Eur J Pediatr (2008) 167:1313–1319 DOI 10.1007/s00431-008-0733-y

ORIGINAL PAPER

Reliability of urine collection pads for routine and metabolic biochemistry in infants and young children Patricia M. Crofton & Neil Squires & D. Fraser Davidson & Paul Henderson & Sepideh Taheri

Received: 6 February 2008 / Accepted: 31 March 2008 / Published online: 21 May 2008 # Springer-Verlag 2008

Abstract The aim of our study was to evaluate whether specifically designed urine collection pads give reliable results for routine and metabolic biochemistry tests in paediatric urine. Urine collected by bag or clean-catch from infants and children <2 yrs without metabolic disorders was divided into two aliquots, one of which was added to a collection pad, incubated for 15 min at 37°C (simulating in vivo collection conditions), then recovered by aspiration. Urine from adults with phaeochromocytoma and aqueous solutions of catecholamines were similarly treated. Routine, catecholamine, and metabolic analyses were performed on pad/non-pad aliquots. Selected metabolic analyses were also performed on pad/non-pad urine from patients with diagnosed inborn errors and urine containing added metabolites to simulate metabolic disorders. Routine tests (urea, electrolytes, creatinine, osmolality, calcium:creatinine, phosphate:creatinine, magnesium:creatinine, urate:creatinine [n=32], oxalate:creatinine [n=10]), and catecholamines (n= 12) showed good or acceptable concordance with no clinically significant pad/non-pad differences. Metabolic tests in infants and children without metabolic disorders all

P. M. Crofton (*) : N. Squires Department of Paediatric Biochemistry, Royal Hospital for Sick Children, Edinburgh EH9 1LF, UK e-mail: [email protected] D. F. Davidson Biochemistry Department, Crosshouse Hospital, Kilmarnock, UK P. Henderson : S. Taheri Department of Medical Paediatrics, Royal Hospital for Sick Children, Edinburgh, UK

showed good pad/non-pad concordance for amino acids (n=10), organic acids (n=12), and glycosaminoglycans (n= 8). In patients with metabolic disorders (phenylketonuria [n= 1], homozygous/heterozygous cystinuria [n=3], mucopolysaccharidoses II [n=2] and III [n=1], organic acid disorders [n=6]) and urine containing added orotic acid to simulate urea cycle disorders, there was also good pad/non-pad concordance for diagnostic urinary metabolites. No extraneous organic acids were eluted from the pads. Sugar chromatography showed identical staining intensity in pad/ non-pad samples. In conclusion, urine collection pads give reliable results for a wide range of routine and metabolic biochemistry tests in urine from paediatric patients. Keywords Urine . Collection pad . Biochemistry . Catecholamines . Inborn errors of metabolism

Introduction Urine biochemistry tests are often performed on urine samples collected from children in hospital or in outpatient clinics. However, it is often difficult to collect urine in infants or very young children. To circumvent this problem, specifically designed urine collection pads may be used which can be put inside a nappy. These pads resemble a female sanitary pad in size and appearance but are made of plain cotton fibre to facilitate recovery of urine. It has previously been demonstrated that these pads give reliable results for most routine biochemistry tests in adult urine [4]. However, paediatric urine may have a different composition from adult urine, and the pads have not yet been fully validated for use in infants and young children. Furthermore, the pads have not yet been evaluated for metabolic biochemistry tests.

1314

The aims of this study were (1) to confirm that the pads give reliable results for routine biochemistry tests in urine from infants and young children, and (2) to evaluate whether reliable metabolic biochemistry results are also obtained in urine collected using these pads.

Methods Samples Random, untimed urine samples from infants and young children below 2 years of age with a variety of clinical conditions were collected by bag or as a clean catch specimen (direct collection of free-flowing urine into a sterile pot) and sent to the laboratory either for routine biochemistry tests (n=32), oxalate (n=10), or for metabolic screening tests (amino acids n=10, organic acids n=12, and glycosaminoglycans n=8), as required for their clinical management. None of these patients had inborn errors of metabolism. Surplus urine (minimum 15 ml) remaining after these laboratory tests were completed was stored at – 20°C until analysis. In addition, surplus urine from infants, older children, and adults with diagnosed inborn errors of metabolism was stored at –20°C or –70°C until analysis. To evaluate recovery of catecholamines from the pads, we used (i) surplus urine samples from five adult patients with diagnosed phaeochromocytoma, (ii) five quality control/assessment urine samples, and (iii) two aqueous standards containing catecholamines. To determine whether exogenous organic acids were eluted from the pads under various conditions of pH likely to be encountered in paediatric urine, we prepared aqueous solutions of pH 5, 6, 7, and 7.5. To evaluate recovery of orotic acid from the pads, we added orotic acid to orotic acid-free urine to achieve final orotic acid concentrations corresponding to borderline-high and markedly elevated excretion. To evaluate recovery of sugars from the pads, we added (i) glucose, sucrose, and xylose, and (ii) fructose, galactose, and lactose to separate aliquots of a single urine sample containing no detectable reducing substances (by Clinitest; Bayer Diagnostics, Bridgend, UK) to produce urinary sugar concentrations of 1.0, 2.0, 5.0, and 10.0 mmol/l. Application and recovery from pads All urine and other aqueous samples were divided into two aliquots, of which one 10 ml aliquot was poured onto a Newcastle Urine Collection Pad (Uricol, Ontex, Corby, UK). The pad was immediately placed inside a sealed plastic bag and incubated for 15 min at 37°C to simulate enclosure in a child’s nappy. Urine was then recovered by

Eur J Pediatr (2008) 167:1313–1319

aspiration from the pad using a syringe according to the manufacturer’s instructions. Analytical methods Routine biochemistry The following analytes were measured on a Modular P analyser (Roche Diagnostics, Lewes, UK) using Roche reagents according to manufacturer’s instructions: creatinine (modified Jaffé alkaline picrate), urea (kinetic urease method), sodium, potassium and chloride (ion-selective electrodes), calcium (cresolphthalein complexone), phosphate (molybdenum blue), magnesium (xylidyl blue), and urate (uricase). Osmolality was measured by freezing point depression on an Advanced MicroOsmometer 3MO (Vitech Scientific, Sussex, UK). Oxalate was measured by an oxalate oxidase method (Trinity Biotech, Co., Wicklow, Ireland) on an Ortho Fusion 5.1 FS analyser. At the concentrations typically found in urine from infants and young children, between-assay coefficients of variation (CVs) were: creatinine 10% at 1 mmol/l, urea 3% at 100 mmol/l, sodium 7.5% at 20 mmol/l, potassium 2% at 30 mmol/l, chloride 4% at 20 mmol/l, calcium 2.5% at 1 mmol/l, phosphate 2% at 10 mmol/l, magnesium 12% at 1 mmol/l, urate 1% at 1 mmol/l, oxalate 2% at 100 μmol/l, and osmolality 0.5% at 200 mmol/kg. Catecholamines Catecholamines were measured by highperformance liquid chromatography with coulometric detection [2]. The upper limits of the analytical range were approximately: 80 μmol/l for vanillylmandelic acid (VMA), 120 μmol/l for homovanillic acid (HVA), 4.0 μmol/l for dopamine, 2.0 μmol/l for adrenaline and noradrenaline, and 4.0 μmol/l for free metadrenaline and free normetadrenaline. Between-assay CVs were: VMA 5% at 76 μmol/l, HVA 6% at 76 μmol/l, dopamine 5% at 3.3 μmol/l, adrenaline 8% at 0.27 μmol/l, noradrenaline 9% at 1.3 μmol/l, free metadrenaline 6% at 1.8 μmol/l, and free normetadrenaline 7% at 4.8 μmol/l. Metabolic biochemistry Amino acids were measured by ion-exchange chromatography on a Biochrom 30 Amino Acid Analyser (Biochrom, Cambridge, UK) with ninhydrin detection. Between-assay CVs ranged from 4% to 9% for the individual aminoacids measured. Organic acids were extracted from urine using ethyl acetate, converted to trimethylsilyl derivatives, and analysed by gas chromatographymass spectrometry (Agilent Technologies, Stockport, UK). Methylmalonic acid was quantitated in relation to pimelic acid as internal standard, using a stored calibration curve (between-assay CV 17%). Orotic acid was quantitated using a 15N2 orotic acid internal standard (between-assay CV 8%). Glycosaminoglycans were quantitated using a manual dimethylene blue method and chondroitin sulphate A

Eur J Pediatr (2008) 167:1313–1319

1315

standards [3]. Between-assay CV was 4%. Qualitative separation of individual glycosaminoglycans was carried out by Alcian Blue extraction followed by one-dimensional electrophoresis and visual inspection of staining intensity [9]. Sugars were analysed by thin-layer chromatography with visual inspection of staining intensity [6].

staining intensity (glycosaminoglycans, sugars) or by detailed comparison of profiles (organic acids).

Results Routine biochemistry

Data analysis Calcium, phosphate, magnesium, urate, oxalate, aminoacids, glycosaminoglycans, methylmalonic acid, and orotic acid were expressed as ratios to creatinine to conform with the units used for paediatric reference ranges in random urine samples. Other analytes were expressed in concentration units. Data generally showed a non-Gaussian distribution and were expressed as median and range. For quantitative analyses, paired ‘pad’ and ‘non-pad’ results were compared using Passing and Bablok regression [7], Bland-Altman bias plots with 95% limits of agreement [1], and Wilcoxon signed-rank tests. Passing and Bablok regression formally tests for agreement between two clinical methods when imprecision occurs in both variables, and is not unduly biased by extreme values. Bland-Altman bias or difference plots are also recommended for comparisons between two clinical methods and have been summarised here as mean bias (pad–non-pad) and 95% limits of agreement (mean pad/non-pad difference ± twice the standard deviation of the differences). The Analyse-it software package, version 1.71, was used for all statistical calculations. For qualitative results, the concordance between paired results was assessed by visual inspection of

Routine biochemistry was analysed in urine from 32 infants and young children. Median patient age was 4.7 months (range, 0.2–20 months). Table 1 shows the median and range of concentrations of routine biochemistry analytes and also the regression line, mean bias, and 95% limits of agreement between paired pad and non-pad urine samples. Agreement between pad and non-pad urine was good or acceptable for all routine biochemistry tests. Although pad urine gave results that were statistically lower compared to paired non-pad urine for creatinine, urea, potassium, chloride, and osmolality, these differences were quantitatively insignificant (Table 1). For most routine assays, 95% limits of agreement were as expected based on the analytical imprecision of the individual assays. However, osmolality, calcium, magnesium, and oxalate showed more variability in the bias than would be predicted from analytical imprecision alone. Calcium and magnesium also showed slightly greater decreases in pad urine, although not enough to affect clinical interpretation. Catecholamines Catecholamines were analysed in urine from five patients with phaeochromocytoma, five quality control samples, and

Table 1 Routine biochemistry in 32 paired pad and non-pad samples Test

Creatinine, mmol/l Urea, mmol/l Sodium, mmol/l Potassium, mmol/l Chloride, mmol/l Osmolality, mOsm/kg water Calcium, mmol/mmol creatinine Phosphate, mmol/mmol creatinine Magnesium, mmol/mmol creatinine Urate, mmol/mmol creatinine Oxalateb, μmol/mmol creatinine a

Non-pad

Bland-Altman bias plots

Median

Range

Slope

Intercept

Mean bias (pad – non-pad)

95% limits of agreement

1.1 82 20 27 22 199 1.1 7.2 0.8 0.7 75

0.3–6.8 23–449 5–76 5–195 4–152 55–849 0.03–3.2 0.0–21.5 0.0–2.1 0.1–3.1 22–285

1.000 0.996 1.000 0.990 1.000 0.997 0.954 1.040 0.922 0.881 1.098

0.0 −1 0 0 0 −2 0.0 −0.3 0.0 +0.1 −5

0.0 * −3 * 0 −1 ** −1* −6 ** −0.1 ** 0.0 −0.1* 0.0 −7

−0.2 to +0.1 −13 to +7 −3 to +3 −4 to +2 −4 to +3 −29 to +17 −0.6 to +0.4 −1.3 to +1.3 −0.6 to +0.4 −0.3 to +0.2 −71 to +57

Regression line between pad (y) and non-pad (x) urine Oxalate was measured on a sub-set of 10 samples Pad versus non-pad: *P<0.05, ** P<0.01 (Wilcoxon signed-rank test)

b

Passing and Bablok Regressiona

1316

Eur J Pediatr (2008) 167:1313–1319

Table 2 Catecholamines (μmol/l) in 12 paired pad and non-pad samples Sample

Patients 1 2 3 4 5 Quality control 1 2 3 4 5 Aqueous standards 1 2

VMA

HVA

Non-pad Pad

>80 >80 >80 59 52

Dopamine

Adrenaline

Noradrenaline

Free metadrenaline

Free normetadrenaline

Non-pad Pad Non-pad Pad

Non-pad Pad

Non-pad Pad

Non-pad Pad

Non-pad

Pad

>80 >80 >80 58 53

19 35 33 15 14

28 35 33 15 14

1.02 1.31 NA 0.90 1.19

0.94 1.32 NA 0.87 1.25

0.25 0.34 NA 0.39 0.61

0.27 0.39 NA 0.37 0.59

0.20 0.16 NA 1.13 0.99

0.19 0.15 NA 1.08 1.04

>4.0 >4.0 NA 1.43 1.99

>4.0 >4.0 NA 1.43 2.22

>4.0 >4.0 NA 1.70 1.92

>4.0 >4.0 NA 1.67 2.14

12 33 50 72 26

13 33 41 69 24

15 30 47 78 33

15 38 36 76 27

1.08 >4.0 >4.0 3.10 NA

1.11 >4.0 >4.0 3.36 NA

0.08 0.19 0.35 0.81 NA

0.09 0.21 0.35 0.65 NA

0.22 0.68 1.23 1.13 NA

0.22 0.72 1.13 1.18 NA

0.09 0.98 1.88 1.78 NA

0.10 0.96 1.88 1.94 NA

0.28 >4.0 >4.0 >4.0 NA

0.30 >4.0 >4.0 >4.0 NA

54 NA

53 NA

52 NA

51 NA

NA 1.87

NA 1.85

NA 2.02

NA NA 1.97 1.94

NA NA 1.90 1.93

NA 1.94

NA 2.06

NA 2.10

NA not analysed either because of an analytical problem (patient 3) or because not present (aqueous standards and quality control 5)

two aqueous standards. Agreement between pad and nonpad urine pairs was generally excellent for all catecholamines over a wide range of concentrations (Table 2). Aminoacids Aminoacids were analysed in urine from ten infants and young children without inborn errors of metabolism. Median age was 6.7 months (range, 0.1–24 months). Table 3 shows the median and range of concentrations in relation to creatinine for those aminoacids that were >10 μmol/mmol

creatinine in all urine samples. There were negligible differences between paired pad and non-pad urine samples for most aminoacids (Table 3). 95% limits of agreement were all within the limits expected based on the combined analytical imprecision of amino acid and creatinine assays. A 10-day-old infant newly diagnosed with phenylketonuria had a phenylalanine of 217 μmol/mmol creatinine in pad urine compared with 202 μmol/mmol in paired non-pad urine. Three patients aged 1.7, 10, and 34 years with heterozygous/ homozygous cystinuria showed good agreement between pad and non-pad results for cystine and dibasic aminoacids, all

Table 3 Selected aminoacids (μmol/mmol creatinine) in ten paired pad and non-pad urine samples from paediatric patients who did not have an inborn error of amino acid metabolism Amino acid

Taurine Aspartate Threonine Serine Glutamine Glycine Alanine Lysine Histidine

Non-pad

Passing and Bablok regressiona

Bland-Altman bias plots

Median

Range

Slope

Intercept

Mean bias (pad–non-pad)

95% limits of agreement

134 38 41 133 85 438 158 47 176

20–899 21–113 13–296 69–419 25–450 137–3074 50–424 28–141 38–447

1.013 1.045 1.017 1.026 0.978 1.004 1.006 0.961 0.967

−4 −1 −1 −4 +1 −5 −1 −1 +2

+3 0 −1 −2 −4 −11 −2 −5 −12

−36 to +41 −8 to +8 −10 to +9 −23 to +18 −31 to +23 −88 to +67 −25 to +21 −21 to +11 −57 to +33

b

c

For aspartate, one sample was excluded from analysis because of a chromatography artefact. Glutamate, cystine, methionine, leucine, valine, isoleucine, tyrosine, phenylalanine, ornithine, arginine, and proline were excluded from statistical analysis because most samples had low or undetectable levels. There were no clinically significant discrepancies between pad and non-pad urine samples for these aminoacids. a Regression line between pad (y) and non-pad (x) urine b The two samples with the highest glycine excretion were from neonates c Pad versus non-pad: P<0.05 (Wilcoxon signed-rank test)

Eur J Pediatr (2008) 167:1313–1319

1317

Table 4 Cystine and dibasic aminoacids (μmol/mmol creatinine) in three paired pad and non-pad urine samples from patients with heterozygous/ homozygous cystinuria Age (yrs)

1.7 10.0 34.2

Cystine

Lysine

Ornithine

Arginine

Non-pad

Pad

Non-pad

Pad

Non-pad

Pad

Non-pad

Pad

172 425 228

140 359 252

1386 1170 535

1091 1142 623

97 218 55

75 215 64

43 687 127

34 666 140

within the range that would be expected based on analytical imprecision (Table 4). Glycosaminoglycans Glycosaminoglycans were analysed in urine from eight infants and young children without inborn errors of metabolism. Median age was 12 months (range, 0.1–24 months). Median glycosaminoglycan excretion was 19.3 mg/mmol creatinine (range, 12.9–36.9 mg/mmol, all within the appropriate reference range for age). The Passing and Bablok regression line between pad (y) and non-pad (x) urine had a slope of 0.997 and an intercept of –1.0 mg/mmol. The BlandAltman bias plot showed a mean bias of –0.6 mg/mmol, with 95% limits of agreement –4.3 to +3.2 mg/mmol, within the range expected based on analytical imprecision. Table 5 shows total glycosaminoglycan excretion and the individual glycosaminoglycans present in excess in paired pad and non-pad urine samples from two patients with mucopolysaccharidosis II (Hunter) and one patient with mucopolysaccharidosis III (San Filippo). Total glycosaminoglycan concentrations and staining intensity of individual glycosaminoglycans on electrophoresis showed good concordance between pad and non-pad urine pairs for patients with mucopolysaccharidoses. Organic acids Extracts of water and aqueous solutions with pH 5–7.5 showed no extraneous organic acid peaks after application to and recovery from pads compared with non-pad

solutions. Even after prolonged incubation of water with the pads (60 min at 37°C), no extraneous organic acid peaks were observed. Organic acids were analysed in urine from 12 infants and young children without inborn errors of metabolism. Median age was 6.7 months (range, 0.1–24 months). One sample showed a mild dicarboxylic aciduria consistent with resolving ketosis, and one sample showed slight ketonuria. Seven urine samples contained drug metabolites but had otherwise normal organic acid profiles: penicillin (four samples, including the sample with dicarboxylic aciduria), ibuprofen (three samples), paracetamol (two samples), and a large peak of homovanillic acid in a sample from a patient on dopamine treatment. The remaining four samples had entirely normal organic acid profiles. All 12 urine samples showed close concordance in terms of the organic acids present and peak size between pad and non-pad urine pairs. Urine from five patients with diagnosed inborn errors affecting organic acid metabolism and one patient with atypical glutaric aciduria of unknown aetiology were analysed after application to and recovery from the pads (Table 6). The organic acid profiles, including the abnormal metabolites present in excess, were closely concordant in terms of peak size, with no discernible difference between pad and non-pad pairs. Two separate urine samples from the patient with methylmalonic aciduria had methylmalonic acid concentrations of 360/337 and 383/386 μmol/mmol creatinine in non-pad/pad urine pairs, respectively, which were well within expected limits based on analytical imprecision. Two normal urine samples enriched with added orotic acid to simulate borderline-high and markedly elevated levels of

Table 5 Glycosaminoglycans in paired pad and non-pad urine samples from three patients with mucopolysaccharidoses II (Hunter) or III (San Filippo) Mucopolysaccharidosis

II (Hunter) II (Hunter) III (San Filippo)

Age (yrs)

5.6 NR 7.7

Total GAGs (mg/mmol creat.)

GAGS present in excess

Non-pad

Pad

Ref. range

Non-pad

Pad

77.4 67.2 20.7

insuff 65.1 15.1

<11 <10

DS++, HS+ DS++, HS+ HS+

DS++, HS+ DS++, HS+ HS+

GAGs glycosaminoglycans, DS dermatan sulphate, HS heparan sulphate, insuff insufficient, NR not recorded

1318

Eur J Pediatr (2008) 167:1313–1319

Table 6 Diagnostic organic acids present in excess in urine samples from six patients with organic acid disorders Diagnosis

Age (yrs)

Organic acids present in excessa

MCAD deficiencyb

-

Phenylketonuriac

0.03

Methylmalonic aciduriad SSADH deficiency Glycerol kinase deficiencye Glutaric aciduriaf

4.3 6.8 5.8 1.4

Octanoic, 5-hydroxyhexanoic, 7-hydroxyhexanoic, adipic, suberic, sebacic, octenedioic, decenedioic, 3-hydroxydecenedioic, 3-hydroxysebacic, 3-hydroxydodecenedioic, hexanoyl-glycine, suberyl-glycine Phenylacetic, 2-hydroxyphenylacetic, 3-phenyllactic, 4-hydroxyphenylacetic, 3-phenylpyruvic, 4-hydroxyphenyllactic Methylmalonic 4-hydroxybutyric, 4,5-dihydroxyhexanoic (erythro and threo) Glycerol Glutaric

MCAD medium-chain acyl-CoA dehydrogenase, SSADH succinic semi-aldehyde dehydrogenase All organic acids listed were present in similar amounts in both pad and non-pad urine b Pooled, diluted urine from patients with MCAD deficiency c Sample collected pretreatment d Defect in synthesis of adenosylcobalamin cofactor for methylmalonyl-CoA mutase e Contiguous gene deletion resulting in Duchenne muscular dystrophy and glycerol kinase deficiency f Patient with persistent glutaric aciduria without excessive excretion of 3-hydroxybutyric acid and with normal blood glutarylcarnitine, in whom glutaric aciduria types 1 and 2 were ruled out by analysis of glutaryl-CoA dehydrogenase and tritium release assays of fatty acid oxidation on cultured skin fibroblasts a

orotic acid excretion had closely concordant orotic acid concentrations of 4.9/4.7 and 42.2/42.4 umol/mmol creatinine, respectively, in non-pad/pad urine pairs. Sugars The staining intensity of the thin layer chromatography bands corresponding to glucose, sucrose, xylose, fructose, galactose, and lactose at concentrations of 1.0, 2.0, 5.0, and 10.0 mmol/l showed good concordance between pad and non-pad urine pairs.

Discussion A previous study using the same collection pads demonstrated that routine biochemistry results were similar in pad and non-pad urine obtained from adult volunteers [4]. However, urine obtained from infants and young children has a different composition from adult urine, particularly in relation to lower concentrations of sodium and creatinine. It also has higher excretion of those urinary constituents expressed as a ratio to creatinine, owing to the lower excretion of creatinine in infants compared with adults. Our study using urine from infants and young children confirmed that the pads also gave good or acceptable results for routine biochemistry over the concentration range encountered in this age group. The slightly lower and more variable calcium, magnesium, and oxalate results obtained in pad urine were probably due to variable precipitation on storage because it was not possible to put pre-acidified urine through the pads as this damaged the

pads. This would not be a problem for fresh urine collected directly from an infant as no significant precipitation occurs within the first few hours after collection [5]. All catecholamines showed good agreement between pad and non-pad urine, including samples with very high catecholamine excretion, confirming that use of the pads will give reliable results in children with suspected neuroblastoma or phaeochromocytoma, consistent with previous observations [4]. The pads have not hitherto been evaluated for metabolic tests in children being investigated for inborn errors of metabolism. The tests included in our study are those that are widely used as first-line investigations for a range of inborn errors of metabolism. Such disorders are usually manifested by fairly marked abnormalities. Minor deviations in excretion have little clinical significance. We observed good quantitative concordance for aminoacids between pad and non-pad urine, spanning the ranges of amino acid excretion seen in infants with and without inborn errors affecting amino acid metabolism or excretion. There was also good quantitative concordance for total glycosaminoglycans, both in infants without metabolic disorders and in patients with mucopolysaccharidoses. In the latter group, the qualitative appearance and staining intensity of the glycosaminoglycans present in excess were identical in pad and non-pad urine. Organic acid profiles were virtually identical in pad and non-pad urine, both in infants without metabolic disorders and in patients with various organic acid disorders. In particular, hexanoyl- and suberyl-glycine, the specific metabolites found in mediumchain acyl-CoA dehydrogenase deficiency, and orotic acid, elevated in several urea cycle disorders, showed excellent concordance between pad and non-pad urine. We did not

Eur J Pediatr (2008) 167:1313–1319

detect any extraneous or interfering organic acids eluted from the pads, even after prolonged incubation. Sugars showed identical staining intensity in pad and non-pad urine samples over a range of concentrations, confirming that use of the pads would give reliable results for infants with suspected galactosaemia or hereditary fructose intolerance. Our study design could not reproduce exactly the conditions of possible contamination that may occur when the pads are in contact with an infant’s skin. Previous studies, in which pads were used to collect urine from children to diagnose or rule out urinary tract infection, have demonstrated a high rate of sample contamination by faecal/perineal flora resulting in heavy mixed growth [8]. Although this can be minimised by changing the pad every 30 minutes until urine is passed, clean-catch or supra-pubic aspiration remains the collection method of choice for this purpose. In our study, however, children were not being evaluated for suspected urinary tract infection, and we focussed specifically on biochemistry rather than microbiological investigations. Nevertheless, even for routine and metabolic biochemistry, it is important to adhere strictly to the manufacturer’s instructions to minimise possible bacterial or other contamination and prevent evaporative loss: clean genital area and bottom with soap and water, dry front to back, do not apply cream or talcum powder, check the pad every 10 minutes, apply a new pad if not wet within 30 minutes, and start again if the pad is soiled. In conclusion, this study has confirmed that the pads give reliable results for routine biochemistry in urine from infants and young children and for catecholamines in urine from patients with phaeochromocytoma. It has also established for the first time that the pads give reliable results for a wide range of metabolic tests in patients with and without inborn errors of metabolism.

1319 Acknowledgements We are very grateful to many colleagues in the Department of Paediatric Biochemistry, Royal Hospital for Sick Children, Edinburgh for the routine biochemistry assays and for their expertise in performing assays of aminoacids, organic acids, and mucopolysaccharides. We also thank Tamara Hanson in the Department of Clinical Biochemistry, Western General Hospital for the oxalate assays and Linda Griffiths in the Biochemistry Department, Crosshouse Hospital, Kilmarnock for the catecholamine assays.

References 1. Bland JM, Altman DG (1986) Statistical methods for assessing agreements between two methods of clinical measurement. Lancet 1(8476):307–310 2. Davidson DF (2002) Phaeochromocytoma with normal urinary catecholamines: the potential value of urinary free metadrenalines. Ann Clin Biochem 39:557–566 3. de Jong JGN, Wevers RA, Liebrand-van Sambeek R (1992) Measuring urinary glycosaminoglycans in the presence of protein: an improved screening procedure for mucopolysaccharidoses based on dimethyldimethylene blue. Clin Chem 38:803–807 4. Macfarlane PI, Ellis R, Hughes C, Houghton C, Lord R (2005) Urine collection pads: are samples reliable for urine biochemistry and microscopy? Pediatr Nephrol 20:170–179 5. McConnell W, Vasikaran SD (2002) Spot urine analysis: acidification does not increase calcium recovery. Ann Clin Biochem 39: 64–65 6. Menzies IS, Seakins JWT (1976) Sugars. In: Smith I, Seakins JWT (eds) Chromatographic and electrophoretic techniques, vol 1. Paper and thin layer chromatography, 4th edn. Heinemann Medical Books, pp 183–217 7. Passing H, Bablok W (1983) A new biometrical procedure for testing the equality of measurements from two different analytical methods. J Clin Chem Clin Biochem 21:709–720 8. Rao S, Bhatt J, Houghton C, Macfarlane P (2003) An improved urine collection pad method: a randomised clinical trial. Arch Dis Child 89:773–775 9. Wessler E (1968) Analytical and preparative separation of acidic glycosaminoglycans by electrophoresis in barium acetate. Analyt Biochem 26:439–444