A N N A L E S

U N I V E R S I TAT I S M A R I A E C U R I E - S K L O D O W S K A
L U B L I N P O L O N I A
SECTIO DDD 2011 VOL.XXIV,N3,
Copyright2011MedicalUniversityinLublin
Chair and Department oI Laboratory Diagnostics, Medical University oI Lublin
IWONA KAZNOWSKA-BYSTRYK
7KH DXWRPDWHG KHPDWRORJ\ DQDO\]HUV
Automatyczne analizatory hematologiczne
INTRODUCTION
The complete blood count (CBC) and diIIerential leukocyte count (DLC) are the backbone oI any
laboratory evaluation, not only used by general practitioners, but also by other medical specialties. They
provide valuable inIormation about the blood and to some extent about the bone marrow, which is the
blood-Iorming tissue. The CBC and DLC are used to diagnose anemia, to identiIy acute and chronic
illnesses, bleeding tendencies, and white blood cell disorders, Ior example leukemia. During the last
quarter century, blood cell analysis has progressed Irom the use oI labor-intensive manual procedures
to the use oI highly automated instruments. Modern hematology analyzers provide new additional
parameters enabling earlier and more precise diagnosis. Moreover, they incorporate fow cytometry,
robotics and expert system technologies. ThereIore, it is crucial that obtained parameters are clear Ior
clinicians, and the results completely used.
DEVELOPMENT OF AUTOMATION - ELECTRICAL IMPEDANCE METHODS
Hematology blood cell analyzers have become increasingly sophisticated during the past decade.
Until the mid-1950s, laboratories perIormed hematology testing manually, thus making it time consuming
and labor intensive. The frst automated analyzer was introduced in 1956 by Wallace Coulter. This was
a single channel instrument that revolutionized the tedious and imprecise manual chamber counting
methods |1|. The frst hematology analyzers were based upon the principle oI impedance (i.e. resistance
to current fow). In this method a small sample oI the blood is aspirated into a chamber and diluted with
an isotonic saline solution. The analyzer prepares two dilutions: one lysed and the other unlysed. AIter
the frst dilution is measured, the white blood cells (WBC) and hemoglobin (Hb) values are displayed
on the screen oI the instrument; meanwhile, the analyzer processes the second dilution, which measures
red blood cells (RBC), hematocrit (Hct), mean corpuscular volume (MCV) and platelet count (PLT)
|2, 11|. A small portion oI the diluted fuid in each bath is allowed to fow past a small aperture. An
electrical current is produced in each aperture by two electrodes, one on the inside and the other on the
outside oI the aperture. The saline solution is responsible Ior conducting current between the electrodes.
The cells move through the aperture, then displace a volume oI electrolyte equal to its size. The cell acts
as an electrical resistor, and impedes the fow oI current. This produces a voltage pulse, the magnitude
7
64
oI which is proportional to the size oI the cell. Instrument electronics are adjusted to discriminate
voltage pulses produced by diIIerent cells (called thresholds). For example, the threshold Ior counting
a RBC is equivalent to a cell volume oI 36 f or higher. Voltage pulses that are equivalent to volumes oI
220 f are counted as platelets. Although the methods directly determining RBC, WBC and PLT counts
diIIer, the 50-year-old impedance procedure continues to be the most Irequently used, either alone or
in combination with the others. Small hematology analyzers based on this method are Iound in small
laboratories, and are also used in hospitals as the backups Ior larger analyzers |2, 11|.
HEMOGLOBIN, HEMATOCRIT, RBC COUNT AND RBC PARAMETERS
Automated counting methods Ior RBC have been based originally on electrical impedance or later
on light-scattering techniques. The hemoglobin concentration is measured optically using the solution
in the WBC bath. The lysing agent contains potassium cyanide that reacts with the hemoglobin to Iorm
cyanmethemoglobin. The color intensity, measured in separate cuvette, is read spectrophotometrically
at 540 nm and is proportional to the concentration oI hemoglobin. Some instruments use a modifed
cyanmethemoglobin procedure, and others use cyanide-Iree colorimetric determinations |5|. The Hct
is a test that measures the volume oI blood in percent that is comprised oI the RBC. Automated cell
counters calculate the Hct by multiplying the RBC count by the MCV. The three main RBC indices
are used to determine the average size and hemoglobin content oI the RBC, and they help to determine
the cause oI anemia. MCV is the average size oI the red blood cells expressed in Iemtolitres (f). The
MCV is measured by electronic cell counters, usually by dividing the summation oI the cell volumes
by the RBC count. Mean corpuscular hemoglobin (MCH) is the average amount oI hemoglobin inside
an RBC expressed in picograms. The MCH is calculated by dividing the hemoglobin concentration by
the RBC count. Mean corpuscular hemoglobin concentration (MCHC) is the average concentration
oI hemoglobin in the RBC. It is calculated by dividing the hemoglobin by the Hct. Some analyzers
calculate additional parameters to screen Ior various types oI anemia. The mean oI the RBC distribution
histogram, based on electrical impedance, is the MCV, and the coeIfcient oI variation, or sometimes
the standard deviation, is the red cell distribution width (RDW). The RDW provides some insight and
quantifcation into the variation in red cell size, or anisocitosis |2|.
Some analyzers (e.g. Advia, Siemens) have generated characteristic red cells cytograms,
representing the red cell volume and hemoglobin concentration (V/HC). On the V/HC cytogram , Hb
concentration is plotted along the x-axis and the cell volume along the y-axis. It is easy to interpret,
Ior the blood samples with normal red cells show most RBC to Iall in the central square oI cytogram,
because oI their normal size and Hb content |5|.
THE PLATELET COUNT
The platelet count (PLT) is most oIten measured by impedance counting. Using this principle PLT
and RBC which are both analyzed in the same channel, are discriminated according to their volume and
volume histograms are generated next. For PLT, the histogram generates a log curve iI the distribution oI
PLT volume fts that oI a (log) normal distribution: eventually all particles located under the ftted curve
are considered as PLT. Mean PLT volume ranges Irom 6 to 10 f, but impedance-type counters analyze
particles ranging Irom 2 to 20 f and, according to the ftted curve, the upper threshold that discriminates
Iwona Kaznowska-Bystryk
65
PLT Irom RBC may either be at 36 f or may vary automatically depending on the characteristics oI
individual blood sample (Sysmex). Instrument fags are triggered Ior cases corresponding to inability to
separate clearly PLT Irom RBC |10|.
On laser-type analyzer (Abbott, Bayer, others), each particle passes through a laser beam and scatters
light that is detected by a photodiode. The amount oI light scattered (at one, two or even Iour angles Ior
some analyzers), is proportional to the area and thereIore to the volume oI the particle. PLT are indentifed
on a scatter histogram based on their volume and reIractive index values. Some analyzers (e.g. Abbott)
provide up to three counts on the same dilution iI required, corresponding to optical, impedance, and
immunological counts (CD61). Other analyzers (e.g. Sysmex) may determine, iI required, an optical PLT
count together with the reticulocyte count, aIter the use oI RNA fuorescent stain |10|.
An abnormal mean platelet volume or platelet histogram indicates that morphological platelet
abnormalities are present and the platelets should be observed Irom a stained blood flm to characterize
the abnormality. Manual evaluation is perIormed when the PLT is very low, platelet clumping is
observed, or abnormally large (giant) platelets are present. OIten these abnormalities and others, such
as cryoglobulinemia, cell Iragmentation (hemolysis), and microcytic RBC are signaled by abnormal
RBC and PLT indices and abnormal population fags |1|.
It is important to note that the most oIten abnormal result (about 1.0-2.0 Ior hospitalized patients)
is pseudothrombocytopenia. Spurious thrombocytopenia occurs in several circumstances related to the
presence oI ethylenediamine tetra-acetic acid (EDTA) used as the anticoagulant. Mechanism oI EDTA-
dependent platelet agglutination is related to circulating antibodies directed against normally hidden epitopes
in the glycoprotein alpha IIb/beta IIIa complex Irom PLT membrane exposed only in the presence oI EDTA.
Other spuriously low PLT counts may be related to EDTA, including PLT rosetting around white blood
cells (satellitism) and PLT-WBC aggregates |10|. Spurious increase oI PLT count may be related to several
situations, including Iragmented red blood cells, cytoplasmic Iragments oI nucleated cells, cryoglobulins,
bacteria or Iungi, and lipids. Flags generated in several oI these situations alert the operator on possible
abnormal fndings and may identiIy the problem. Analysing only PLT parameters is not suIfcient: in many
situations the WBC diIIerential scattergram is oI crucial help Ior fagging |1, 10, 11|.
DIFFERENTIAL LEUKOCYTE COUNT (DLC)
Automated diIIerential leukocyte count (DLC) implementation is a more complicated process and has
been through several diIIerent technologies. The frst automated analyzer used cell volume to provide a
three-part diIIerential analysis: neutrophil, lymphocyte and monocyte cell counts |2, 6|. The voltage pulses
produced by the WBC depend upon the size oI the cell and its nuclear density. ThereIore, the pulses may
be analyzed to diIIerentiate between the types oI WBC Iound. For example, lymphocytes are the smallest
WBC and comprise the lower end oI the size scale. Monocytes, prolymphocytes, and immature granulocytes
comprise the central area oI the WBC histogram, and mature granulocytes comprise the upper end. Many
small hematology instruments use the three-part diIIerential WBC separation because it is inexpensive
and generally reliable. These analyzers are intended Ior low-volume laboratories, or as a backup Ior larger
laboratories. It should be remembered that identifcation oI monocytes, immature myeloid cells and nucleated
red blood cells (NRBC) by those instruments is diIfcult. In general, a peripheral smear review is necessary to
validate the automated diIIerential generated by the instrument |1, 3, 6, 11|.
The automated hematology analyzers
66
Current hematology instruments combine laser technology, impedance, radio Irequency,
direct current, optimized temperatures and volumes, and staining in various ways to
maximize sensitivity and specifcity Ior the automated DLC. As a result oI improvements in
the technologies, the accuracy, precision and standardization oI leukocyte diIIerentials have
increased substantially. Examples oI multichannel counters which have been introduced
since 1995 include those produced by Coulter-Beckman Electronics, Sysmex, Abbott, ABX
Diagnostics and Bayer-Technicon (now Siemens) |2, 6|.
The Coulter-Beckman STKS, like other instruments, includes a 5-parameter DLC reportable in both
relative percentage and absolute numbers. The analyzer uses the volume, conductivity and light scatter
(VCS) technology to generate the DLC. The V measures cell volume and number by impedance, the
C measures the nuclear size and density oI each cell with high-Irequency current, and the S scattering
Irom a laser source measures internal structure, granularity, and surIace characteristics oI cells as well
as provides inIormation on the shape and structure oI individual cells |2, 6, 8|.
The Sysmex hematology analyzers simultaneously measure both cell size and intracellular inIormation
by using the radio Irequency/direct current method (RF/DC). This method uses direct current combined
with radio Irequency current to count and classiIy cell types. Choosing the appropriate Irequency allows
Ior detection oI certain WBC Ieatures. Another channel detects immature WBC, where a specifc reagent
lyses the cytoplasm oI normal WBC, but leaves immature WBC intact to be counted |2, 3|.
The Abbott hematology analyzers use both impedance and laser optical technology. Placed in a
single fle by hydrodynamic Iocusing and laminar fow, cell stream through a fow cell Ior counting and
analysis. WBC are counted and classifed by laser light-scattering data, using a multi-angle polarized
scatter and separation technique. The angle oI scattering is a Iunction oI cell size, reIractive index,
nuclear shape, nuclear-cytoplasmic ratio, and granularity. One oI the recent Cell-Dyn instrument uses
optical scatter and fuorescence technology with an argon laser to count NRBC and to separate DLC
parameters. The next Cell-Dyn instrument oIIers reticulocyte, CD4/CD8, and CD64 counts |2, 6|.
The Advia (Bayer, now Siemens) DLC uses light scatter readings oI cell size and peroxidase activity
aIter exposure to a peroxidase - chromogen reagent to enumerate the neutrophils, eosinophils, monocytes,
and lymphocytes. The basophils are counted in a diIIerent fow cytometric operation; a lysing reagent
is added, stripping the other WBC oI their cytoplasm, and leaving the basophils intact and countable.
Another cluster oI abnormal WBC, designated as large unstained cells (LUC), sometimes results. These
are peroxidase-negative cells that are either large lymphocytes, virocytes, or blast cells |2, 6|.
The hematology analyzer oI ABX Diagnostics oIIers a 5-part automated DLC count by
combining impedance, light transmission, cytochemistry, and fuoro-fow cytometry into their Double
Hydrodynamic Sequential System. This approach combines a primary Iocused fow to measure
impedance and a second Iocused fow to detect light |2|.
Generally, analyzers reduced the need to manually veriIy abnormal or fagged patient results by
combining diIIerent methods to maximize sensitivity and specifcity. This has improved the accuracy in
counting low-incidence cell such as NRBC, percentage oI hypochromic cells ( HYPO), and various
immature cells, thus expanding the list oI tests done by hematology analyzers. The development oI
other tests, such as the reticulocyte, CD4, CD8, and CD64 counts will obscure the line between the
immunology and hematology testing environments |5, 6|.
Iwona Kaznowska-Bystryk
67
The automated hematology instruments provide an accurate and precise total neutrophil count but
do not report band numbers. In some clinical situations this inIormation is very necessary, Ior example
band neutrophil counts are helpIul Ior the detection oI neonatal sepsis. Instrument fagging Ior samples
with increased band has been characterized as unreliable, and one way Ior the physician to obtain a
band count is to request a manual review. It may be manual scan oI flm in which a technologist reviews
at least 10 microscopic felds to confrm the absence oI signifcant abnormality bands or more precise,
time-consuming diIIerential count |8|.
When leukocyte aggregation occurs, the hematology analyzer produces spurious and low WBC
count. A demonstrated increase in the VCS parameters (Coulter VCS technology), such as neutrophil
volume distribution width (NDW) may be used as an adjunct indicator Ior leukocyte aggregation
along with the unstable WBC counts or unexplained low WBC counts |9|. The manual diIIerential
is the accepted reIerence method Ior evaluating an automated DLC, but poses problems in evaluating
the automated monocyte and basophil results. The correlation between the automated and manual
diIIerential Ior neutrophils, lymphocytes, and eosinophils is described as good, but because oI low
numbers in manual counts, there is poor agreement between the manual and automated basophil and
monocyte counts |1, 6|.
For the most accurate and reproducible hematological results, whole blood specimens should be
analyzed as soon as possible aIter collection. Because the size and scatter properties oI white cells
change as a blood sample ages, automated analysis oI the diIIerential become more unreliable with
time. New technologies Ior the perIormance oI automated diIIerentials have reduced many oI the
problems associated with automated diIIerential counts dependent on the scatter properties only. On a
24-hour old sample, the MPV and MCHC should be excluded Irom the report. Because the HCT, MCV
and RDW parameters are consistently elevated due to the sample`s age, even at 24 hours, those results
should also be excluded |4|.
THE AUTOMATED TECHNOLOGY GOOD AND BAD POINTS
Hematology instrument operations consist oI computation, CBC analysis, and other analyses such
as DLC. On many large analyzers, the onboard computer does computation, analysis; it controls all
hardware and soItware operations, and contains a database oI stored patient data and quality control
results |2, 6|. Some instruments have incorporated sophisticated Ieatures such as delta-checking oI data
Irom the same patient, and the displaying oI electronic cell-sizing data via multi-color scattergram plots.
The blood cells scattergrams provide a wealth oI inIormation about the DLC, however, these scattergrams
have not gained popularity among clinicians. Generally, an experienced hematology technologist can
link an abnormal WBC scattergram to various clinical conditions or interIering substances. In addition
to reviewing scattergrams, the technologist must rely on the discriminatory fagging oI suspect WBC
parameters by the instrument. These are subroutines developed by each vendor in an attempt to
accurately detect WBC abnormalities. In addition to these in vivo abnormalities, other abnormal results
can be produced by a variety oI Iactors. Pseudoleukopenia can result Irom prolonged specimen storage
at room temperature in excess oI 24 hours, or exposure to excessive temperatures. Pseudoleukopenia
also may occur secondary to cell clumping, which may be antibody related, EDTA dependent or, rarely,
secondary to mucopolysaccharide produced by tumors, especially adenocarcinomas |2, 10|. Erroneous
The automated hematology analyzers
68
leukocyte counts can be seen in leukemic patients undergoing chemotherapy. In such situations, the
cells oIten are more Iragile than normal and disintegrate during their pressurized passage through the
instrument counting orifce.
New hematology systems include tests, such as reticulocyte counts and fow cytometric CD4/CD8
counts, and CD64. Enhancements in fow cytometry make it possible Ior these new instruments to also
count immature (blast) cells and NRBC |2|.
Since the number and characteristic oI the reticulocytes in the peripheral blood refect the activity
oI the bone marrow, reticulocyte counting has become a Iundamental part oI the evaluation oI patients
with hematopoietic disease. In contrast to microscopic counting, automated techniques oI reticulocyte
enumeration are more precise, accurate, objective, and cost-eIIective, since 30 000 or more cells can
be accurately evaluated in very short period oI time. A variety oI RNA-specifc fuorescent dyes have
been utilized Ior automated reticulocyte enumeration and some hematology analyzers utilize optical
light scatter analysis to perIorm reticulocyte analysis on specimens stained with new methylene blue or
other dyes. Measurements oI reticulocyte and other parameters are still under investigation, but there is
extensive evidence that these parameters are useIul in the accurate classifcation oI anemia patients, and
monitoring patients receiving rEPO or recovering Irom chemotherapy or bone transplantation. Although
economic consideration limit dedicated reticulocyte analyzers to laboratories with large reticulocyte
sample volumes, the recent trend to incorporate reticulocyte enumeration into routine capacity oI the
hematology analyzers will make automated reticulocyte analysis increasingly common |7|.
Some large instruments have expert systems computer-based programs to automate the labor-
intensive application oI complex review criteria. Expert systems can be programmed to initiate refex
testing, return specimens, or make and stain slides according to user-defned rules. Flagging properties
have also been incorporated into these analyzers to alert the operator to the existence oI abnormal cell
morphology and the presence oI abnormal cells which the analyzer is unable to count, as well as certain
sample characteristics, Ior example platelet clumps, which may cause incorrect results to be produced |1|.
Early automated diIIerential analyzers suIIered Irom excessive fagging oI samples Ior manual review,
with some instruments being virtually unusable because they fagged a high proportion oI either normal or
slightly abnormal specimens. Over the years, as the capabilities and perIormance oI automated analyzers
have improved, the respective roles oI the automated analyzer and the complementary procedures have
changed. It is recognized that the automated systems are superior Ior counting oI WBC, RBC and PLT and
Ior diIIerential counting oI WBC Ior well-characterized (matured) cell types, whereas visual microscopy
is superior Ior diIIerentiating cell based on nuances oI cytological Ieatures, especially Ior immature cells
|1|. Microscopic smear review is the main complementary procedure , the decision as to whether smear
review is necessary Ior each sample plays a major role in hematology laboratory costs, productivity and
speed oI reporting |1, 6|. There is little uniIormity among diIIerent laboratories oI criteria Ior action.
Recognizing the long-standing need Ior generally accepted guidelines (rules), which could be applied
to criteria Ior review oI CBC and diIIerential results Irom automated hematology analyzers, Barnes and
co-workers |1| described 41 rules oI the automated analyzer and complementary procedures - manual
action, most commonly smear review. They include rules Ior frst-time samples analysis, as well as rules
Ior repeating samples within 72 hours Irom their taking Irom a patient, Ior example, iI any plates PLT 100
x109/l or ~1000 x109/l and frst time analysis, then reviewing a slide is required |1| .
Iwona Kaznowska-Bystryk
69
To conclude, the automated hematology analyzers provide valuable inIormation regarding common
hematological conditions. It is important that operators and end users have an understanding oI the
parameters while interpreting their numerical and graphical data. This can enhance the diagnostic utility
oI automated data.
The automated hematology analyzers are characterized by a good sensitivity and specifcity,
reliability, high-quality patient results and saIety. In general, their perIormance has been excellent and
is now generally accepted by the medical community.
SUMMARY
The complete blood count (CBC) and diIIerential leukocyte count (DLC) are the backbone oI any
laboratory evaluation, not only used by general practitioners, but also by other medical specialties. The
CBC and DLC are used to diagnose anemia, to identiIy acute and chronic illnesses, bleeding tendencies,
and white blood cell disorders. During the last quarter century, blood cell analysis has progressed Irom
the use oI labor-intensive manual procedures to the use oI highly automated instruments. Modern
hematology analyzers provide new additional parameters enabling earlier and more precise diagnosis.
Moreover, they incorporate fow cytometry, robotics and expert system technologies.
Sensitive fags now allow the identifcation oI abnormal cells or spurious counts, but only the most
sophisticated hematology analyzers have optimal fagging and more simple analyzers, especially those
without a WBC diIIerentiation scattergram, do not possess the same sensitivity Ior detecting anomalous
results. ThereIore it is crucial that obtained results are clear Ior clinicians, and the parameters obtained
completely used.
The study presents the current state oI knowledge about hematology analyzer, concerning principles
oI their operation, their advantages and limitations, as well as the provided parameters.
Keywords: automated hematology analyzer, complete blood count, diIIerential leukocyte count
STRESZCZENIE
MorIologia krwi obwodowej i ronicowanie leukocytow stanowi podstaw w badaniach laboratoryjnych
zarowno dla lekarza pierwszego kontaktu, jak rownie dla lekarzy innych specjalnoci. Wykorzystywane
s w diagnostyce anemii, w rozpoznawaniu ostrych i przewleklych schorze, diagnostyce krwawie oraz w
zaburzeniach ukladu bialokrwinkowego. W ostatnim cwiercwieczu dokonal si ogromny postp w badaniu
komorek krwi, od uywania metod manualnych, a do stosowania wysoko zautomatyzowanych aparatow.
Nowoczesne analizatory hematologiczne dostarczaj nowych dodatkowych parametrow pozwalajcych
na wczeniejsz i dokladniejsz diagnoz. Ponadto, wyposaone s w cytometri przeplywow, robotyk
i eksperckie systemy technologiczne. Jedynie najbardziej zaawansowane analizatory posiadaj wraliwe
systemy fagowania, pozwalajce na identyfkacj komorek nieprawidlowych i zaIalszowanego zliczania,
za prostsze analizatory nie posiadaj takiej czuloci w wykrywaniu nieprawidlowoci. Dlatego kluczowe
jest, aby uzyskane parametry byly zrozumiale dla klinicysty, a otrzymane wyniki w pelni wykorzystane.
Praca przedstawia stan wiedzy na temat analizatorow hematologicznych, zasad ich dzialania, zalet i
ogranicze oraz dostarczanych parametrow.
6RZD NOXF]RZH analizatory hematologiczne, morIologia krwi, ronicowanie leukocytow
The automated hematology analyzers
70
REFERENCES
1. Barnes P. W., McFadden S. L., Machin S. J., et al.: The international consensus group Ior hematology
review: suggested criteria Ior action Iollowing automated CBC and WBC diIIerential analysis. Lab
Hematol., 11, 2005.
2. Chapman M.: Hematology analyzers oIIer new technology and user-Iriendliness. Lab Med, 31, 3,
2000.
3. Gopal S., Rappaport E. S., Koss W., et al: Applicability oI the Sysmex pocH-100i analyzer Ior
satellitate laboratories. Labmedicine, 36, 8, 2005.
4. Hill V. L., Simpson V. Z., Higgins J.M., et al.: Evaluation oI the perIormance oI the Sysmex XT-2000i
hematology analyzer with whole blood speciments stored at room temperature. Labmedicine 40, 12,
2009.
5. Kakkar N., Makkara M.: Red cell cytograms generated by an Advia 120 automated hematology
analyzer: characteristic patterns in common hematological conditions, Labmedicine, 40, 9, 2009.
6. Koenn M. E,, Kirby B. A., Cook L. L., et al.: Comparison oI Iour automated hematology analyzers.
Clin Lab Sci, 14, 4, 2001.
7. Riley R. S., Ben-Ezra J. M., Tidwell A.: Reticulocyte enumeration: past and present. Lab Med, 10,
32, 2001.
8. Senzel L. B., Spitzer E. D., DeCristoIaro J. D.: A streamlined laboratory process Ior obtaining band
counts in hospitalized neonates. Labmedicine, 41, 5, 2010.
9. Yang D., Guo X., Chen Y., et al.: Leukocyte aggregation in vitro as a cause oI pseudoleukopenia.
Labmedicine 39, 2, 2008.
10. Zandecki M., Genevieve F., Gerard J., et al.: Spurious counts and spurious results on hematology
anayzers: a review. Part I: platelets. Int J Lab Hematol., 29, 1, 2007.
11. Zandecki M., Genevieve F., Gerard J., et al.: Spurious counts and spurious results on hematology
anayzers: a review. Part II: white blood cells, red cells, hemoglobin, red cell indices and reticulocytes.
Int J Lab Hematol., 29, 1, 2007.
Iwona Kaznowska-Bystryk