAE in package inserts, most tests have a target value of acceptable total analytical error. This is the value of the total error that, in the consensus of the medical community, must be achieved or surpassed to deliver a test value that is sufficiently close to the true value to enable the correct clinical decision.
The new EP21 provides guidance for healthcare professionals so that they can select the appropriate limits for allowable total error relative to a performance goal for clinical utility – giving them a more relevant and useful assessment of a particular diagnostic.
Twenty years ago, large lab-based tests (which are considered the gold standard) had TAE values as high as 30-40% from true value. Today, we are developing portable instruments that can be used by the patient’s side that have TAE as low as only a few percent points from true value. The reduction in the TAE has significant positive implications for patient care… and gets to the heart of “precision medicine.”
This progress in reducing TAE has been transformative on two levels. First, healthcare professionals no longer have to choose speed or performance when evaluating assays. Today, they can demand – and obtain – both. Second, progress leads to more progress. Now, many teams across different disciplines involved in developing new diagnostic platforms, from R&D to design, are challenged to strive for even greater performance because the market demands it.
In addition, these higher standards are being applied to point-of-care tests, as we’ve seen with the recent CLIA waiver of molecular tests, which would have been unimaginable ten years ago. On the production side, diagnostic companies are fully automating work streams, reducing human error and enhancing efficiencies.
We are seeing progress like this in other areas of the diagnostics sector, and it’s leading to better-performing tests reaching the market. And this, in turn, helps healthcare professionals to better care for their patients and improve their outcomes.