| Abstract|| |
Context: Monitor 100® (Electa Lab, Italy) is a newly developed automated method for measurement of erythrocyte sedimentation rate (ESR). Aims: The aim of our study was to compare the ESR values by Monitor 100® against the standard Westergren method. Patients and Methods: This cross-sectional study was conducted at a Level I trauma care center on 200 patients. The samples taken were as per the recommendations charted out by International Council for Standardization in Hematology (ICSH) for comparing automated and manual Westergrens method. Statistical Analysis Used: Bland and Altman statistical analysis was applied for evaluating Monitor 100® against the conventional Westergren method. Results: The analysis revealed a low degree of agreement between the manual and automated methods especially for higher ESR values, mean difference -11.2 (95% limits of agreement, -46.3 to 23.9) and mean difference -13.4 (95% limits of agreement-58.9 to 32.1) for 1 and 2 hours, respectively. This discrepancy which is of clinical significance was less evident for ESR values in the normal range <25 mm/hour (-7.7 mean of difference; -18.9 to 3.5 limits of agreement). Conclusions: The fully automated system Monitor 100® for ESR measurement tends to underestimate the manual ESR readings. Hence it is recommended that a correction factor be applied for the range of ESR values while using this equipment. Further studies and validation experiments would be required.
Keywords: Comparison, erythrocyte sedimentation rate, Monitor 100® , Westergren method
|How to cite this article:|
Subramanian A, Rangarajan K, Pandey RM, Gandhi JS, Sharma V, Bhoi SK. Evaluation of an automated erythrocyte sedimentation rate analyzer as compared to the Westergren manual method in measurement of erythrocyte sedimentation rate. Indian J Pathol Microbiol 2011;54:70-4
|How to cite this URL:|
Subramanian A, Rangarajan K, Pandey RM, Gandhi JS, Sharma V, Bhoi SK. Evaluation of an automated erythrocyte sedimentation rate analyzer as compared to the Westergren manual method in measurement of erythrocyte sedimentation rate. Indian J Pathol Microbiol [serial online] 2011 [cited 2018 Jan 22];54:70-4. Available from: http://www.ijpmonline.org/text.asp?2011/54/1/70/77328
| Introduction|| |
The erythrocyte sedimentation rate (ESR) is still widely used in clinical practice as an indicator of inflammation, infection, trauma, or malignant disease and several billion tests are being performed in clinical laboratories throughout the world.  The most satisfactory method of performing the test was introduced by Westergrens in 1921.  Although the method lacks specificity, it can be effective in determining prognosis, as in Hodgkin's disease or prostatic cancer, and for monitoring disease activity as in Rheumatoid arthritis. ,, Despite its advantages, the risk to the medical staff regarding contact with blood specimens leading to blood borne infection is very high.  The original method recommended by International Council for Standardization in Hematology (ICSH) is based on that of Fahraeus and Westergren. , Subsequently, modifications of this reference method were made and ICSH guidelines now allow for the use of alternative ESR techniques provided that comparability with the Westergren method is achieved. 
Many newer and safer methods have evolved to determine ESR accurately without added risks. Monitor 100® (Electa lab, Italy) is an automated technique for measuring ESR. The greatest advantage with this method is that it can give the ESR readings in 30 minutes of 100 patients with all the temperature corrections at 18°C using infrared barriers which are not seen with the usual standardized methods for ESR. However there is no such report regarding the validity of ESR measurement using the Monitor 100® . Therefore, the aim of the present study was to compare the performance of Monitor 100® , an automated ESR analyzer with the gold standard manual Westergren method.
| Patients and Methods|| |
It was a cross-sectional study done on routine hemogram samples over a period of 7 months from June 2008 to December 2008. An approval from the Institutional Ethics Committee was obtained. Patients presenting to the outpatient department were randomly selected and after getting an informed consent from the patient or patient's attendant, total of 458 samples were collected. All the ESR tests were carried out within 3 hours from the time of collection.
Patients from both sexes and all age groups were included in the study. No controls were included in the study. These patients had hematocrit more than or equal to 30% and less than or equal to 36%.
Blood collected by vein puncture taking more than 30 seconds and with excessive venous stasis were excluded from the study. Blood samples which were not in proper proportions to the anticoagulant, strongly lipemic, hyperbilirubinemic, hemolyzed samples were also excluded. Finally, samples from patients with low hemoglobin, i.e., with hematocrit less than 30% and more than 36% were not included.
Quality control was performed by testing one blood sample on each day. Standardized control samples with known normal and abnormal values were selected with a PCV between 0.30 and 0.36 and ESR was performed by both the standard method and by the automated Westergren ESR method. The test results did not differ from that obtained by the standard method by more than 12 mm (Westergren) (ICSH recommendations) and hence was under satisfactory control.
Under all aseptic precautions, samples were collected from the antecubital vein using a 10-ml syringe with 24G needle. Four milliliter of blood sample was drawn in the two special 2-ml EDTA vacutainers containing 1.5 mg/ml of EDTA and mixed immediately five times.
Conventional Westergrens Method
In this method, a disposable, plastic tube with a bore size of 2.55 mm and a length of 230 mm, vertically aligned, open at both ends was used. The pipette was filled with K3 EDTA anticoagulated venous blood to a height of at least 200 mm. The sedimentation occurring at 60 minutes and 120 minutes from the beginning of the test was noted in mm/hour equivalent to the Westergren ESR.
Monitor 100® (Electa Lab, Italy)
The blood was drawn into special MONOSED vacutainers of Monitor 100® (1.6 ml, 120 mm long, 6 mm diameter) with 1.28 ml automatic draw containing 0.32 ml of 3.2% sodium citrate. The blood citrate mix reaches up to a maximum length of 60 mm from the bottom of the tube. The minimum level up to which blood has to be filled is up to a length of 50 mm from the bottom of the tube. Monitor 100® processes the sample only if the blood level is between these two limits. After mixing, the samples were promptly transferred to the analyzer and individual patient identity was given to prevent any preanalytical errors. The ESR reading was taken through a 45-mm high window 2 mm above the maximum sample level. The Monitor 100® has the advantage of giving the results of 100 samples in 30 minutes (equivalent to 1 hour Westergren reading) and 60 minutes (equivalent to 2 hour Westergren reading).
Reading Principle of Monitor 100®
One hundred infrared barriers vertically cover 100 test tube positions. At 0.2 mm intervals, all 100 positions on the reading plate are analyzed at the same time. As soon as the reading plate comprising hundred pairs of infrared rays begins to rise, the indicating system intercepts any position occupied by samples containing the right level of blood. After approximately 3 minutes, the actual analysis begins. The computer records the zero time of each sample at regular intervals of 3 minutes for a total of 30 or 60 minutes. The instrument automatically converts the temperature to 18°C (Manley Table) and gives the reading in 30 minutes.
For comparability according to ICSH guidelines, ESR was measured by the routine Westergren method and also on an aliquot of the same specimen undiluted in a Westergren sedimentation tube specified by ICSH under strictly controlled conditions. The results for the routine method are related to the undiluted ESR as follows:
Routine Westergren ESR (in mm/hour) = (undiluted Westergren ESR × 0.86) - 12.
Any ESR system under evaluation should give a result within 12 mm of this expected Westergren ESR.
Statistical analysis was performed using SPSS 16.0 for Windows (SPSS Inc., Chicago, IL, USA). Evaluation of Monitor 100® method was done as described by Bland and Altman.  Manual Westergren method was considered as the reference method (Gold standard). Therefore mean value for ESR with both the methods is plotted against the difference between the Westergrens and the Monitor 100® . The 95% limits of agreement were calculated as d ± 1.96 SD where d = mean difference between the two measurements; and SD = standard deviation of differences.
| Results|| |
Of the total 458 samples collected, 200 samples were within the recommended ICSH hematocrit range (≤36% and ≥30%) and hence included in the present study. The rest of the samples, i.e., 258 samples, were outside the range, and therefore were excluded from the study. Out of these 200 samples, 79 samples were within the reference range used in our hospital (0-25 mm/hour), while 121 samples had higher ESR values of more than 25 mm/hour. Agreement between the results obtained by manual and automated method for 1 and 2 hours is shown in [Figure 1] and [Figure 2]. The results obtained with the reference method were plotted against the difference between the reference and the automated method separately for 1- and 2-hour values.
The mean difference between the two methods and 95% limits of agreement at 1 hour was found to be -11.2±35.1 (95% limits of agreement, -46.3 to 23.9). The same samples at 2 hours showed the mean difference between the two methods to be -13.4±45.5 (95% limits of agreement, -58.9 to 32.1) [Table 1]. Thus we estimate that the 1 hour ESR readings for 95% of subjects as measured by the automated method will be 46.3 mm/hour below the manual method or 23.9 mm/hour above it. Similarly, the estimated ESR readings measured at 2 hours by the automated method for 95% of subjects will be 58.9 mm/hour below the manual method or 32.1 mm/hour above the reference method. This was unacceptable for clinical interpretation since there was a marked discrepancy between the reference and the automated methods. This variation was particularly evident for samples with high ESR readings greater than 25mm/hour. Hence for samples with high manual ESR values, the mean difference was estimated to be -13.4 and the limits of agreement was (-57.3, 30.5) which was markedly different from the corresponding values -7.7 mean of difference; -18.9 to 3.5 limits of agreement for ESR values less than 25 mm/hour. Thus samples with high ESR values vary considerably around the mean difference compared with samples which had normal ESR readings. We recommend a correction factor to rectify the discrepancy in the results while using this equipment [Table 2].
|Table 1: Mean difference in ESR readings as measured by manual and automated methods |
Click here to view
|Figure 1: Bland and Altman analysis of the comparison between reference and automated method at 1 hour, average difference 11.2; 95% limits of agreement are from 23.9 to -46.3.|
Click here to view
|Figure 2: Bland and Altman analysis of the comparison between reference and automated method at 2 hours, average difference 13.4; 95% limits of agreement are from 32.1 to -58.9.|
Click here to view
| Discussion|| |
Erythrocyte sedimentation rate, though nonspecific, is still commonly used as an indicator of inflammation and infection. The gold standard technique for measuring ESR is the Westergrens method. However, it is fraught with many disadvantages.  Attempts to introduce automated systems for measuring ESR have been in vogue since many years. The modifications introduced include the use of unopened blood collection tubes, vacuum-controlled aspiration of the sample providing correct dilution with the anticoagulant, and automated mixing. ,,,, Due to the rise in blood borne diseases such as hepatitis B and HIV, safety precautions against contamination of laboratory personnel are a must and using an automatic Westergrens such as Monitor 100® is one such step. 
Many new automated systems have been introduced since 1990s and have been evaluated for performance with each other as well as with the gold standard Westergrens method. Some of them include Ves-matic 60 (Menarini Diagnostics S.r.l. Milan, Italy), Sediscan (Becton Dickinson, Meylan Cedex, France), Sedimatic (Technicon international Inc, Tokyo, Japan), TEST 1 automated ESR analyzer (Alifax S.p.A, Polverara, Italy), and others. Although these automated techniques offer more benefits in terms of reduced biohazard risks, speedy processing time, and quicker results, it is essential to validate these equipments against the standard Westergrens method to enable routine use and to replace the standard ESR method in any hospital setting.
Monitor 100® is a newly developed automated method which can give the ESR readings in 30 minutes (equivalent to 1 hour Westergren) of 100 patients with all the temperature corrections at 18°C using infrared barriers which are not seen with the usual standardized methods for ESR.  The Westergren's ESR reading at 1 hour correlated with 30 minutes reading of automated analyzer and Westergren's reading at 2 hour correlated with the automated reading at 1 hour. The automated ESR analyzer does not provide an equivalent reading for 30 minutes Westergren method and hence comparison with 30 minutes Westergren reading was not done. The added advantage of Monitor 100® is that there is no external influence on the final reading such as temperature, contaminating dust particles, tilting of tube, and ratio of diluents. The number of samples that can be processed with this method is higher (maximum of 100 samples can be processed at a time) than the manual method with the additional benefit that samples can be added in between. This device has not been validated previously in India. To the best of our knowledge, this is the first study comparing the efficacy of the Monitor 100® with the standard Westergren method.
In the present study, the results obtained with the automated technique were compared with the gold standard Westergrens method using the agreement analysis of Bland and Altman.  Statistical analysis such as linear regression and correlation, though commonly used in the evaluation of new equipment, is usually not considered to be appropriate for validating equipments. Agreement analysis is a more sensitive method than the correlation coefficient for comparison between the two methods. Bland and Altman analysis revealed marked discrepancy between the two methods especially for values on the higher side. The mean of the difference between the two methods (i.e., bias -11.2 mm/hour) for the 1 hour values and -13.4 mm/hour for the 2 hour values were clinically significant and hence unacceptable due to marked discrepancy in the readings between the two methods. This discrepancy was not evident for normal ESR values less than 25 mm/hour (-7.7, mean of difference; -18.9, 3.5 limits of agreement).
Such discrepancies with the ESR automated analyzers have also been shown previously by various authors. ,,, In a study by Caswell et al,  the automated system of Ves-matic (Diesse Diagnostica Senese, Milan, Italy) showed low agreement values with the Westergren ESR method of the ICSH.  In contrast, various other authors showed a good correlation and agreement results with the same method. ,, Low agreement results were also shown with the TEST 1 automated analyzer (Alifax S.p.A, Polverara, Italy) by Plebani et al. , However, validation studies of TEST 1 analyzer comparing with the reference method published by other authors revealed high agreement values. ,,, In a similar analysis of the fully automated SEDI system TM (Becton Dickinson, Meylan Cedex, France), Alfadhli et al,  showed low-agreement results with the higher ESR values and comparable results in the normal ESR range. We obtained similar results with marked variation in the high ESR range and comparable results in the normal ESR range. More recently, newer automated systems for measuring ESR have shown comparably good agreement results enabling their use in clinical laboratories with a high workload as well as for emergency laboratories. ,
The use of Bland and Altman analysis for evaluating the agreement between the two methods not only assesses the mean of the difference (d) between the two methods (i.e., bias) but also the limits of agreement by calculating the standard deviation of the differences (d ± 2SD). Only when the difference (d ± 2SD) does not affect the clinical interpretation, the two methods can be used interchangeably. In other words, the probability of obtaining similar results with the two methods is doubtful; hence the difference in the results obtained with both the methods should be clinically acceptable. Only then, can we justify the new method replacing the standard method in the hospital. It should be kept in mind that these device evaluation and method agreement studies are basically done with the desire to know how much hope we should place on these equipments for making important decisions in patient management. Hence it is suggested that a correction factor for higher ESR values be incorporated while using Monitor 100® . We also suggest a table of correction factors for the range of Monitor 100® ESR values to be incorporated while using this equipment. Further, more studies on validation would be required.
| Conclusions|| |
On comparing the manual and automated methods for measuring ESR, marked discrepancy in the ESR results was noted for high ESR values. However, this was not evident for normal ESR values. Thus the automated system tended to underestimate the manual readings for ESR values on the higher range which is clinically unacceptable. Hence it is recommended that a correction factor be applied for the range of ESR values while using this equipment. Further studies and validation experiments would be required.
| References|| |
|1.||Plebani M, Piva E. Erythrocyte sedimentation rate: Use of fresh blood for quality control. Am J Clin Pathol 2002;117:621-6. |
|2.||Westergren A. Studies of the suspension stability of the blood in pulmonary tuberculosis. Acta Med Scand 1921;54:247-82. |
|3.||User manual. Monitor 100. Forli, Italy: Electra Lab s.r.l.; 2005. |
|4.||Brigden ML. Clinical utility of the ESR: Am Fam physician 1999;60:1443-50. |
|5.||Bull BS, Chien S, Dormandy JA, Lewis SM. Guidelines for selection of laboratory tests for monitoring the acute phase response. J Clin Pathol 1988;41:1203-12. |
|6.||Fahraeus R. The suspension stability of blood. Acta Med Scand 1921;55:1-228. |
|7.||International council for Standardization in hematology. ICSH recommendation for measurement of ESR. J Clin Pathol 1993;46:198-203. |
|8.||Bland JM, Altman DG. Statistical method for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10. |
|9.||Atas A, Cakmak A, Soran M, Karazeybek H. Comparitive study between the Ves-Matic and Micro erythrocyte sedimentation rate method. J Clin Lab Anal 2008;22:70-2. |
|10.||Patton WN, Meyer PJ, Stuart J. Evaluation of sealed vacuum extraction method (Seditainer) for measurement of erythrocyte sedimentation rate. J Clin Pathol 1989;42:313-7. |
|11.||Besson I, Kinder M, Jou JM, Vives Corron JL. The evaluation of three automatic systems for determining the velocity of globular sedimentation. Sangre 1995;40:103-7. |
|12.||Plebani M, De Toni S, Sanzari MC, Bernardi D, Stockreiter E, et al. The TEST1 automated system: A new method for measuring the erythrocyte sedimentation rate. Am J Clin Pathol 1998;110:334-40. |
|13.||Plebani M, Piva E, Sanzari MC. Method comparison of automated systems for the erythrocyte sedimentation rate- the author's reply. Am J Clin Pathol 1999;112:723-4. |
|14.||Fernandez de Castro M, Fernandez Calle P, Viloria A, Larrocha C, Jimenez MC. Evaluation of a fully automated system for determination of the erythrocyte sedimentation rate. Sangre 1989;34:1-9. |
|15.||Caswell M, Stuart J. Assessment of Diesse Ves-matic 20 automated system for measuring erythrocyte sedimentation rate. J Clin Pathol 1991;44:946-9. |
|16.||AlFadhli SM, Al-Awadhi AM. Comparison of erythrocyte sedimentation rate measurement by the automated SEDI system and conventional Westergren method using the Bland and Altman statistical method. Med Princ Pract 2005;14:241-4. |
|17.||Fares AK. Evaluation of the Ves-matic 20 -An automated system for the determination of the erythrocyte sedimentation rate. Middle East Laboratory 2001;4:19-23. |
|18.||Giavarina D, Dall'Olio G, Soffiati G. Method comparison of automated systems for the erythrocyte sedimentation rate. Am J Clin Pathol 1999;112:721-2. |
|19.||Arezzini C, Ricci A. Method comparison of automated systems for the erythrocyte sedimentation rate-more study needed [Letter]. Am J Clin Pathol 1999;112:722-4. |
|20.||Romero A, Muñoz M, Ramírez G. Length of sedimentation reaction in blood: A comparison of the test 1 ESR system with the ICSH reference method and the sedisystem 15. Clin Chem Lab Med 2003;41:232-7. |
|21.||De Jonge N, Sewkaransing I, Slinger J, Rijsdijk JJ. Erythrocyte sedimentation rate by the Test-1 analyzer Clin Chem 2000;46:881-2. |
|22.||Ajubi NE, Bakker AJ, Van den Berg GA. Determination of the length of sedimentation reaction in blood using the TEST 1 system: Comparison with the Sedimatic 100 method, turbidimetric fibrinogen levels, and the influence of M-proteins. Clin Chem Lab Med 2006;44:904-6. |
|23.||Ozdem S, Akbas HS, Donmez L, Gultekin M. Comparison of TEST 1 with SRS 100 and ICSH reference method for the measurement of the length of sedimentation reaction in blood. Clin Chem Lab Med 2006;44:407-12. |
|24.||Arikan S, Akalin N. Comparison of the erythrocyte sedimentation rate measured by the Micro Test 1 sedimentation analyzer and the conventional Westergren method. Ann Saudi Med 2007;27:362-5. |
|25.||Mahlangu JN, Davids M. Three-way comparison of methods for the measurement of the erythrocyte sedimentation rate. J Clin Lab Anal 2008;22:346-52. |
Department of Lab Medicine, Jai Prakash Narayan Apex Trauma centre, AIIMS, New Delhi - 110 029
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2]
[Table 1], [Table 2]