| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
Department of Clinical Chemistry, Leyenburg Hospital, PO Box 40.551, 2504 LN The Hague, The Netherlands
a Author for correspondence. Fax 31-70-3592191.
To the Editor:
We evaluated the analytical performance of the Test-1 automated analyzer (Alifax S.p.A) for the assessment of the erythrocyte sedimentation rate (ESR). In this analyzer, 150-µL samples are aspirated, rotated slowly for 2 min, and put into the system, which measures the erythrocyte concentration in a capillary through a photometer at a wavelength of 950 nm. For each sample, a sedimentation curve is obtained, which is transformed to comparable Westergren values. The first result is obtained in <3 min, and subsequent results are obtained every 30 s (1)(2).
The Test-1 method was compared with both the manual Westergren reference method and an automated Westergren method (Starrsed; Charles Goffin Medical Systems) (3). Sodium citrate-anticoagulated blood samples were used for both Westergren methods because this is the current practice in our laboratory. For the Test-1, tripotassium EDTA-anticoagulated blood samples were used. Patient samples (n = 105) with a Westergren ESR of 2–120 mm/h were used to compare the three methods. Passing and Bablok regression analysis was used for comparisons of the three methods (4)(5), and the Student t-test was used to test for differences between test results among the three methods. The day-to-day reproducibility of the Test-1 method was assessed by measuring 20 samples 10 times. The maximum permissible storage time was assessed by measuring the ESR in 10 samples after 0, 1, 2, 3, 4, 5, 24, and 48 h. The samples were stored at room temperature up to 5 h, after which they were stored at 4 °C.
A good correlation was found between the Test-1 system and the
Westergren reference method: the correlation coefficient was 0.97;
intercept, -0.15 mm/h [95% confidence interval (CI), -0.64 to
0.48]; slope, 0.83 (95% CI, 0.76–0.88); Sy|x =
5.33; n = 105. Compared with the Starrsed method, the correlation
coefficient was 0.97; intercept, -0.48 mm/h (95% CI, -0.87 to 0.29);
slope, 0.91 (95% CI, 0.86–0.97); Sy|x = 5.06;
n = 105 (Fig. 1
A). Although the difference plot shows a slight increase in test
difference toward higher ESR values (Fig. 1B
), no statistically
significant difference was found among ESR results obtained with the
three ESR methods. The bias between TEST-1 and the manual Westergren
method was -2.7 mm/h (agreement limits, -33 to 8 mm/h). Concerning
reproducibility, the mean interassay CV was 7.9%. The ESR values for
samples stored for up to 24 h were, on average, 4% lower than the
initial values. However, after storage for 48 h, ESR values were,
on average, 24% lower than the initial values. Therefore, we consider
the maximum permissible storage time to be 24 h.
|
In conclusion, the Test-1 offers a fast determination of the ESR with acceptable intraassay variability and good correlation with traditional Westergren methods. The bias and agreement limits are comparable to those found by others (6)(7). The use of samples with EDTA as anticoagulant instead of sodium citrate has two possible advantages: the avoidance of a dilution step and consequent analytical errors, and the use of one sample for multiple hematologic analyses (e.g., ESR and complete blood count) (2).
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. T M Raijmakers, P. H M Kuijper, D. L Bakkeren, and H. L Vader The effect of paraproteins on the erythrocyte sedimentation rate: a comparison between the StarrSed and TEST 1 Ann Clin Biochem, November 1, 2008; 45(6): 593 - 597. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
