Vox Sanguinis (2014) 107, 343–350 © 2014 International Society of Blood Transfusion DOI: 10.1111/vox.12165

ORIGINAL PAPER

Evaluation of two-step haemoglobin screening with HemoCue for blood donor qualification in mobile collection sites A. G
omez-Sim
on, E. M. Plaza, J. M. Torregrosa, F. Ferrer-Mar
ın, I. S
anchez-Guiu, V. Vicente, M. L. Lozano & J. Rivera Centro Regional de Hemodonacion, Servicio de Hematologıa y Oncologıa Me dica, IMIB, Universidad de Murcia, Murcia, Spain

Background and Objectives Inaccuracy of fingerstick haemoglobin compromises donor’s health and losses blood donations. We evaluated the benefit of double haemoglobin screening with HemoCue. Study Design and Methods Blood donors underwent fingerstick screening by HemoCue and were driven for donation if capillary haemoglobin was within the regulatory range. Those failing were drawn venous blood and donated if their venous haemoglobin determined with HemoCue was acceptable. Results Of 276 605 donor clinic visits, 10 011 (3
6%) were assessed by two-step haemoglobin screening using HemoCue, because of low (n = 9444) or high (n = 567) capillary haemoglobin. Among these, 2561 (25
6%) were deemed eligible [recovered donations]. The recovery rate was 23
8% and 55
0% among donors presenting with low and high capillary haemoglobin, respectively. In both categories of attempted donations, capillary and venous haemoglobin with HemoCue correlated significantly in recovered donors (R2  0
5–0
7) but not in deferred visits (R2 < 0
15). Venous haemoglobin with HemoCue and by haematological analyzer significantly correlated in all donations attempts (R2  0
7). Donors presenting with low capillary haemoglobin showed small bias between capillary and venous haemoglobin by HemoCue (-2
4 – 6
2 g/l), fingerstick haemoglobin and venous haemoglobin with counter (1
3 – 7
3 g/l), and venous haemoglobin with HemoCue and counter (3
7 – 3
9 g/l). This bias was slightly greater in donors with high capillary haemoglobin (-7
5 – 7
8, 13
7 – 7
5, and 6
2 – 7
5, respectively). Double haemoglobin screening by HemoCue reached an accuracy of 87
3% for qualifying donors presenting with low fingerstick haemoglobin.

Received: 13 December 2013, revised 16 April 2014, accepted 8 May 2014, published online 11 June 2014

Conclusions Double haemoglobin measurement with HemoCue [fingerstick and venous blood if required] is feasible and allows a significant recovery of blood donations. Key words: blood donation, blood donors, capillary haemoglobin, HemoCue, venous haemoglobin.

Introduction Assessment of haemoglobin level is a requirement aiming to safeguard the health of blood donors and ensure the quality of blood units for transfusion, and the current

Correspondence: Jos
e Rivera, Centro Regional de Hemodonaci
on, C/Ronda de Garay s/n, Murcia 30003, Spain E-mail: [email protected]

haemoglobin cut-off values set by the European Union are 125 g/l for women and 135 g/l in men [1]. Low haemoglobin is the most common cause of donor deferral worldwide, particularly among women [2–6]. As temporary deferral may discourage donors to return, improper donation rejection due to low haemoglobin causes not only the current loss of blood units, but also possible shortage of blood supply in future [7–10]. The determination of haemoglobin in a venous sample by an automatic haematological counter is the method of

343

344 A. G
omez-Sim
on et al.

reference, but its use is not feasible in blood collection at mobile units. Portable haemoglobinometers are used worldwide for donor haemoglobin evaluation on the donor side, as they are fast, easy to use and provide a quantitative measurement within a wide range of haemoglobin concentrations [11–15]. We have previously demonstrated that, in our routine working conditions, measurement of capillary haemoglobin with the HemoCue system is superior to the classical gravimetric test copper sulphate in detecting individuals with anaemia, but a significant proportion of donors with acceptable haemoglobin concentration are inappropriately rejected [16]. Differences in the level of haemoglobin in capillary and venous blood might account, at least in part, for the misclassification of blood donors using HemoCue [17, 18]. In this study, we have evaluated a two-step method for the assessment of haemoglobin in blood donors with HemoCue, consisting of a first measurement in capillary blood followed by the determination of haemoglobin in venous blood samples in those donors who failed the fingerstick screening.

Subjects and Methods This study included all donor clinic visits at the mobile blood collection sites in the Region of Murcia (south-east Spain) during a 5-year period (September 2008–September 2013). Eligibility for donation was assessed according to the Spanish (RD 1088/2005) and European regulatory criteria [1]. Each donor clinic visit received a single number according to the ISBT 128 codification system and was linked to the donor’s identification number to ensure traceability. Electronic records of donations and donors [demographic data, physical and analytical examinations, history of donations, informed consent, etc.] were managed using the Hematos IIG software (MedInfo, Nice, France). A two-step haemoglobin screening procedure was used for qualifying donors. First, all donors underwent finger€ stick screening with HemoCue 301 (HemoCue AB Angelholm, Sweden), as described [16]. They were eligible for donation if the capillary haemoglobin level was within our established cut-off values, that is, between 125 and 165 g/ l for women and between 135 and 185 g/l for men. The lower haemoglobin limits were set according to current European standards [1], while the upper haemoglobin cutoff for men and women was an internal criterion incorporated in our centre based on recommendations from the Spanish Ministry of Health to exclude donors with polyglobulia. Donors displaying capillary haemoglobin values outside these cut-off ranges were asked to be drawn a venous blood sample in an EDTA tube (Becton Dickinson, Madrid, Spain) by standard phlebotomy. After mixing by repetitive mild inversion, blood aliquots from the collection tubes were placed with a pipette onto plastic films

and used to determine the haemoglobin level using the same HemoCue. These donors were considered eligible for donation if their venous haemoglobin concentration fulfilled the above regulatory levels and were labelled as ‘recovered donors’, or were definitively rejected for donation. In all these recovered and deferred donors, venous haemoglobin was also determined, in the same EDTA tubes within 12 h of collection, by an automatic haematology analyser (Coulter Electronics, Hialeah, FLA). These values were considered as the ‘real’ donor’s haemoglobin. Data are shown as mean – SD unless otherwise stated. The Kolmogorov–Smirnov test was used to check for the normal distribution of continuous data. Haemoglobin values were compared by paired or unpaired Student’s t-test and Pearson’s correlation test as indicated. Agreement between venous haemoglobin concentrations measured with HemoCue and electronic counter was also evaluated according to the Bland–Altman approach [19]. Cohen’s kappa statistic (k) [20] was calculated to analyse agreement between HemoCue and haematological analyzer in the categorization of donors as eligible or ineligible according to their venous haemoglobin concentration. The statistical SPSS 15.0 software (SPSS, Inc., Chicago, IL, USA) was used.

Results General characteristics of donors and donations A total of 276 605 donor clinic visits were attended during the study period, 10
1% from new donors. According to age, 19
2% were younger than 30 years, 55
0% from 30 to 50, and 25
8% over 50 years. Overall, males were more prone to donate blood than women (54
1% vs. 45
9%), but more women offered to donate in the subgroup of younger’s donors (