study done order of draw for potassium

[Guest guest]

H pylori antibody titres in serum, plasma and successively thawed specimens

we excluded two subjects with high titres, 1900

and 3600, well above the cut off value of 500 in

the repeated measures analysis of this effect.

The average change in titre (using untransformed

data) detected among the remaining

subjects was -0.31 per thaw (se = 5.7,

p = 0.96). The maximum probable change per

thaw was estimated (from the 95% CI) at

-1 1.5. Hence, after three thawings it would be

possible for the titre to drop by as much as

34.5. This would cause the sensitivity of the

test to decrease only slightly (estimated reduction

1.3% based on the receiver operator characteristic

(ROC) curve produced by the manufacturer).

This small decrease from 92.5% to

91.2% in the proportion of H pylori positive

subjects detected by the assay is not of any epidemiological

significance.

For eradication trials we were interested in

the effect of thawings on serum titres among all

subjects who were H pylori positive and who

would normally be eligible for treatment. For

these subjects the mean titre was 1245 and the

average drop in titre detected by repeated

measures analysis was -75.8 per thaw (se =

33.7 p = 0.04). The maximum probable drop

per thaw was 11.4%, which may be of clinical

significance.

Discussion

Our results suggest that either plasma or

serum, and specimens that have been stored

frozen and subsequently thawed on several

occasions, can be used for epidemiological

research where a small additional misclassification

rate can be tolerated. Specimens which

have thawed accidentally during storage or

transportation and been refrozen may also be

usefully tested for Hpylori status.

Among all H pylori positive subjects, the

change in titre per thaw was statistically significant

and the estimated drop could be as

extreme as 11.4% per thaw. This needs to be

taken into account when using a drop in titre as

an indicator of eradication of Hpylori infection

during follow up of treatment. Successful

eradication of Hpylori infection is usually associated

with a drop in titre of at least 60% of the

pretreatment value, which generally occurs

within 12 months.3 It is possible that multiple

thawing of specimens could make a significant

contribution to such a drop in titre. Thawing

baseline and follow up specimens once only

and testing all specimens concurrently would

be prudent.

We thank Mr C Pearce and staff of the Department of Microbiology,

Dorevitch Pathology, Ballarat Base Hospital, for their

technical assistance and use of facilities.

This study was funded by the Shepherd Foundation.

The Department of Public Health and Community Medicine

is a member of the n Public Health Consortium.

1 Granberg C, Mansikka A, Lehtonen O-P, Kujari H,

Gronfors R, Nurmi H, et al. Diagnosis of Helicobacter

pylori infection by using Pyloriset EIA-G and EIA-A for

detection of serum immunoglobulin G (IgG) and IgA antibodies.

J Clin Microbiol 1993;31:1450-3.

2 Bland JM, Altman DG. Statistical methods for assessing

agreement between two methods of clinical measurement.

Lancet 1986;i:307-10.

3 Kosunen TU, Seppala K, Sarna S, Sipponen P. Diagnostic

value of decreasing IgG, IgA, and IgM antibody titres after

eradication of H. pylori. Lancet 1992;339:893-5.

_F Clin Pathol 1996;49:1019-1020

The order of draw of blood specimens into

additive containing tubes does not affect

potassium and calcium measurements

Department of

Medicine and

Therapeutics,

Western Infirmary,

Glasgow Gll 6NT

A Majid

D C Heaney

N Padmanabhan

Department of

Biochemistry

R Spooner

Correspondence to:

Dr A Majid,

Department of Neurology,

Washington University

School of Medicine,

Campus Box 8111,

660 S Euclid Avenue,

St Louis,

MO 63110-1093

USA.

Accepted for publication

4 September 1996

A Majid, D C Heaney, N Padmanabhan, R Spooner

Abstract

The effect of order of draw when taking

blood into tubes containing additive was

investigated in 47 medical inpatients; 12 of

these patients acted as a control group.

The samples were analysed in the order in

which they were withdrawn. The results of

potassium and calcium concentrations

did not differ significantly between

groups. Manufacturers recommend a specific

order of draw when taking blood

using vacuum based blood collection systems,

which are routinely used in many

hospitals. The results of this study, however,

show that order of draw has no effect

on calcium or potassium concentrations.

(7 Clin Pathol 1996;49:1019-1020)

Keywords: order of draw, venepuncture, Vacutainer system,

potassium concentration, calcium concentration.

The importance of correct collection and handling

of blood specimens has been highlighted

before.1 2 Calam and have suggested

that the order of draw of blood specimens into

tubes containing additive may affect measured

serum potassium and calcium concentrations.3

In their small (n = 5) report they suggest that if

1019

Majid, Heaney, Padmanabhan, Spooner

blood is drawn initially into potassium-EDTA

bottles before bottles containing no anticoagulant,

then abnormally low calcium and abnormally

high potassium values will be obtained. It

was suggested that potassium-EDTA from the

first tube contaminated the second tube. However,

venepuncture was noted to be difficult in

all their subjects.

As a result of the above report, Becton Dickinson

(Cowley, Oxford, UK) who manufacture

the Vacutainer system, recommend an order of

draw for collecting blood specimens into additive

containing bottles as follows: (a) tubes

containing no anticoagulant; ( citrate; ©

heparin; (d) potassium-EDTA; and (e) tubes

containing fluoride-oxalate. This recommendation

is rarely adhered to in clinical practice.

Critical therapeutic decisions are therefore

made on the assumption of accurate biochemical

results.

In our larger study, the effect of order of

draw on measured potassium and calcium

concentrations was formally investigated.

Methods

Using a 21 gauge needle, antecubital blood

samples were withdrawn from a single vein

within two minutes of a tourniquet being

applied, from 47 medical inpatients whose

informed permission had been obtained.

Twelve patients acted as a control group,

having blood withdrawn sequentially into three

vacuum tubes containing no anticoagulant

(Vacutainer, Becton Dickinson). The other 35

patients acted as the trial group. Blood was

drawn sequentially into a tube containing no

anticoagulant, a potassium-EDTA tube, and

then a second tube containing no anticoagulant.

The samples were analysed in the order

they were withdrawn, using conventional biochemical

analysis on an Olympus AU5200

(Eastleigh, UK) series analyser.

Paired t tests were used to assess significance

of any differences between non-anticoagulated

sample pairs within the two groups. Unpaired t

tests were used to compare the differences

observed in the control group with those in the

trial group.

Results

The results are shown in table 1 and exclude

one patient in whom a difficult venepuncture

was associated with ballooning at the site of

venepuncture. In this subject, both analytes lay

more than 5 SD from the mean of the trial

group. While a slight rise in potassium concentration

and a slight lowering of calcium

concentration was seen in the trial group, this

Table I Mean concentration differences between first and

final blood samples for control and trial subjects. Results are

expressed as mean difference (SD)

n Potassium Calcium

(mmolNl) (mmolNl)

Control 12 0.025 (0.205) 0.014 (0.041)

Trial 34 0.015 (0.131) -0.011 (0.044)

Table 2 T tests between paired and unpaired analyte

concentrations

Potassium Calcium

Control pre v post p = 0.68 p = 0.26

Trial pre vpost p = 0.52 p = 0.15

Control v trial p = 0.87 p = 0.09

was not significantly different from that found

in the control group (table 2).

The results obtained are in line with laboratory

internal quality control data. Within batch

imprecision (SD) for potassium was 0.16

mmol/l and 0.042 mmol/l for calcium.

Conclusion

The results of this study show that success or

otherwise of venepuncture is more important

in determining serum calcium and potassium

results than the order of draw. Our results show

that the order of draw of blood specimens into

additive containing tubes does not affect

potassium and calcium measurements. We

suggest that the high potassium and the low

calcium measurements of other studies were

due to local factors rather than the order of

draw. Local tissue damage during a difficult

venepuncture induces release of potassium

from damaged cells and this high extracellular

potassium leads to depolarisation of local cells

causing calcium to flood into them, resulting in

high potassium and low calcium measurements.

We conclude that the order of draw does not

effect the potassium and calcium measurements

but difficult venepuncture may result in

high potassium and low calcium concentrations

as a result of local factors.

We are grateful for the advice of Professor J L Reid and Dr J

Frater, Department of Biochemistry, and Department of Medicine

and Therapeutics, Western Infirmary, Glasgow.

1 Young DS, Pestaner LC, Gibberman V. Effects of drugs on

clinical laboratory tests. Clin Chem 1975;21:1-432D.

2 Winston S. Collection and preservation of specimens. Stand

Methods Clin Chem 1965;5:1-17.

3 Calam RR, MH. Recommended 'Order of Draw'

for collecting blood specimens into additive containing

tubes. Clin Chem 1982;28:1399.

1020