Overview

Red Cell Distribution Width, Revisited

Benie T. Constantino, ART, MLT(CSMLS), SH(ASCP)I

ABSTRACT with other CBC parameters, such as the histogram, mean corpuscular
volume (MCV), and peripheral blood film analysis, RDW is frequently
The red blood cell distribution width (RDW), as part of an automated used to interpret aberrations in red cell morphology. This article
complete blood count (CBC), is a routinely available parameter describes and discusses the different methods for measuring red
on hematology analyzers. This parameter is the most commonly blood cell dispersion and explains the reasons for the inconsistencies
reported index of the variation in red cell volume and can be used to observed when interpreting RDW results.
detect subtle degrees of anisocytosis. It is one of the most studied
parameters; however, some earlier studies have shown overlap or Keywords: red cell distribution width, mean corpuscular volume,
discrepancy in the interpretations of these results. RDW is computed histogram
directly from the red blood cell (RBC) histogram and expressed as
coefficient of variation (CV) or standard deviation (SD). In conjunction

The original, classic Price-Jones studies1,2 on the het- The RDW is the most commonly reported index on the
erogeneity of red blood cell diameters have paved the variation or degree of anisocytosis in red cell volume.
way in establishing red cell anisocytosis as a means of The impedance and flow cytometric counters provide
evaluating red cell morphology in health and disease. this parameter, which is directly calculated from the RBC
Price-Jones quantifies anisocytosis as a coefficient of histogram.5,6,7 The RDW is derived from pulse-height
variation (CV) of red blood cell size in a Price-Jones analysis and can be expressed as standard deviation
curve, which serves to determine the presence and (SD) in femtoliters (fL), or as CV in the percentage of the
degree of red cell volume heterogeneity. Heterogeneity measurements of the red cell volume (Figure 1 and Fig-
generally has been identified by qualitative inspection ure 2).8 The earliest method provided by the hematology
of peripheral blood film (PBF). However, because of the analysis instruments to measure red cell variations is the
technical difficulty in measuring red cell variability from RDW-CV.3 During the past 3 decades, the clinical useful-
the PBF, recently designed automated particle counters ness of this parameter has been studied in a wide range
have allowed rapid and precise quantitation of volume of hematological disorders.9-14 Although many investiga-
heterogeneity.3,4 Measured as a CV, red blood cell distri- tors9,14,15,16,17 have advocated for the use of RDW along
bution width (RDW) is now an integral part of automated with MCV in delineating the probable cause(s) of each
complete blood count (CBC) analysis and is available on of the categories of micro-, macro-, and normocytic
all automated hematology analyzers. anemias, others11,12,13,18 have doubted the usefulness of
these parameters in the evaluation of micro- and macro-
cytosis. In one study,11 55% of patients with thalassemia
(thal) trait showed normal RDW or homogeneous micro-
DOI: 10.1309/LMZ1GKY9LQTVFBL7 cytosis, in contrast with the results of another study,9
which claimed 96% accuracy in such categorization. In
Abbreviations yet another study,18 the mean RDW values were higher
CV, coefficient of variation; PBF, peripheral blood film; RDW, red blood in thal trait than in patients with iron deficiency. Further,
cell distribution width; CBC, complete blood count; SD, standard devia- some authors9,10 have stated that all patients with vitamin
tion; fL, femtoliters; thal, thalassemia; IDA, iron-deficiency anemia;
retic, reticulocyte; homo, homogeneous; hetero, heterogeneous B12 deficiency have increased RDW-CV, compared with
the results of another study,13 which showed that a sig-
Hematology Department, CML HealthCare Inc, Mississauga, nificant proportion of patients with the same deficiency
Ontario, Canada
had normal RDW. Recently, an additional parameter,
*To whom correspondence should be addressed: RDW-SD, has been introduced on the Beckman Coulter
E-mail: [email protected] LH 780 Hematology Analyzer (Beckman Coulter Inc,

e2 Lab Medicine  Spring 2013  |  Volume 44, Number 2 www.labmedicine.com

 Overview

A B
68.26% of total
distribution area
(1 SD)
(1 SD)

80 110 200 L1 L2 200

Figure 1
Most recent hematology analyzers have provided 2 methods to calculate the red blood cell distribution width (RDW), namely, the
RDW–coefficient of variation (RDW-CV) and the RDW–standard deviation (RDW-SD), as follows:
RDW-CV (%) Particle Size Distribution5,19
A. Beckman Coulter, Inc method* B. Sysmex Corporation method*

1 SD L 2 – L1
RDW-CV = ──── × 100 RDW–CV = ───── × 100
MCV L 2 + L1

1 SD = MCV × (RDW-CV) L1 = MCV – 1 SD

100 L 2 = MCV + 1 SD

*indicates that these 2 methods are equivalent─they produce the same result using different formulas. Beckman Coulter, Inc, is located
in Brea, CA; Sysmex Corporation, in Kobe, Japan. MCV indicates mean cell volume; L, location of ±1 SD of distribution area of the curve.

100%

Figure 2
Red blood cell distribution width standard deviation (RDW-SD)
20%
particle size distribution. RDW-SD is the arithmetic width of the
distribution curve measured at the 20% frequency level. 80 110 200
RDW-SD

Brea, CA),5 which includes the measurement of red cell a distribution curve with a normal width may produce
size variation. Due to the existing inconsistencies in the a high RDW. On the contrary, a wide RBC distribution
initial interpretations using RDW-CV results, this method curve in patients with markedly raised MCV may still
needs to be revisited and/or reevaluated. This article generate a normal RDW number. In other words, micro-
also describes and discusses the different methods for cytosis tends to elevate the RDW-CV simply by lowering
measuring red cell dispersion and explains the reasons the denominator of the ratio. Conversely, macrocytosis,
for inconsistency in interpreting RDW-CV results. by increasing the MCV in the denominator, may offset
the change or increase in the width of the curve, thereby
The RDW-CV and RDW-SD are measures of the disper- producing a normal RDW-CV.
sion of data around the mean. The more spread apart
the data, the higher the SD. Although both methods use By contrast, the RDW-SD is a direct measure of RDW
SD to measure the degree of anisocytosis, they measure taken at the 20% frequency level of the histogram
cell variations differently. (Figure 2). The information or particles below the 20%
scale of the curve, however, are excluded to avoid in-
The RDW-CV, as shown in Figure 1, measures disper- terference in the RDW computation. These particles
sion by means of a ratio formula of 1 SD to the MCV. include aperture artifacts, cell coincidence, doublets,
Because it is a ratio, changes in the SD (width) or MCV triplets, and agglutinates on the right side of the
will influence the results. For example, a low MCV and curve and electrical interference, platelet clumps, and

www.labmedicine.com Spring 2013  |  Volume 44, Number 2  Lab Medicine e3

Overview

A B

80 110 200 80 110 200

C D E

80 110 200 80 110 200 80 110 200

F G H

80 110 200 80 110 200 80 110 200

I J K

80 110 200 80 110 200 80 110 200

Figure 3
Histograms showing graphical distribution of red cell volume density (key hematological features of these histograms are summarized
in Table 1).

megathrombocytes on the left. Because this method includes the measurement of the important abnormal
is independent of the MCV and is considered to be the small and large cells outside ±1 SD, this value repre-
absolute measure of dispersion regarding measurement, sents the genuine morphological and/or pathological
far beyond ±1 SD or across the MCV, it is the better and status of the patients in question. Moreover, because it
more reliable measure of RBC variability, particularly is a direct measure across the RBC distribution curve,
in highly abnormal conditions. Because the RDW-SD one can roughly estimate the spread of the distribution
encompasses the entire spectrum of MCV values and of cells by examining the histogram (Figure 3).

e4 Lab Medicine  Spring 2013  |  Volume 44, Number 2 www.labmedicine.com

 Overview

Nevertheless, it must be pointed out that the RDW-CV results of the RDW-CV and RDW-SD value. Because of
method has distinct advantages and disadvantages for these inconsistencies, the interpretations of RDW-CV
interpreting the degree of red cell anisocytosis. It is im- results must be correlated with the histogram and PBF
portant for the entire laboratory staff to understand the to ensure an accurate interpretation of results and their
RDW-CV and RDW-SD. One advantage of RDW-CV is use in patient care.
that it shows better correlation as an indicator of aniso-
cytosis when the MCV is in the low-normal range and It must be emphasized, however, that MCV is an average
when anisocytosis is difficult to detect, such as in mild value and a measure of central tendency; thus, it does
iron deficiency. It is also a useful method to monitor the not reflect the heterogeneity of the red cell populations.
effect of iron therapy and/or blood transfusion and to Its value is influenced by natural and biological condi-
differentiate iron-deficiency anemia (IDA) from uncompli- tions, sometimes compounded by coexisting patho-
cated thal trait. Therefore, the method works effectively logical conditions that often lead to a great variation in
when the red cell distribution spread is within ±1 SD. the MCV, rendering it an unreliable, if not misleading,
measurement on occasion. Thus a normal, lowered,
In view of the foregoing information, it is logical that or raised MCV cannot always be equated to normo-,
technologists examining peripheral blood film (PBF) micro-, or macrocytosis, respectively, often requiring
should have ready access to RDW-CV and RDW-SD PBF validation.21 Simply put, because MCV is an aver-
results for comparative review and correlation of results. age value, it tends to represent predominating cell sizes.
For example, if uniform or general microcytosis is pres-
ent, the MCV will be low. However, if a dimorphic or dual
red cell population is present, this population may not
Factors That May Influence always be accurately reflected in the numerical data,
due to the antagonistic effect of micro- and macrocyto-
RDW-CV sis on the MCV.22
Three important factors that may affect the RDW-CV,
namely, the method (formula) itself, the sample condi-
tion, and the sample population. Sample Condition
The age of the sample is an important factor to consider
because of the instability of RDW at room temperature.
The Formula: (RDW-CV = 1 SD ÷ MCV × 100) Although one study23 shows no change in RDW values
Using 1 SD as the numerator and MCV as the denomi- after 6 to 24 hours of storage, others4,24 show significant
nator in an equation is a fundamentally flawed strategy changes when samples are stored for longer than 6
and has limitations that may influence the value of the hours. Particularly, any changes in the width or MCV due
results. The 1 SD value imposed by the formula weakens to sample storage may not considerably influence the in-
and restricts its capability to measure red cell dispersion terpretation of results. However, the microcytosis of thal
as high as 1 SD only.20 As a result, other small and large trait shows greater variability in width, thus leading to a
abnormal cells outside the ±1 SD range are excluded slightly higher RDW-CV, which could be exaggerated in
from the calculation; hence, we observe the compara- cases of improper storage conditions. A slightly higher
tively reduced RDW-CV number compared with the MCV (ie, macrocytosis) may be a problem but usually
RDW-SD (Figure 3, section K). points to a normal RDW-CV. Because the SD and MCV
are affected by this condition, results should be inter-
Likewise, the MCV, which serves as the denominator, preted cautiously.
can greatly influence the results depending on its mean
value and the value of the width because it is a valua-
tion of ratio and proportion. For example, in Figure 3, Sample Population
sections G and H, although the mean values of both The characteristics and nature of the populations
MCVs are almost the same, their widths in the curve are tested may affect the RDW results. For example, altera-
different; hence, we witness an inconsistent or varying tions in the red cell size (resulting in anisocytosis) are
RDW result. Another interesting example is illustrated in a progressive, well-recognized feature of nutritional
Figure 3, part J, in which a highly disproportionate ratio deficiencies, such as iron deficiency and megaloblas-
between the MCV and width can profoundly affect the tic anemia. During disease states, both disorders may

www.labmedicine.com Spring 2013  |  Volume 44, Number 2  Lab Medicine e5

Overview

show an inadequate reticulocyte response to anemia. difference in the retic count.17 Likewise, in macrocytic
When treated appropriately, an effective reticulocyte anemia, the characteristic homogeneity or heterogeneity
response with an increased RDW ensues. Thus, treated of the populations tested may also influence the results.
and untreated nutritional deficiency can influence
RDW-CV results. Analysis and Interpretations of Histograms
in Figure 3
Also, some samples from patients with hemoglobi- The red cell histograms presented for illustration of the
napathic manifestations, particularly the β-thal trait RDW results were randomly selected to represent vari-
population, may exhibit reticulocytosis. Although β-thal ous common hematological conditions (Figure 3 and
trait usually manifests as homogeneous microcytic red Table 1). The histogram reflect the native size of the
cell populations, the presence of polychromasia from red cells or any particles in the red cell size range and
reticulocytes in the peripheral blood may considerably compare the red cell size to the relative cell numbers. A
change the RDW value, which may overlap with an IDA histogram may be interpreted by examining the shape
interpretation. A polychromatophilic red cell population of the data distribution. Because the RDW and MCV
of 1% to 3%, which is equivalent to an average of 2% to are determined from the histogram, their interpretations
5% reticulocyte (retic) count,25,26 may affect the RDW by must always be correlated with the data depicted in the
slightly increasing it. Most thal traits have high red cell histogram and confirmed microscopically.
counts of greater than 5.5 × 1012/L or erythrocytosis11,17;
with a retic count of 3%, this combination of slight To understand and appreciate the RDW, one should
polychromasia and erythrocytosis can generate signifi- know not only the methods and factors that affect it but
cant anisocytosis. (Although anisocytosis may also be also the definition or distinction between homogeneous
caused by destruction of red blood cells by the spleen, (homo) and heterogeneous (hetero) red cell populations
especially in β-thal major, the mechanism of anisocyto- based on the following categories: morphological (PBF),
sis in β-thal trait is different. The major morphological graphical (via histogram), and numerical data (RDW). A
changes in β-thal major are so obvious that moderate homo red cell population is usually considered when the
to marked anisopoikilocytosis with very high RDW is not red cells are morphologically of the same size, graphi-
unusual.) The absolute retic count, then, would be 5.5 cally narrow on the histogram, and numerically within a
× 1012/L × 0.03 = 165 × 109/L. Any spurious cell popula- normal RDW range. Whereas a hetero population of red
tions that comprise greater than 1% of the total red cell cells usually varies in size (2 or more populations), the
mass will influence the RDW-CV and red cell volume histogram shows a wide curve, and the RDW is elevated
histogram.9,23 Thus, a possible explanation for the differ- (Table 2). Sometimes, a homo red cell population may
ence in RDW between thal trait and IDA may lie in the show a high RDW-CV and/or the hetero red cell population

Table 1. Key Hematological Features of Various Histographic Images Shown in Figure 3

Subsection Hg, RBC Count, 1 SD, MCV, RDW-CV, RDW-SD, SF

of Figure 3 120-160 (g/L) 4.4-5.5 (× 1012/L) 10.5-15.0 (fL)a 80-99 (fL) 11.5-15.5 (%) 39-47 (fL) (µg/L)b Condition

A 142 4.62 12.1 89.7 13.5 39.4 N N

B 127 5.00 13.9 78.5 17.8 49.0 11 IDA
C 100 5.07 13.0 64.1 20.3 45.9 6 IDA
D 83 3.93 14.1 68.3 20.7 48.1 5 IDA
E 141 6.48 11.5 69.6 16.5 38.1 450 Thal trait
F 103 4.64 10.0 69.9 14.3 35.0 115 Thal trait
G 133 3.77 13.2 102.3 12.9 44.2 ND Unknown
H 131 3.77 19.4 102.9 18.9 66.5 ND Unknown
I 126 3.09 16.1 118.3 13.6 54.3 ND Unknown
J 159 3.21 25.2 141.5 17.8 98.0 ND Unknown
K 77 2.14 38.3 111.7 34.3 120.0 ND Vitamin B12 deficiency

Abbreviations: Hg, hemoglobin; RBC, red blood cell; SD, standard deviation; MCV, mean corpuscular volume; RDW, red cell distribution width; CV, coefficient of variation;
SF, serum ferritin; N, normal; IDA, iron-deficiency anemia; thal, thalassemia; ND, not determined.
a
1 SD was obtained from 100 normal complete blood count results in the laboratory.
b
Males: n = 27-260 µg/L; females: n = 11-145 µg/L.

e6 Lab Medicine  Spring 2013  |  Volume 44, Number 2 www.labmedicine.com

 Overview

Table 2. Differences Between Homogeneous and Heterogeneous Red Cell Populations

Category Homogeneous Heterogeneous

Histogram (graphical) Narrow or compact Wide

RDW-CV and RDW-SD (numerical) Normal range Increased
Morphological (PBF) Single population 2 or more red cell populations or anisopoikilocytosis

Abbreviations: RDW-CV, red blood cell distribution width – coefficient of variation; SD, standard deviation; PBF, peripheral blood film.

will show a normal RDW-CV. For this reason, the inter- usually displays hetero microcytosis with marked aniso-
pretations of red cell heterogeneity should be consid- poikilocytosis and a very high RDW. Because thal trait
ered in their full context or in conjunction with graphical is a hereditary disorder, the red cell size in the disorder
and morphological findings. is usually reduced and uniform, and the width of the
base of the curve is narrowed; hence, no anisocytosis is
Heterogeneous Microcytosis (Figure 3, parts present. Although the β-thal trait usually manifests homo
B through D) microcytosis, some samples may show heterogeneity as
The most common condition associated with hetero a consequence of reticulocytosis, as discussed previ-
microcytosis is IDA. In untreated IDA, the red cell vol- ously herein.
ume and morphology often show progressive changes,
from normochromic-normocytic to slightly hypochro- Homogeneous Macrocytosis (Figure 3, parts
mic–microcytic during early onset of iron deficiency to G and I)
markedly hypochromic-microcytic in later stages. All the Macrocytosis with an MCV higher than 100 fL (and
changes in the cohort of red cells, each with different normal RDW) is found in 3% to 5% of adult patients, a
mean values, are reflected accordingly in the RDW-CV, significant proportion of whom may not have anemia.29
as shown in Figure 3, parts B through D, and in Table 1. Despite the elevated MCV value, the RDW-CV and
In states of iron deficiency, the MCV is reduced and the RDW-SD in Figure 3, part G are within normal ranges.
width of the base of the curve is somewhat increased This occurs because the width and the MCV are pro-
(increased anisocytosis). The width of the base of the portionately or slightly increased. Also, because the size
curve (or the coefficient of variability) gives an indication and distribution differences between normocytic and
of the degree of inequality in the size of the red cells. macrocytic red cells are minimal, the red cell distribution
When IDA persists for longer than 4 months (the normal curve with normal width (homo) can be observed even
circulating red cell life span is 100-120 days), the original though the red cell population is hetero, due to the pres-
normochromic-normocytic red cells are replaced by ence of the 2 normocytic and macrocytic red cell popu-
new cells that are strikingly hypochromic and microcytic. lations that will be reflected in the curve.
Consequently, most of the circulating red cells may be-
come homo hypochromic-microcytic red cells and may As in Figure 3, part G, even with a high MCV, the RDW-
yield a misleading result.11 Nevertheless, patients show- CV in Figure 3, part I is also normal. This occurs be-
ing severe anemia with hemoglobin levels of less than cause the high MCV offsets the effect of the increased
90 g/L and having hypochromic-microcytic red cells are width; the RDW-SD, however, is increased. As explained
most likely (ie, 80%) to have iron deficiency.27,28 Thus, earlier herein, RDW-SD is measured directly from the
early stages of iron deficiency are manifested in a hetero curve; in this case, the curve has a wide appearance.
normocytic population, and mid- to late-stage IDA will
manifest as a hetero microcytic population. Overlapping and inconsistent results are clearly high-
lighted by these 2 illustrations. In Figure 3, part G, the
Homogeneous Microcytosis (Figure 3, parts histographic image shows an overlap of homo features
E and F) because of the narrow spread of the curve (normal
In β-thal, both the heterozygous and the homozygous RDW-CV) and hetero features due to the presence of
forms exhibit microcytosis. Whereas the former may normocytic and macrocytic red cells. However, in Fig-
often be misdiagnosed and treated as IDA,10 the lat- ure 3, part I, an inconsistent result consisting of normal
ter are easily detectable morphologically. β-thal major RDW-CV, or homo, and abnormal RDW-SD, or hetero,

www.labmedicine.com Spring 2013  |  Volume 44, Number 2  Lab Medicine e7

Overview

can be observed. Also, the overlapping presence of a However, when viewed microscopically, the red cell vol-
single population of macrocytosis (homo) and increased ume accurately reflects the macrocytic heterogeneity of
width (hetero) can be observed in the curve. The most the red cell populations.
likely explanations for the discrepancy of these results
are the proportion of the ratio in the calculation of RDW-
CV and the instrument’s algorithm.
Discussion
These results are complicated, confusing, and ambigu-
ous because they can be interpreted as homo or hetero. The RDW measurement, by itself, is meaningless. It
Consequently, this may explain why in certain earlier becomes useful only when it is compared with the refer-
studies9,13 on megaloblastic anemia, normal or different ence ranges and/or correlated with the red cell histo-
RDW-CV results were obtained despite the high MCV. gram and confirmed microscopically.

Although the RDW is one of the most studied and fre-

Heterogeneous Macrocytosis (Figure 3, Parts quently used CBC parameters, it is not well understood.
G Through K) The failure to understand the overlap and inconsisten-
Figure 3, parts G through K, display the different ap- cies in interpreting RDW-CV results suggest that some
pearances of heterogeneity in macrocytosis.13,20,30 Note fundamental knowledge is lacking in current under-
that in these illustrations, the results of the RDW-CV and standing of the method. This occurs because most
RDW-SD have no pattern; they range from normal to studies show only the effect of the method in the clinical
slightly elevated to moderately and markedly elevated. interpretations of results; the cause(s) of discrepant re-
Clearly, the RDW-SD is relatively higher than the RDW- sults is (are) not well explained. To interpret the RDW-CV
CV. Figure 3, parts G and I, which were discussed previ- properly, its derivation must also be understood.
ously in the homo macrocytosis section, are included in
this category because they have hetero features as well. The main reason for the discrepant interpretations of
RDW-CV results lies in the method itself, specifically,
Figure 3, parts G and H, represent strong examples of the way dispersion is measured by the formula: 1 SD
contrasting RDW results. Despite their almost similar divided by MCV. Because this method is a ratio, any
MCV values, their varying RDW values are notable. In proportionate or disproportionate increase or decrease
Figure 3, part G, the RDW-CV and RDW-SD are within of the SD or MCV will affect the results. In short, any
normal range compared with Figure 3, part H, in which proportionate increase or decrease of SD and MCV will
the RDW-CV and RDW-SD are increased. The reasons produce a normal value, whereas any disproportionate
for the discrepant results are the proportionate and dis- increase or decrease of SD or MCV will lead to a normal
proportionate ratios of the width and MCV, respectively. or increased RDW-CV value.
These contrasting RDW values further reinforce the im-
portance of understanding the cause and effect of the Moreover, the 1 SD restriction imposed by the formula
RDW methods on the results. weakens its usefulness and/or limits its capability to
measure red cell dispersion as high as 1 SD only. As
In Figure 3, parts J and K, despite the huge differences a result, some important abnormal small and large
in the MCV values—the former shows 141.5 fL and the cells in varying degrees of anisocytosis and/or poikilo-
latter, 111.7 fL—the similarities of their RDW-SD values cytosis outside the ±1 SD range are excluded in the
are notable: both are grossly elevated. However, the estimation. Hence, the value obtained is sometimes
inconsistent results of slightly elevated RDW-CV and not a genuine representation of red cell morphologi-
markedly raised RDW-SD are illustrated in Figure 3, cal aberrations or a true measure of dispersion. It ap-
part J. These results clearly demonstrate the weakness pears, however, that the higher the RDWs, the higher
or dubious effects of a ratio formula on the RDW-CV the degrees of anisopoikilocytosis. Thus, in summary,
values. The RDW-CV results are comparatively lower the reasons for the discrepant initial interpretations
than the RDW-SD values; the reasons for the discrepant of results include the disproportionate ratio of 1 SD
outcomes include the disproportionate ratio of the width to MCV; the 1 SD measurement limit imposed by the
and MCV, the constraining effect of 1 SD imposed by formula; and the methods of measuring red cell
the formula, and the methods of measuring dispersion. dispersions, namely, RDW-CV and RDW-SD.

e8 Lab Medicine  Spring 2013  |  Volume 44, Number 2 www.labmedicine.com

 Overview

The definitions of hetero and homo, as presented herein, 8. Bate I, Bain BJ. Basic haematological techniques. In: Bates I, Bain
BJ, Lewis SM, eds. Dacie and Lewis Practical Hematology. 9th edn.
assume that interpretations of red cell variations can London: Churchill Livingstone; 2001:19-46.
be accomplished under 3 sets of conditions, namely, 9. Bessman JD, Gilmer PR Jr, Gardner FH. Improved classification of
numerical value (RDW), graphical judgment (ie, via his- anemias by MCV and RDW. Am J Clin Pathol. 1983;80:322-326.

togram), and morphological findings. Based on these 10. Karnad A, Poskitt TR. The automated complete blood cell count:
Use of the red blood cell volume distribution width and mean platelet
terms, distinguishing heterogeneity from homogeneity volume in evaluating anemia and thrombocytopenia. Arch Intern Med.
seems simple; however, in certain cases, particularly 1985;145:1270-1272.
those depicted in Figure 3, parts I-K, the results are 11. Flynn MM, Reppun TS, Bhagavan NV. Limitations of red blood cell
distribution width (RDW) in evaluation of microcytosis. Am J Clin
mixed, highly hetero, and confusing or hard to interpret. Pathol. 1986;85:445-449.
Heterogeneity can be determined from RDW results; 12. Brittenham GM, Koepke JA. Red blood cell volume distributions and
however, if RDW-CV is used only to interpret anisocyto- the diagnosis of anemia: help or hindrance? Arch Pathol Lab Med.
1987;1146-1148.
sis, it is most likely, in certain cases such as Figure 3,
13. Saxena S, Weiner JM, Carmel R. Red cell distribution width in
part I (or in the results of an earlier study13) that discrep- untreated pernicious anemia. Am J Clin Pathol. 1988;89:660-663.
ant results may show (or be misinterpreted as) normal 14. Bessman JD. Heterogeneity of red cell volume: quantitation,
values. For this reason, when interpreting abnormal red clinical correlations, and possible mechanisms. John Hop Med J.
1980;146:226-230.
cell dispersions, use of the RDW-CV in conjunction with
15. Fossat C, David M, Harle JR, et al. New parameters in erythrocyte
the RDW-SD is highly recommended. counting. Value of histograms. Arch Pathol Lab Med. 1987;111:1150-
1154.
The relationship among the RDW, histogram, and 16. McClure S, Custer E, Bessman JD. Improved detection of early iron
PBF, along with the MCV, cannot be overemphasized. deficiency in nonanemic subjects. JAMA. 1985;253:1021-1023.
17. Roberts GT, El Badawi SB. Red cell distribution width index in some
Whereas some technologists rely heavily on the RDW
hematologic diseases. Am J Clin Pathol. 1985;83:262-226.
when interpreting red cell dispersion, most of the tech- 18. Morgan DL, Peck SD. The use of red cell distribution width in the
nologists prefer to base their interpretations on morpho- detection of iron deficiency in chronic hemodialysis patients. Am J
logical findings. In some laboratories, histograms are not Clin Pathol. 1988;89:513-515.

used when interpreting red cell dispersion; however, be- 19. Troubleshooting Guide: Sysmex SE–Series Automated Hematology
Systems. Document no. MKT-40-1010. Sysmex America Inc.; 2004.
cause the RDW and MCV are determined via histogram,
20. Constantino BT. The red cell histogram and the dimorphic red cell
it is prudent to correlate the results with those derived population. LabMed. 2011;42:300-308.
from histogram and to confirm these results microscopi- 21. Waters HM, Seal LH. A systematic approach to the assessment of
cally. In all these illustrations, therefore, only by careful erythropoiesis. Clin Lab Haematol. 2001;23:271-263.

examination of the histogram, knowledge of the possible 22. Kakkar N, Makkar M. Red cell cytograms generated by an ADVIA 120
Automated Hematology Analyzer: characteristic patterns in common
causes of abnormal RDWs, and careful correlation with hematological conditions. LabMed. 2009;40:549-555.
peripheral blood morphologic findings can a correct di- 23. Park KI, Kim KY. Clinical evaluation of red cell volume distribution
agnosis be derived for use in patient health care. LM width (RDW). Yonsei Med J. 1987;28:282-290.
24. Hill VL, Simpson VZ, Higgins JM, et al. Evaluation of the performance
of the Sysmex XT-2000i Hematology Analyzer with whole blood
specimens stored at room temperature. LabMed. 2009;40:709-718.
25. Gulati G. Blood Cell Morphology Grading Guide. 3rd edn. Chicago:
References American Society for Clincal Pathology Press; 2009:14-15.
26. Escobar MC, Rappaport ES, Tipton P, Balentine P, Riggs MW.
1. Price-Jones C. The variation in sizes of red cells. Brit Med J. Reticulocyte estimate from peripheral blood smear: a simple,
1910;2:1418. fast, and economical method for evaluation of anemia. LabMed.
2. Price-Jones C. Red Blood Cell Diameters. London: Oxford University 2002;33:703-705.
Press; 1933. 27. Constantino BT. The evaluation and differentiation of hypochromic
3. Rowan RM, Fraser C, Gray JH, McDonald GA. The Coulter Counter microcytic red blood cells of thalassemia trait and iron deficiency.
Model S-Plus–the shape of things to come. Clin Lab Haematol. Can J Med Lab Sci. 1999;61:112-121.
1977;1:29-40. 28. Okuno T, Chou A. The significance of small erythrocytes. Am J Clin
4. Cornbleet PJ, Beuchel J. Red cell distribution width on the Coulter Pathol. 1975;64:48-52.
Model S-Plus. Am J Med Tech. 1983;49:865-867. 29. Brigden ML. A systematic approach to macrocytosis. Sorting out the
5. Beckman Coulter LH 780 On Line IBO72641. Miami Lakes: Beckman causes. Postgrad Med. 1995;97:171-184.
Coulter Education Center; 2007. 30. Rowan RM. Blood Cell Volume Analysis—A New Screening
6. ADVIA 120 Hematology System. Technology and Cytogram Technology for the Hematologist. London: Albert Clark and
Interpretation. Tarrytown: Bayer Diagnostics; 2005. Company; 1983:43-55.

7. CELL DYN Sapphire TM System Operator Manual. Abbott Park:

Abbott Laboratories; 2005.

www.labmedicine.com Spring 2013  |  Volume 44, Number 2  Lab Medicine e9