![]()
The haematology samples were collected in diffe-
rent types of EDTA-tubes in the participating labo-
ratories. Results from samples collected in tubes
with liquid K
3
-EDTA are affected by a dilution effect.
Better agreement between the analyte means from
different laboratories can be achieved if the results
are corrected for this dilution effect. Furthermore,
an increasing number of laboratories are using
tubes with dry K
2
-EDTA, which are recommended
already 1993 by the ICSH (The International Council
for Standardization in Haematology). Accordingly,
all results were corrected, to the expected values in
undiluted samples, before the reference intervals
were calculated centrally by EQUALIS. The calcula-
tions, of the reference intervals with simple nonpa-
20
Half of the laboratories partici-
pating in the clinical chemistry
part of NORIP also measured
the common haematology
analytes in samples from the
reference persons. The decision
to include the haematology in
the NORIP was first made in
Finland and since then joined
all participating laboratories in Sweden and Iceland,
one third of the Danish and one laboratory in
Norway this part of the project.
The haematology part of the project can be loo-
ked upon as an addition to the original NORIP. The
reference persons are the same and the information
from the questionnaire to them is the same. This
description focuses on the differences compared to
the clinical chemistry.
At the same time as the samples for the clinical
chemistry analyses were collected from the reference
person, venous samples with EDTA as anticoagulant
were also drawn. These samples, from altogether
about 1 800 reference persons, were analysed as rou-
tine samples on the haematology instruments in the
participating laboratories. The analytical data, met-
hod data and reference person data were submitted to
the central database where it is stored in a MS Access
relational database administered by Gunnar Nordin
and Arne Mårtensson at EQUALIS, Uppsala.
Different criteria for exclusion of data have been
considered. However, data exclusion has been kept
to a minimum. Data have been excluded only for
the following reasons:
• For laboratories with more than one series of
data from different instruments, only one seri-
es is included into the calculations.
• Single results outside mean ± 5
s
(
s
- standard
deviation) have been excluded. This exclusion
was repeated until no more outliers were found
(twice needed). In total 28 results out of about
21 000 were excluded.
| 2 | 2003
Klinisk Biokemi i Norden
Reference intervals for haematology analytes
Preliminary results from the Nordic Reference Interval
Project (NORIP)
Arne Mårtensson (arne.martensson@equalis.se) EQUALIS, SE-751 09 Uppsala, Sweden
