ORIGINAL ARTICLE Comparison of three noninvasive methods for hemoglobin screening of blood donors Sergey Ardin,1* Melanie Störmer,1* Stela Radojska,1 Larissa Oustianskaia,1 Moritz Hahn,2 and Birgit S. Gathof 1
BACKGROUND: To prevent phlebotomy of anemic individuals and to ensure hemoglobin (Hb) content of the blood units, Hb screening of blood donors before donation is essential. Hb values are mostly evaluated by measurement of capillary blood obtained from fingerstick. Rapid noninvasive methods have recently become available and may be preferred by donors and staff. The aim of this study was to evaluate for the first time all different noninvasive methods for Hb screening. STUDY DESIGN AND METHODS: Blood donors were screened for Hb levels in three different trials using three different noninvasive methods (Haemospect [MBR Optical Systems GmbH & Co. KG], NBM 200 [LMB Technology GmbH], Pronto-7 [Masimo Europe Ltd]) in comparison to the established fingerstick method (CompoLab Hb [Fresenius Kabi GmbH]) and to levels obtained from venous samples on a cell counter (Sysmex [Sysmex Europe GmbH]) as reference. The usability of the noninvasive methods was assessed with an especially developed survey. RESULTS: Technical failures occurred by using the Pronto-7 due to nail polish, skin color, or ambient light. The NBM 200 also showed a high sensitivity to ambient light and noticeably lower Hb levels for women than obtained from the Sysmex. The statistical analysis showed the following bias and standard deviation of differences of all methods in comparison to the venous results: Haemospect, −0.22 ± 1.24; NBM, 200 −0.12 ± 1.14; Pronto-7, −0.50 ± 0.99; and CompoLab Hb, −0.53 ± 0.81. CONCLUSION: Noninvasive Hb tests represent an attractive alternative by eliminating pain and reducing risks of blood contamination. The main problem for generating reliable results seems to be preanalytical variability in sampling. Despite the sensitivity to environmental stress, all methods are suitable for Hb measurement.
D
onors’ eligibility to donate blood and blood components needs to be determined by a physical determination, a health history assessment, and hemoglobin (Hb) analysis before blood donation.1-4 These requirements serve for donors’ security to prevent phlebotomy of anemic individuals and also for quality assurance of blood products, but the Hb level to determine donor acceptability is still controversial. Current European cutoff levels of donor Hb are 12.5 g/dL for women and 13.5 g/dL for men while it is 12.5 g/dL for both male and female donors in the United States.1,2,5,6 The method of choice for predonation screening for several decades has been performed with a portable photometer reading absorbance of hemolyzed blood in microcuvettes obtained by fingerstick because this method is fast, cost-effective, and easy to perform,7-9 but the reliability of capillary Hb estimation is controversial because capillary and venous Hb levels are different.7,10-14 While the fingerstick blood drop reflects the content of blood from various loop capillaries and small arterioles and venules in the finger, the venous sample reflects the blood coursing through the veins, heart, and arteries.14 Belardinelli and colleagues7 demonstrated that fingerstick samples often gave falsely low evaluations due to sample
ABBREVIATIONS: LOA = limits of agreement; SDD = standard deviation of the difference; SLS = sodium lauryl sulfate. From 1Transfusion Medicine, University Hospital of Cologne; and the 2Institute of Medical Statistics, Informatics, and Epidemiology, University of Cologne, Cologne, Germany. Address reprint requests to: Melanie Störmer, Transfusion Medicine, University Hospital of Cologne, Kerpener Strasse 62, 50937 Cologne, Germany; e-mail: [email protected]. *These authors contributed equally Received for publication February 26, 2014; revision received June 30, 2014, and accepted July 8, 2014. doi: 10.1111/trf.12819 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
ARDIN ET AL.
contamination with tissue fluid or incomplete filling of the cuvette, which leads to deferral of the prospective blood donor. Moreover, the fingerstick is reported by blood donors as one of the worst parts of blood donation.8,9 For example, Bednall and Bove15 reported that in one study 17% of participants reported that they are averse to blood donation because they disliked the fingerstick test.15,16 Therefore, noninvasive Hb screening methods may represent an attractive alternative by eliminating pain, reducing stress and infection risk.7 Recently, new noninvasive methods have been developed, using the principle of reflection spectroscopy,7,17,18 which exempts the donor fingerstick, makes the procedure more comfortable, and could contribute to greater donor retention.19 Therefore, the aim of this study was to evaluate the use of three noninvasive methods for Hb screening in healthy blood donors in comparison to an invasive method and venous blood sample as reference and to assess whether they have the potential to enhance the retention of blood donors.
was obtained from prospective donors for donation of samples for validation use. As shown in Fig. 1 blood donors were tested in three consecutive trials by the Haemospect (March 21 and 22, 2013, MBR Optical Systems GmbH & Co. KG, Wuppertal, Germany), the NBM 200 (February 14 and 15, 2013, LMB Technology GmbH, Schwaig, Germany), or the Pronto-7 (February 19 and 20, 2013, Masimo Europe Ltd, Puchheim, Germany) in comparison to the conventionally used invasive method using the CompoLab Hb (Fresenius Kabi GmbH, Bad Homburg, Germany) to measure Hb levels on capillary samples obtained by a fingerstick lancing technique. For all trials the Hb levels were compared to the values obtained from the Sysmex KX21n (Sysmex Europe GmbH, Norderstedt, Germany) on venous samples that served as reference. Finally, statistical analysis was performed and the usability was assessed by the operators with a specifically designed questionnaire.
Noninvasive methods
MATERIALS AND METHODS Study design The evaluation was carried out in three independent study periods due to the terms of the manufacturers at the Blood Donation Centre of the Institute for Transfusion Medicine of the University Hospital of Cologne. Informed consent
The Haemospect is a hand-held device running on battery power based on transcutaneous reflection spectroscopy. It uses a button sensor that adheres to the palm side of the finger of the nondominant hand.18 A sensor head placed on the skin projects a white light into the underlying tissue (0.5-0.9 mm) via a waveguide. Some of the projected light is absorbed by the various components of tissue, while
Fig. 1. Study design: comparison of invasive and noninvasive Hb methods. Blood donors were screened for Hb levels in three independent periods using three different noninvasive methods (Haemospect, NBM 200, and Pronto-7) in comparison to the established fingerstick method (CompoLab Hb) and to Hb levels obtained from venous samples on a cell counter (Sysmex) as reference. Finally, the usability of the noninvasive methods was assessed by a specially developed questionnaire. 2
TRANSFUSION Volume **, ** **
NONINVASIVE Hb SCREENING OF BLOOD DONORS
some of it is reflected back to the device. Finally, the spectrometer breaks the light down into its separate wavelengths and an electronic evaluation unit connected to the system analyzes it. The quantitative value is visualized on the display of the device.20,21 The NBM 200 device, an occlusion spectrophotometer-based Hb monitor, operates via a ring-shaped sensor that is fitted on the donors’ finger and applies pressure so that the blood flow is temporarily occluded. Optical elements perform a measurement of the light transmitted through the finger, which is projected by glass-fiber wires in the sensor and the values are analyzed and displayed.7 The Pronto-7 is a hand-held noninvasive pulse CO oximeter that measures Hb spectroscopically. Multiple wavelengths (500-1300 nm) of light are transmitted through a reusable finger sensor placed on the ring finger of the subjects’ nondominant hand. Based on the light attenuation characteristics, the device measures and calculates Hb, as well as spectroscopic O2, pulse rate, and perfusion index.7,17,22
Invasive methods for capillary and venous Hb level determination
cleaning, visualization of results, time to result, and donors’ feedback (Fig. 2). School marks were given where 1 represents the best and 6 the worst mark. The questionnaires were completed immediately after each completed method examination at the end of the second day.
Statistical analysis Statistical analysis of the three separate but consecutive studies was performed by computer software (Microsoft Office Excel 2003, Microsoft Corp., Redmond, WA). Descriptive statistics (mean and standard deviation [SD]) as well as Bland-Altman plots27 were used for analysis. The Bland-Altman’s approach measures the strength of relation by a plot of the difference between methods against their mean value.7
RESULTS Statistical analysis The study period using the Haemospect included 121 potential blood donors, 60 men and 61 women (mean age, 33 ± 12 years). The results are summarized in Tables 1 and 2 (descriptive statistics), Table 3 (donor rejection rates),
The CompoLab Hb was used for invasive photometric measuring of capillary Hb levels. For this purpose the donors’ finger is cleaned with alcohol and picked with a lancet. A drop of blood (approx. 10 μL) is drawn into a microcuvette by capillary action, placed into the device, and photometrically analyzed within 10 to 60 seconds.23 For venous Hb level determination, the Sysmex KX21n was used. The Sysmex is an automated blood cell counter intended for in vitro diagnostic use in clinical laboratories.24 The measurement of Hb concentration is based on the sodium lauryl sulfate (SLS)-Hb method. SLS dissolves lipoproteins of the cell membrane of the red blood cells to release Hb and convert it into SLS-Hb. The concentration of SLS-Hb is measured as light absorbance at 555 nm.25,26 The test was performed as stated in the manufacturer’s manual.25
Usability test Operators evaluated the usability with a specially designed questionnaire with respect to visual appearance, handling,
Fig. 2. Questionnaire for application assessment of noninvasive Hb methods. The users were asked to assess the noninvasive methods using a questionnaire and grading the methods according to school marks 1 (very good) to 6 (unsatisfactory). Volume **, ** **
TRANSFUSION
3
ARDIN ET AL.
TABLE 1. Descriptive statistics of noninvasive and invasive methods for Hb measurements* Donor characteristics Haemospect Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL) NBM 200 Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL) Pronto-7 Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL)
Noninvasive method
Capillary, invasive (CompoLab Hb)
Venous (Sysmex)
121 (60 men, 61 women) 33 ± 12 13.92 ± 1.13
13.65 ± 1.29
14.14 ± 1.19
27 (19-58) 13.24 ± 0.86
12.92 ± 1.14
13.45 ± 0.93
35 (20-69) 14.62 ± 0.93
14.40 ± 0.97
14.84 ± 1.01
120 (78 men, 42 women) 36 ± 14 14.88 ± 1.50
14.35 ± 1.08
15.00 ± 1.22
27 (19-66) 13.50 ± 1.31
13.39 ± 0.83
13.87 ± 0.84
31 (19-68) 15.61 ± 0.99
14.86 ± 0.81
15.60 ± 0.93
110 (68 men, 42 women) 34 ± 14 13.99 ± 1.34
14.04 ± 1.17
14.50 ± 1.10
27 (20-66) 12.94 ± 1.10
13.22 ± 1.10
13.76 ± 1.02
30 (18-68) 14.64 ± 1.04
14.55 ± 0.91
14.95 ± 0.88
* Data are reported as mean ± SD or median (range).
the other hand 10 donors (8.26%) would have been accepted if Haemospect had been the reference method. The meaParameter Haemospect NBM 200 Pronto-7 CompoLab Hb surement was performed in 25 ± 2 Total seconds. On the basis of a first prestudy Donors 121 120 110 351 trial, the software of the Haemospect Bias ± SDD (g/dL) −0.22 ± 1.24 −0.12 ± 1.14 −0.50 ± 0.99 −0.53 ± 0.81 Men needed to be optimized and this period Donors 60 78 68 206 was repeated. Bias ± SDD (g/dL) −0.22 ± 1.26 0.01 ± 1.08 −0.31 ± 0.92 −0.54 ± 0.76 The study period using the NBM 200 Women Donors 61 42 42 145 included 120 potential blood donors, 78 Bias ± SDD (g/dL) −0.22 ± 1.22 −0.36 ± 1.22 −0.82 ± 1.02 −0.52 ± 0.89 men and 42 women (mean age, 36 ± 14 years). The Bland-Altman plot showed a bias of −0.12 g/dL, a 1.14 SDD, and 95% LOA from −2.35 to 2.11 (Fig. 3, Table 2). With the venous Hb and Fig. 3 (Bland-Altman plots). The Bland-Altman plot level as reference standard, the accordance of acceptable for the Haemospect and venous measurement showed a and unacceptable donors between the Sysmex and the bias (difference of mean values) of −0.22 g/dL, a 1.24 SD of NBM 200 was 91.7% (Table 3). Three donors (2.5%) would the differences (SDD), and 95% limits of agreement (LOA) be classified as unacceptable due to the too low Hb level from −2.64 to 2.21 (Fig. 3, Table 2). Using the venous from the venous sample but seven donors (5.8%) by using (Sysmex) Hb level as reference standard, the percentage of the NBM 200. No correspondent unacceptable classificadonors correctly screened (accordance of acceptable and tions occurred. Therefore, the seven donors would be unacceptable donors between Sysmex and test method) falsely rejected. Predominantly lower Hb levels were by Haemospect was 88.4% as shown in Table 3. Fourteen obtained by the NBM 200 especially for women (Table 3). donors (11.6%) would be classified as unacceptable due to Taking the measurements lasted 82 ± 6 seconds. low Hb levels from the venous sample but only eight The study period using the Pronto-7 included 110 donors (6.6%) by using the Haemospect (Table 3). Four potential blood donors, 68 men and 42 women (mean age, correspondent unacceptable classifications occurred. On TABLE 2. Mean of differences between test method and venous reference results (bias ± SDD)
4
TRANSFUSION Volume **, ** **
NONINVASIVE Hb SCREENING OF BLOOD DONORS
TABLE 3. Predonation screening using invasive and noninvasive Hb methods according to the venous reference value* Donors Total Screened donors Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§ Accordance (%)|| Women Screened Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§ Men Screened Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§
Haemospect
NBM 200
Pronto-7
CompoLab Hb
121 14 (11.57) 8 (6.61) 4 (3.31) 4 (3.31) 10 (8.26) 88
120 3 (2.50) 7 (5.83) 0 7 (5.83) 3 (2.50) 92
110 5 (4.55) 18 (16.36) 4 (3.64) 14 (12.73) 1 (0.91) 86
351 22 (6.27) 28 (7.98) 11 (3.13) 17 (4.48) 11 (3.13) 92
61 9 (14.74) 6 (9.84) 3 (4.92) 3 (4.92) 6 (9.84)
42 1 (2.38) 6 (14.29) 0 6 (14.29) 1 (2.38)
42 4 (9.52) 9 (21.43) 3 (7.14) 6 (14.29) 1 (2.38)
145 14 (9.66) 21 (14.48) 8 (5.52) 13 (8.97) 6 (4.14)
60 5 (8.33) 2 (3.33) 1 (1.67) 1 (1.67) 4 (6.67)
78 2 (2.56) 1 (1.28) 0 1 (1.28) 2 (2.56)
68 1 (1.47) 9 (13.24) 1 (1.47) 8 (11.76) 0
206 8 (3.88) 7 (3.40) 3 (1.46) 4 (1.94) 5 (2.43)
* Data are reported as number (%). † Donor was rejected due to a low Hb level of the test method that could be confirmed by a low venous Hb level (
BACKGROUND: To prevent phlebotomy of anemic individuals and to ensure hemoglobin (Hb) content of the blood units, Hb screening of blood donors before donation is essential. Hb values are mostly evaluated by measurement of capillary blood obtained from fingerstick. Rapid noninvasive methods have recently become available and may be preferred by donors and staff. The aim of this study was to evaluate for the first time all different noninvasive methods for Hb screening. STUDY DESIGN AND METHODS: Blood donors were screened for Hb levels in three different trials using three different noninvasive methods (Haemospect [MBR Optical Systems GmbH & Co. KG], NBM 200 [LMB Technology GmbH], Pronto-7 [Masimo Europe Ltd]) in comparison to the established fingerstick method (CompoLab Hb [Fresenius Kabi GmbH]) and to levels obtained from venous samples on a cell counter (Sysmex [Sysmex Europe GmbH]) as reference. The usability of the noninvasive methods was assessed with an especially developed survey. RESULTS: Technical failures occurred by using the Pronto-7 due to nail polish, skin color, or ambient light. The NBM 200 also showed a high sensitivity to ambient light and noticeably lower Hb levels for women than obtained from the Sysmex. The statistical analysis showed the following bias and standard deviation of differences of all methods in comparison to the venous results: Haemospect, −0.22 ± 1.24; NBM, 200 −0.12 ± 1.14; Pronto-7, −0.50 ± 0.99; and CompoLab Hb, −0.53 ± 0.81. CONCLUSION: Noninvasive Hb tests represent an attractive alternative by eliminating pain and reducing risks of blood contamination. The main problem for generating reliable results seems to be preanalytical variability in sampling. Despite the sensitivity to environmental stress, all methods are suitable for Hb measurement.
D
onors’ eligibility to donate blood and blood components needs to be determined by a physical determination, a health history assessment, and hemoglobin (Hb) analysis before blood donation.1-4 These requirements serve for donors’ security to prevent phlebotomy of anemic individuals and also for quality assurance of blood products, but the Hb level to determine donor acceptability is still controversial. Current European cutoff levels of donor Hb are 12.5 g/dL for women and 13.5 g/dL for men while it is 12.5 g/dL for both male and female donors in the United States.1,2,5,6 The method of choice for predonation screening for several decades has been performed with a portable photometer reading absorbance of hemolyzed blood in microcuvettes obtained by fingerstick because this method is fast, cost-effective, and easy to perform,7-9 but the reliability of capillary Hb estimation is controversial because capillary and venous Hb levels are different.7,10-14 While the fingerstick blood drop reflects the content of blood from various loop capillaries and small arterioles and venules in the finger, the venous sample reflects the blood coursing through the veins, heart, and arteries.14 Belardinelli and colleagues7 demonstrated that fingerstick samples often gave falsely low evaluations due to sample
ABBREVIATIONS: LOA = limits of agreement; SDD = standard deviation of the difference; SLS = sodium lauryl sulfate. From 1Transfusion Medicine, University Hospital of Cologne; and the 2Institute of Medical Statistics, Informatics, and Epidemiology, University of Cologne, Cologne, Germany. Address reprint requests to: Melanie Störmer, Transfusion Medicine, University Hospital of Cologne, Kerpener Strasse 62, 50937 Cologne, Germany; e-mail: [email protected]. *These authors contributed equally Received for publication February 26, 2014; revision received June 30, 2014, and accepted July 8, 2014. doi: 10.1111/trf.12819 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
ARDIN ET AL.
contamination with tissue fluid or incomplete filling of the cuvette, which leads to deferral of the prospective blood donor. Moreover, the fingerstick is reported by blood donors as one of the worst parts of blood donation.8,9 For example, Bednall and Bove15 reported that in one study 17% of participants reported that they are averse to blood donation because they disliked the fingerstick test.15,16 Therefore, noninvasive Hb screening methods may represent an attractive alternative by eliminating pain, reducing stress and infection risk.7 Recently, new noninvasive methods have been developed, using the principle of reflection spectroscopy,7,17,18 which exempts the donor fingerstick, makes the procedure more comfortable, and could contribute to greater donor retention.19 Therefore, the aim of this study was to evaluate the use of three noninvasive methods for Hb screening in healthy blood donors in comparison to an invasive method and venous blood sample as reference and to assess whether they have the potential to enhance the retention of blood donors.
was obtained from prospective donors for donation of samples for validation use. As shown in Fig. 1 blood donors were tested in three consecutive trials by the Haemospect (March 21 and 22, 2013, MBR Optical Systems GmbH & Co. KG, Wuppertal, Germany), the NBM 200 (February 14 and 15, 2013, LMB Technology GmbH, Schwaig, Germany), or the Pronto-7 (February 19 and 20, 2013, Masimo Europe Ltd, Puchheim, Germany) in comparison to the conventionally used invasive method using the CompoLab Hb (Fresenius Kabi GmbH, Bad Homburg, Germany) to measure Hb levels on capillary samples obtained by a fingerstick lancing technique. For all trials the Hb levels were compared to the values obtained from the Sysmex KX21n (Sysmex Europe GmbH, Norderstedt, Germany) on venous samples that served as reference. Finally, statistical analysis was performed and the usability was assessed by the operators with a specifically designed questionnaire.
Noninvasive methods
MATERIALS AND METHODS Study design The evaluation was carried out in three independent study periods due to the terms of the manufacturers at the Blood Donation Centre of the Institute for Transfusion Medicine of the University Hospital of Cologne. Informed consent
The Haemospect is a hand-held device running on battery power based on transcutaneous reflection spectroscopy. It uses a button sensor that adheres to the palm side of the finger of the nondominant hand.18 A sensor head placed on the skin projects a white light into the underlying tissue (0.5-0.9 mm) via a waveguide. Some of the projected light is absorbed by the various components of tissue, while
Fig. 1. Study design: comparison of invasive and noninvasive Hb methods. Blood donors were screened for Hb levels in three independent periods using three different noninvasive methods (Haemospect, NBM 200, and Pronto-7) in comparison to the established fingerstick method (CompoLab Hb) and to Hb levels obtained from venous samples on a cell counter (Sysmex) as reference. Finally, the usability of the noninvasive methods was assessed by a specially developed questionnaire. 2
TRANSFUSION Volume **, ** **
NONINVASIVE Hb SCREENING OF BLOOD DONORS
some of it is reflected back to the device. Finally, the spectrometer breaks the light down into its separate wavelengths and an electronic evaluation unit connected to the system analyzes it. The quantitative value is visualized on the display of the device.20,21 The NBM 200 device, an occlusion spectrophotometer-based Hb monitor, operates via a ring-shaped sensor that is fitted on the donors’ finger and applies pressure so that the blood flow is temporarily occluded. Optical elements perform a measurement of the light transmitted through the finger, which is projected by glass-fiber wires in the sensor and the values are analyzed and displayed.7 The Pronto-7 is a hand-held noninvasive pulse CO oximeter that measures Hb spectroscopically. Multiple wavelengths (500-1300 nm) of light are transmitted through a reusable finger sensor placed on the ring finger of the subjects’ nondominant hand. Based on the light attenuation characteristics, the device measures and calculates Hb, as well as spectroscopic O2, pulse rate, and perfusion index.7,17,22
Invasive methods for capillary and venous Hb level determination
cleaning, visualization of results, time to result, and donors’ feedback (Fig. 2). School marks were given where 1 represents the best and 6 the worst mark. The questionnaires were completed immediately after each completed method examination at the end of the second day.
Statistical analysis Statistical analysis of the three separate but consecutive studies was performed by computer software (Microsoft Office Excel 2003, Microsoft Corp., Redmond, WA). Descriptive statistics (mean and standard deviation [SD]) as well as Bland-Altman plots27 were used for analysis. The Bland-Altman’s approach measures the strength of relation by a plot of the difference between methods against their mean value.7
RESULTS Statistical analysis The study period using the Haemospect included 121 potential blood donors, 60 men and 61 women (mean age, 33 ± 12 years). The results are summarized in Tables 1 and 2 (descriptive statistics), Table 3 (donor rejection rates),
The CompoLab Hb was used for invasive photometric measuring of capillary Hb levels. For this purpose the donors’ finger is cleaned with alcohol and picked with a lancet. A drop of blood (approx. 10 μL) is drawn into a microcuvette by capillary action, placed into the device, and photometrically analyzed within 10 to 60 seconds.23 For venous Hb level determination, the Sysmex KX21n was used. The Sysmex is an automated blood cell counter intended for in vitro diagnostic use in clinical laboratories.24 The measurement of Hb concentration is based on the sodium lauryl sulfate (SLS)-Hb method. SLS dissolves lipoproteins of the cell membrane of the red blood cells to release Hb and convert it into SLS-Hb. The concentration of SLS-Hb is measured as light absorbance at 555 nm.25,26 The test was performed as stated in the manufacturer’s manual.25
Usability test Operators evaluated the usability with a specially designed questionnaire with respect to visual appearance, handling,
Fig. 2. Questionnaire for application assessment of noninvasive Hb methods. The users were asked to assess the noninvasive methods using a questionnaire and grading the methods according to school marks 1 (very good) to 6 (unsatisfactory). Volume **, ** **
TRANSFUSION
3
ARDIN ET AL.
TABLE 1. Descriptive statistics of noninvasive and invasive methods for Hb measurements* Donor characteristics Haemospect Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL) NBM 200 Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL) Pronto-7 Donors Age (years) Total Hb (g/dL) Women Age (years) Hb (g/dL) Men Age (years) Hb (g/dL)
Noninvasive method
Capillary, invasive (CompoLab Hb)
Venous (Sysmex)
121 (60 men, 61 women) 33 ± 12 13.92 ± 1.13
13.65 ± 1.29
14.14 ± 1.19
27 (19-58) 13.24 ± 0.86
12.92 ± 1.14
13.45 ± 0.93
35 (20-69) 14.62 ± 0.93
14.40 ± 0.97
14.84 ± 1.01
120 (78 men, 42 women) 36 ± 14 14.88 ± 1.50
14.35 ± 1.08
15.00 ± 1.22
27 (19-66) 13.50 ± 1.31
13.39 ± 0.83
13.87 ± 0.84
31 (19-68) 15.61 ± 0.99
14.86 ± 0.81
15.60 ± 0.93
110 (68 men, 42 women) 34 ± 14 13.99 ± 1.34
14.04 ± 1.17
14.50 ± 1.10
27 (20-66) 12.94 ± 1.10
13.22 ± 1.10
13.76 ± 1.02
30 (18-68) 14.64 ± 1.04
14.55 ± 0.91
14.95 ± 0.88
* Data are reported as mean ± SD or median (range).
the other hand 10 donors (8.26%) would have been accepted if Haemospect had been the reference method. The meaParameter Haemospect NBM 200 Pronto-7 CompoLab Hb surement was performed in 25 ± 2 Total seconds. On the basis of a first prestudy Donors 121 120 110 351 trial, the software of the Haemospect Bias ± SDD (g/dL) −0.22 ± 1.24 −0.12 ± 1.14 −0.50 ± 0.99 −0.53 ± 0.81 Men needed to be optimized and this period Donors 60 78 68 206 was repeated. Bias ± SDD (g/dL) −0.22 ± 1.26 0.01 ± 1.08 −0.31 ± 0.92 −0.54 ± 0.76 The study period using the NBM 200 Women Donors 61 42 42 145 included 120 potential blood donors, 78 Bias ± SDD (g/dL) −0.22 ± 1.22 −0.36 ± 1.22 −0.82 ± 1.02 −0.52 ± 0.89 men and 42 women (mean age, 36 ± 14 years). The Bland-Altman plot showed a bias of −0.12 g/dL, a 1.14 SDD, and 95% LOA from −2.35 to 2.11 (Fig. 3, Table 2). With the venous Hb and Fig. 3 (Bland-Altman plots). The Bland-Altman plot level as reference standard, the accordance of acceptable for the Haemospect and venous measurement showed a and unacceptable donors between the Sysmex and the bias (difference of mean values) of −0.22 g/dL, a 1.24 SD of NBM 200 was 91.7% (Table 3). Three donors (2.5%) would the differences (SDD), and 95% limits of agreement (LOA) be classified as unacceptable due to the too low Hb level from −2.64 to 2.21 (Fig. 3, Table 2). Using the venous from the venous sample but seven donors (5.8%) by using (Sysmex) Hb level as reference standard, the percentage of the NBM 200. No correspondent unacceptable classificadonors correctly screened (accordance of acceptable and tions occurred. Therefore, the seven donors would be unacceptable donors between Sysmex and test method) falsely rejected. Predominantly lower Hb levels were by Haemospect was 88.4% as shown in Table 3. Fourteen obtained by the NBM 200 especially for women (Table 3). donors (11.6%) would be classified as unacceptable due to Taking the measurements lasted 82 ± 6 seconds. low Hb levels from the venous sample but only eight The study period using the Pronto-7 included 110 donors (6.6%) by using the Haemospect (Table 3). Four potential blood donors, 68 men and 42 women (mean age, correspondent unacceptable classifications occurred. On TABLE 2. Mean of differences between test method and venous reference results (bias ± SDD)
4
TRANSFUSION Volume **, ** **
NONINVASIVE Hb SCREENING OF BLOOD DONORS
TABLE 3. Predonation screening using invasive and noninvasive Hb methods according to the venous reference value* Donors Total Screened donors Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§ Accordance (%)|| Women Screened Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§ Men Screened Rejected by Sysmex Rejected Rejected correctly† Rejected falsely too low‡ Accepted falsely too high§
Haemospect
NBM 200
Pronto-7
CompoLab Hb
121 14 (11.57) 8 (6.61) 4 (3.31) 4 (3.31) 10 (8.26) 88
120 3 (2.50) 7 (5.83) 0 7 (5.83) 3 (2.50) 92
110 5 (4.55) 18 (16.36) 4 (3.64) 14 (12.73) 1 (0.91) 86
351 22 (6.27) 28 (7.98) 11 (3.13) 17 (4.48) 11 (3.13) 92
61 9 (14.74) 6 (9.84) 3 (4.92) 3 (4.92) 6 (9.84)
42 1 (2.38) 6 (14.29) 0 6 (14.29) 1 (2.38)
42 4 (9.52) 9 (21.43) 3 (7.14) 6 (14.29) 1 (2.38)
145 14 (9.66) 21 (14.48) 8 (5.52) 13 (8.97) 6 (4.14)
60 5 (8.33) 2 (3.33) 1 (1.67) 1 (1.67) 4 (6.67)
78 2 (2.56) 1 (1.28) 0 1 (1.28) 2 (2.56)
68 1 (1.47) 9 (13.24) 1 (1.47) 8 (11.76) 0
206 8 (3.88) 7 (3.40) 3 (1.46) 4 (1.94) 5 (2.43)
* Data are reported as number (%). † Donor was rejected due to a low Hb level of the test method that could be confirmed by a low venous Hb level (
