Newborn screening blood spot analysis in the UK: influence of spot size,punch location and haematocrit

Abstract

Objective

In dried blood spot analysis, punch location and variations in applied sample volume and haematocrit can produce different measured concentrations of analytes. We investigated the magnitude of these effects in newborn screening in the UK.

Methods

Heparinized blood spiked with thyroid stimulating hormone (TSH), phenylalanine, tyrosine, leucine, methionine, octanoyl carnitine (C8), and immunoreactive trypsinogen (IRT) was spotted onto filter paper: (i) at a constant haematocrit of 50% at various volumes, and (ii) at a range of haematocrits using a constant volume. Subpunches (3.2 mm) of the dried blood spots were then analysed.

Results

Compared with a central punch from a 50 µL blood spot with 50% haematocrit, 10 µL spots can have significantly lower measured concentrations of all analytes, with decreases of 15% or more observed for leucine, methionine, phenylalanine, and tyrosine. Punching at the edge of a spot can increase measured concentrations up to 35%. Higher haematocrit decreased measured TSH and C8 yet increased amino acids and IRT by 15% compared with 50% haematocrit. Lower haematocrits had the opposite effect, but only with higher concentrations of some analytes.

Conclusions

Differences in blood spot size, haematocrit and punch location substantially affect measured concentrations for analytes used in the UK newborn screening programme, and this could affect false positive and negative rates. To minimize analytical bias, these variables should be controlled or adjusted for where possible.

Keywords

Newborn screening dried blood spot blood spot quality haematocrit punch location filter paper

Introduction

In 1969, Hill and Palmer noted sources of analytical bias in dried blood spot measurements,¹ including variability in commercially available blood spot calibrants for phenylalanine and the haematocrit of the blood. For the same volume of applied blood, a decreasing area of filter paper is covered with increasing haematocrit in the sample,² and this can affect measured analyte concentrations.³ Batch to batch variation of filter paper contribute to analytical imprecision⁴ (though this has been reduced by the development of blood absorption criteria for filter papers by the Clinical and Laboratory Standards Institute in the USA), as does the location of the punch⁵ for methods employing non-volumetric blood application, such as newborn screening, in which only a portion of the spot is analysed.

Identification of disorders by newborn blood spot screening typically relates whole blood measurements to biochemical concentrations in serum. For example, serum and blood spot phenylalanine concentrations are expected to be very similar. However, the diameter of the patient's blood spot and haematocrit may differ significantly from that of assay calibrants, and the extent of the variance of whole blood from the plasma measurement may be unrecognized. Sadler and Lynskey⁶ reported in 1979 that plasma volume varied by applied blood volume, with less plasma in a punch from a smaller compared with a larger volume of blood applied. They observed that TSH, located solely in plasma, measured less in a standard centre punch from smaller spots compared with a central or edge punch from a larger blood spot.

Newborn screening in England and Northern Ireland utilizes heelprick blood applied non-volumetrically to 10 mm circles on PE 226 filter paper,⁷ and most commonly a 3.2 mm diameter punch size for analysis. The volume of blood to fill each of the four circles on these cards will be different from that for larger diameter circles used in the USA and other countries. Blood spot external quality assessment (EQA) and performance evaluation material for newborn screening from the United States Centers for Disease Control (CDC) is used widely around the world, including in England. However, it is unclear whether bias exists in each country's analytical results if their assays are set up and calibrated with different circle diameters from those from the CDC.

A haematocrit of approximately 50–55% is generally anticipated in infants during the first few days of life. Calibration materials for neonatal screening are therefore created with this or a similar haematocrit.^8,9 However, variation in the haematocrits of 5–8 day old babies' blood as screened in the UK will exist and it is not known how much this might influence the outcome of screening tests and, if variations are observed, whether correction can be made for these changes.

Dried blood spot analysis has many advantages compared with utilizing liquid whole blood samples.¹⁰ It uses significantly smaller volumes of blood, does not require centrifugation facilities, tends to stabilize many analytes on drying, and permits inexpensive mailing of specimens. However, the impact of variable haematocrit on results, which should ideally be within 15% of the ‘true' result,¹¹ is a stumbling block. Sampling matrices which mitigate the haematocrit effect are being developed.^12,13 If efficacy of these matrices is proven, they might be of value in improving accuracy not only of drug testing results but also of newborn blood spot results.

Little is known about the effect of blood spot size, spot punch location, and haematocrit on the performance of newborn screening blood spot analysis in England and Northern Ireland. We investigated the magnitude of these effects in relation to screening for congenital hypothyroidism, phenylketonuria, cystic fibrosis, Medium-chain acyl-CoA dehydrogenase deficiency homocystinuria, and maple syrup urine disease.

Methods

Reagents

L-phenylalanine, L-methionine, L-tyrosine, L-leucine, and octanoyl carnitine (C8) were purchased from Sigma-Aldrich (Gillingham, Dorset, UK). The Second International Reference Preparation of human thyroid stimulating hormone (TSH), 80/558, was obtained from the National Institute for Biological Standard (South Mimms, Hertfordshire, UK).

Preparation and analysis of blood spots

Blood was collected with full consent from a single donor (a member of the research team) into lithium heparin tubes (BD, Oxford, UK). Endogenous concentrations of the analytes of interest and haematocrit were measured by standard laboratory assays in routine use at Birmingham Children’s Hospital (see Table 1 for performance characteristics). TSH and IRT were measured using AutoDELFIA® fluoroimmunoassay (Perkin Elmer, Seer Green, Beaconsfield, UK); C8, leucine, methionine, phenylalanine and tyrosine were measured by underivatized electrospray ionization tandem mass spectrometry using a Waters Quattro instrument (in-house developed method). This involved direct extraction from dried blood spot subpunches with a methanolic solution containing deuterated internal standards. For analyte quantitation, specific ion transitions were detected using multiple reaction monitoring and the ratio of analyte to internal standard used to calculate the concentration. The laboratory participates in EQA schemes and all methods were performing satisfactorily.

Table 1.

Method and performance characteristics for blood spot newborn screening assays.

Analyte	Method details	Analytical range	Example CV (data from relevant time period)
TSH	Perkin Elmer AutoDELFIA® Neonatal hTSH; solid-phase, two-site fluoroimmunometric assay performed on an automated analyser	1 – 280 mU/L	8.9% at15.4 mU/L
IRT	Perkin Elmer AutoDELFIA® Neonatal IRT; solid-phase, two-site fluoroimmunometric assay performed on an automated analyser	10 – 430 ng/mL	5.9% at 66 ng/mL
Phenylalanine	In-house method; underivatized tandem mass spectrometry	Up to 3000 µmol/L	9.8% at 204 µmol/L
Tyrosine	In-house method; underivatized tandem mass spectrometry	Up to 3000 µmol/L	9.7% at 272 µmol/L
Leucine	In-house method; underivatized tandem mass spectrometry	8.19–2993 µmol/L	13.4% at 309 µmol/L
Methionine	In-house method; underivatized tandem mass spectrometry	3.41–2091 µmol/L	10.9% at 69 µmol/L
C8	In-house method; underivatized tandem mass spectrometry	Up to 67 µmol/L	7.2% at 0.52 µmol/L

CV = coefficient of variation. CVs were calculated from >100 values, and were from the time period when this study was undertaken.

Haematocrit was measured on a Sysmex XE-2100 Automated Hematology System. Aliquots of blood were spiked with the chosen analytes at concentrations representative of endogenous concentrations, an analytical alert defined by the laboratory or an individual to be referred for investigation (Table 2).

Table 2.

Analyte concentrations used for all experiments.

Level	TSH	IRT	C8	Leucine	Methionine	Phenylalanine	Tyrosine
Low	NA	NA	0.09	200	18	75	72
Medium	4.2	87	0.34	292	56	216	226
High	11.6	97	0.68	577	97	408	418

Concentrations were representative of a healthy individual (low), analytical alert (medium), or an individual to be referred for investigation (high). Endogenous levels of TSH and IRT in the donor were too low to measure using the autoDELFIA™ fluoroimmunoassay. TSH units are mIU/L whole blood, IRT is ng/mL, while others are in µmol/L. Both IRT concentrations tested are above the cystic fibrosis screening cut-offs but represent values likely to be seen in affected neonates.

To investigate the effect of punch location and spot volume on newborn screening results, blood of a constant 50% haematocrit was spotted onto Perkin Elmer 226 cards (Lot Number 100535/311433) in four different volumes (10 µL, 20 µL, 50 µL, and 75 µL), n = 20 for each volume at each analyte concentration. Cards were punched as per Figure 1A and each analyte measured. TSH was measured from a separate punch to other analytes due to different methodology. To investigate the effect of haematocrit, blood was manipulated using removal or addition of donor plasma for each analyte concentration, to create haematocrit values of 30%, 40%, 45%, 50%, and 60%. 50 µL of blood was then spotted onto the cards, n = 6 for each haematocrit at each analyte concentration. Cards were punched as per Figure 1B and each analyte measured.

Figure 1.

Experimental design to investigate the effect of spot volume, punch location (A) and haematocrit (B) on newborn screening results. Punch locations shown by black dots. All spot volumes were kept constant at 50 µL. Haematocrit values used for the separate IRT experiment were 25%, 50% and 65%. A: volumes in µL are volume of blood added to the card. All haematocrits constant at 50%. B: percentage values refer to haematocrit of blood applied to the card.

In a separate experiment, 60 mL of blood from a different donor (research team member, full consent provided) was adjusted to three different haematocrits; 25%, 50%, and 65%, and spiked with serum from an anonymous adult patient with pancreatitis containing high levels of immunoreactive trypsinogen (IRT) to create two IRT concentrations (Table 2). Based on the initial IRT measurements, calculated amounts of this serum were added to the whole blood to provide the desired concentrations. The effects of punch location and spot volume were investigated as above, and IRT measured. Blood spots were also produced using samples with haematocrits of 25% and 65%, and IRT concentrations were measured from a central punch of a 50 µL spot and compared with 50% haematocrit.

Collection of haematocrit data

Haematocrits for screened babies were collected from hospital records at Birmingham Children's Hospital.

Analysis of blood spot homogeneity

Using bespoke proprietary software, blood spot cards were scanned with a Cardscan™ device supplied by Luminex. Blood migration across spots was assessed at different sites by recording light transmission through a 3.2 mm circular area (equivalent to a standard punch size) at the centre and edge of blood spots.

Statistical Analysis

One-way ANOVA with post-hoc Tukey test, linear regression and paired t-test analyses were performed using SPSS Statistics (IBM) and GraphPad Prism software. For all experiments statistical analysis was performed on absolute (ie. not percentage change) values.

Results

Effect of sample volume and punch location

For newborn screening in the UK, single heelprick non-volumetric blood drops are applied to fill 10 mm diameter circles onto the blood spot card.⁷ This equates to approximately 35–50 µL of blood, however it is common for the circles to be incompletely filled or overfilled. Figure 2 shows that a reduction in the volume of blood applied to the card from 50 µL to 10 or 20 µL significantly reduced the measured analyte value, in some cases up to 20%. This reduction was observed at all three analyte concentrations except TSH, where the difference was only observed at the higher TSH level (Figure 2C). For IRT, the lower of the two concentrations tested showed the largest difference due to small spot size (Figure 2B and C). Using larger 75 µL spots generally increased measured values of all analytes, although this effect was more variable and did not reach significance.

Figure 2.

Effect of spot size and punch location on newborn screening results at concentrations representative of an unaffected individual (A), laboratory alert (B) and an individual referred for further investigation (C). N = 20 for each data point. Data displayed as percentage change (y axis) in analyte concentration when compared with a 50 µL centre punch at a haematocrit of 50%. Statistical analysis performed using one-way ANOVA with post-hoc Tukey test using SPSS Statistics (IBM). * = p < 0.05, ** = p < 0.01 and *** = p < 0.001. Error bars represent +/− standard error of the mean. Leu = Leucine, Met = Methionine, Phe = Phenylalanine & Tyr = Tyrosine.

In punches taken from the edge of a 50 µL or 75 µL blood spot, measured analyte concentrations were significantly increased (up to 35%) (Figure 2) for all analytes tested. The effect was greater with larger blood spots at higher concentrations of analyte. TSH and C8 were most affected by peripheral punches. For IRT, a significant increase was seen with the lower of the two concentrations when a peripheral punch of a 75 µL spot was taken, however at the higher concentration no significant differences were seen, although the trend was the same.

Effect of haematocrit

The haematocrit of the newborn test population blood is taken to be fairly consistent (about 50–55%), but data from Birmingham Children’s Hospital show that in babies admitted for heart surgery, liver disease, or attending follow up of positive PKU or tyrosinaemia screening tests, the haematocrit can deviate markedly from this average with both higher and lower values (Figure 3). Figure 4 shows that with increasing haematocrit TSH and C8 measured analyte concentrations decreased, while leucine, methionine, phenylalanine and tyrosine increased. This effect becomes increasingly significant at higher analyte concentrations (Figure 4B & C) and linear regression analysis showed highly significant correlations (Table 3). However, when compared on a pair-wise basis with a haematocrit value of 50%, only large alterations in haematocrit were shown to be significant for certain analytes, most noticably TSH and C8 (Figure 4C). The low level of significance attributed to some of the changes in analyte concentrations with differing haematocrits may be explained by the low numbers of replicates at each data point (n = 6) when compared with the previous experiment on spot size and punch location (n = 20).

Figure 3.

Haematocrits of babies in hospital. Haematocrits for hospitalized infants routinely identified as requiring newborn screening and not known to have been transfused, n = 39.

Figure 4.

Effect of haematocrit on newborn screening results at concentrations representative of an unaffected individual (A), laboratory alert (B) and an individual referred for further investigation (C). (D) The effect of haematocrit on measured IRT concentrations in blood spots. N = 6 for each data point in A-C, n = 20 for D. Data displayed as percentage change (y axis) in analyte concentration when compared with a 50 µL centre punch at a haematocrit of 50%. Statistical analysis performed using one-way ANOVA with post-hoc Tukey test. * = p < 0.05, ** p = < 0.01, *** = p < 0.001. Error bars represent +/− standard error of the mean.

Table 3.

R² and p values for linear regression analysis of each measurand with increasing haematocrit.

	Normal		Laboratory Alert		Clinical Referral
Analyte	R²	p	R²	p	R²	p
TSH	NA	NA	0.815	<0.001	0.863	<0.001
C8	0.08	0.181	0.661	<0.001	0.734	<0.001
Leu	0.398	<0.001	0.015	0.572	0.36	0.002
Met	0.297	0.006	0.27	0.009	0.581	<0.001
Phe	0.759	<0.001	0.31	0.005	0.805	<0.001
Tyr	0.337	0.003	0.283	0.008	0.651	<0.001

Linear regression performed using SPSS Statistics (IBM)

In a separate experiment, IRT concentrations were measured using spots from blood with 25% and 65% haematocrit and compared with 50% haematocrit samples. Measured IRT levels were not affected by a reduction in haematocrit but were very significantly increased, by 10–15%, when the haematocrit was 65%, at both IRT concentrations tested (Figure 4D).

Assessment of blood spot parameters using Cardscan™

To investigate the effect of blood spot size, cards were scanned using the Cardscan™ device. Light transmission through circular areas matching the punch size used for measurement of analytes showed that as blood spot areas increase, transmission of light at both central and edge locations decreases (Figure 5A and Table 4), suggesting relative compaction of red cells when larger blood volumes are applied. Comparing central transmission at different blood spot volumes, all differences were highly significant, with p < 0.001 showing reduced light transmission with increased blood spot volume, except 50 µL and 75 µL where p was <0.05. Similarly, decreasing edge transmissions with increasing volume were also highly significant, p < 0.001, again except for 50 µL compared with 75 µL, where no significant difference was shown (Table 4). For all blood volumes tested, light transmission was significantly greater at the edge of the spot compared with the centre (Figure 5A). Finally, we tested the effect of haematocrit on light transmission through the spots. Unsurprisingly, higher haematocrit values resulted in lower light transmission at the centre of spots (Figure 5B), and also at peripheral punch locations (data not shown).

Figure 5.

Assessment of blood spot homogeneity using a Cardscan™ device. Blood spot cards were scanned and parameters recorded using provided software. (A) Effect of spot size and punch location on light transmission through the spot. Within the software, a 3.2 mm diameter circle was used to assess light transmission at the centre and edge of each spot (n = 60 for each spot size). (B) Effect of haematocrit on blood spot transmission. Transmission values for a 3.2 mm diameter circle at the centre of the spot were recorded (n = 36 for each haematocrit). Data analysed using one-way ANOVA with post-hoc Tukey test, and paired t-test to compare central and edge transmission values for each spot.

Table 4.

Comparison of light transmission for different blood spot volumes and punch locations.

Tukey's Multiple Comparison Test	Significant? P < 0.05?	Summary
10 µL vs 20 µL centre	Yes	***
10 µL vs 50 µL centre	Yes	***
10 µL vs 75 µL centre	Yes	***
20 µL vs 50 µL centre	Yes	***
20 µL vs 75 µL centre	Yes	***
50 µL vs 75 µL centre	Yes	*
10 µL vs 20 µL edge	Yes	***
10 µL vs 50 µL edge	Yes	***
10 µL vs 75 µL edge	Yes	***
20 µL vs 50 µL edge	Yes	***
20 µL vs 75 µL edge	Yes	***
50 µL vs 75 µL edge	No	ns

One-way ANOVA with post-hoc Tukey test performed using GraphPad Prism software

Discussion

We have shown here that blood spot punch location can significantly affect the measured concentration of an analyte. This has previously been reported for phenylalanine,¹⁴ TSH⁶, and a large selection of metabolites,¹⁵ but not using the 10 mm spot size used in England. IRT and TSH are plasma components with negligible presence in red cells,⁶ and C8 and methionine are present at higher amounts in plasma compared with red cells.¹⁶ Their concentration in blood spots will reflect the proportion of plasma in the punch; if higher fractions of plasma are present at the edge of the spot, this could explain the higher measured values we observed in peripheral punches. Our observation that light transmission through blood spots is greater at the edge, regardless of spot size, supports this theory. However, leucine, tyrosine and phenylalanine are present at similar levels in red cells and plasma,¹⁶ so this does not explain the higher concentrations found for these analytes with peripheral punches.

Various punch machines, with differing capabilities, are used in England. The Perkin Elmer Multipuncher™ used in this study permits punching four 3.2 mm discs from both circular and oval blood spots. One punch, at least, is more likely to be central in the spot, and other(s) more peripheral, leading to greater analytical variation than taking one central punch for analysis. This may influence the triplicate mean calculated to categorize abnormal newborn screening test results, if confirmatory punches are taken from two central positions in other spots when the initial punch was peripheral.

Blood spot volume was also shown here to have large effects on measured concentrations of newborn screening analytes. In general a decrease was observed when smaller volumes were used to make the blood spots. This could represent greater migration of the sample once it is applied the filter paper,¹⁷ and therefore a punch from a smaller spot will contain less blood. Cardscan™ transmission data presented here support this, with higher transmission values obtained from smaller spots.

The range of haematocrits we observed in a hospitalized (ie. sick) newborn English population at age 5 days or older supports Jopling's¹⁸ reference ranges of 35–60% (mean 47%) in the term infant and 34–55% (mean 44%) in the preterm (29–34 weeks gestation), each at age 5 days. This demonstrates that haematocrits in newborn blood spots can differ widely from the 50–55% of calibration and EQA materials. The average haematocrit for a healthy 5 day old baby may ordinarily be less than 55%, and a lower haematocrit for calibration and EQA materials for this age could be more appropriate. Our observed increasing trend in leucine, methionine, phenylalanine, and tyrosine concentrations with increasing haematocrit concord with Holub's work,¹⁵ and may be due to changes in plasma volume²⁰ or reduced sample spreading on the filter paper.²

The decreasing concentration with increasing haematocrit observed for C8 differs from Holub’s¹⁵ findings, and merits substantiation. We know of no studies to date describing the effect of haematocrit on measured blood spot TSH and IRT concentrations. Our Cardscan™ data confirm that alterations in haematocrit substantially affect the properties of the blood spot, possibly partly due to reduced migration of the blood in the filter paper.² A more extensive study with larger numbers of samples is needed to confirm this finding.

The blood spots in this study were produced in a laboratory environment using a dispensing pipette, which is likely to differ substantially from sample collection in the community where non-homogenous blood spots can be produced (eg. layering of spots). This will introduce additional confounding factors and could affect measured analyte concentrations if non-ideal spots are used for sampling.

Effects of blood spot size, punch location, and haematocrit could lead to false positive or false negative screening results, with consequent inappropriate investigations or missed diagnoses. It is difficult to control precisely blood spot size when collecting heel prick specimens from neonates, although detailed guidelines are available to midwives.⁷ The laboratory can assist by applying agreed blood spot acceptance criteria, and preferentially selecting appropriately sized blood spots for punching. The screening laboratory also controls punch location, but given the number of punches required, they cannot all be central.

Attempts to correct blood spot test results for haematocrit have been explored by measurement of blood spot potassium,¹⁹ light reflectance,²⁰ and using data from HPLC haemoglobin screening.²¹ The Cardscan™ device also has this potential.²² Our data show that variation in light transmission through the blood spot correlates with differences in measured analyte concentrations. Blood spot transmission could, therefore, be used as a surrogate marker to correct for punch location and for haematocrit, and appropriate correction could then applied to the results. This should lead to improved accuracy of newborn screening and fewer false positive and false negative results.

Limitations of the study

It would have been preferable to test alternative concentrations of IRT closer to the cut offs used in UK newborn screening programmes. Patient serum was used to spike donor blood from a different individual, and this may have affected the sample in non-discernable ways. Six replicates were tested at each haematocrit for all analytes except IRT; a more extensive study with more replicates and using more than one blood donor could confirm our findings.

Footnotes

Acknowledgements

Russell Denmeade and Deborah Finnerty (Newborn Screening Laboratory, Birmingham Children’s Hospital) coordinated the blood spot punching and analysis for all analytes. Luminex Corporation kindly provided the Cardscan™ device and associated software on loan.

Declaration of competing interests

None

Funding

Partial financial support was received from the UKNSPC for reagent use.

References

Hill

Palmer

. Filter paper blood collection and punching as a means of quantification. Clin Chem 1969; 15(5): 381–9.

O'Broin

. Influence of hematocrit on quantitative analysis of “blood spots” on filter paper. Clin Chem 1993; 39(6): 1354–5.

O'Mara

Hudson-Curtis

Olson

. The effect of hematocrit and punch location on assay bias during quantitative bioanalysis of dried blood spot samples. Bioanalysis 2011; 3(20): 2335–47.

Slazyk

Phillips

Therrell

Jr. . Effect of lot-to-lot variability in filter paper on the quantification of thyroxin, thyrotropin, and phenylalanine in dried-blood specimens. Clin Chem 1988; 34(1): 53–8.

Arends, J. Neonatal screening using dried blood spotted on filter paper: methodological factors which contribute to imprecision. In: Neonatal Screening Symposium, New York, 1987, pp.549–50. Elsevier Publishing Co.

Sadler

Lynskey

. Blood-spot thyrotropin radioimmunoassay in a screening program for congenital hypothyroidism. Clin Chem 1979; 25(6): 933–8.

UK National Screening Committee. Guidelines for Newborn Blood Spot Sampling. UK Newborn Screening Programme Centre, Department of Health, 2012.

Chace

Adam

Smith

. Validation of accuracy-based amino acid reference materials in dried-blood spots by tandem mass spectrometry for newborn screening assays. Clin Chem 1999; 45(8 Pt 1): 1269–77.

Dhondt

Loeber

Elvers

. Preparation of the first European working standard for phenylalanine determination in dried blood spots. J Med Screen 1998; 5(2): 63–6.

10.

Lehmann

Delaby

Vialaret

. Current and future use of “dried blood spot” analyses in clinical chemistry. Clin Chem Lab Med 2013; 51(10): 1897–909.

11.

Biopharmaceutics Coordinating Committee. Guidance for Industry: Bioanalytical Method Validation. US Department of Health and Human Services Food and Drug Administration, 2001.

12.

Hudson, WC, Yong, B, Arora, R, et al. Investigation of Varying Blood Hematocrit level in Dried Blood Spotting. Agilent Technologies Inc., 2011.

13.

Mengerink, Y, Mommers, JHM, and Qiu, J. Multilayered woven manufacture and use of the multilayer woven manufacture as carriers for dried matrix spot applications. Patent Application PCT/EP2012/076688, 2013.

14.

Adam

Alexander

Smith

. Recoveries of phenylalanine from two sets of dried-blood-spot reference materials: prediction from hematocrit, spot volume, and paper matrix. Clin Chem 2000; 46(1): 126–8.

15.

Holub

Tuschl

Ratschmann

. Influence of hematocrit and localisation of punch in dried blood spots on levels of amino acids and acylcarnitines measured by tandem mass spectrometry. Clin Chim Acta 2006; 373(1–2): 27–31.

16.

Chaleckis

Ebe

Pluskal

. Unexpected similarities between the Schizosaccharomyces and human blood metabolomes, and novel human metabolites. Mol Biosyst 2014; 10(10): 2538–51.

17.

Mei

Alexander

Adam

. Use of filter paper for the collection and analysis of human whole blood specimens. J Nutr 2001; 131(5): 1631S–6S.

18.

Jopling

Henry

Wiedmeier

. Reference ranges for hematocrit and blood hemoglobin concentration during the neonatal period: data from a multihospital health care system. Pediatrics 2009; 123(2): e333–7.

19.

Capiau

Stove

Lambert

. Prediction of the hematocrit of dried blood spots via potassium measurement on a routine clinical chemistry analyzer. Anal Chem 2013; 85(1): 404–10.

20.

Miller

Poston

Rutan

. An On-card Approach for Assessment of Hematocrit on Dried Blood Spots which Allows for Correction of Sample Volume. J Anal Bioanal Techniques 2013; 4(162): 162–162.

21.

Orsini

Yeman

Caggana

. Semi-quantitative method for determination of hematocrit in dried blood spots, using data collected in HPLC hemoglobin variant testing. Clin Chim Acta 2010; 411(11–12): 894–5.

22.

Watts, S, de Hora, MR, Wester, DR, Green, D. Application of the Luminex CardScan™ in Improving Quality in Newborn Screening. In: Joint Meeting of the Newborn Screening and Genetic Testing Symposium and International Society for Neonatal Screening, Atlanta, Georgia, 2013.