| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Diagnostics Consultancy Desk, 5631 AH Eindhoven, The Netherlands, Fax 31-40-2908621, E-mail r.m.lequin{at}planet.nl
To the Editor:
With great interest I read the "Point/Counterpoint" articles by Drs. Krouwer (1) and Kristiansen(2). In this regard, I would like to raise three issues: two analytical and one clinical.
With respect to the first analytical issue: laboratory experts and external quality assurance system organizers advocate that laboratory measurements be highly accurate, i.e., unbiased and precise. Although manufacturers of routine measurement procedures are usually able to solve most of the problems of imprecision, they are often unable to solve problems related to measurement bias. In my opinion, these Point/Counterpoint articles put undue pressure on the manufacturers to "improve the quality of their products". The professionals in laboratory medicine should provide guidance to the manufacturers to help solve the problems of accuracy in a fair and transparent manner.
Quantities measured or determined in laboratory medicine are now divided into two categories: (A) those that are traceable to SI (n 
100; well-defined chemical compounds) and (B) those not traceable to SI (n >500) (3). For category A, traceability is (or can be) assured, although some experts may argue that this is not yet applicable to all of these quantities. Measurement results of category A compounds in patient samples should be appropriately accompanied by the uncertainty of measurement [preferably according to the Guide to the Expression of Uncertainty in Measurement (GUM) to fulfill internationally agreed requirements].
However, very serious problems arise with category B, a very large group of quantities consisting primarily of (glyco)proteins and measured by means of immunochemical techniques. The crucial point is that for category B, we cannot speak of unbiased results of measurement because we have rarely adequately defined the "measurand" in the patients’ biological fluids. The "true value or true concentration" of the quantity of category B in patient samples simply is unknown.
The manufacturer responsible for calibration of its product calibrators against "reference materials of higher order", if and when available (3)(4), faces a dilemma. These reference materials of higher order are sometimes available under the aegis of WHO. However, the WHO Expert Committee on Biological Standardization never defined the measurand in biological fluids. Consequently, these materials should be regarded as "surrogates" for the relevant quantity in human or animal biological fluids.
Nevertheless, manufacturers use these materials for calibration of their "master calibrator lot", but the uncertainty budget (GUM) of the values assigned to these WHO reference materials is unknown. Assuming that these uncertainty budgets become known in the future, that uncertainty should then be further propagated through the calibration hierarchy down to the results reported for a patient’s sample (3).
The second analytical issue is the question of whether the available and used reference materials for category A and category B are commutable with the quantity (quantities) in the biological fluids? This question merits attention because the validity of calibration and other measurement exercises depends on it. However, this aspect is hardly ever addressed, for example, in WHO documents.
The biased results of measurements of category B quantities and the issue of commutability for category A as well as for category B analytes have important impacts on the uncertainty of measurement results.
The third issue relates to clinical decision-making: what is the effect on clinicians and the clinical decision process of reporting a measurement result for a patient’s sample with the uncertainty budget calculated according to, e.g., GUM? Will clinicians understand it? Will it improve their efficacy and efficiency? GUM was elaborated by representatives of a host of international organizations; it is meant to be applicable to all scientific measurements, be they physical or chemical. The concept of uncertainty of measurement in laboratory medicine was incorporated in the "traceability" document (3), and it plays a role in the obtaining of accreditation by medical laboratories (5)(6). It then is important to ask whether medical associations in, for example, the US, Europe, and Japan have been consulted on this matter? If so, what are the problems perceived by our medical colleagues? As far as I am aware, to date no such attempt has been made. If that is true, are not laboratorians merely satisfying analytical and metrologic requirements? Should we not defer reporting the uncertainty of measurements of patients’ samples until it is accepted clinically as useful and beneficial to patient care?
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
