aging stability of complete blood count and white blood cell differential parameters analyzed by...
TRANSCRIPT
Aging stability of complete blood count and white blood
cell differential parameters analyzed by Abbott CELL-DYN
Sapphire hematology analyzerP. HEDBERG, T. LEHTO
INTRODUCTION
The recent trends towards large centralized laborato-
ries, and changes in laboratory organizations, have
brought redistribution activity to a new perspective.
Laboratories now test specimens that have been dis-
patched over a long distance; as a result, testing is
often delayed by 12–24 h or more after venipuncture.
Although laboratories should still give reliable results,
excessive delays in processing might affect the reliabil-
ity, accuracy and imprecision analysis. Cellular ele-
ments are known to have limited stability in
ethylenediaminetetraacetic acid (EDTA)-anticoagulat-
ed blood (Buttarello, 2004). Also, the different behav-
ior of automated counters using impedance and
optical methods may have an effect, and this should
be taken into account (Wood et al., 1999).
Of the three EDTA salts used for the anticoagula-
tion of blood specimens for hematological testing,
potassium salts are the most readily soluble (England
et al. 1993). K3EDTA is dispensed as a liquid and thus
causes a slight dilution of the specimen. This salt also
affects the red blood cell size at increased concen-
trations and on storage than the dipotassium salt.
Department of Clinical Chemistry,
Laboratory, Oulu University
Hospital, University of Oulu, Oulu,
Finland
Correspondence:
Pirjo Hedberg, Laboratory, Oulu
University Hospital, PO Box 500,
FI-90029 OYS, Oulu, Finland.
Tel.: +358 8 3155453;
Fax: +358 8 3154409;
E-mail: [email protected]
doi:10.1111/j.1751-553X.2007.01009.x
Received 31 May 2007; accepted
for publication 8 October 2007
Keywords
Automated blood cell counts,
hematology, hematology analyzer,
stability, WBC differential, WBC
flagging
SUMMARY
This study presents the results of an aging stability study of complete
blood count (CBC) and leukocyte differential parameters using the
Abbott CELL-DYN Sapphire hematology analyzer. Stability studies
showed no substantial change in CBC parameters up to 24–48 h at
+23 ± 2 �C (room temperature), except for optical platelet count
(PLTo). For specimens aged over 24, the value of impedance platelet
count yielded more reliable results than the routine PLTo. White blood
cell (WBC) differential parameters, except eosinophils, were stable for
up to 48 h at +23 ± 2 �C. CBC parameters were stable for 72 h, except
mean platelet volume, which slightly increased between 48 and 72 h,
at +4 �C. WBC differentials were stable 48–72 h, with a slight decrease
observed in absolute neutrophils and lymphocytes at +4 �C.
ORIGINAL ARTICLE INTERNATIONAL JOURNAL OF LABORATORY HEMATOLOGY
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96 87
Therefore, in 1993, the International Council for
Standardization in Haematology recommended that
K2EDTA should be used as the anticoagulant of choice
in specimen collection for blood cell counting and
sizing (England et al. 1993). However, the cellular ele-
ments in hematology specimens have limited stability
in EDTA-anticoagulated blood during storage at room
temperature or at +4 �C. Of these two temperatures,
conservation of specimens at +4 �C is known to
improve the stability of some complete blood count
and white blood cell differential parameters, but there
have not been any systematic studies that take into
account all the modern instruments or that considered
normal specimens, as well as a wide variety of pathol-
ogies, where a delay in analysis might give different
results (Buttarello, 2004).
In order to decide whether to accept or reject the
aged specimen, laboratory staff needs to be familiar
with the changes known to occur in blood speci-
mens during storage. For reticulocyte counts, the
National Committee for Clinical Laboratory Stan-
dards (NCCLS) H44-A protocol (National Committee
for Clinical Laboratory Standards, 1997) recom-
mended that analyses should be performed within
6 h of collection if the specimen is kept at room
temperature; conserving specimens at +4 �C stabilizes
parameters for 72 h. Also, for the differential leuko-
cyte count, it was recommended that smears should
be prepared for microscopic analysis within 4 h
(National Committee for Clinical Laboratory Stan-
dards, 1996).
Several studies have been published during recent
years that include data concerning the stability of
complete blood count and automated differential
count using different analyzers (Warner & Reardon,
1991; Briggs, Harrison & Grant, 1999; Wood et al.,
1999; Walters & Garrity, 2000; Gulati et al., 2002).
Our study investigated the stability of complete
blood count (CBC), reticulocyte and white blood cell
(WBC) differential parameters in EDTA-anticoagulated
blood at room temperature (+23 ± 2 �C) and at +4 �C.
The WBC viability fraction (WVF) and leukocyte flag-
ging changes were also followed during these condi-
tions. The measurements were undertaken using an
Abbott CELL-DYN Sapphire hematology analyzer (Ab-
bott Laboratories, Diagnostic Division, Abbott Park, IL,
USA) at Oulu University Hospital Laboratory, Oulu,
Finland.
MATERIALS AND METHODS
Analytical methods
The Abbott CELL-DYN Sapphire Hematology System
uses multi-angle polarized scatter separation (MAPSS)
and focused-flow impedance technologies, combined
with three-color fluorescent flow cytometry. This sys-
tem has a fully automated reticulocyte analysis with
immature reticulocyte fraction (IRF), a five-part WBC
differential, fluorescent DNA staining of nucleated red
blood cell absolute count (NRBC), optical and imped-
ance platelet measurement and fully automated
monoclonal antibody testing for CD 3/4/8 and CD61.
A cyanide-free method is used to measure hemoglobin
(HGB) colorimetrically.
Samples
For these studies, residual fresh (<2 h) EDTA-antico-
agulated samples submitted for routine full blood cell
counts were used. All samples were drawn into Bec-
ton Dickinson Vacutainer K2EDTA tubes (Becton
Dickinson Vacutainer, Cat. No. 388010, Becton Dick-
inson Systems, Plymouth, UK). Sample selection crite-
ria were used only to ensure a representative balance
of normal and abnormal parameter ranges. Data
exclusions were only made when there was clear evi-
dence of incomplete or suboptimal sample aspiration
or processing faults. Samples were not selected accord-
ing to any special treatment ward. All samples were
processed anonymously.
Aging stability
Every sample was analyzed at time point of <2 h
(baseline measurement). Twenty-five different sam-
ples were maintained after the first measurement at
room temperature (+23 ± 2 �C) and were reanalyzed
after 6-, 24-, 48- and 72-h storage. The specimens
randomly collected from routine workflow for these
studies included hematologically abnormal samples.
Abnormal cells may have a different stability from
normal cells. Three of the samples had an abnormal
WBC, with immature granulocytes (IG) being flagged
at the baseline measurement. In the second study, 40
samples were divided into four aliquots and stored
in polystyrene tubes (75 · 12 mm, ref. 55.476.005
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
88 P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS
Sarstedt, D-51588) at +4 �C. One aliquot from each
sample was analyzed after 6, 24, 48 and 72 h. However,
it is not known how aliquoting into nonblood collec-
tion tubes could affect any analysis. It is difficult to
maintain specimens at +4 �C when using automated
sample loading. Samples stored at +4 �C were initially
cold, taken from the refrigerator and then reanalyzed
after a brief warming period (1–5 min) at 23 ± 2 �C. All
of the time measurement groups (6, 24, 48 and 72 h)
included pathologic specimens according to the data
attained from the first measurements. At baseline
measurements, six of the samples in the group stored
for 6 h at +4 �C were flagged as abnormal, with one
variant lymphocyte (VARLYM), three IG, one uniden-
tified fluorescent population (FP?) or NRBC and one
BLAST flags. Of the samples stored for 24 h, nine
were abnormal (five IG and four BAND flags) and
eight of the samples in the group stored for 48 h were
abnormal (five VARLYM, one IG, one FP? or NRBC
and one BAND flags). Of the samples stored for 72 h,
eight were abnormal (four IG, one FP? or NRBC and
three BLAST flags). In the third study, 20 different
samples (six were abnormal from the beginning; four
IG, one BLAST and one FP? or NRBC flags) were
stored at +4 �C in between the measurements (6 h,
24 h and 48 h) and warmed to room temperature
before analysis (altogether four cycles). Additionally,
the stability of the leukocyte flagging was studied. The
suspect population flags BAND, IG, BLAST, VARLYM
and FP? were followed up, as were any positive NRBC
results. Stability profiles were derived from longitudi-
nal comparisons of parameter changes.
Statistical analysis
Statistical processing was performed using personal
computer EXCEL software. The statistical significance of
the differences between the means was assessed by
Student’s t-test as appropriate. P < 0.05 was consid-
ered statistically significant.
RESULTS AND DISCUSSION
Aging stability at +23 ± 2 �C (room temperature)
Changes observed in the CBC results of blood speci-
mens stored at +23 ± 2 �C are summarized in Table 1.
At this temperature, the WBC, RBCi (impedance mea-
surement of red blood cells), RBCo (optical measure-
ment of RBC), HGB, MCH (mean corpuscular
hemoglobin), RDW (red cell distribution width) and
PLTi (impedance measurement of PLT) were found to
be stable over the time, with mean percentage
changes of less than ±10%. However, a statistically
significant change (P < 0.05) was found for RDW at
48- and 72-h storage. The MCV (mean corpuscular
volume), HCT (hematocrit) and MPV (mean platelet
volume) increased over time, producing mean per-
centage changes over 10% within 72 h (10.6%,
11.1% and 11.1%, respectively). The statistically sig-
nificant changes were found at 24-, 48- and 72-h stor-
age for MCV, at 48- and 72-h storage for HCT and at
72-h storage for MPV. The MCHC (mean corpuscular
hemoglobin concentration) trended downward over
time, with change of 13.1% within 48 h. Statistically
significant changes were seen at all storage time
points. There was a proportional 12.1% decrease at
24-h storage, 19.8% decrease at 48-h storage and
24.4% decrease at 72-h storage for PLTo (optical mea-
surement of PLT). The R% (reticulocyte percent) and
RETC (reticulocyte absolute count) and IRF (imma-
ture reticulocyte fraction) fell slightly at 72 h, but not
at earlier time points, at +23 ± 2 �C. A statistically
significant change (P < 0.05) was found for R% at
24- and 72-h storage.
The results of the automated differential count
for specimens stored at +23 ± 2 �C are summarized
in Table 1. Storage of specimens at +23 ± 2 �Cresulted in a fast decrease in WVF (leukocyte viabil-
ity fraction) (P < 0.05 at 6-, 24-, 48- and 72-h stor-
age). After 24 h, the change was already )29.4%
and tended to decrease further with time. Storage of
specimens for 6 h resulted in decreased absolute
basophil counts ()31.1%). In contrast, prolonged
storage up to 72 h showed an increase in basophils.
Basophils represented the smallest fraction of all the
WBC. The standard deviations of the percentage
changes of the absolute basophil counts were very
high. Also, there was a poor analytical day-to-day
precision of the basophil counting (data not shown).
Additionally, one of the samples in this study with
very low absolute basophil counts measured after
the 6-h storage differed from the first measurement
more than the other samples thus giving a high
overall bias%. If this sample was not included in
the analysis, the decrease in the absolute counts of
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS 89
basophils would be 2%. Eosinophil counts had
decreased by 78.8% during 24-h storage and
showed a progressive decline over the time. Statisti-
cally significant (P < 0.05) changes were found for
eosinophils at 24-, 48- and 72-h storage. Also the
changes over 10% were seen for basophils and
monocytes at 72 h.
The stability of NRBC values were assessed and
any NRBC value from the analyzer was considered
as a positive result with discrimination level of
>0%. None of the samples analyzed at baseline, at
6-h and 24-h time points had a positive NRBC value.
Four of the samples analyzed after 48- and 72-h
storage had positive NRBC values (48 h: 0.35, 0.25,
0.27, 0.25 · 109/l and 72 h: 0.54, 0.44, 0.23,
0.39 · 109/l).
Aging stability at +4 �C
The effects of +4 �C storage on the CBC and automated
WBC differential parameters are shown in Table 2 and
Figures 1 and 2. Prolonged storage of specimens for up
to 72 h at +4 �C revealed stability of the CBC parame-
ters, reticulocyte parameters and monocyte counts
within 10% of their original values. However, a statisti-
cally significant change (P < 0.05) was seen in the
MCHC at 48 and 72 h. A decrease in PLTo at +23 ±2 �Cwas corrected by storage at +4 �C. A small increase in
MPV was seen during storage (up to 11.8%). A statisti-
cally significant change (P < 0.05) was observed for
MPV at 24-, 48- and 72-h storage. A substantial
decrease in the absolute counts of eosinophils and in
WVF with storage at +23 ± 2 �C was not observed
Table 1. Changes of CBC parameter and WBC differential values induced by storage of blood at +23 ± 2 �C. Data are
presented as means of parameter values, changes (% from the original value, which is 100%) at different time points
and standard deviation (SD) of changes in parenthesis. Statistically significant changes (P < 0.05; Student’s t-test)between the means are marked with asterisks (*)
<2 h 6 h 24 h 48 h 72 h
Mean Mean (%/SD) Mean (%/SD) Mean (%/SD) Mean (%/SD)
WBC, ·109/l 9.67 9.61 (99.3/2.6) 9.52 (98.2/2.4) 9.42 (96.8/4.2) 9.47 (97.7/2.5)
RBCi, ·1012/l 3.96 3.96 (100.0/0.8) 3.95 (99.8/1.1) 3.98 (100.7/0.7) 3.99 (100.71.0)
RBCo, ·1012/l 4.06 4.05 (99.9/0.6) 4.04 (99.6/0.8) 4.05 (99.5/0.8) 4.06 (100.0/0.6)
Hemoglobin, g/l 121 120 (99.1/1.2) 120 (99.7/0.7) 121 (99.6/0.9) 121 (99.7/0.6)
Hematocrit, l/l 0.36 0.37 (101.2/0.9) 0.39 (105.6/1.1) 0.40 (109.6/1.1)* 0.41 (111.1/1.3)*
MCV, fl 92.2 93.4 (101.3/0.2) 97.8 (105.8/0.6)* 101.5 (109.1/0.8)* 103.1 (110.6/0.8)*
MCH, pg 30.8 30.5 (99.1/1.2) 30.7 (99.7/1.3) 30.4 (98.8/1.0) 30.4 (98.8/1.1)
MCHC, g/l 334 326 (97.7/1.4)* 314 (93.6/1.4)* 299 (88.7/1.5)* 295 (86.9/1.5)*
RDW, %CV 13.2 13.3 (101.1/0.9) 13.8 (104.8/1.1) 14.1 (106.7/1.4)* 14.3 (107.9/1.8)*
PLTo, ·109/l 259 249 (95.8/2.2) 248 (95.8/2.2) 203 (80.2/6.4) 208 (75.6/8.4)
PLTi, ·109/l 252 262 (103.2/8.1) 262 (103.2/8.1) 268 (106.0/8.7) 268 (105.8/10.3)
MPV, fl 8.4 9.0 (102.1/2.5) 8.5 (100.6/3.0) 9.2 (104.6/10.9) 9.7 (111.1/9.1)*
RETC, ·109/l 75.6 74.7 (98.0/3.3) 68.2 (90.6/5.5) 72.2 (97.7/8.0) 65.6 (86.1/9.3)
R% 1.8 1.8 (97.8/3.4) 1.6 (90.4/5.9)* 1.7 (97.3/7.7) 1.6 (85.2/8.8)*
IRF 0.35 0.34 (95.1/11.0) 0.34 (94.1/9.3) 0.33 (91.9/10.2) 0.30 (79.0/17.3)
WVF 0.992 0.987 (98.8/1.0)* 0.789 (70.6/18.9)* 0.568 (13.8/39.4)* 0.441 ()41.6/54.3)*
Neutrophils, ·109/l 6.60 6.67 (101.5/3.3) 6.77 (103.2/3.7) 6.68 (101.4/5.5) 6.74 (102.5/4.8)
Lymphocytes, ·109/l 1.94 1.85 (93.4/7.6) 1.80 (91.3/9.4) 1.81 (92.0/10.5) 1.92 (97.9/10.9)
Monocytes, ·109/l 0.80 0.77 (96.1/9.2) 0.72 (91.2/13.2) 0.79 (91.8/21.9) 0.69 (64.0/41.3)
Eosinophils, ·109/l 0.28 0.28 (95.7/16.6) 0.17 (21.2/66.7)* 0.10 ()188.2/228.1)* 0.07 ()440.5/454.1)*
Basophils, ·109/l 0.04 0.05 (66.3/87.9) 0.05 (91.1/77.4) 0.05 (98.4/64.7) 0.05 (110.2/53.8)
WBC, white blood cell; RBCi, impedance measurement of red blood cells; RBCo, optical measurement of red blood
cell; HGB, hemoglobin; HCT, hematocrit; MCV, mean corpuscular volume; MCHC, mean cell hemoglobin concentra-
tion; RDW, red cell distribution width; PLTo, optical measurement of platelets; PLTi, impedance measurement of plate-
lets; MPV, mean platelet volume; WVF, white blood cell viability fraction, RETC, reticulocyte count; R%, percentage
reticulocytes; IRF, immature reticulocyte fraction.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
90 P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS
Tabl
e2.
Ch
an
ges
of
CB
Cpara
mete
ran
dW
BC
dif
fere
nti
alvalu
es
indu
ced
by
stora
ge
ofblo
od
at
+4
�C.D
ata
are
pre
sen
ted
as
mean
sofpara
mete
rvalu
es,
chan
ges
(%fr
om
the
ori
gin
al
valu
e,
wh
ich
is100%
)at
dif
fere
nt
tim
epoin
tsan
dSD
sof
chan
ges
inpare
nth
esi
s.Th
est
ati
stic
all
ysi
gn
ifica
nt
chan
ges
(P<
0.05;Student’s
t-test)betweenthemeansare
marked
withasterisks(*)
Batc
h1
Batc
h2
Batc
h3
Batc
h4
<2
h6
h<
2h
24
h<
2h
48
h<
2h
72
h
Mean
Mean
(%/S
D)
Mean
Mean
(%/S
D)
Mean
Mean
(%/S
D)
Mean
Mean
(%/S
D)
n=
40
(n=
20
inR
ETC
,R
%an
dIR
F)
WB
C,
·10
9/l
9.8
59.8
7(1
00.5
/1.8
)8.9
28.7
6(9
7.1
/5.8
)8.8
68.9
7(1
00.5
/4.6
)8.0
27.6
7(9
4.6
/4.5
)
RB
Ci,
·10
12/l
3.6
93.6
9(9
9.9
/0.6
)3.7
33.7
5(1
00.5
/1.5
)3.8
23.9
1(9
9.6
/0.6
)3.9
63.9
3(9
8.3
/4.0
)
RB
Co,
·10
12/l
3.7
83.7
8(1
00.1
/0.5
)3.8
33.8
3(9
9.9
/1.7
)3.9
33.9
1(9
9.6
/0.7
)4.1
04.0
6(9
8.1
/3.9
)
Hem
oglo
bin
,g/l
114
114
(99.4
/0.5
)116
118
(100.9
/1.8
)118
118
(100.3
/0.6
)124
123
(98.9
/3.1
)
Hem
ato
crit
,l/
l0.3
40.3
4(9
9.7
/0.7
)0.3
50.3
5(1
01.4
/1.9
)0.3
60.3
6(1
01.6
/1.0
)0.3
70.3
7(9
9.9
/3.8
)
MC
V,
fl93.4
93.3
(99.8
/0.3
)93.7
94.5
(100.9
/0.7
)93.4
94.7
(101.3
/0.3
)93.5
95.1
(101.7
/0.9
)
MC
H,
pg
31.1
30.9
(99.5
/0.7
)31.2
31.3
(100.5
/1.2
)31.1
31.1
(99.7
/1.1
)31.2
31.4
(100.4
/1.4
)
MC
HC
,g/l
332
331
(99.6
/0.9
)333
331
(99.5
/1.2
)333
328
(98.6
/1.3
)*334
330
(98.8
/1.7
)*
RD
W,
%C
V13.4
13.4
(99.9
/0.5
)13.4
13.3
(99.1
/1.7
)13.1
13.1
(99.5
/0.9
)12.9
12.8
(99.2
/1.3
)
PLTo,
·10
9/l
262
268
(102.2
/2.5
)286
268
(91.5
/15.9
)354
338
(94.8
/5.0
)294
298
(98.8
/3.0
)
PLTi,
·10
9/l
275
277
(100.7
/2.9
)301
286
(92.8
/14.5
)375
373
(98.4
/6.7
)308
317
(100.3
/3.5
)
MPV
,fl
8.3
8.5
(101.9
/2.1
)8.1
8.6
(106.4
/3.4
)*7.7
8.7
(110.2
/3.5
)*8.0
9.0
(111.8
/4.2
)*
RE
TC
,·1
09/l
73.5
74.5
(100.4
/5.9
)97.2
96.3
(95.5
/12.5
)76.0
75.4
(98.8
/5.7
)86.1
83.3
(95.5
/6.4
)
R%
2.2
2.2
(100.7
/5.7
)2.9
2.8
(92.0
/18.7
)2.1
2.1
(98.8
/5.7
)2.5
2.4
(94.3
/6.1
)
IRF
0.4
20.4
3(1
03.7
/5.5
)0.4
30.4
4(9
9.4
/13.2
)0.3
80.3
8(1
02.1
/6.8
)0.4
40.4
5(1
04.0
/5.3
)
WV
F0.9
92
0.9
80
(98.3
/2.2
)0.9
88
0.9
5(9
5.4
/4.6
)*0.9
90
0.9
10
(90.5
/5.4
)*0.9
82
0.8
00
(74.9
/16.1
)*
Neu
troph
ils,
·10
9/l
7.5
77.6
3(1
01.2
/2.5
)6.6
66.5
3(9
4.4
/12.4
)6.1
26.1
8(1
00.7
/3.9
)5.1
84.8
5(8
9.5
/13.2
)
Lym
ph
ocy
tes,
·10
9/l
1.2
71.2
1(9
4.5
/6.9
)1.4
01.4
2(9
5.5
/13.9
)1.6
21.5
8(9
2.7
/14.8
)1.9
01.8
8(9
3.9
/18.1
)
Mon
ocy
tes,
·10
9/l
0.9
30.8
4(1
00.8
/8.0
)0.6
30.6
7(1
01.2
/16.1
)0.7
90.9
0(1
08.4
/13.0
)0.6
90.7
2(1
01.6
/12.2
)
Eosi
noph
ils,
·10
9/l
0.1
40.1
6(1
09.4
/19.7
)0.2
50.1
8(7
5.6
/56.2
)0.3
10.3
1(9
8.4
/17.7
)0.2
40.2
3(9
4.7
/14.9
)
Baso
ph
ils,
·10
9/l
0.0
30.0
4(6
4.4
/88.5
)0.0
30.0
4(8
1.3
/81.3
)0.0
40.0
3(4
4.0
/125.6
)0.0
40.0
4(9
2.6
/74.0
)
WB
C,
wh
ite
blo
od
cell
;R
BC
i,im
pedan
cem
easu
rem
en
tof
red
blo
od
cell
s;R
BC
o,
opti
cal
measu
rem
en
tof
red
blo
od
cell
;H
GB
,h
em
oglo
bin
;H
CT,
hem
ato
crit
;M
CV
,m
ean
corp
usc
ula
rvolu
me;
MC
HC
,m
ean
cell
hem
oglo
bin
con
cen
trati
on
;R
DW
,re
dce
lldis
trib
uti
on
wid
th;
PLTo,
opti
cal
measu
re-
men
tof
pla
tele
ts;
PLTi,
impedan
cem
easu
rem
en
tof
pla
tele
ts;
MPV
,m
ean
pla
tele
tvolu
me;
RE
TC
,re
ticu
locy
teco
un
t;R
%,
perc
en
tage
reti
culo
cyte
s;IR
F,
imm
atu
rere
ticu
locy
tefr
act
ion
WV
F,
wh
ite
blo
od
cell
via
bilit
yfr
act
ion
.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS 91
during storage at +4 �C. However, a statistically significant
change was seen for WVF at 24-, 48- and 72-h storage.
A slight decrease in the absolute counts of lymphocytes
and absolute counts of neutrophils was present with
the storage at +4 �C, which was not seen at +23 ± 2 �C.
One sample analyzed at baseline and at 6 h had a
positive NRBC value in both measurements (0.88 and
1.00 · 109/l, respectively). None of the samples ana-
lyzed at baseline and 24 h had a positive NRBC value.
Five of the samples in the group analyzed at baseline
and 48 h had positive NRBC values in the latter mea-
surements (0.30, 0.30, 0.30, 0.25 and 0.18 · 109/l).
Finally, four of the samples in the group analyzed at
baseline and 72 h had positive NRBC values at 72 h
(0.47, 0.44, 0.24 and 0.57 · 109/l).
Aging stability at refrigeration and +23 ± 2 �C (room
temperature) cycles
To further evaluate the effect of storage at +4 �C and
rewarming to +23 ± 2 �C, 10 normal specimens were
assayed at baseline, 6, 24 and 48 h after collection
(Table 3). The samples were same for all the time-
point measurements. The only changes from the origi-
nal values at 6 h were seen with absolute basophil
counts ()24.6%). They progressively decreased 150%
from the original value at 72 h. The statistically signif-
icant change (P < 0.05) was observed at 48 h. A pro-
gressive loss of lymphocytes was also seen during
cycles up to 48 h. All the other parameters were sta-
ble under these conditions, reaching a less than 10%
change from the original values. However, a statisti-
cally significant change was observed for WVF at 24
and 48 h. Positive NRBC values were detected with
one sample at each measuring time point (0.85, 0.88,
0.95 and 1.08 · 109/l, respectively) and with two
samples 48 h (0.28 and 0.26 · 109/l, respectively).
CELL-DYN Sapphire flagging stability
Additional analysis of CELL-DYN Sapphire flagging
stability (Table 4) revealed better storage conditions
5060708090
100110120130
WBC
RBCi
RBCoHGB
HCTMCV
MCH
MCHCRDW
PLToPLTi
MPV
WBC
RBCi
RBCoHGB
HCTMCV
MCH
MCHCRDW
PLToPLTi
MPVW
BCRBCi
RBCoHGB
HCTMCV
MCH
MCHCRDW
PLToPLTi
MPV
WBC
RBCi
RBCoHGB
HCTMCV
MCH
MCHCRDW
PLToPLTi
MPV
% c
han
ge
fro
m in
itia
l val
ues
6 h
5060708090
100110120130
% c
han
ge
fro
m in
itia
l val
ues
24 h
5060708090
100110120130
% c
han
ge
fro
m in
itia
l val
ues
48 h
5060708090
100110120130
% c
han
ge
fro
m in
itia
l val
ues
72 h
Figure 1. Changes of CBC parameter values induced by storage of blood at +4 �C. Data are presented as mean
changes (% from the original value, which is 100%) at different time points (error bars represent the ±SDs). WBC,
white blood cell; RBCi, impedance measurement of red blood cells; RBCo, optical measurement of red blood cell;
HGB, hemoglobin; HCT, hematocrit; MCV, mean corpuscular volume; MCHC, mean cell hemoglobin concentration;
RDW, red cell distribution width; PLTo, optical measurement of platelets; PLTi, impedance measurement of plate-
lets; MPV, mean platelet volume.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
92 P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS
for samples at +4 �C. IG flag appeared at 24 h at both
temperatures. BAND and IG flags also appeared at
24 h and the number of samples with IG flag progres-
sively increased over time at both temperatures.
BLAST cell flagging did not appear to be affected by
storage time or temperature.
DISCUSSION
Manufacturers of automated analyzers and published
literature often quote that blood specimens, kept at
either at room temperature or at +4 �C for up to 24 h,
generally yield reliable results for complete blood
count and automated differential count.
Recently, Muller et al. (2006), in a multicenter
evaluation of Abbott CELL-DYN Sapphire, studied the
aging stability with 10 normal EDTA-anticoagulated
samples at +4 �C and room temperature. They found
that proportions of nonviable leukocytes (predomi-
nantly neutrophils) progressively increased between
12- and 72-h storage at room temperature compared
with +4 �C, where viability remained above 90% for
up to 36 h. At both temperatures, the absolute WBC
count remained stable. Relatively little variation was
seen in the analysis of leukocyte population stability
stored at +4 �C up to 72 h, whereas at room tempera-
ture all parts of the leukocyte differential were stable
for up to 48 h, with the exception of a significant
decline (statistical processing was performed using
ANALYSE-IT Software, Leeds, UK) in the eosinophil frac-
tion after 12-h storage. Storage changes at +4 �C did
not exceed 10% during the 72 h period for RBC, reti-
culocytes and platelet count. However, at room tem-
perature, all parameters were stable for up to 24 h
with subsequent progressive changes exceeding 10%
being seen with the MPV (increase), immature reticu-
locyte fraction (decrease), percentage reticulocytes
(decrease) and optical platelet count (decrease). A less
significant trend for increasing MCV values after
hours of storage was noted. They also followed the
flagging stability of the CELL-DYN Sapphire and they
noted that, with the exception of IG flag (room
WVF
NEU#
LYMF#
MONO#
EOS#
BASO#
WVF
NEU#
LYMF#
MONO#
EOS#
BASO#
WVF
NEU#
LYMF#
MONO#
EOS#
BASO#
0
50
100
150
200
% c
han
ge
fro
m in
itia
l val
ues
6 h 24 h
020406080
100120140160180
% c
han
ge
fro
m in
itia
l val
ues
0
20
40
60
80
100
120
140
160
% c
han
ge
fro
m in
itia
l val
ues
48 h 72 h
0
50
100
150
200
WVF
NEU#
LYMF#
MONO#EOS#
BASO#
% c
han
ge
fro
m in
itia
l val
ues
Figure 2. Changes of WBC differential parameter values induced by storage of blood at +4 �C. Data are presented as
mean changes (% from the original value, which is 100%) at different time points (error bars represent the ±SDs).
WVF, white blood cell viability fraction; NEU, neutrophil; LYMP, lymphocyte; MONO, monocytes; EOS, eosino-
phils; BASO, basophils.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS 93
temperature), reliable flagging could be obtained for
up to 24 h, independent of storage temperature.
Flagging efficiencies of hematology analyzers have
also been found to be associated with total WBC
counts, with the flagging sensitivities being rather
poor in leukocytopenic samples and highest in sam-
ples with more than 10 · 109 WBC/l (Ruzicka et al.,
2001). In contrast, the flagging specificities were
found to be poorer with rising WBC counts. Leuko-
cytopenic and thrombocytopenic specimens have also
been found to have the greatest day-to-day mean per-
centage changes at room temperature and the widest
day-to-day standard deviations compared with the
changes observed for specimens with normal or high
initial respective counts (Gulati et al., 2002). Further-
more, the platelet counts of thrombocytopenic speci-
mens tend to increase over time, and samples with
higher initial platelet count tend to decrease.
Although the precision of new automated hemato-
logy analyzers have been improved, the precision per-
formance of each parameter has to be considered
when finalizing the results. In particular, the precision
of WBC count and differential count have been found
to be different during storage at room temperature
and at +4 �C (Wood et al., 1999). Also, the precision
might be different in different concentration levels. In
our analytical evaluation of CELL-DYN Sapphire (data
not shown), we found that between-day precisions
were higher for monocytes (17.8% at 0.2 · 109/l vs.
7% at 1.7 · 109/l), eosinophils (19.5% at 0.1 · 109/l
vs. 6.2% at 0.5 · 109/l) and basophils (42% at
0.1 · 109/l vs. 10% at 0.7 · 109/l) in the low concen-
trations compared with high concentration levels. Also
the imprecision of PLTo and RETC were �4.6% in
our study, thus may have had some effect on the
results of these stability studies.
Table 3. Changes of CBC parameter and WBC differential values induced by storage of blood at +4 �C and rewarming
before analyzing. Data are presented as means of parameter values, changes (% from original value, which is 100%) at
different time points and SDs of change percentages in parenthesis. The statistically significant changes (P < 0.05;Student’s t-test) between the means are marked with asterisks (*)
<2 h 6 h 24 h 48 h
Mean Mean (CV%/SD) Mean (CV%/SD) Mean (CV%/SD)
n = 20
WBC, ·109/l 9.30 9.31 (100.3/1.2) 9.26 (99.2/1.0) 9.10 (98.0/2.8)
RBCi, ·1012/l 3.72 3.73 (100.1/0.6) 3.72 (99.9/0.9) 3.73 (100.2/0.8)
RBCo, ·1012/l 3.80 3.80 (100.1/0.5) 3.78 (99.5/0.8) 3.78 (99.5/0.8)
Hemoglobin, g/l 116 115 (99.4/0.4) 173.2 (105.5/9.9) 116 (100.1/0.4)
Hematocrit, l/l 0.35 0.35 (99.7/0.6) 0.35 (101.0/0.9) 0.35 (101.2/0.9)
MCV, fl 94.1 93.7 (99.6/0.3) 95.1 (101.1/0.2) 95.1 (101.0/0.2)
MCH, pg 31.3 31.0 (99.3/0.7) 31.3 (100.2/0.9) 31.2 (100.0/0.9)
MCHC, g/l 332 331 (99.7/0.8) 329 (99.1/1.1) 328 (98.9/1.0)
RDW, %CV 13.4 13.4 (99.7/0.7) 13.5 (100.1/0.6) 13.4 (99.4/1.2)
PLTo, ·109/l 198 202 (102.0/2.6) 194 (98.1/2.9) 194 (98.0/4.5)
PLTi, ·109/l 207 211 (101.9/3.3) 205 (98.2/3.9) 204 (96.6/8.7)
MPV, fl 8.82 9.0 (102.1/2.5) 9.35 (105.7/2.5) 9.63 (108.5/2.4)
WVF 0.990 0.970 (97.2/4.6) 0.975 (98.6/0.9)* 0.950 (96.1/1.5)*
Neutrophils, ·109/l 6.95 7.04 (101.6/1.8) 7.06 (101.6/1.3) 6.99 (100.9/2.7)
Lymphocytes, ·109/l 1.30 1.22 (92.5/6.2) 1.15 (84.9/8.3) 1.10 (79.8/11.6)
Monocytes, ·109/l 0.87 0.85 (98.5/7.5) 0.88 (100.2/8.5) 0.86 (98.4/8.7)
Eosinophils, ·109/l 0.15 0.17 (108.4/18.5) 0.15 (100.3/15.7) 0.14 (92.3/19.6)
Basophils, ·109/l 0.03 0.04 (75.8/64.1) 0.03 (20.3/129.9) 0.02 ()48.3/191.6)*
WBC, white blood cell; RBCi, impedance measurement of red blood cells; RBCo, optical measurement of red blood
cell; HGB, hemoglobin; HCT, hematocrit; MCV, mean corpuscular volume; MCHC, mean cell hemoglobin concentra-
tion; RDW, red cell distribution width; PLTo, optical measurement of platelets; PLTi, impedance measurement of plate-
lets; MPV, mean platelet volume; WVF, white blood cell viability fraction.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
94 P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS
In general, if we only contemplate the stability of
the parameter changes to be acceptable within ±10%
of the original values, not from the clinically signifi-
cant point of view (MCV increasing over time etc.),
the CBC parameters were stable for up to 24–48 h at
room temperature, except PLTo.
In the CELL-DYN Sapphire system, thrombocytes
are measured by both PLTi and PLTo and the analyzer
creates an alert if the difference between PLTo and
PLTi exceeds the present limit. For specimens aged
over 24 h, the PLTi could be used instead of the rou-
tine PLTo to yield more reliable results. WBC differen-
tial parameters, except eosinophils, were stable to
48 h at +23 ± 2 �C, confirming the results of Muller
et al. (2006). CBC parameters were stable at 72 h,
except MPV, which slightly increased between 48 and
72 h, at +4 �C. WBC differentials were stable at
48–72 h, a slight decrease was seen in absolute neu-
trophils, lymphocytes and lymphocyte percents. The
imprecision of basophils using the CELL-DYN Sap-
phire is high, reflecting that the cell types that occur
in low numbers can yield possible unreliable results in
the stability studies. Differences in the sample size and
sample type between our study and that of Muller
et al. (2006) may at least partly account for the differ-
ent findings. Additionally, the exact calibration of the
WBC differential channels may affect the results,
especially the flagging efficiencies. For example, in
our results, the BAND flag appeared as early as 24 h,
but Muller et al. (2006) reported that this flag
appeared at 36 h. However, the accuracy of the flags
should be checked by microscopic review by experi-
enced observers, which was not undertaken in our
study. Also, a fault of our study was that the changes
were not observed in the low and high levels, for
example, using leukopenic and thrombocytopenic
specimens.
From the review of the literature and according to
our results, we conclude that CBC and differential
parameters should be analyzed as soon as possible
after collection. If analysis is delayed for a longer per-
iod of time specimens should preferably be main-
tained at +4�C. The date and time collection should
be clearly indicated and, in cases where a delay in
analysis does not enable accurate results, the corre-
sponding parameter should be omitted and substituted
by a comment.
ACKNOWLEDGEMENTS
We would like to thank Oulu University Hospital Lab-
oratory staff and Dr Veli Isomaa for the valuable help.
Table 4. Stability of CELL-DYN Sapphire flagging
alerts. Data indicate the number of samples with flags
at each time point at +23 ± 2 �C and at +4 �C
Flag <2 h 6 h 24 h 48 h 72 h
+23 ± 2 �C, n = 25
VARLYM 0 1 0 0 13
IG 3 6 23 25 22
FP? or NRBC 0 1 4 16 23
BAND 0 3 12 17 14
BLAST 0 0 0 0 0
+4 �C,
n = 40
Batch 1
(n = 40)
Batch 2
(n = 40)
Batch 3
(n = 39)
Batch 4
(n = 30)
<2 h 6 h <2 h 24 h <2 h 48 h <2 h 72 h
VARLYM 1 0 0 6 5 6 0 19
IG 3 6 5 23 1 23 4 21
FP? or NRBC 1 1 0 13 1 24 1 26
BAND 0 6 4 21 1 24 3 18
BLAST 1 3 0 3 0 0 0 1
VARLYM, variant lymphocyte; IG, immature granulo-
cytes; FP?, fluorescent population; NRBC, nucleated red
blood cell absolute count.
REFERENCES
Briggs C., Harrison P. & Grant D. (1999)
Performance evaluation of the Sysmex
XE-2100� automated haematology
analyzer. Sysmex Journal International
9, 113–119.
Buttarello M. (2004) Quality specification
in haematology: the automated blood
cell count. Clinica Chimica Acta 346,
45–54.
England J., Rowan R., van Assenfelft O.,
Bull B., Coulter W., Fujimoto K.,
Groner W., Richardson-Jones A., Klee
G., Koepke A., Lewis S., McLaren C.,
Shinton N., Tatsumi N. & Verwilghen R.
for the International Council for Stan-
dardization in Haematology. (1993).
Recommendations for the International
Council for Standardization in Haema-
tology for ethylenediamine-tetraacetic
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS 95
acid anticoagulation of blood for blood
cell counting and sizing. American
Journal of Clinical Pathology 100, 371–
372.
Gulati G.L., Hyland L.J., Kocher W. &
Schwarting R. (2002) Changes in auto-
mated complete blood cell count and
differential leukocyte counts results inc-
uded by storage of blood at room tem-
perature. Archives of Pathology and
Laboratory Medicine 126, 336–342.
Muller R., Mellors I., Johannessen B.,
Aarsand A., Kiefer P., Hardy J., Kendall
R. & Scott C.S. (2006) European multi-
center evaluation of the Abbott Cell-
Dyn Sapphire hematology analyzer.
Laboratory Hematology 12, 15–31.
National Committee for Clinical Labora-
tory Standards (1996).Reference Leuko-
cyte Differential Count (Proportional)
and Evaluation of Instrumental Meth-
ods; Approved Standard. H20A. Wayne,
PA, NCCLS.
National Committee for Clinical Labora-
tory Standards. (1997) Methods for
Reticulocyte Counting (Flow Cytometry
and Supravital Dyes). Approved Guide-
line. H44-A. Wayne, PA, NCCLS.
Ruzicka K., Veitl M., Thalhammer-Scher-
rer R. & Schwarzinger I. (2001) The
new hematology analyzer Sysmex XE-
2100. Performance evaluation of a
novel blood cell differential technology.
Archives of Pathology and Laboratory
Medicine 125, 391–396.
Walters J. & Garrity P. (2000) Perfor-
mance evaluation of the Sysmex XE-
2100 hematology analyzer. Laboratory
Hematology 6, 83–92.
Warner B.A. & Reardon D.M. (1991) A
field evaluation of the Coulter STKS.
American Journal of Clinical Pathology
95, 207–217.
Wood B.L., Andrews J., Miller S. & Sa-
bath D.E. (1999) Refrigerated storage
improves the stability of the complete
blood cell count and automated differ-
ential. American Journal of Clinical
Pathology 112, 687–695.
� 2008 The Authors
Journal compilation � 2008 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2009, 31, 87–96
96 P. HEDBERG AND T. LEHTO AGING STABILITY OF AUTOMATED HEMATOLOGY PARAMETERS