Sage Journals: Discover world-class research

Abstract

In veterinary medicine, point-of-care testing (POCT) techniques have become popular, since they provide immediate results and only small amounts of blood are needed. However, their accuracy is controversial. Pigs are often used for research purposes and accurate measurement of haemoglobin (Hb) is important during invasive procedures. The aim of this study was to evaluate two different Hb POCT devices in neonatal pigs. A prospective study with 57 pigs of 3–6 weeks of age, weighing 4.1–6.2 kg (median 5.1 kg) was performed. Fifty-seven blood samples were analysed for Hb using a conductivity-based and a photometrical POCT device and compared with a photometrical reference method. Statistical analysis was performed with Bland–Altman analysis, Spearman correlation and Passing–Bablok regression analysis. Hb values ranged from 32 to 108 g/L (median 80 g/L) using the reference method. The bias of the photometrical method (HemoCue^®) to the reference method was −1 g/L, with limits of agreement (LOA) of −7 to 6 g/L. The conductivity-based method (i-STAT^®) had a bias of −15 g/L with LOA from −24 to −6 g/L. There was a significant association between protein values and the bias of i-STAT versus CellDyn (r ² = 0.27, P < 0.05) but not with the bias of HemoCue versus CellDyn (r ² = 0.001, P = 0.79). The lower the protein values were, the lower the Hb values were measured by the i-STAT. The conductivity-based measurement of Hb constantly underestimated Hb values, whereas the photometrical method demonstrated a better accuracy and is therefore more reliable for on-site measurement of Hb in pigs.

Keywords

Quality assurance/control pigs diagnostic assay haematology

Point-of-care testing (POCT) systems have been introduced into clinical practice to provide analytical results more rapidly and independently from hospital laboratory services, to allow for shorter therapeutic response intervals and to ease therapy control in the operating theatre, intensive care units or emergency departments.^1,2 An important feature of modern laboratory bedside analysers is the small amount of blood requested for determination of several laboratory parameters. Modern POCT systems have the potential to markedly reduce iatrogenic diagnostic blood loss and to reduce the need for blood transfusions.² Nevertheless, adequate evaluation of the device in the expected population regarding species, age and concurrent medical disease should be performed before its clinical use.³

The HemoCue (HemoCue B-Hemoglobin, HemoCue AB, Wetzikon, Switzerland) and the i-STAT (i-STAT-1, Abbott AG, Baar, Switzerland) devices represent two modern techniques for bedside measurement of blood haemoglobin (Hb) content. Both systems have been evaluated in humans^4–10 and in various animal species^11–21 and were judged to be clinically useful. Most authors judged the analytical performance to be sufficient in both systems even if the i-STAT measurements are often measuring too low, especially if hypoproteinaemia is present.

Pigs are often used for research purposes and evaluation of Hb concentration is important especially for the calculation of haemodynamic parameters as well as during invasive procedures where blood loss and need for transfusion have to be quantified.²² Additionally, iron deficiency and concurrent anaemia are common in neonatal pigs.^23,24 Therefore, monitoring of Hb is important for evaluation of efficiency of nutritional or parenteral iron supplementation.^25,26 Although the HemoCue²⁷ and the i-STAT analyser²⁸ have been used as decision-guiding tools for Hb in pig research, they have not yet been validated in this species. The aim of the current study was to evaluate the accuracy of the HemoCue and the i-STAT analyser for measurement of Hb in young pigs in comparison with a reference method for Hb measurement.

Materials and methods

Animals

After obtaining approval from the Cantonal Veterinary Office (No. 150/2009), 57 healthy female and male pigs from 3 to 6 weeks of age weighing 4.1–6.2 kg (median 5.1 kg) and undergoing different kinds of studies in our laboratory requiring anaesthesia were included in this investigation. Health status was assessed by history and visual examination. All pigs were housed in a conventional farm in farrowing crates of 7.5 m² on straw with 75% of solid floor and soil for rotting. They had free access to food and water until one hour before induction of anaesthesia.

The animals were maintained in surgical anaesthesia throughout the whole study and died because of an overdose of local anaesthetics during a concurrent study investigating toxic effects thereof. If animals recovered from the overdose they were euthanized using an overdose of pentobarbital (Eskonarkon^®, Streuli Pharma AG, Uznach, Switzerland) intravenously.

HemoCue

The HemoCue haemoglobinometer measures the Hb concentration from a 10 µL blood sample that is drawn by capillary action into a disposable microcuvette.¹⁰ The erythrocytes are lysed and the free Hb is conjugated to azidemethaemoglobin²⁹ that is photometrically measured at 570 nm. The additional measurement at 880 nm corrects for turbidity caused by marked lipaemia or leukocytosis in the sample.

i-STAT

The i-STAT POCT consists of a portable analyser and different self-contained disposable cartridges that are available for the measurement of electrolytes, blood chemistry and blood gases.⁴ The CG8+ cartridge includes measurement of Hb, as well as blood gases and electrolyte analysis. The required amount of blood is 65 µL. The cartridges have to be stored at 2–8°C and have to be passively warmed to room temperature before use. The i-STAT system measures the haematocrit (Hct) with conductimetry, calculates the Hb with the fixed equation of Hct/3 = Hb and displays this value. The conductivity of blood is dependent on electrolyte and protein concentrations.⁶ The i-STAT device corrects for changes in sodium concentrations but not for abnormal protein values, as it is not able to measure protein concentration within the current system.

Reference instrument

The CellDyn 3500 (Abbott AG, Baar, Switzerland) was used as the reference instrument. It is a well-established haematology instrument, which is used worldwide in veterinary laboratories. Haemoglobin concentration is measured using a cyanide-free spectrophotometry method. Haematocrit values are computed from direct measurements of red blood cell counts and mean cell volume. The CellDyn was calibrated in the Clinical Laboratory so the haematocrit results agreed with manual PCV results for pigs. Three levels of commercially available control blood were run daily prior to analysis of patient samples. The CellDyn showed very low inter-assay imprecision. Coefficient of variation was 1.4% in level 1 (mean Hb 76 g/L), 1% in level 2 (mean Hb 123 g/L) and 3.3% in level 3 (mean Hb 158 g/L).

Experimental set-up

Anaesthesia was induced via the inhalational route using sevoflurane by face mask. During the induction period, each animal was held tightly in the arms of one examiner without prior sedation, while another held an appropriate-sized face mask to the pig's muzzle. Once the pig was unconscious, a venous access was established and the trachea intubated. The lungs were artificially ventilated using pressure-controlled ventilation. Anaesthesia was maintained with sevoflurane in oxygen and air 1:1. No neuromuscular blocking agents were used. Monitoring consisted of pulseoximetry, three-lead electrocardiograph (ECG), invasive arterial blood pressure measurement after cannulation of the carotid artery by a cut-down procedure, end-tidal gas analysis (sevoflurane, O₂, CO₂) and rectal temperature control.

In each pig, 3 mL of blood were withdrawn from the arterial catheter after discarding the first 5 mL. The HemoCue microcuvette and the i-STAT cartridge were filled directly with fresh whole blood and measured immediately. Care was taken that no air bubbles were present in the systems. The measurement takes less than 60 s with the HemoCue and 120 s with the i-STAT. The remaining blood was submitted to the clinical pathology laboratory in K₃-ethylenediaminetetraacetic acid (EDTA) and in lithium heparinized tubes (Sarstedt AG, Sevelen, Switzerland) for photometric determination of Hb concentration and Hct measurement with the Coulter method, both performed in an automatic counter (CellDyn 3500). The total plasma protein concentration was determined by the biuret method using an automated analyser (Cobas Integra 800, Roche Diagnostics, Rotkreuz, Switzerland). The protein calibrator was prepared from human serum (c.f.a.s. calibrator, Roche). For internal quality control, two levels of a commercially available control serum were run daily prior to analysis (Precinorm U, Roche). Quarterly, external quality control was performed by analysing serum from the Quality Control Centre Switzerland. All devices underwent standard services biannually.

Part of the studies included significant blood loss including repeated blood sampling resulting in haemodilution due to volume therapy to keep the animals normovolaemic. This led to a broad range of Hb and protein values as it occurs under clinical condition in sick animals.

To perform intra-assay precision analysis two samples of porcine blood were each analysed 10 times with both POCT devices under the same condition as described above.

Statistical analysis

Hb values measured by the HemoCue and displayed by the i-STAT were compared with standard laboratory Hb measurement with Bland–Altman analysis, Spearman correlation and Passing–Bablok regression analysis. Hct values measured with the i-STAT were additionally compared with the Hct measurements of the automatic counter using the same methods.

The bias of the POCT methods to the reference method was compared with the protein content with linear regression. This was performed by comparing the difference of each pair of the relative POCT method and the reference method to the corresponding protein value.

To quantify the clinical importance of the bias, the sensitivity and specificity of the methods for the detection of anaemia (defined as Hb < 80 g/L with CellDyn 3500) were assessed. Continuous parametric variables are presented as mean (±standard deviation, SD) and non-parametric variables as median (range), respectively. Statistical significance was declared at P ≤ 0.05. Data management, statistical analysis and graphs were performed with Prism 4.0b (GraphPad Software, La Jolla, CA, USA) and Excel for Macintosh (Ofﬁce X, Microsoft, Redmond, WA, USA).

Results

In total, 57 blood samples were analysed. Results of the analysis of Hb, Hct and plasma protein are presented in Table 1.

Table 1
Measured haemoglobin (Hb), haematocrit (Hct) and protein values from 57 samples of porcine blood

Analysis Median Range

CellDyn Hb (n = 57) 80 g/L 32–108 g/L

HemoCue Hb (n = 57) 81 g/L 27–112 g/L

i-STAT Hb (n = 56) 68 g/L 34–92 g/L

CellDyn Hct (n = 56) 24.6 L/L 9.6–37.4 L/L

i-STAT Hct (n = 56) 20.4 L/L 10.2–27.6 L/L

Protein (n = 57) 36 g/L 25–55 g/L

Intra-assay precision of the two POCT was high and is presented in Table 2.

Table 2
Results of intra-assay precision analysis of haemoglobin (Hb) in two samples of porcine blood

Sample 1 Sample 2

Method Mean (g/L) Standard deviation (g/L) Coefficient of variation (%) Mean (g/L) Standard deviation (g/L) Coefficient of variation (%)

HemoCue 49 1 2.2 82 1 1.1

i-STAT 37 1 3.4 67 1 1.9

The HemoCue measured similar values to the reference method (Table 3 and Figures 1a and 2a). The i-STAT constantly measured lower Hb values (Table 3 and Figures 1b and 2b) and was not able to quantify Hb values in one sample (32 g/L CellDyn).

Table 3
Results of the Bland–Altman analysis, Spearman correlation and Passing–Bablok analysis among three different measurement methods of haemoglobin (Hb) and haematocrit (Hct)

Group Bias LOA R Intercept with 95% CI Slope with 95% CI

HemoCue/CellDyn Hb (n = 57) −1 g/L −7 to 6 g/L 0.98 −0.92 (−1.18 to −0.59) 1.11 (1.07 to 1.15)

i-STAT/CellDyn Hb (n = 56) −15 g/L −24 to −6 g/L 0.95 −0.47 (−1.20 to 0.03) 0.88 (0.81 to 0.96)

i-STAT/CellDyn Hct (n = 56) −4.8 L/L −8.4 to −1.2 L/L 0.96 −1.32 (−3.48 to 0.82) 0.86 (0.77 to 0.94)

LOA: limits of agreement; CI: confidence interval

Figure 1
Bland–Altman bias plots for comparison of the two haemoglobin (Hb) analysing techniques with the standard Hb measurement technique (CellDyn). (a) HemoCue. (b) i-STAT. The solid line indicates bias and the dashed lines indicate ±2 standard deviations around the bias (HemoCue: n = 57; i-STAT: n = 56); overlying dots are plotted only once

Figure 2
Passing–Bablok regression plot for comparison of the two haemoglobin (Hb) analysing techniques with the standard Hb measurement technique (CellDyn). (a) HemoCue (n = 57). (b) i-STAT (n = 56). The thin black line is the line of identity (y = x) and the thick black line is the line of best fit

There was a significant association between protein values and the bias of i-STAT versus CellDyn (r ² = 0.27, P < 0.05) (Figure 3). The lower the protein values were, the lower the Hb values were measured by the i-STAT. There was no relationship between protein values and the bias of HemoCue versus CellDyn (r ² = 0.001, P = 0.79) (Figure 3).

Figure 3
Linear regression scatter plot of the difference of HemoCue–CellDyn Hb (n = 57) and i-STAT–CellDyn Hb (n = 56) versus protein values. The thick black lines are the lines of best fit

The sensitivity of both POCT to detect anaemia (Hb < 80 g/L) was high (Table 4). The specificity of the i-STAT was low compared with the high specificity of the HemoCue.

Table 4
Diagnosis of anaemia (haemoglobin <80 g/L) with three different methods in 57 samples of porcine blood

Anaemia CellDyn HemoCue i-STAT

True positive 29 28 29

False positive 0 0 17

True negative 28 28 11

False negative 0 1 0

Sensitivity 97% 100%

Specificity 100% 39%

Discussion

The present study evaluated the Hb measurement of the i-STAT and the HemoCue POCT devices against a standard laboratory technique in neonatal pigs over a wide range of Hb values. Good agreement between the photometrical method (HemoCue) and the reference instrument was observed. In contrast to this, the conductivity-based method (i-STAT) underestimated Hb values, with a strong negative bias of 15 g/L. In this population, the conductimetry method did not provide adequate measurement of Hb.

Conductimetry for Hct measurement has been evaluated in human medicine in a broad range of patients, and Stott et al.³⁰ concluded that it provides accurate measurements for physiologically normal subjects, but not for some intensive care and surgical patients. The method becomes inaccurate when serum electrolytes and/or protein concentrations deviate from the norm, as for example in haemodilution^6,31 occurring during surgery.

Newer instruments, such as the i-STAT, are able to measure electrolytes and correct for the deviation, usually with an algorithm based on the sodium concentration.³⁰ However, adjusted values are still not as accurate as bedside photometry (HemoCue).³¹ Hct cannot be corrected automatically for abnormal protein values (<65 g/L, >80 g/L), since plasma proteins cannot be measured with the i-STAT device. For every decrease of 10 g/L in protein concentration, the conductimetric device measures a 1–1.3 L/L lower Hct.^30,31 In cardiopulmonary bypass (CPD) patients, plasma is diluted and Hct measurements with conductimetry are underestimating the real value because of hypoproteinaemia. The i-STAT has a so-called CPD correction mode,³² which corrects for haemodilution. This mode should not only be activated in CPD, but also in all anaesthetized patients.²² However, the automatic CPD correction of the values may not be appropriate as hypoproteinaemia varies among individuals.

In contrast to this, the photometric HemoCue method is not affected by alterations in plasma protein and electrolyte content in humans³³ and is also an accurate measurement of neonatal Hb.³⁴ Also in our group of neonatal pigs this device provided accurate measurement as shown before in different animal species.^11,12,17,18 This is not surprising as the principle of measuring Hb in lysed erythrocytes applies to all species.¹¹ Erythrocyte size and conformation can be very variable between different species.³⁵ Thus, a measurement relying on conductimetry of the intact cells like the i-STAT is more susceptible to errors when used in different species. In cats, horses, mice and chickens a significant underestimation of Hct was shown to a similar extent to that found in the current study.^{14,16,19–21} Only Looney et al.¹⁶ excluded an influence of protein values as all horses included in their study had normal plasma protein values.

The conductivity of blood is greatly dependent on the volume fraction of erythrocytes. The small size of equine erythrocytes (5.5 µm) and the resulting decreased volume fraction leads to an increased conductivity of equine samples compared with canine or human blood and to decreased Hct readings.¹⁶ Cats have a similar size of erythrocytes (5.6 µm) like horses.³⁵ This could explain the decreased Hb/Hct readings of the i-STAT found in this species, although this study did not exclude low protein values as another possible reason for the underestimation.¹⁴ In mice, the Hct was also underestimated especially in anaemic mice.²⁰ Mice have an erythrocyte size of around 7 µm, which is slightly smaller than the size of human erythrocyte.³⁵ In contrast to these findings, Hct was constantly measured 4 L/L higher with the i-STAT in elephant seals than with the reference method.¹⁵ The volume fraction of RBC is higher in elephant seals than in human beings³⁶ and results in decreased conductivity and therefore increased Hct values measured with the i-STAT.

Pigs have an erythrocyte size smaller but similar to humans (7 µm). If young pigs are anaemic they show a microcytosis typical for iron deficiency. This would explain in part the underestimation detected in the present study. However, the bias between the methods was correlated with the protein values. Pigs were under anaesthesia and showed a significant hypoproteinaemia, which was probably the most important factor for the measured bias between the methods. This did not affect the HemoCue measurement in this study and the bias showed no dependency on plasma protein values as reported before.³³

If protein values of measured samples were available, a correction for deviations in plasma proteins could be performed. As many pathological conditions have an influence on proteinaemia, changes cannot be excluded under clinical conditions without appropriate determination of protein levels.

In most veterinary studies, the accuracy of both monitors was judged to be clinically sufficient without adequate definition thereof.^{14–16,19,21} If samples of healthy non-anaemic animals are tested, the difference of 4–5 L/L bias in Hct seems negligible, but the POCT needs to be accurate enough to diagnose pathological conditions such as anaemia or haemoconcentration. It was proposed that the limits of agreement (LOA) should lie within ±5 g/L to detect if anaemia is present or transfusion therapy is effective.^5,37 The limits of the HemoCue LOA are a bit wider than required but nevertheless are still classified as adequate accuracy.⁵ However, the i-STAT bias and LOA are far from being accurate, not only in our study but also in most other veterinary studies.^{14–16,19–21}

Another way to define if a POCT is distinguishing pathological samples from normal ones is to calculate the sensitivity and the specificity of the method to detect anaemia.^17,38 Both devices had a very high sensitivity, but this was clouded by the specificity of only 39% of the i-STAT, with 17 false-positive samples. This could lead to many wrong conclusions and unnecessary transfusions in clinical and research practices.

None of the previous studies performed with the i-STAT in veterinary medicine tried to quantify the importance of the bias of the respective POCT monitor.^{14–16,19–21} The high correlation values between the tested devices and the reference method show that the criticism of Bland and Altman³⁹ against correlation and regression analysis to test for accuracy of various methods is highly justified. Correlation and regression analysis measures the degree of the linear relationship, not the degree of agreement.

In our study we aimed to mimic the clinical situation and one single sample was measured once. However, loading multiple microcuvettes and averaging the results may have led to better accuracy, since variability would have been reduced.^7,9 However, the intra-assay precision performed with porcine blood showed a very good precision.

Another influence might be the use of fresh whole blood in the POCT monitors and the use of K₃-EDTA anticoagulated blood in the reference methods. The use of wet heparinized samples is well known to influence scattering of the result,⁵ and might have influenced our reference values that were obtained with tubes containing liquid K₃-EDTA. As the laboratory samples were also only measured once from one sample, this could have led to more variability than reported for the reference method and therefore an increased standard deviation and LOA of the mean bias.

The clinical and research implications of our study results are that if i-STAT is used to measure pen-side blood values, all low readings of Hb need to be reconfirmed with a more robust method such as photometry or alternatively Hct could be controlled with the microhaematocrit method. These methods are neither sensitive to protein or electrolyte changes nor to species differences in erythrocyte size and conformation.

The conductivity-based measurement of Hb (i-STAT) constantly measured lower Hb values whereas the photometrical HemoCue demonstrated a good accuracy compared with standard laboratory assessment.

Analysis	Median	Range
CellDyn Hb (n = 57)	80 g/L	32–108 g/L
HemoCue Hb (n = 57)	81 g/L	27–112 g/L
i-STAT Hb (n = 56)	68 g/L	34–92 g/L
CellDyn Hct (n = 56)	24.6 L/L	9.6–37.4 L/L
i-STAT Hct (n = 56)	20.4 L/L	10.2–27.6 L/L
Protein (n = 57)	36 g/L	25–55 g/L

	Sample 1	Sample 2
HemoCue	49	1	2.2	82	1	1.1
i-STAT	37	1	3.4	67	1	1.9

Group	Bias	LOA	R	Intercept with 95% CI	Slope with 95% CI
HemoCue/CellDyn Hb (n = 57)	−1 g/L	−7 to 6 g/L	0.98	−0.92 (−1.18 to −0.59)	1.11 (1.07 to 1.15)
i-STAT/CellDyn Hb (n = 56)	−15 g/L	−24 to −6 g/L	0.95	−0.47 (−1.20 to 0.03)	0.88 (0.81 to 0.96)
i-STAT/CellDyn Hct (n = 56)	−4.8 L/L	−8.4 to −1.2 L/L	0.96	−1.32 (−3.48 to 0.82)	0.86 (0.77 to 0.94)

Anaemia	CellDyn	HemoCue	i-STAT
True positive	29	28	29
False positive	0	0	17
True negative	28	28	11
False negative	0	1	0
Sensitivity		97%	100%
Specificity		100%	39%

References

Schimke

. Quality and timeliness in medical laboratory testing. Anal Bioanal Chem 2009;393:1499–504

Kost

, Ehrmeyer

, Chernow

, The laboratory-clinical interface: point-of-care testing. Chest 1999;115:1140–54

Briggs

, Carter

, Lee

, Sandhaus

, Simon-Lopez

, Corrons

JLV

. ICSH guideline for worldwide point-of-care testing in haematology with special reference to the complete blood count. Int J Lab Hematol 2008;30:105–16

Erickson

, Wilding

. Evaluation of a novel point-of-care system, the i-STAT portable clinical analyzer. Clin Chem 1993;39:283–7

Gehring

, Hornberger

, Dibbelt

, Accuracy of point-of-care-testing (POCT) for determining hemoglobin concentrations. Acta Anaesthesiol Scand 2002;46:980–6

Hopfer

, Nadeau

, Sundra

, Makowski

. Effect of protein on hemoglobin and hematocrit assays with a conductivity-based point-of-care testing device: comparison with optical methods. Ann Clin Lab Sci 2004;34:75–82

Jaeger

, Ashbury

, Adams

, Duncan

. Perioperative on-site haemoglobin determination: as accurate as laboratory values? Can J Anaesth 1996;43:795–8

Lardi

, Hirst

, Mortimer

, McCollum

. Evaluation of the HemoCue for measuring intra-operative haemoglobin concentrations: a comparison with the Coulter Max-M. Anaesthesia 1998;53:349–52

Rippmann

, Nett

, Popovic

, Seifert

, Pasch

, Spahn

. HemoCue, an accurate bedside method of hemoglobin measurement? J Clin Monit 1997;13:373–7

10.

von Schenck

, Falkensson

, Lundberg

. Evaluation of ‘HemoCue’, a new device for determining hemoglobin. Clin Chem 1986;32:526–9

11.

Callan

, Giger

, Oakley

, Scotti

, Shofer

. Evaluation of an automated-system for hemoglobin measurement in animals. Am J Vet Res 1992;53:1760–4

12.

Chevalier

, Posner

, Ludders

, French

, Erb

, Gleed

. Accuracy and precision of a point-of-care hemoglobinometer for measuring hemoglobin concentration and estimating packed cell volume in horses. J Am Vet Med Assoc 2003;223:78–83

13.

Coulson

, Januszkiewicz

, Dodd

, Ripple

. The cardiorespiratory effects of diazepam-ketamine and xylazine-ketamine anesthetic combinations in sheep. Lab Anim Sci 1989;39:591–7

14.

Grosenbaugh

, Gadawski

, Muir

. Evaluation of a portable clinical analyzer in a veterinary hospital setting. J Am Vet Med Assoc 1998;213:691–4

15.

Larsen

, Haulena

, Grindem

, Gulland

. Blood values of juvenile northern elephant seals (Mirounga angustirostris) obtained using a portable clinical analyzer. Vet Clin Pathol 2002;31:106–10

16.

Looney

, Ludders

, Erb

, Gleed

, Moon

. Use of a handheld device for analysis of blood electrolyte concentrations and blood gas partial pressures in dogs and horses. J Am Vet Med Assoc 1998;213:526–30

17.

Magona

, Walubengo

, Anderson

, Olaho-Mukani

, Jonsson

, Eisler

. Portable haemoglobinometers and their potential for penside detection of anaemia in bovine disease diagnosis: a comparative evaluation. Vet J 2004;168:343–8

18.

Posner

, Erb

, Gleed

. The HemoCue for point-of-care hemoglobin measurement and packed cell volume estimation in cats. J Vet Emerg Crit Care 2005;15:22–5

19.

Steinmetz

, Vogt

, Kastner

, Riond

, Hatt

J-M

. Evaluation of the i-STAT portable clinical analyzer in chickens (Gallus gallus). J Vet Diagn Invest 2007;19:382–8

20.

Tinkey

, Lembo

, Craig

, West

, Van Pelt

. Use of the i-STAT portable clinical analyzer in mice. Lab Anim (NY) 2006;35:45–50

21.

Tschudi

. [Evaluation of the portable blood analyser i-STAT]. Schweiz Arch Tierheilkd 1998;140:507–12

22.

Myers

, Browne

. Point of care hematocrit and hemoglobin in cardiac surgery: a review. Perfusion 2007;22:179–83

23.

Kernkamp

HCH

, Clawson

, Ferneyhough

. Preventing iron-deficiency anemia in baby pigs. J Anim Sci 1962;21:527–32

24.

Ullrey

, Miller

, West

, Oral and parenteral administration of iron in the prevention and treatment of baby pig anemia. J Anim Sci 1959;18:256–63

25.

Peters

, Mahan

. Effects of neonatal iron status, iron injections at birth, and weaning in young pigs from sows fed either organic or inorganic trace minerals. J Anim Sci 2008;86:2261–9

26.

Rincker

, Hill

, Link

, Rowntree

. Effects of dietary iron supplementation on growth performance, hematological status, and whole-body mineral concentrations of nursery pigs. J Anim Sci 2004;82:3189–97

27.

Drobin

, Kjellstrom

, Malm

, Hemodynamic response and oxygen transport in pigs resuscitated with maleimide-polyethylene glycol-modified hemoglobin (MP4). J Appl Physiol 2004;96:1843–53

28.

Loepke

, Priestley

, Schultz

, McCann

, Golden

, Kurth

. Desflurane improves neurologic outcome after low-flow cardiopulmonary bypass in newborn pigs. Anesthesiology 2002;97:1521–7

29.

Vanzetti

. An azide-methemoglobin method for hemoglobin determination in blood. J Lab Clin Med 1966;67:116–26

30.

Stott

RAW

, Hortin

, Wilhite

, Miller

, Smith

, Landt

. Analytical artifacts in hematocrit measurements by whole-blood chemistry analyzers. Clin Chem 1995;41:306–11

31.

McNulty

, Sharkey

, Asam

, Lee

. Evaluation of STAT-CRIT hematocrit determination in comparison to coulter and centrifuge – the effects of isotonic hemodilution and albumin administration. Anesth Analg 1993;76:830–4

32.

Connelly

, Magee

, Kiessling

. The use of the i-STAT portable analyzer in patients undergoing cardiopulmonary bypass. J Clin Monit 1996;12:311–15

33.

McNulty

, Sharkey

, Schieren

. Bedside hemoglobin measurements – sensitivity to changes in serum-protein and electrolytes. J Clin Monit 1994;10:377–81

34.

Rechner

, Twigg

, Davies

, Imong

. Evaluation of the HemoCue compared with the Coulter STKS for measurement of neonatal haemoglobin. Arch Dis Child 2002;86:188–9

35.

Gregory

. Nucleotypic effects without nuclei: genome size and erythrocyte size in mammals. Genome 2000;43:895–901

36.

Wickham

, Bauersachs

, Wenby

, Sowemimo-Coker

, Meiselman

, Elsner

. Red cell aggregation and viscoelasticity of blood from seals, swine and man. Biorheology 1990;27:191–204

37.

Rajamaki

. Limits of performance for measurement of blood hemoglobin concentration, number concentration of erythrocytes, and erythrocyte volume fraction in inter-laboratory trials – results of the Finnish proficiency testing program in hematology. Scand J Clin Lab Invest 1980;40:563–6

38.

Munoz

, Romero

, Gomez

, Manteca

, Naveira

, Ramirez

. 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

39.

Bland

, Altman

. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10

	Sample 1			Sample 2
Method	Mean (g/L)	Standard deviation (g/L)	Coefficient of variation (%)	Mean (g/L)	Standard deviation (g/L)	Coefficient of variation (%)
HemoCue	49	1	2.2	82	1	1.1
i-STAT	37	1	3.4	67	1	1.9

Evaluation of two devices for point-of-care testing of haemoglobin in neonatal pigs

Abstract

Keywords

Materials and methods

Animals

HemoCue

i-STAT

Reference instrument

Experimental set-up

Statistical analysis

Results

Discussion

References