Available online at www.medicinescience.org
ORIGINAL RESEARCH
Medicine Science 2018; ( ):
Measurement uncertainty in biochemical parameters
Selda Telo, Dilara Kaman
Firat University Faculty of Medicine, Department of Biochemistry and Clinical Biochemistry, Elazig, Turkey
Received 23 January 2018; Accepted 09 March 2018
Available online 18.04.2018 with doi: 10.5455/medscience.2018.07.8769
Copyright © 2018 by authors and Medicine Science Publishing Inc.
Abstract
Measurement uncertainty is a quality indicator that is used to show the distribution level of a test result in a laboratory. The aim of this study is to calculate the measurement
uncertainty of ten biochemical parameters investigated in our laboratory and compare the obtained values with the permissible total error values (% TEa), which various
authorities determine for these tests. The measurement uncertainty of glucose, ALT, AST, urea, creatinine, sodium, potassium, albumin, total protein and total cholesterol
were investigated in the biochemical auto analyzer (Siemens Advia 2400). Fifty randomly selected 2 level (low and high) internal quality assessment results and external
quality assessment of six-month data, the first six months of 2016, were calculated for each test. Nordtest guide was used for the measurement uncertainty calculation.
The measurement uncertainties were calculated as 3.9%, 6.72%, 3.4%, 8.06%, 9.06%, 6.08%, 5.02%, 4.98%, 4.96% and 7.22% for glucose, ALT, AST, urea, creatinine,
albumin, total protein and total cholesterol, respectively. The calculated measurement uncertainties were found to be lower than the CLIA’88 and RiLi-BEAK % TEa
values, except for sodium. Calculated % TEa values for creatinine, sodium, total protein and albumin were found to be higher than Fraser % TEa values. A test result or
a measurement is not sufficiently effective when reliability is not assessed. Laboratories should calculate the measurement uncertainty for each parameter. They should
give the results that do not exceed the targeted % TEa values. Clinicians should be informed about the measurement uncertainties so that they consider it while evaluating
patient results.
Keywords: Measurement uncertainty, biochemical tests, quality control
Medicine Science
International
Medical Journal
1
Introduction
Uncertainty is an interval that contains the values given together
with a result of a measurement or a result of a test and the values
with the possibility to be attributed to the measurement magnitude
[1]. The paradigm of measurement uncertainty, which emerged
during the early 20th century, claims that uncertainty is not a
drawback but a property of the result of the measurement [2]. The
term measurement uncertainty covers the information about the
real value of the result and the factors that influence the result.
In order to estimate the uncertainty of a measurement, random
and systematic errors should be clarified and the effect of each
major component on the measurement uncertainty should be
determined by suitable methods [3]. Measurement uncertainty
is a quantitative indicator of an obtained result. In other words,
it demonstrates the extent to which the result represents the
real value [4]. Measurement uncertainty is significant for the
evaluation of method’s availability for clinical implementations
and for the comparison of similar types of results. An extensive
evaluation of the components that contributes to the measurement
uncertainty also reveals the points to which attention should be
*Coresponding Author: Selda Telo , Firat University Faculty of Medicine,
Department of Biochemistry and Clinical Biochemistry, Elazig, Turkey
E-mail: drseldatelo@hotmail.com
paid in order to verify the measurement procedure of a test method
[5]. When the measurement of uncertainty is presented together
with the result, it provides an information regarding the quality
of the measurement to the users of the result [4]. Calculation of
the measurement uncertainty and its presentation along with
the measurement demonstrates the borders within which the
measurement is placed, and the level of reliability and quality
of the measurement. Thus, measurement uncertainty is required
while making medical decisions, comparing measurement results,
deciding the conformity within limits[6]. A number of factors,
such as matrix effect, interferences, reference materials, mass
and volume uncertainties, environmental factors, method and
procedure of the measurement and the operative, contribute to the
measurement uncertainty [1,5-7].
Clinical Laboratory Implementation Amendments’ 1988 (CLIA
’88) criteria specify the maximum limit of error that is allowed
legally for the measured material in European countries [8]. Just as
CLIA, RiLi-BAEK criteria in Germany are available for the total
error measurements [9]. The aim of this study is to calculate the
measurement uncertainty values for each of the ten-biochemical
parameters, which are used in our laboratory, by making use of
the internal and external quality control data and compare these
calculated values with the total error allowed percentage (%TEa)
values of CLIA, RiLi-BAEK and Fraser.