Original article
J Clin Pathol: first published as 10.1136/jclinpath-2014-202693 on 19 January 2015. Downloaded from http://jcp.bmj.com/ on June 7, 2023 at Gerstein Science Information Centre Serials
▸ Additional material is
published online only. To view
please visit the journal online
(http://dx.doi.org/10.1136/
jclinpath-2014-202693).
University Department of
Chemistry, Medical School
University Hospital Sestre
milosrdnice, Zagreb, Croatia
Correspondence to
Lora Dukic, University
Department of Chemistry,
Medical School University
Hospital Sestre milosrdnice,
Vinogradska c 29, Zagreb
10 000, Croatia;
lora.dukic@gmail.com
Received 25 September 2014
Revised 3 December 2014
Accepted 15 December 2014
Published Online First
19 January 2015
To cite: Dukic L,
Simundic A-M. J Clin Pathol
2015;68:288–291.
288
Short-term and long-term storage stability
of heparin plasma ammonia
Lora Dukic, Ana-Maria Simundic
ABSTRACT
Aims Ammonia is an extremely unstable analyte and
requires special attention during sampling, transport and
storage. The aim of this study was to evaluate the
stability of ammonia in lithium-heparin plasma during
short-term (at +4°C) and long-term (at −20°C) storage.
Methods Twenty plasma samples were used for short-
term stability assessment. Each sample was divided into
five aliquots and stored in stoppered tubes at +4°C, for
1, 2, 3, 4 and 24 h from initial testing. Fifteen plasma
samples were used for long-term stability assessment.
Each sample was divided into eight aliquots and stored
in stoppered tubes at −20°C for 3, 24, 48 h and 1, 2,
4, 8 and 12 weeks from initial testing. Ammonia
concentration was determined on a Beckman Coulter
AU2700 chemistry analyser using Randox ammonia
enzymatic UV method. Bias was calculated from initial
value for each time point and compared with quality
specifications defined by Royal College of Pathologists of
Australasia.
Results The average bias exceeded the total allowable
error after storage of samples for 1 h at +4°C and 3 h
at −20°C.
Conclusion Ammonia is not stable during storage at
+4°C and −20°C in lithium-heparinised plasma and
should therefore be analysed immediately.
INTRODUCTION
Ammonia determination is clinically significant in
various pathological states such as in patients with
symptoms of neuromuscular and cerebral distur-
bances caused by hepatopathy, oncology patients
receiving aggressive chemotherapy and patients
taking the antiepileptic drug valproic acid.1 2 In neo-
nates and children, ammonia is often requested when
metabolic disturbance is suspected.2 Guidelines for
detection and management of hyperammonaemia
suggest that levels of ammonia up to 200 μmol/L are
associated with acquired conditions such as sepsis,
chemotherapy or liver dysfunction, while ammonia
levels higher than 200 μmol/L indicate metabolic
disorders.3
Ammonia is an extremely unstable analyte and
requires special attention during sampling, trans-
port and storage. It has been recommended that
heparinised or EDTA plasma samples be kept in an
ice bath after sampling, centrifuged immediately on
arrival at the laboratory, aliquoted and analysed
within 15–30 min, for accurate analysis of
ammonia.4 Guidelines for the investigation of
hyperammonaemia even recommend repeated
ammonia sampling to exclude possible artefactual
cause of high ammonia values.3 Unfortunately,
ammonia testing is not universally available and
delayed analysis due to prolonged sample transport
Dukic L, et al. J Clin Pathol 2015;68:288–291. doi:10.1136/jclinpath-2014-202693
time may sometimes be carried out. However,
existing original reports about short-term and long-
term ammonia stability are inconsistent. According
to the manufacturer of our assay (Randox
Laboratories Ltd, Crumlin, UK), if plasma is sepa-
rated within 30 min, ammonia is stable for 2 h at
+2 to +8°C.5 Some authors even recommend
long-term plasma storage at −70°C.6
The aim of our study was to verify the manufac-
turer’s declaration by assessing the stability of
ammonia in lithium-heparin plasma. Stability was
assessed during standard storage conditions in our
laboratory for short-term (at +4°C) and long-term
(at −20°C) sample storage.
MATERIALS AND METHODS
From December 2013 until May 2014, routine
Section. Protected by copyright.
patient samples for which ammonia testing was
requested were collected. Sampling was performed in
4.5 mL Lithium Heparin plasma tubes (Greiner
Bio-One GmbH, Kremsmünster, Austria) in a single
venipuncture without stasis. On sampling, whole
blood was transported on ice immediately to the bio-
chemistry laboratory. On delivery to the laboratory,
the specimen was centrifuged for 10 min at 1800g in
a benchtop centrifuge (Rotofix 32, Hettich,
Tuttlingen, Germany). Specimens with visible haem-
olysis, icteria and lipaemia were not included in the
study. Plasma was immediately separated from cells
and 300 μL aliquot was processed for ammonia ana-
lysis on a Beckman Coulter AU2700 chemistry ana-
lyser (Beckman Coulter, Tokyo, Japan) using Randox
ammonia enzymatic UV method (Randox
Laboratories Ltd, Crumlin, UK). Daily quality
control was performed using Ammonia/Ethanol
Control Level 1 and 2 (Randox Laboratories Ltd,
Crumlin, UK). Within-run precision coefficients of
variation declared by the manufacturer were 3.6%,
2.5% and 2.8% for the ammonia concentrations
57.2, 167.0 and 305.0 μmol/L, respectively. Our own
laboratory data for precision on patient samples are:
within-run precision 14.4% and 3.7% at concentra-
tions 24.3 and 102.5 μmol/L, respectively.
Following initial testing, remaining plasma was
immediately divided in several portions of 300 μL,
according to the protocol described below.
Twenty plasma samples were used for short-term
stability assessment. Each sample was divided in
five aliquots and stored in stoppered tubes at +4°C,
for 1, 2, 3, 4 and 24 h from initial testing.
Fifteen plasma samples were used for long-term
stability assessment. Each sample was divided in
eight aliquots and stored in stoppered tubes at
−20°C for 3, 24, 48 h and 1, 2, 4, 8 and 12 weeks
from initial testing. After thawing, plasma

Figure 1 The increase of ammonia concentrations during short-term
storage of lithium-heparin plasma aliquots in stoppered tubes at +4°C
(N=20).
specimens were visually inspected and centrifuged for 10 min
on 1800g, if necessary.
All analyses in this study were carried out in triplicate. For
this study, only residual patient samples were used. The study
was approved by the Hospital Ethical Committee.
Statistical analysis
For statistical analysis, median of triplicate measurements was
used, as recommended by CLSI.7 For each measurement, bias
from initial value was calculated according to the following
formula:
Bias ¼ median at specified time point  median at initial value
Bias was expressed in absolute values and in percentages.
Calculations were made with Microsoft Excel 2003 (Microsoft
Corporation, Washington, USA). Measurements were expressed
as median and IQR. Median differences between serial measure-
ments were compared by non-parametric Friedman test for
serial measurements. p Value <0.05 was used as a significance
level. Statistical analysis was carried out using the MedCalc stat-
istical software V.13.1.2.0 (MedCalc Software, Mariakerke,
Belgium).
Average bias values were compared with specifications for
allowable error from RCPA (Royal College of Pathologists of
Australasia).8 For ammonia values ≤50 μmol/L, defined allowable
error was considered ±5 μmol/L and for values ≥50 μmol/L
allowable bias was ±20%.
RESULTS
The median initial concentrations of ammonia of plasma samples
used in short-term (at +4°C) stability assessment (N=20) was
Table 1 Short-term stability of ammonia in lithium-heparin plasma aliquots during storage for 1, 2, 3, 4 and 24 h at +4°C in stoppered tubes
(N=20)
Initial value 1h
Ammonia concentration (μmol/L) 36.3 (28.5 to 67.8) 42.5 (37.3 to 72.4)
median and IQR
Bias (μmol/L) median and 95% CI / 4.6*(3.8 to 7.1)
Bias (%) / 18
*All data points that did not meet RCPA (Royal College of Pathologists of Australasia) criteria.
Dukic L, et al. J Clin Pathol 2015;68:288–291. doi:10.1136/jclinpath-2014-202693
Original article
J Clin Pathol: first published as 10.1136/jclinpath-2014-202693 on 19 January 2015. Downloaded from http://jcp.bmj.com/ on June 7, 2023 at Gerstein Science Information Centre Serials
36.3 (IQR 28.5–67.8) μmol/L. For plasma samples used for long-
term stability assessment (at −20°C; N=15), the median of initial
concentration was 18.7 (IQR 16.9–25.4) μmol/L.
Figure 1 and table 1 show the median ammonia concentra-
tions and respective bias (absolute and relative) in samples
stored at +4°C during different time points. After 1 h of plasma
storage at +4°C, observed bias had already exceeded RCPA cri-
teria for total allowable error (CI for estimated bias exceeded
the limit of acceptance). There was a gradual increase in
ammonia concentration during the first 4 h of storage, followed
by almost twofold increase of initial concentration of ammonia
after 24 h of storage. Median concentrations between serial
measurements in samples stored at +4°C differed significantly
(p<0.001).
Figure 2 and table 2 contain data for samples stored at −20°C
followed over 12 weeks, at different time points. There was a
statistically significant difference between median concentrations
of ammonia across different time intervals during long-term
storage at −20°C ( p<0.001). The median bias had already
exceeded criteria for total allowable error after 3 h of storage at
−20°C after initial testing.
DISCUSSION
Our study shows that measurement of ammonia in aliquoted
and stoppered plasma samples stored for only 1 h in a refriger-
ator at +4°C and stored for 3 h at −20°C results in unacceptable
Section. Protected by copyright.
bias. Therefore, neither short-term nor long-term storage of
heparin plasma at +4°C and −20°C, respectively, is possible for
reliable ammonia determination.
Stability under various storage conditions has been previously
studied by several authors. Howanitz et al6 have studied the effect
of storage of sodium heparin plasma at +4°C, +22°C, −20°C and
−70°C. The observed increases of ammonia were much smaller
than in our study. Ammonia increased only 5.6% from the initial
value in plasma stored for 1 h at 4°C (in our study the increase was
18%). After 24 h of storage at −20°C the increase observed by
Howanitz et al was 18%, whereas we already observed an increase
of 39% after 3 h of storage at −20°C. However, their study was
carried out on healthy volunteers who have generally lower con-
centrations of ammonia, whereas we studied short-term ammonia
stability on hospital inpatients with various diseases, whose initial
concentrations were substantially higher. Moreover, Howanitz
et al studied samples stored in glass tubes and we used plastic
tubes. It has been reported that differences in blood collection
tube material and tube components may affect sample quality.9
De Fonseca-Wollheim studied the rate of increase of ammonia
concentration in dipotassium EDTA plasma stored at 0°C, +20°
C and +37°C at six different time points between 0 and
90 min. The mean ammonia increase in that study was 1.3-fold
at +20°C and 6.5-fold at +37°C in relation to ammonia con-
centration measured in a sample stored at 0°C, as reference.
This study also assessed the stability of ammonia at −38°C.
2h 3h 4h 24 h
43.0 (35.4 to 74.4) 49.2 (36.3 to 77.3) 46.8 (33.5 to 78.2) 71.7 (46.4 to 104.8)
8.2*(4.1 to 8.9) 8.4*(6.3 to 10.2) 9.3*(7.1 to 13.3) 24.6*(21.1 to 43.6)
20 25 26 78
289
 Original article

J Clin Pathol: first published as 10.1136/jclinpath-2014-202693 on 19 January 2015. Downloaded from http://jcp.bmj.com/ on June 7, 2023 at Gerstein Science Information Centre Serials
36.2* (22.1 to 41.5)
51.2 (42.3 to 57.5)
12 weeks

220
25.0* (19.7 to 31.8)
39.2 (35.5 to 56.9)
Long-term stability of ammonia in lithium-heparin plasma aliquots during storage for 3, 24 and 48 h and 1, 2, 4, 8 and 12 weeks at −20°C in stoppered tubes (N=15)
8 weeks

171
21.1* (12.1 to 25.6)
39.1 (31.4 to 46.1)
Figure 2 The increase of ammonia concentrations during long-term
storage of lithium-heparin plasma aliquots in stoppered tubes at −20°C

4 weeks
(N=15).

136
14.0* (12.3 to 21.8)
37.3 (27.7 to 44.1)
Interestingly, the authors did not observe a significant increase
of ammonia after 24 h of storage at −38°C.10 The stability at
−20°C was, unfortunately, not studied in their work.

2 weeks
Stability of ammonia was also studied in equine EDTA

104
plasma. The increase of ammonia was significant after 24 h of

Section. Protected by copyright.

storage at +4°C, whereas ammonia was stable for 21 days if
stored at −20°C.11 It has to be pointed out that equine and

8.8* (5.7 to 13.2)

25.4 (17.6 to 25.5)
human matrices have differences, which may explain observed
differences in stability of ammonia in the specimens. Moreover,

1 week
the study assessed the stability of ammonia in EDTA plasma,
whereas in our study heparin plasma was used. This might at

59
least partially explain the observed differences.

4.4* (2.6 to 11.3)

22.3 (19.8 to 29.7)
There are several reasons for such high in vitro instability of
ammonia in human plasma. Ammonia concentration in hepari-
nised plasma is pH-dependent.12 Substantial changes in pCO2
caused by inappropriate sampling and storage conditions cause
48 h

changes in the rate of ammonia formation in heparinised blood

46
specimens.13 When laboratory glassware was used to test
19.5 (15.4 to 29.1)

ammonia, great difficulty was encountered owing to contamin-

4.2*(2.6 to 6.4)

*All data points that did not meet RCPA (Royal College of Pathologists of Australasia) criteria.
ation by ammonia-containing detergents. Lots of issues related
to specimen collection and determination of ammonia were
resolved by introduction of more sophisticated specimen collec-
24 h

25

tion devices and disposables, however, special precautionary

measures should still be respected. Exposure to cigarette smoke
23.4 (20.6 to 27.2)

5.6* (2.3 to 7.0)

pollution causes false elevated ammonia values, because

ammonia-forming compounds are added routinely to tobacco in
order to enhance flavour and maximise the quantity of free
nicotine in the smoke.14 Moreover, ammonia is continuously
3h

39

generated in the dipotassium EDTA plasma sample as a result of

the production of ammonia caused by γ-glutamyltransferase
18.7 (16.9 to 25.4)

hydrolysis of glutamine.15
Information on adequate short-term and long-term storage
Initial value

conditions for accurate ammonia measurement is important for

routine everyday work as well as in research settings. The prea-
/

nalytical phase is recognised as the most vulnerable part of the

total testing process and it is therefore extremely important to
Bias (μmol/L) median and
(μmol/L) median and IQR
Ammonia concentration

recognise and properly manage all potential sources of variabil-

ity of laboratory test results.16 17
As already pointed out, accurate information on short-term
storage stability is necessary for laboratories that receive plasma
Table 2

Bias (%)

samples from distant laboratories where ammonia testing is not

95% CI

available. Moreover, according to ISO standard 15 189, each

laboratory is obliged to define storage protocols for different
290 Dukic L, et al. J Clin Pathol 2015;68:288–291. doi:10.1136/jclinpath-2014-202693

analytes.18 Long-term ammonia sample stability is also import-
ant in the research setting, where samples need to be stored for
longer periods of time.
Hyperammonaemia is a severe clinical condition owing to
central nervous system abnormalities such as seizures, drowsi-
ness and apnoea. In children and neonates, hyperammonaemia
(ammonia concentration >100 and 150 mmol/L, respectively) is
highly suggestive of potential inherited deficiencies of urea
cycle enzymes.19 Differential diagnosis and timely management
of patients presenting with hyperammonaemia are critical for
patient outcome.20 A false increase of ammonia may mask
other potentially dangerous pathological processes in the
central nervous system and lead to misdiagnosis, or, alterna-
tively, lead to additional diagnostic workup to investigate the
cause of the hyperammonaemia and cause delayed diagnosis.
The frequency of false positive findings of ammonia due to
preanalytical errors has been estimated to account for almost
48% of all cases of hyperammonaemia in children.21 It is
therefore absolutely necessary not to delay the analysis of
ammonia for longer than 15–30 min. All deviations from the
recommended procedure might lead to false increases of
ammonia concentration, cause diagnostic errors and jeopardise
patient safety.
Experimental data on short-term and long-term storage of
plasma ammonia differ substantially in their design and method-
ology. Various authors have studied different time points at dif-
ferent temperatures. We have explored stability of ammonia in
heparin plasma stored at conditions most commonly used in
routine laboratories (+4°C and −20°C). One possible limitation
of our study is that ammonia concentrations in wide ranges
were not included. We believe that our results may help labora-
tory professionals and researchers to define their optimal
storage protocols during sample transport as well as provide
evidence-based acceptance criteria.
In conclusion, although manufacturers may contend that
ammonia is stable for 2 h at +2 to +8°C if plasma is separated
within 30 min, we were not able to confirm this in our study.
Therefore, manufacturers’ claims on heparinised plasma storage
stability need revision. Our results indicate that ammonia
cannot be reliably measured if stored in lithium-heparinised
plasma tubes for 1 h at +4°C and for 3 h at −20°C.
Take home messages
▸ Ammonia is an extremely unstable analyte and requires
special attention during sampling, transport and storage.
▸ We have assessed the stability of ammonia in
lithium-heparin plasma during short-term (at +4°C) and
long-term (at −20°C) storage.
▸ Lithium-heparin plasma ammonia is not stable when stored
for 1 h at +4°C and for 3 h at −20°C.
Dukic L, et al. J Clin Pathol 2015;68:288–291. doi:10.1136/jclinpath-2014-202693
Original article
J Clin Pathol: first published as 10.1136/jclinpath-2014-202693 on 19 January 2015. Downloaded from http://jcp.bmj.com/ on June 7, 2023 at Gerstein Science Information Centre Serials
Handling editor Tahir S Pillay
Contributors LD as first author and A-MS as coauthor have contributed in data
collection and analysis, design and writing of the manuscript.
Funding This study was supported by the Ministry of Science, Education and
Sports, Republic of Croatia, project number: 134-1340227-0200.
Competing interests None.
Ethics approval Medical School University Hospital Sestre milosrdnice Ethical
Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1 Newman DJ, Price CP. Nonprotein nitrogen metabolites. In: Burtis CA, Ashwood ER,
eds. Tietz fundamentals of clinical chemistry. 5th edn. Philadelphia, PA: W.B.
Saunders Company, 2001:417–19.
2 Thomas L, ed. Ammonia. In: Clinical laboratory diagnostics: use and assessment of
clinical laboratory results. 1st edn. Frankfurt: TH-Books Verlagsgesellschaft,
1998:186–92.
3 UK National Metabolic Biochemistry Network. Guidelines for the investigation of
hyperammonemia. http://www.metbio.net/docs/MetBio-Guideline_
AMUP100834-21-07-2010.pdf
4 Nikolac N, Omazic J, Simundic AM. The evidence based practice for optimal sample
quality for ammonia measurement. Clin Biochem 2014;47:991–5.
5 Randox Ammonia (NH3) Enzymatic UV-Method. Package insert. Randox
Laboratories Ltd. REV 001
6 Howanitz JH, Howanitz PJ, Skrodzki CA, et al. Influences of specimen
processing and storage conditions on results for plasma ammonia. Clin Chem
1984;30:906–8.
7 CLSI. EP07-A2—Interference Testing in Clinical Chemistry; approved guideline. 2nd
Section. Protected by copyright.
edn. Wayne, PA: Clinical Laboratory Standards Institute, 2005.
8 Royal College of Pathologists of Australasia Allowable Limits of Performance.
http://www.westgard.com/rcpa-australasian-quality-requirements.htm#alcohol (accessed
1 Jun 2014).
9 Bowen RA, Remaley AT. Interferences from blood collection tube components on
clinical chemistry assays. Biochem Med (Zagreb) 2014;24:31–44.
10 da Fonseca-Wollheim F. Preanalytical increase of ammonia in blood specimens from
healthy subjects. Clin Chem 1990;36:1483–7.
11 Lindner A, Bauer S. Effect of temperature, duration of storage and sampling
procedure on ammonia concentration in equine blood plasma. Eur J Clin Chem Clin
Biochem 1993;31:473–6.
12 da Fonseca-Wollheim F. The influence of pH and various anions on the distribution
of NH+4 in human blood. Eur J Clin Chem Clin Biochem 1995;33:289–94.
13 da Fonseca-Wollheim F, Heinze KG. The influence of pCO2 on the rate of ammonia
formation in blood. Eur J Clin Chem Clin Biochem 1992;30:867–9.
14 Seeman JI, Carchman RA. The possible role of ammonia toxicity on the exposure,
deposition, retention, and the bioavailability of nicotine during smoking. Food Chem
Toxicol 2008;46:1863–81.
15 da Fonseca-Wollheim F. Deamidation of glutamine by increased plasma
γ-glutamyltransferase is a source of rapid ammonia formation in blood and plasma
specimens. Clin Chem 1990;36:1479–82.
16 Simundic AM. Preanalytical phase—an updated review of the current evidence.
Biochem Med (Zagreb) 2014;24:6.
17 Lippi G, Becan-McBride K, Behúlová D, et al. Preanalytical quality improvement:
in quality we trust. Clin Chem Lab Med 2013;51:229–41.
18 Fuentes-Arderiu X. Clinical sample stability and measurement uncertainty. Clin
Chem Lab Med 2014;52:e37–8.
19 National organization for rare disorders (NORD). The physician guide to urea cycle
disorders. Danbury, 2012.
20 Auron A, Brophy PD. Hyperammonemia in review: pathophysiology, diagnosis, and
treatment. Pediatr Nephrol 2012;27:207–22.
21 Maranda B, Cousineau J, Allard P, et al. False positives in plasma ammonia measurement
and their clinical impact in a pediatric population. Clin Biochem 2007;40:531–5.
291
Sažetak

Cilj: Amonijak je izrazito nestabilan analit koji zahtjeva posebnu pažnju tijekom uzorkovanja,
transporta i pohrane. Cilj ovog istraživanja bio je utvrditi stabilnost amonijaka u uzorku litij-
heparinizirane plazme tijekom kratkoročne (+4°C) i dugoročne pohrane (-20°C).
Metode: Za procjenu kratkoročne stabilnosti koristili smo plazmu od dvadeset ispitanika.
Svaki uzorak je razdijeljen u 5 alikvota i pohranjen u začepljene epruvete na +4°C na 1, 2, 3, 4
i 24 sata od početne analize. Za procjenu dugoročne stabilnosti korišteno je petnaest
uzoraka plazme. Svaki uzorak je razdijeljen u 8 alikvota koji su pohranjeni u začepljene
epruvete na -20°C na 3, 24 i 48 sati i 2, 4, 8 i 12 tjedana od početne analize. Koncentraciju
amonijaka smo odredili na biokemijskom analizatoru Beckman Coulter AU 2700 koristeći
Randox Ammonia enzimatsku UV metodu. Odstupanje od inicijalne koncentracije u svakoj
vremenskoj točki je uspoređeno s unaprijed definiranim kriterijima prihvatljivosti prema
Royal College of Pathologists of Australasia.
Rezultati: Prosječno odstupanje bilo je veće od kriterija prihvatljivosti prilikom pohrane
uzoraka tijekom sat vremena na +4°C i 3 sata na -20°C.
Zaključci: Amonijak nije stabilan tijekom pohrane na +4°C i -20°C u litij-hepariniziranoj plazmi
i stoga ga treba odmah analizirati.

Disclaimer: This abstract has been translated and adapted from the original English-language
content. Translated content is provided on an "as is" basis. Translation accuracy or reliability
is not guaranteed or implied. BMJ is not responsible for any errors and omissions arising
from translation to the fullest extent permitted by law, BMJ shall not incur any liability,
including without limitation, liability for damages, arising from the translated text.