Asian Journal of Transfusion Science
Home About Journal Editorial Board Search Current Issue Ahead of print Back Issues Instructions Subscribe Login  Users: 754 Print this page  Email this page Small font sizeDefault font sizeIncrease font size 


 
ORIGINAL ARTICLES Table of Contents   
Year : 2020  |  Volume : 14  |  Issue : 1  |  Page : 49-53
Comparison of the accuracy of capillary hemoglobin estimation and venous hemoglobin estimation by two models of HemoCue against automated cell counter hemoglobin measurement


1 Department of Transfusion Medicine and Blood Bank, AIIMS, Rishikesh, Uttarakhand, India
2 Department of Pathology and Laboratory Medicine and In-charge, Blood Bank, AIIMS, Rishikesh, Uttarakhand, India

Click here for correspondence address and email

Date of Submission12-Jul-2017
Date of Acceptance30-Nov-2017
Date of Web Publication24-Jul-2020
 

   Abstract 

Background: HemoCue point of care devices has been extensively used in screening for anemia in blood banking. HemoCue can estimate hemoglobin (Hb) both from venous as well as capillary blood. However, the suitability of HemoCue Hb estimation in donor selection is unclear.
Aims: The aims of this study were to evaluate variance of difference in Hb measurement in capillary HemoCue estimation as compared to venous HemoCue estimation from automated cell counter and to assess accuracy of two different HemoCue models (201 and 301) against automated cell counter Hb measurements in both capillary as well as venous blood.
Materials and Methods: HemoCue 201 and 301 were evaluated by a comparison of methods study against Sysmex XP-100 three-part analyzer at a blood bank of a tertiary care hospital in Uttarakhand, India, in 2017. Assessment for anemia of 115 donors was done initially by capillary Hb by a convenience sampling to 2 instruments from 2 different models of HemoCue (total of 4 instruments). Venous blood collected was analyzed by Sysmex XP-100 and all HemoCue analyzers.
Results: For capillary method, bias ranged from −0.97 to −0.37 g/dL, upper limit of agreement (LOA) ranged from 0.72 to −1.06 g/dL, and lower LOA ranged from −2.65 to −1.79 g/dL. For venous method, bias ranged from −0.03 to −0.24 g/dL, the upper LOA ranged from 0.81 to −1.07 g/dL, and lower LOA ranged from −1.04 to −0.57 g/dL. Thus, capillary HemoCue estimation exhibited greater bias as well as wider LOA. Variance of the differences from automated counter was significantly lower for venous HemoCue comparison compared to capillary HemoCue estimation (P < 0.001 for each instrument).
Conclusion: Errors in capillary sampling of blood show the extent to which preanalytical errors can influence results in point-of-care devices. We suggest augmentation of any blood bank-based Hb screening process based just on capillary sampling to be augmented by a properly selected venous sampling to reduce deferral for a false-positive screen of anemia.

Keywords: Capillary, hemoglobin, HemoCue, variance, venous

How to cite this article:
Jain A, Chowdhury N. Comparison of the accuracy of capillary hemoglobin estimation and venous hemoglobin estimation by two models of HemoCue against automated cell counter hemoglobin measurement. Asian J Transfus Sci 2020;14:49-53

How to cite this URL:
Jain A, Chowdhury N. Comparison of the accuracy of capillary hemoglobin estimation and venous hemoglobin estimation by two models of HemoCue against automated cell counter hemoglobin measurement. Asian J Transfus Sci [serial online] 2020 [cited 2020 Oct 25];14:49-53. Available from: https://www.ajts.org/text.asp?2020/14/1/49/290650





   Introduction Top


Predonation hemoglobin (Hb) screening for blood donors is an essential procedure in blood banking. It is essential that a rapid, yet accurate method for Hb screening is used that can reject donors having low Hb to ensure donor health and maintain quality of blood components. Yet, it is also essential that prospective donors with adequate Hb not be falsely rejected so that supply of blood is not compromised. A rapid portable method is also essential for use in outdoor blood camps.

The HemoCue point of care devices has been extensively used in the screening for anemia in clinical situations.[1],[2] This method has been shown superior to the copper sulfate method as a donor Hb screening method.[3],[4] Although this method has been found satisfactory by some users, others have found HemoCue to result in a high number of inappropriate donor rejections.[5],[6],[7] HemoCue can estimate Hb both from venous as well as capillary blood. It is unclear Hb from which mode of blood collection was validated in some method comparison studies. Furthermore, the issue of instrument-to-instrument variability has also not been adequately reported.

Keeping in mind the above limitations in the existing literature, the present study was carried out in order to assess the accuracy of two different HemoCue models (model 201 and 301) against automated Hb measurements in both capillary as well as venous blood.


   Materials and Methods Top


This study was initially conceived as a routine quality assessment of HemoCue instruments received and in use in our blood bank after preliminary validation. This study follows the Helsinki guidelines,[8] and we obtained our institutional ethical committee approval to publish our quality assessment findings.

Our blood bank has two instruments each of two HemoCue models 201 and 301 (hereafter, the individual instruments are referred to as a201, b201, a301, and b301, respectively). This study was carried out in the blood bank of a tertiary health care hospital in Uttarakhand, India, from February to March 2017. One hundred and fifteen prospective blood donors, after passing the medical checkup and found otherwise satisfactory according to Indian national regulatory guidelines,[9] were assessed for anemia initially by capillary blood Hb by a convenience sampling to two instruments from different models of HemoCue. Proper instructions of the user's manual were followed. Each donor was pricked once and the third and fourth drop of capillary blood was tested on two separate instruments after discarding initial two drops of blood. More than two instruments could not be tested on a single prick because of the difficulty in obtaining enough capillary blood for all four instruments after a single prick. Once found satisfactory, the donors were directed for donation. In such donors, venous blood was collected from the diversion pouch in K2 EDTA vials and sent for hematological analysis. In the case of prospective donors being found to have low Hb on predonation capillary Hb assessment, venous blood samples in K2 EDTA vials were taken to confirm or reject the findings, with the donor being temporarily deferred until the annulment of the capillary Hb report by an automated analyzer on the venous sample.

The venous blood thus collected was analyzed by Sysmex XP-100 three-part analyzer. Hb from the venous blood was also measured by all the HemoCue analyzers

Comparison between the different methods was carried out by the Bland–Altman limits of agreement.[10] The following comparisons were undertaken:

  1. Separate comparisons of the capillary Hb estimation by the four different HemoCue instruments versus the automated Hb measurement by Sysmex XP-100
  2. Separate comparisons of the venous Hb estimation by the four different HemoCue instruments versus the automated Hb measurement by Sysmex XP-100. Before reporting the results, the differences between the different methods compared were visualized graphically by Bland–Altman plots and scatterplots
  3. Comparison of the variance of the differences between the capillary Hb estimation and the hematology analyzer readings to the variance of the differences between the venous Hb estimation by HemoCue and the hematology analyzer readings by the Levene's test.


All statistical analysis was carried out by the R statistical environment[11] and NCSS 11 statistical software (NCSS, LLC, Kaysville, Utah, USA).[12]


   Results Top


The basic statistical descriptives of the Hb measurements by the different methods by the different instruments are given in [Table 1]. The mean difference along with the limits of agreement between the different instruments and the automated Hb estimate is given in [Table 2]. The above results show that there is significant under-reporting of Hb values by the capillary method of HemoCue relative to the automated counts. The bias is much reduced when Hb is measured from the venous sample by the HemoCue. The limits of agreement are also narrower (indicating lesser random error) for the venous estimation of Hb by HemoCue [Figure 1]. The variance of the differences from the automated counter is significantly lower for the venous HemoCue estimation as compared to the capillary HemoCue estimation (P < 0.001 for each instrument by the Levene's test).
Table 1: The summary statistics of the reported hemoglobin by various methods by the instruments used

Click here to view
Table 2: The mean difference and the Bland–Altman limits of agreement for the various HemoCue instruments and hemoglobin estimation method versus the Sysmex analyzer

Click here to view
Figure 1: Variance of difference in HemoCue capillary versus venous hemoglobin measurement from automated counter hemoglobin measurement

Click here to view


The number of hemoglobin measurements which exceeded the acceptable total error of 7% for each mode of hemoglobin measurement for each instrument was also calculated. The number of such unacceptable errors in hemoglobin estimates by capillary sampling was 11 (out of 57, i.e., 19.30%) and 10 (out of 58, i.e., 17.24%) for the two instruments of HemoCue Model 301 and 23 (out of 55, i.e., 41.82%) and 21 (out of 58, i.e., 36.21%) for the two instruments of HemoCue model 201. The number of unacceptable errors in venous sampling was 1 (out of 115, i.e., 0.87%) and 1 (out of 115, i.e., 0.87%) for the two instruments of HemoCue model 301 and 2 (out of 115, i.e., 1.74%) and 1 (out of 115, i.e., 0.87%) for the two instruments of HemoCue model 201.


   Discussion Top


The HemoCue devices with estimation of Hb from venous blood were found satisfactory in the present study, with acceptable shift and narrow limits of agreement. The Clinical Laboratory Improvement Amendments standards specify a total error of <7% for Hb estimation to be acceptable;[13]<2% of the measurements taken from venous blood in all the HemoCue instruments showed a >7% difference from the Sysmex analyzer.

The Hb estimations from capillary sampling in all the HemoCue instruments, however, are unsatisfactory for the blood bank, showing unacceptable bias as well as wide limits of agreement. About 17%–19% of the measurements taken by capillary sampling in HemoCue 301 and 36%–42% of the measurements taken by capillary sampling in HemoCue 201 showed difference of >7% compared to the Sysmex analyzer.

If the findings in the present study can be slightly extrapolated, we would come to the inevitable conclusion that a lot of false-positive deferrals for anemia have taken place due to the negative bias of the capillary sampled HemoCue devices relative to Sysmex. This places a lot of burden on the blood supply. Therefore, we suggest that capillary sampled Hb from the HemoCue instruments be used with caution. Similar findings were found in a Spanish study, due to which they recommended a two-step strategy for Hb screening with the HemoCue.[5] Briefly, a two-step strategy entails re-testing of the venous blood if and when the prospective donor fails to have satisfactory Hb by capillary sampled blood. Since the venous sampled Hb by HemoCue is satisfactory for blood bank screening, this can be an ideal strategy in the present scenario. A criticism of this strategy is that an additional phlebotomy has to be done even before blood donation and may lead to increased donor noncompliance; however, this cost has to be weighed against the risk of loss of a donor anyway due to a false-positive screen of low Hb.

The errors in the capillary sampling of blood show the extent to which preanalytical errors can influence the results in point-of-care devices. The satisfactory results with venous sampling coupled with the unsatisfactory results with the capillary sampled Hb, supporting other studies,[5],[7] show the difficulty in implementing a satisfactory point of care diagnostic or screening device. Even though the analytical quality of the devices seems to be satisfactory, preanalytical errors seem to be the primary determinants in the determination of quality or lack thereof of HemoCue devices. There may be variability in the amount of blood expressed, contamination with tissue fluids, errors in collection as well as selection of site for the lancet puncture, all of which may lead to significant errors in capillary blood sampling. Therefore, special attention needs to be directed toward training and proper use of the technique. However, the present evaluation used a single trained person for evaluation of Hb to reduce the error variability; the errors found in spite of such a precaution accentuates the problem of preanalytical errors even further.

This study was carried out in a blood bank setting; the unsatisfactoriness of capillary measurement for screening of anemia may or may not hold in a community setting. That evaluation needs to take the particular problem in question, and whether sensitivity or specificity or both of anemia diagnosis is more important. In a community setting needing a high sensitivity followed by confirmatory investigation and rapid treatment of anemia, a high sensitivity may take precedence over specificity, and capillary sampling may still be found adequate.


   Conclusion Top


Hb estimation after venous sampling by HemoCue models 201 and 301 is accurate and suitable for Hb screening in the blood bank. Hb estimation after capillary sampling should be used more cautiously; we suggest augmentation of any blood bank-based Hb screening process based just on capillary sampling to be augmented by a properly selected venous sampling to reduce deferral for a false-positive screen of anemia.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
   References Top

1.
Sanchis-Gomar F, Cortell-Ballester J, Pareja-Galeano H, Banfi G, Lippi G. Hemoglobin point-of-care testing: The HemoCue system. J Lab Autom 2013;18:198-205.  Back to cited text no. 1
    
2.
Cohen AR, Seidl-Friedman J. HemoCue system for hemoglobin measurement. Evaluation in anemic and nonanemic children. Am J Clin Pathol 1988;90:302-5.  Back to cited text no. 2
    
3.
Gómez-Simón A, Navarro-Núñez L, Pérez-Ceballos E, Lozano ML, Candela MJ, Cascales A, et al. Evaluation of four rapid methods for hemoglobin screening of whole blood donors in mobile collection settings. Transfus Apher Sci 2007;36:235-42.  Back to cited text no. 3
    
4.
Boulton FE, Nightingale MJ, Reynolds W. Improved strategy for screening prospective blood donors for anaemia. Transfus Med 1994;4:221-5.  Back to cited text no. 4
    
5.
Gómez-Simón A, Plaza EM, Torregrosa JM, Ferrer-Marín F, Sánchez-Guiu I, Vicente V, et al. Evaluation of two-step haemoglobin screening with HemoCue for blood donor qualification in mobile collection sites. Vox Sang 2014;107:343-50.  Back to cited text no. 5
    
6.
Muñoz M, Romero A, Gómez JF, Manteca A, Naveira E, Ramírez G, et al. Utility of point-of-care haemoglobin measurement in the HemoCue-B haemoglobin for the initial diagnosis of anaemia. Clin Lab Haematol 2005;27:99-104.  Back to cited text no. 6
    
7.
Bahadur S, Jain S, Jain M. Estimation of hemoglobin in blood donors: A comparative study using hemocue and cell counter. Transfus Apher Sci 2010;43:155-7.  Back to cited text no. 7
    
8.
General Assembly of the World Medical Association. World medical association declaration of Helsinki: Ethical principles for medical research involving human subjects. J Am Coll Dent 2014;81:14-8.  Back to cited text no. 8
    
9.
National AIDS Control Organization. Standards for Blood Banks and Blood Transfusion Services. New Delhi: National AIDS Control Organization, Ministry of Health and Family Welfare, Government of India; 2007.  Back to cited text no. 9
    
10.
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.  Back to cited text no. 10
    
11.
RR Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2014.  Back to cited text no. 11
    
12.
NCSS, LLC. NCSS 11 Statistical Software. Kaysville, Utah, USA: NCSS, LLC; 2016.  Back to cited text no. 12
    
13.
Westgard QC. CLIA Requirements for Analytical Quality – Westgard; 2009. Available from: https://www.westgard.com/clia.htm. [Last accessed on 2016 Dec 24].  Back to cited text no. 13
    

Top
Correspondence Address:
Dr. Ashish Jain
Quarter No: A-20, Pashulok Barrage Colony, Rishikesh, Uttarakhand
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/ajts.AJTS_93_17

Rights and Permissions


    Figures

  [Figure 1]
 
 
    Tables

  [Table 1], [Table 2]



 

Top
 
  Search

  
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Email Alert *
    Add to My List *
* Registration required (free)  


    Abstract
   Introduction
    Materials and Me...
   Results
   Discussion
   Conclusion
    References
    Article Figures
    Article Tables

 Article Access Statistics
    Viewed334    
    Printed27    
    Emailed0    
    PDF Downloaded0    
    Comments [Add]    

Recommend this journal

Association Contact us | Sitemap | Advertise | What's New | Feedback | Copyright and Disclaimer

2006 - Asian Journal of Transfusion Science | Published by Wolters Kluwer - Medknow
Online since 10th November, 2006