Comments on General Nordic Recommendations about Quality Control
and Quality Assurance in Clinical Chemistry
Prepared by the Committee on Quality Control
(NKK) of the Nordie Society of Clinical Chem–
istry.
INTRODUCTION
These recommendations are a revised version
of recommendations originally presented in a
Report of an extensive Nordie Project on
"Quality Control in Clinical Chemistry-Ef–
forts to Find an Effective Strategy" supported
by Nordkem and published in 1984 (ref. 1).
Further arguments for issuing general Nordic
recommendations have been provided in an
editorial review in the Scand J Clin Lab lnvest
(ref. 2). Later documents under productian by
NCCLS and ECCLS and focusing specific areas
(ref.
3-5)
provide further motivations for this
type of recommendations.
PARAGRAPH l
The laboratory itself must feel the responsibility
for the quality of the analyses and the investiga–
tians performed. The laboratory knows the pur–
poses of these analyses and investigations and
should specify its quality in terms of prohability
ranges for imprecision and inaccuracy. This is a
troublesame and tedious process that will re–
quire time and support from the committees of
the Nordic Society of Clinical Chemistry and
from Nordkem but ought to be started as soon
as possible.
Quality of a clinical laboratory is more than
goodprecision and accuracy. This broader con–
cept has been defined by the term "Good Labo–
ratory Practice" and includes e.g. short request–
report times.
It
is important that the laboratory
keeps its customers informed not only of its
analytical quality but also of other issues con–
cerning its service.
PARAGRAPH 2
It
is the obligation of the laboratory to update
the arsenal of laboratory methods, instruments
24
and equipment. This is done in order to provide
medically necessary and meaningful data of
high quality to a low cost and to be able to
add-when appropriate-relevant interpreta–
tions (ref. 6) for diagnosis, treatment, preven–
tion and research. The laboratory must be
aware that it is a unit within the world-wide
medical community and that it ought to pro–
mate transferability of clinical laboratory data.
With respect to this responsibility careful cali–
bration is important.
It
includes using pure and
weil characterized calibration material and
avoiding random errors. Current practice to
measure the calibrator once in each series and
to refer all measurements of specimens from
patients to that calibration is suboptimal (ref.
7), as the error of the calibration propagates to
all estimates on patients' specimens.
Another not recommended current practice is
to reconstitute lyophilized serum by a volumet–
ric procedure. Because poor pipetting invalves
errors for one or more days, the reconstitution
volume should be checked on a balance.
PARAGRAPH 3
The difficulty in designing an interna! quality
control system is the balance between the re–
quest of the clinical practice for rapid reporting
and the need for many control observations, as
many observations improve the estimate of the
quality parameters and the prohability of error
detection. Due to the function of the laboratory
the effort to optimize the quality control system
may lead to samewhat different solutions.
It
is, however, important for all types oflabo–
ratories that the control system provides the
different types of staff members with clear and
easily understood information for the decisions
which these employees have to make.
Gross errors in the process of handling speci-
Klinisk Kemi i Norden
