See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/50196146

Erythrocyte Sedimentation Rate (ESR) Measured by the Streck ESR-Auto Plus Is

Higher Than With the Sediplast Westergren Method: A Validation Study

Article in American Journal of Clinical Pathology · March 2011

DOI: 10.1309/AJCP48YXBDGTGXEV · Source: PubMed

CITATIONS READS

30 6,979

5 authors, including:

Qian-Yun Zhang
University of New Mexico
18 PUBLICATIONS 493 CITATIONS

SEE PROFILE

All content following this page was uploaded by Qian-Yun Zhang on 01 June 2014.

The user has requested enhancement of the downloaded file.

 Hematopathology / Higher ESR Using Streck ESR-Auto Plus

Erythrocyte Sedimentation Rate (ESR) Measured by the

Streck ESR-Auto Plus Is Higher Than With the Sediplast
Westergren Method
A Validation Study
Bharathi Vennapusa, MD,1 Laura De La Cruz, MD,1 Hemangini Shah, MLT(ASCP),2
Veronica Michalski, MT(ASCP),2 and Qian-Yun Zhang, MD, PhD1,2

Key Words: Erythrocyte sedimentation rate; Streck ESR-Auto Plus; Sediplast; Westergren; Validation

DOI: 10.1309/AJCP48YXBDGTGXEV

Abstract PROOF The erythrocyte sedimentation rate (ESR) is the rate of

The erythrocyte sedimentation rate (ESR) is a sedimentation of RBCs and is used often as a nonspecific
nonspecific indicator of disease activity and is often measure in monitoring disease activity and assisting in the
used by clinicians in assisting diagnosis and follow-up diagnosis of many inflammatory disorders. The ESR is not
of many inflammatory disorders. The Westergren a well-understood phenomenon and has been described as
method is considered the standard method for occurring in 3 phases: RBC aggregation, precipitation, and
measuring ESR. Recently, many automated instruments packing. RBC aggregation is a critical factor for the sedi-
have become available to address laboratory safety and mentation and is facilitated by the presence of certain plasma
time efficiency. We validated the Streck ESR-Auto Plus proteins called agglomerins, which include fibrinogen, IgM,
instrument (Streck, Omaha, NE) using the Sediplast and α2-macroglobulin. Any factors affecting these 3 phases,
(Polymedco, Cortlandt Manor, NY) Westergren method including those in the number and shape of RBCs and plasma
as the reference method. Blood samples collected in viscosity, can affect the sedimentation rate. The ESR is
113 EDTA tubes were transferred into Sediplast vials expressed as millimeters per hour and varies between age
and Streck high-altitude vacuum tubes for measuring groups and sexes.
the ESR. There was good correlation between the The test for measuring the ESR was first described
Sediplast and Streck methods (0.95) using Pearson by Westergren and Fahraeus1 and is well known as the
correlation. The y-intercept was at 6.5 using linear Westergren method. There have not been many changes to the
regression, indicating systematic bias with a mean originally described method, and all subsequently introduced
difference of 7.13 using the t test (P < .0001). We methods have been evaluated against Westergren method as
modified our reference ranges to rectify the systematic the standard. The International Council for Standardization
bias found during validation. in Haematology recommended the Westergren method as the
method of choice. The Westergren method uses citrate, a liq-
uid-based anticoagulant, resulting in dilution of blood, inac-
curacies of which significantly affect the ESR. Modifications
of the reference method, also called the modified Westergren
method, are also widely used and use EDTA in place of citrate
as an anticoagulant. EDTA is a solid-based anticoagulant that
does not cause significant dilution (<1%), reducing the errors
of dilution.1-3 Several modifications to the existing method,
including automated instruments, have been introduced dur-
ing recent years to minimize human contact with blood prod-
ucts and to improve turnaround time.

© American Society for Clinical Pathology Am J Clin Pathol 2011;135:000 1

1 DOI: 10.1309/AJCP48YXBDGTGXEV 1
Vennapusa et al / Higher ESR Using Streck ESR-Auto Plus

The ESR increases with age and is higher in women than stored immediately at 4°C and transported to the UH Lab
in men.4,5 The reference ranges that have been established and within 8 to 10 hours. Testing was performed within 12 hours
traditionally used reflect the effects of age and sex. The higher after sample collection and never exceeded 24 hours. Blood
limits of the reference range for the ESR in men and women samples were normalized to room temperature by incubating
50 years or younger are 15 and 20 mm/h, respectively, where- them for 1 hour before testing. None of the samples received
as they are 20 and 30 mm/h, respectively, for men and women for the micromethod were used for method comparison. The
older than 50 years. These ranges have been confirmed to be EDTA blood was well mixed by manually inverting the tube
applicable to our population using the Sediplast (Polymedco, 10 times before transferring the sample.
Cortlandt Manor, NY) Westergren method and were used as The Sediplast Westergren method was done by transfer-
the reference ranges for the current validation study. ring 0.8 mL of potassium-EDTA blood into a vial containing
The Sediplast ESR system has Westergren and modified a defined amount of 3.8% sodium citrate (0.2 mL) solution.
Westergren methods available for measuring ESR, which use The blood was mixed thoroughly by manually inverting for 10
sodium citrate and EDTA as anticoagulants, respectively. We times. The 150-mm graduated, Sediplast pipette was inserted
had used the Sediplast Westergren method in our laboratory for into the vial by piercing through the diaphragm and pushed
measuring the ESR for many years and used the same as the to the bottom of the vial, followed by placing the vial-pipette
reference method for our validation. The Sediplast Westergren unit on a leveled rack. The blood in the pipette automatically
method is a manual method, requiring considerable time in levels to zero, a function called auto-zeroing or self-zeroing,
sample preparation, personnel presence for measuring ESR in with excess blood flowing into a reservoir at the top of the

PROOF
a timely manner, and contact with human blood. pipette, eliminating errors of manually loading the pipette
The Streck ESR-Auto Plus (Streck, Omaha, NE) is a itself. The ESR was measured exactly at 60 minutes ± 1 min-
random access automated ESR analyzer using a quick mode ute using a timer.
method, by which erythrocyte sedimentation is measured at The blood was also transferred into the high-altitude ESR
30 minutes and extrapolated to 1 hour using a mathematical vacuum tubes containing citrate using disposable pipettes.
formula by the instrument. The sedimentation is measured The tubes were filled manually, and the final level was within
by infrared light, which increases accuracy. Results are auto- 5 mm of the level indicated on the side of the tubes. After
matically printed after the measurement of ESR. Blood can mixing the blood sample on the mixer provided by the Streck
be directly drawn or transferred manually from an EDTA tube manufacturer for 3 minutes, tubes were placed in the Streck
into 1.2-mL safety-coated ESR vacuum tubes containing 3.2% ESR-Auto Plus instrument. After 30 minutes, the measured
sodium citrate as an anticoagulant. The ESR vacuum tubes are ESR was automatically printed out by the instrument.
available in 2 types: high- and low-altitude tubes for laborato- To evaluate the direct collection method using Streck
ries above and below approximately 2,500 ft above sea level, evacuated tubes, blood was drawn into high- and low-altitude
respectively, with varying amounts of vacuum. The instrument tubes from 10 apparently healthy volunteers, and the ESR was
has an incorporated bar-code reader, which, along with connec- measured using the Streck method. The difference between
tion to a laboratory information system, can make it a “walk- the ideal and actual fill levels after blood draw was measured
away” method. We validated the Streck ESR-Auto Plus because in millimeters, with a positive value indicating above and
of the aforementioned advantages and to deal with the increased negative value indicating below the ideal level. Universal
patient load. From here on, these 2 methods would are referred precautions were followed while handling the specimens, and
to as the Sediplast Westergren and Streck methods. tubes were disposed appropriately after testing.
Our data consisted of continuous variables and were
skewed owing to inclusion of a majority of samples within the
reference range. Therefore, statistical analysis was performed
Materials and Methods
using the Pearson correlation, linear regression, and paired
For method comparison, 113 potassium-EDTA blood t test to establish the relationship between the 2 methods.
samples received for routine CBC analysis at the University of Precision of the Streck method was measured by using 10
New Mexico Hospital Laboratory (UH Lab; Albuquerque) and replicate samples of control material with ESRs within and
at the TriCore Reference Laboratories (TRL; Albuquerque) outside the reference range.
were selected. The UH Lab is affiliated with TRL and serves
as a rapid response laboratory. The samples were deidentified
except for age and sex. The ages of patients ranged from 2 to Results
82 years, of whom 53 were males and 60 were females. The The results were grouped into 4 age groups similar to
testing was performed at the UH Lab by 2 resident patholo- traditional reference ranges and included males and females
gists (B.V. and L.D.L.C.). Samples collected at TRL were age 50 years or younger and age older than 50 years. Of

2 Am J Clin Pathol 2011;135:000 © American Society for Clinical Pathology

2 DOI: 10.1309/AJCP48YXBDGTGXEV
 Hematopathology / Original Article

the samples, 48.7% (55/113) were from patients 50 years 35

or younger, including 18.2% (10/55) from the pediatric 30

No. of Samples
age group (<18 years), whereas 51.3% (58/113) were from 25
patients older than 50 years ❚Figure 1❚. The ESR values for 20
all the samples ranged from 0 to 134 mm using the Sediplast 15
Westergren method and from 2 to more than 120 mm using 10
the Streck method. The highest ESR that can be measured 5
0
using the Streck method is 120 mm, and any result more than 0-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90
this value is reported as more than 120 mm. Three samples Age Groups (y)
with an ESR of more than 120 mm/h measured by the
❚Figure 1❚ Distribution of patient age groups for the samples.
Streck method had values of 127, 131, and 134 mm/h using
the Sediplast Westergren method. These 3 samples were
excluded from the correlation analysis because they were 140
above the upper limit of the Streck method. The ESR was 130
120
within the reference range (reference ranges as defined for

Streck Method ESR (mm/h)

110
the Sediplast Westergren/reference method) in 84 samples 100
and outside the range in 29 using the Sediplast Westergren 90
80
method compared with 75 within and 38 outside the range
70

PROOF
using Streck method. Nine samples, which were called as 60
within the reference range by the Sediplast Westergren 50
40
method (range, 7-20 mm/h) were read as outside the refer-
30
ence range by the Streck method (range, 20-32 mm/h). The 20
highest difference was seen in a 73-year-old man whose 10
0
ESR was 7 mm/h by the Sediplast Westergren method and 0 10 20 30 40 50 60 70 80 90 100 110 120130140
28 mm/h by the Streck method (reference range, 0-20 mm/h Sediplast Method ESR (mm/h)
for males older than 50 years).
The Pearson correlation coefficient was 0.954, indicating ❚Figure 2❚ Correlation between Streck and Sediplast
good correlation between the methods. The slope of the regres- Westergren methods. Slope, 1.04; y-intercept, 6.5. ESR,
sion line using linear regression was 1.04 with a y-intercept erythrocyte sedimentation rate.
of 6.5 ❚Figure 2❚. The ESR values obtained with the Streck
method were significantly higher than those obtained by the
Sediplast Westergren method (P < .0001) in all age groups (mean, 10 mm/h); and, for the sample outside the reference
both within and outside the reference ranges using the paired range, an SD of 2.5 (95% CI for SD, 1.7-4.5) and a CV of
t test ❚Table 1❚. The mean difference using the paired t test 3.6% (mean, 68.6 mm/h).
was 7.13. The 95% confidence intervals (CIs) for the means When blood samples were directly drawn from healthy
of both methods were also significantly different for differ- volunteers, the Streck low-altitude tubes were significantly
ent age groups ❚Figure 3❚. The precision analysis showed an underfilled (mean, –9.3 mm) compared with high-altitude
SD of 0.5 (95% CI for SD, 0.3-0.9) for the sample within the tubes (mean, +2.7 mm; P < .0001; paired t test), but the resul-
reference range, with a coefficient of variation (CV) of 4.7% tant ESR values were not significantly different (P = .27).

❚Table 1❚
Erythrocyte Sedimentation Rate Values for Different Age Groups With Old and New Reference Ranges

Males, ≤50 y Males, >50 y Females, ≤50 y Females, >50 y

Manual Automatic Manual Automatic Manual Automatic Manual Automatic

Mean 4 9 5 11 7 13 10 18
SD 3.6 5.6 5.5 7.8 5.7 7.5 5.6 7.7
Mean + 2 SD 11.2 20.2 16 26.6 18.4 28 21.2 33.4
Reference range (mm/h)
Sediplast (old) 0-15 — 0-20 — 0-20 — 0-30 —
Streck (new) — 0-21 — 0-25 — 0-28 — 0-33

© American Society for Clinical Pathology Am J Clin Pathol 2011;135:000 3

3 DOI: 10.1309/AJCP48YXBDGTGXEV 3
Vennapusa et al / Higher ESR Using Streck ESR-Auto Plus

25 Miao7 in China studied the effect of geographic variables

Sediplast method
Streck method
on ESR and showed that the ESR was inversely proportional
20 to the altitude. Because our laboratory is located 1 mile (5,280
feet) above sea level (Albuquerque, NM), the ESR of our
ESR mm/h)

15 patient population is expected to be lower because of reduced

oxygen tension and resultant polycythemia, excluding high
10
altitude as a cause of the systematic bias. Even though we
used samples with hematocrit values within our reference
5
ranges (36%-45% [0.36-0.45]), the International Council for
0 Standardization in Haematology recommends use of dilution
Males <50 Males >50 Females <50 Females >50 when the hematocrit is more than 35% (0.35), which is always
Age Groups (y) the case in our patient population.1,2 The Sediplast Westergren
❚Figure 3❚ Confidence intervals of means for the Sediplast and Streck methods use citrate as an anticoagulant, which
Westergren and Streck methods for different age groups. results in dilution of blood and should correct the ESR owing
ESR, erythrocyte sedimentation rate. to higher hematocrit values. The fill levels of the vials or tubes
for the Sediplast and Streck systems were within 2 mm from
the level indicated, excluding the inaccuracies of filling as a
possible factor.
Discussion

PROOF
Because of the large number of facilities from which
The ESR values obtained from the Streck ESR-Auto we receive samples for testing ESR and the problems that
Plus were significantly higher than those of the Sediplast might be encountered in not obtaining a proper sample in the
Westergren method (P < .0001) with a mean difference of vacuum tubes provided by the manufacturer, we validated
7.13 using the paired t test, indicating that the Streck method the Streck instrument with the manual transfer of samples
measures the ESR higher than the Sediplast Westergren meth- from EDTA tubes. Streck manufactures 2 types of vacuum
od. Even though the results of both methods are correlated at tubes, for high and low altitudes, which have different levels
95% (Pearson correlation), the y-intercept of the regression of vacuum but the same volume of citrate anticoagulant. This
line was 6.5, indicating the presence of systematic bias with explains the difference in the fill level observed between the
the Streck method. We modified our reference ranges of dif- high- and low-altitude tubes when blood samples were direct-
ferent age groups to rectify the systematic bias encountered ly drawn from volunteers. We had initially received Streck
using the Streck ESR-Auto Plus instrument (Table 1). low-altitude tubes by mistake and had initiated validation
This significant difference we observed in our study using those tubes. The Streck low-altitude tubes were signifi-
between the Streck and Sediplast Westergren methods was cantly underfilled compared with the high-altitude tubes, but
also apparent in a recent College of American Pathologists this was not reflected in the ESR values obtained from these
survey for ESR (ESR-A 2009) in which the mean ESRs tubes, which showed no significant difference. This indicates
for the Sediplast Westergren diluted and Streck ESR-Auto that the fill level might not have a significant effect on the
Plus diluted methods for high ESR were 38.2 and 64.2 mm, ESR, but this has to be verified with larger number of patients
respectively (3.7 and 8.6 mm for low ESR), further confirm- or volunteers.
ing our observation.6 The Sediplast Westergren method was The ESR vacuum tubes are stable for 6 days when refrig-
the most common method for measuring ESR, used by 970 erated (2°C-10°C), as demonstrated by the manufacturer. The
laboratories, with the Streck ESR-Auto Plus the second same study also proved the stability of refrigerated EDTA
most common method, used by 660 laboratories. The dif- blood samples for 24 hours.8 Because we used manual trans-
ference observed is much higher with higher levels of ESR fer of EDTA blood samples into the ESR vacuum tubes for
compared with the lower range, indicating the presence of method comparison, we further validated the stability of 5
proportional bias as well. The ESR was also higher when EDTA blood samples by measuring the ESR using the Streck
tested on other automated platforms compared with the method immediately after blood draw and after storing samples
manual platforms. One can argue that the sample sent for between 2°C and 10°C for 24 hours. In both cases, samples
proficiency testing is control material and not a fresh blood were allowed to normalize to room temperature for 1 hour and
sample. In our opinion, the control material is supposed to were manually transferred into high-altitude tubes. There was
have a stable result across testing platforms in contrast with no difference in the ESR measurements made immediately
the actual blood sample, in which there are numerous bio- after blood draw and after refrigerating samples for 24 hours (P
logic and nonbiologic variables affecting the ESR measure- = .78; t test), indicating that the EDTA blood sample is stable
ment that can result in significant differences. for up to 24 hours when stored between 2°C and 10°C.

4 Am J Clin Pathol 2011;135:000 © American Society for Clinical Pathology

4 DOI: 10.1309/AJCP48YXBDGTGXEV
 Hematopathology / Original Article

Normalization of the blood samples to room temperature other automated instruments, indicates that these systematic
is a requirement for the Streck method. To evaluate the effect biases exist among most of the automated testing platforms
of sample temperature on ESR measurement, we measured and need further investigation to determine the factors affect-
the ESR using the Streck method on 2 EDTA blood samples ing this difference and potentially revising the standardization
drawn from a healthy volunteer immediately after the blood of reference intervals for automated instruments.
draw at 37°C, after allowing the samples to normalize to
room temperature for 1 hour, after storing them at 2°C to From the 1Department of Pathology, University of New Mexico;
10°C for 8 hours, followed by testing them with and without and 2TriCore Reference Laboratories, Albuquerque.
normalization to room temperature. In all cases, the samples Address reprint requests to Dr Zhang: Dept of Pathology,
were manually transferred into the high-altitude tubes before MSC08 4640, 1 University of New Mexico, Albuquerque, NM
testing on the Streck instrument. There was no difference 87131.
Acknowledgments: We thank the staff of the University
between the sample measurements, which ranged from 7 to 8 of New Mexico Hospital Laboratory and TriCore Reference
mm/h with an SD of 0.58 (mean, 7.5 mm/h; 95% CI of mean, Laboratories, Albuquerque.
6.58-8.42). The Streck manufacturer proved the instrument’s
performance at different temperatures, with the instrument
itself incubated at different temperatures.9 The stability of the
ESR measurement of a blood sample at different temperatures References
at a constant instrument temperature was not tested. Our study 1. Bull BS, Caswell M, Ernst E, et al. ICSH recommendations
for measurement of erythrocyte sedimentation rate. J Clin

PROOF
indicated that the sample temperature does not have a signifi-
Pathol. 1993;46:198-203.
cant effect on the ESR when using the Streck method.
3. Bull BS, Chien S, Dormandy JA, et al. Guidelines on
We compared the Streck ESR-Auto Plus instrument for selection of laboratory tests for monitoring the acute phase
ESR measurement with the Sediplast Westergren method as response. J Clin Pathol. 1988;41:1203-1212.
the reference method and demonstrated that the 2 methods 3. Koepke JA, Bull BS, Simson E, et al. Reference and selected
are correlated at 95%. The y-intercept was 6.5, indicating the procedure for the erythrocyte sedimentation rate (ESR) test;
Approved standard, 4th ed. Wayne, PA: National Committee
presence of systematic bias, being responsible for the higher for Clinical Laboratory Standards; 2000. NCCLS document
ESR measurement by the Streck method. We investigated H2-A4.
various potential factors that might be responsible for sys- 4. Bottiger LE, Svedberg CA. Normal erythrocyte
tematic bias. The higher altitude of our laboratory’s location sedimentation rate and age. Br Med J. 1967;2:85-87.
should have resulted in lower ESR values, but not higher val- 5. Shearn MA, Kang IY. Effect of age and sex on the erythrocyte
sedimentation rate. J Rheumatol. 1986;13:297-298.
ues, thus eliminating altitude as a responsible factor. Because
6. George TI, Etzell JE, Clarke MR, et al. ESR: A Participant
both test methods use citrate as an anticoagulant, the resulting Summary. College of American Pathologists Proficiency
dilution effectively counters higher hematocrit values as a Surveys, 2009.
factor. The fill level of the ESR vacuum tubes did not result 7. Miao G. Reference value of younger people’s erythrocyte
in significant differences in the ESR, which suggests that fill sedimentation rate and altitude [letter]. J Lab Clin Med.
2004;143:367-368.
level variations due to inaccurate filling might not affect ESR
8. Hunsley B, Reiss R, McCarthy K, et al. Streck ESR-vacuum
measurement by the Streck method. The EDTA blood sample tubes stabilize whole blood for accurate ESR testing for
was confirmed to be stable for up to 24 hours when stored at extended time. Erythrocyte Sedimentation Rate Application
2° to 10°C, excluding EDTA tube instability as a potential Note. Issue 1. http://www.Streck.com/pdf/Papers/Hematology/
Streck_ApplicationNote_ESR_320533.pdf. Accessed
factor. We also tested the EDTA blood sample at different February 13, 2010.
temperatures and proved that sample temperature does not 9. Hunsley B, Reiss R, Miller-Lindholm A. Streck ESR-vacuum
affect ESR measurement using the Streck method. The Streck tubes yield stable erythrocyte sedimentation rates with
ESR-Auto Plus measures ESR higher than the Sediplast dilut- variable temperatures. Erythrocyte Sedimentation Rate
Application Note. Issue 2. http://www.Streck.com/pdf/Papers/
ed method owing to the presence of systematic bias, which Hematology/Streck_ApplicationNote_ESR_320534.pdf.
is not clearly understood at this point. The observation of Accessed February 13, 2010.
similar higher results in the College of American Pathologists’
surveys not only for the Streck ESR-Auto Plus, but also for

© American Society for Clinical Pathology Am J Clin Pathol 2011;135:000 5

5 DOI: 10.1309/AJCP48YXBDGTGXEV 5

View publication stats