Clinical Biochemistry 47 (2014) 92–95
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
The stability of ethanol in unstoppered tubes Andrea Saracevic ⁎, Ana-Maria Simundic, Lora Dukic University Department of Chemistry, University Hospital Centre “Sestre Milosrdnice”, Zagreb, Croatia
a r t i c l e
i n f o
Article history: Received 29 May 2013 Received in revised form 1 October 2013 Accepted 7 November 2013 Available online 15 November 2013 Keywords: Ethanol Storage conditions Sample stability Preanalytical phase
a b s t r a c t Objectives: The exact time frame within which ethanol can be reliably measured in unstoppered tubes is not known. The aim of this study was to investigate the stability of alcohol concentration in unstoppered tubes. Design and methods: 44 samples with alcohol concentration N2.7 mmol/L were included in the study. Measurements were done on Vitros 250 analyzer with original Vitros reagents. After the initial alcohol measurement, each sample was aliquoted into two separate clean tubes (1 mL). One of the aliquoted tubes was stoppered immediately after aliquoting and remained stoppered during the experiment; while the other two tubes (original sample tube and the second aliquoted tube) remained open. During the experiment all three tubes were kept at room temperature. Alcohol concentration was measured at 30 minutes, 1, 2 and 3 hours after the initial measurement in all 3 tubes. The differences between the time intervals for each test tubes were examined using repeated measures Anova or Friedman test. The deviation from the initial concentration was calculated for all three test tubes for each time interval. The calculated deviations were compared with desirable imprecision specifications (DI) according to the RiliBÄK (DI b 9%). Results: We found a statistically significant difference between the initial concentration and the concentration in unstoppered tubes for all the investigated time intervals; however, the DI was exceeded only in the original tube and in the tube B, 3 hours after the initial measurement (−9.2% and −12.6%, respectively). Conclusions: Alcohol concentration can be accurately measured in the unstoppered samples within two hours upon decapping the tube, when stored at room temperature. Longer storage time (N 2 hours) in the unstoppered samples introduces significant bias in alcohol concentration. © 2013 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Alcohol concentration measurement is a frequently requested laboratory analysis for diagnostics, therapeutic monitoring or other purposes [1]. In order to assure an accurate total testing process, besides having a standardized analytical procedure, it is important to know the pre-analytical variables that might influence the results, such as sample type or storage conditions. So far several studies have investigated the stability of alcohol concentration in samples from living human subjects within days from blood collection. In 1983 Winek and Paul [2] analyzed the stability of alcohol in whole blood samples stored in sealed tubes at room temperature (22–29 °C) or in a refrigerator (0–3 °C) for 14 days and found that the stability of ethanol did not change significantly during their experiment. In addition, Penetar and colleagues [3] also investigated the stability of ethanol concentrations. They used plasma samples (EDTA and potassium oxalate/sodium fluoride), serum samples (plain serum and serum with sodium fluoride as additive) and whole blood ⁎ Corresponding author at: University Department of Chemistry, University Hospital Centre “Sestre Milosrdnice”, Vinogradska 29, 10 000 Zagreb, Croatia. Fax: +385 1 37 68 280. E-mail addresses: [email protected] (A. Saracevic), [email protected] (A.M. Simundic), [email protected] (L. Dukic).
samples (EDTA and potassium oxalate/sodium fluoride) that were aliquoted into polystyrene tubes and stored for 10 days at room temperature (25 °C) or in a refrigerator (4 °C). Similar to the results of Winek and Paul, their findings have shown that alcohol concentration does not change significantly regardless of sample type or storage conditions. A different study explored the stability of alcohol concentration within 35 days from the collection and showed that alcohol concentration may decrease in whole blood if stored in a closed tube on 26.7, 32.2 or 37.8 °C, but not in serum or plasma samples when kept in the same conditions [4]. Alcohol concentration decreases in unstoppered samples within the first few hours from the collection and such samples are not suitable for alcohol analysis. Although, to the best of our knowledge, the evidence for the evaporation rate of alcohol from uncapped samples has not yet been provided, it is recommended to keep the tubes stoppered to avoid possible evaporation of ethanol [5]. Therefore, the aim of this study was to investigate the stability of alcohol concentration in unstoppered tubes during the first three hours after measuring the ethanol concentration and removing the stopper from the tube. Materials and methods The study was carried out in the Clinical Institute of Chemistry, University Hospital Center “SESTRE MILOSRDNICE” during February
0009-9120/$ – see front matter © 2013 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clinbiochem.2013.11.006
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95 Table 1 The results of repeated measures Anova analysis. Tube
30 minutes
1 hour
2 hours
3 hours
Original bias; P −1.06 ± 2.17; −1.52 ± 1.95; −2.60 ± 3.04; −2.82 ± 3.04; 0.0275 b0.001 b0.001 b0.001 A bias; P −0.81 ± 1.94; −0.02 ± 1.95; −0.53 ± 2.39; −0.43 ± 1.95; 0.155 1.000 1.000 1.000
93
relationship was examined using Pearson's correlation coefficient if the data were normally distributed, or using Spearman's correlation coefficient in case of non-Gaussian distributions. The statistical analysis was done using the MedCalc® programme, version 11.5.1.0 (F. Schoonjans, Belgium). P values b 0.05 were considered significant. The deviation from the initial concentration for all test tubes for each time interval was calculated using Microsoft Excel (Microsoft office 2003, Microsoft, Redmond, Washington, USA).
and March 2013. Samples with alcohol concentration N 2.17 mmol/L (N0.1 g/L) were included in the study (N = 44). Original sample tubes used in the experiment were Vacuette sample tubes (Greiner Bio-One, Kremsmünster, Austria) either with serum clot activator or lithium heparin, as additives. All alcohol measurements were carried out on Vitros 250 analyzer (Ortho Clinical Diagnostics, Johnson & Johnson Medical S.p.A., Milano, Italy) with original Vitros reagents. According to the manufacturer's instructions ethanol concentration remains stable if the tubes are kept tightly closed at room temperature for 3 hours, at 2–8 °C for one week or frozen for 6 months [6]. After the initial alcohol measurement, each sample was aliquoted into two separate clean tubes (1 mL). One of the aliquoted tubes (tube A) was stoppered immediately after aliquoting and remained stoppered during the experiment; while the other two tubes (original sample tube (tube O) and the second aliquoted tube (tube B)) remained open. Tubes made of polystyrene (PS) used for sample storage during the experiment were manufactured by Starstedt (Nümbrecht, Germany). During the experiment all three tubes were kept at room temperature. The average room temperature in our laboratory is 23.5 °C (min 22.1 °C– max 25.1 °C). The average ambient humidity is 55%. Alcohol concentration was measured at 30 minutes, 1, 2 and 3 hours after the initial measurement in all 3 tubes.
Results
Statistical analysis
Discussion
All data sets have been tested for normality using Kolmogorov– Smirnov test. The differences between the time intervals for each test tubes were examined using repeated measures Anova if the data were normally distributed, or using Friedman test in case of non-Gaussian distributions. Additionally, the deviation from the initial concentration was calculated for all three test tubes for each time interval according to the formula [(Cx − Ci)/Ci)] × 100; where Ci represents the mean value for the initial measurement and Cx the mean value of each tube and each time interval. The calculated deviations were compared with the desirable imprecision (DI) according to the RiliBÄK (desirable imprecision b 9%) [7]. In order to investigate whether the stability shows concentration dependence we examined the relationship between the alcohol concentrations in each tube for each time interval and the deviation from the initial concentration for each time interval and each tube. The
Current recommendations state that samples for alcohol measurements should be tightly closed in order to prevent evaporation into the atmosphere [5]. Despite the fact that there was a statistically significant difference between the initial concentration and the concentration in unstoppered tubes for all the investigated time intervals, the desirable bias was not exceeded during the first two hours of the experiment. A few studies have so far investigated the stability of alcohol in samples from living humans and found that alcohol concentration is stable within days if properly stored. For example, back in 1983 Winek and Paul [2] examined the stability of ethanol in whole blood samples that were taken from participants in uncontrolled conditions (variable additives and anticoagulants in sample tubes) and in controlled conditions (additives: sodium heparin or potassium oxalate with sodium fluoride). The samples were stoppered and stored at room temperature (22–29 °C; number of samples in the three studied groups taken under uncontrolled and controlled conditions with
The mean initial alcohol concentration was 32.47 mmol/L (range: 20.60–84.20 mmol/L), which is equal to 1.49 g/L (range: 0.95–3.88 g/L). The results of repeated measures Anova test are presented in Table 1. There was no statistically significant difference between the initial concentration and the concentrations in the closed tube A for the investigated time intervals. For the unstoppered tubes, a statistically significant difference was observed between the initial concentration and the measured concentrations during the entire experiment. Deviations from the initial concentration exceeded the desirable imprecision in the original tube and in the tube B, 3 hours after the initial measurement. The observed biases were − 9.2% and − 12.6% (biases found for the original tube and the tube B, respectively). Mean alcohol concentrations in the original tube and in the tube B 3 hours after the initial measurement were 28.93 and 27.63 mmol/L (1.33 and 1.27 g/L), respectively. The deviation of alcohol concentration was within the acceptable criteria for the closed tubes during the entire experiment. The decrease of alcohol concentration with time is presented in Fig. 1. The results of the correlation analysis are presented in Table 2. Overall, the stability showed poor or no concentration dependency.
Decrease of alcohol concentration with time
initial value (%)
Percentage of the
105% 100% 95%
original tube
90%
tube A
85%
tube B 80% 0
30
60
120
180
DSI criteria
Time (min) Fig. 1. Decrease of alcohol concentration with time. DSI criteria — desirable specification for imprecision according to the RiliBÄK (desirable imprecision b 9%).
94
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95
Table 2 Results of correlation analysis to test whether alcohol stability shows concentration dependence. Tube
Original
Tube A
Tube B
Time interval
30 minutes 1 hour 2 hours 3 hours 30 minutes 1 hour 2 hours 3 hours 30 minutes 1 hour 2 hours 3 hours
X
Y
Alcohol concentration (mean ± SD)/mmol/L
Bias (mean ± SD)/%
31.47 30.81 29.90 28.86 31.68 32.33 31.89 30.38 31.25 31.03 29.73 27.56
−3.1 −4.7 −7.0 −9.3 −1.7 −0.4 −1.3 −1.0 −2.8 −3.8 −7.5 −12.6
± ± ± ± ± ± ± ± ± ± ± ±
23.00 22.57 21.70 21.48 23.00 24.30 23.65 22.79 22.57 22.57 21.70 20.83
± ± ± ± ± ± ± ± ± ± ± ±
6.0 3.9 4.8 6.1 3.7 3.9 4.1 4.1 4.3 5.1 5.3 7.1
correlation coefficient, r
P
−0.131 0.197 −0.192 0.144 −0.248 0.197 −0.005 −0.085 −0.312 −0.186 −0.105 0.094
0.398⁎⁎ 0.201⁎ 0.212⁎ 0.356⁎ 0.104⁎ 0.201⁎ 0.976⁎ 0.594⁎ 0.039⁎ 0.220⁎ 0.496⁎ 0.554⁎⁎
⁎ Pearson's coefficient. ⁎⁎ Spearman's coefficient.
heparin and oxalate/fluoride as additives were 42, 5 and 5, respectively) or in a refrigerator (0–3 °C; number of samples in the three studied groups taken under uncontrolled and controlled conditions with heparin and oxalate/fluoride as additives were 30, 5 and 5, respectively) for 14 days, whereas ethanol concentration was measured on days 1, 2, 5, 7 and 14. The authors found no difference in alcohol concentrations regardless of sample type or storage conditions. About a decade later, Winek and his colleagues [4] investigated the stability of alcohol concentration in whole blood taken from individuals outside their laboratory (the authors had no control over collection or storage conditions prior sample submission) and pooled human serum samples spiked with absolute ethanol to obtain concentrations of 1.50, 2.05 and 3.20 g/L, which is equal to 32.55, 44.49 and 69.44 mmol/L, respectively (additives added to the pools: potassium oxalate or potassium oxalate with sodium fluoride) stored with stoppers during 35 days on 26.7, 32.2 and 37.8 °C. The alcohol concentration in whole blood decreased by 10–19% after 35 days of storage, while there was no significant change in serum alcohol concentration. Recently, another study, published in 2007, investigated the stability of alcohol in whole blood samples collected into tubes with EDTA or fluoride/oxalate, and urine samples [8]. The researchers have divided the samples into 5 groups (15 samples each) depending on storage conditions and the timing of alcohol measurements, as follows: group 1 — samples stored at room temperature with measurements done on days 1, 7 and 14; group 2 — samples stored at 4 °C and alcohol measured after 1–3 months of storage; group 3 — alcohol measured after more than 3 months of storage at 4 °C; group 4 — samples stored at − 20 °C and alcohol measured after 1–3 months of storage; group 5 — alcohol measured after more than 6 months of storage at −20 °C. Additional two groups of samples (one group of samples taken into EDTA tubes and one group with fluoride/oxalate tubes) were stored at 4 °C up to 3 months to compare changes in alcohol concentration between them. The authors found a significant decrease in alcohol concentration in all groups except in group 4 and no difference between whole blood samples taken with EDTA or fluoride/oxalate if stored up to 3 months at 4 °C. Interestingly, the decrease of 22.4% after storing the samples for 14 days is contrary to what Winek and Paul found back in 1983. The reason for such discrepancy in the results could be the sample number used in these studies or the methods used for ethanol measurements (gas chromatography used by Winek and Paul; Roche commercial test with alcohol dehydrogenase used by Mandic-Radic and colleagues). In addition, a year after Mandic-Radic and co-workers, Penetar, together with his colleagues [3] also examined the effect of storage conditions on alcohol concentration. They measured ethanol concentration in serum (serum without additives and with added sodium fluoride), plasma (EDTA or potassium oxalate/sodium fluoride) and whole blood samples (taken with EDTA or potassium oxalate/sodium
fluoride as additives) stored for 10 days at room temperature (25 °C) or in a refrigerator (4 °C) and found, similarly to Winek and Paul, that neither sample type or storage condition influenced alcohol concentration significantly. In a clinical setting it is important to provide accurate results for alcohol concentration especially when there is a doubt in severe alcohol consumption without clinical symptoms of abuse [9]. Nowadays analytical errors have been reduced due to technical and analytical improvements and the most of erroneous results occur from errors in the extra-analytical phase [10,11]. Since additional requests for analysis are not uncommon in a clinical laboratory testing it is important to know the stability of certain analytes in blood, plasma or serum samples [12,13]. To the best of our knowledge, the evidence for the evaporation rate of alcohol and the exact time frame within which ethanol can be reliably measured in unstoppered tubes from uncapped samples, has not yet been provided. Our study, for the first time, explores whether and for how long, unstoppered tubes, already tested for certain biochemical parameters, are acceptable for alcohol analysis. This study was conducted in a Clinical laboratory within a University Hospital Centre where alcohol is measured for diagnostics or alcohol abstinence monitoring and as such is limited to samples used for medical purposes only. Conclusion Finally, we conclude that samples are acceptable for alcohol determination within 2 hours after removing the stopper, when stored at room temperature. Beyond 2 hours there is an unacceptable bias in alcohol measurement from unstoppered aliquoted samples. Such samples should be rejected for analysis and a new sample should be requested. Acknowledgement This work was supported by the Ministry of Science, Education, and Sports, Republic of Croatia (project number: 134-1340227-0200). References [1] Frajola WJ. Blood alcohol testing in the clinical laboratory: problems and suggested remedies. Clin Chem 1993;39:377–9. [2] Winek CL, Paul LJ. Effect of short-term storage conditions on alcohol concentrations in blood from living human subjects. Clin Chem 1983;29:1959–60. [3] Penetar DM, McNeil JF, Ryan ET, Lukas SE. Comparison among plasma, serum, and whole blood ethanol concentrations: impact of storage conditions and collection tubes. J Anal Toxicol 2008;32:505–10. [4] Winek T, Winek CL, Wahba WW. The effect of storage at various temperatures on blood alcohol concentration. Forensic Sci Int 1996;78:179–85.
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95 [5] Porter WH, Moyer TP. Clinical toxicology. In: Burtis CA, Ashwood ER, editors. Tietz fundamentals of clinical chemistry. St. Louis, Missouri: Saunders Elsevier; 2001. p. 642. [6] Instructions for use; Vitros ALC slides. Version 1.0, Pub. No. MP2-110 Ortho-Clinical Diagnostics, Inc.; 2002. [7] http://www.westgard.com/rilibak.htm/ . [accessed May 10, 2013]. [8] Mandic-Radic S, Dzingalasevic G, Lukovic N. Stability of ethanol in blood and urine samples. J Med Biochem 2007;26:241–4. [9] Okruhlica L, Slezakova S. Clinical signs of alcohol intoxication and importance of blood alcohol concentration testing in alcohol dependence. Bratisl Lek Listy 2013;114:136–9.
95
[10] Simundic AM, Lippi G. Preanalytical phase — a continuous challenge for laboratory professionals. Biochem Med 2012;22:145–9. [11] Lippi G, Becan-McBride K, Behúlová D, Bowen RA, Church S, Delanghe J, et al. Preanalytical quality improvement: in quality we trust. Clin Chem Lab Med 2013;51:229–41. [12] Cuhadar S, Koseoglu M, Atay A, Dirican A. The effect of storage time and freeze–thaw cycles on the stability of serum samples. Biochem Med 2013;23:70–7. [13] Ho B, Ho E. The most common nonconformities encountered during the assessments of medical laboratories in Hong Kong using ISO 15189 as accreditation criteria. Biochem Med 2012;22:247–57.
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
The stability of ethanol in unstoppered tubes Andrea Saracevic ⁎, Ana-Maria Simundic, Lora Dukic University Department of Chemistry, University Hospital Centre “Sestre Milosrdnice”, Zagreb, Croatia
a r t i c l e
i n f o
Article history: Received 29 May 2013 Received in revised form 1 October 2013 Accepted 7 November 2013 Available online 15 November 2013 Keywords: Ethanol Storage conditions Sample stability Preanalytical phase
a b s t r a c t Objectives: The exact time frame within which ethanol can be reliably measured in unstoppered tubes is not known. The aim of this study was to investigate the stability of alcohol concentration in unstoppered tubes. Design and methods: 44 samples with alcohol concentration N2.7 mmol/L were included in the study. Measurements were done on Vitros 250 analyzer with original Vitros reagents. After the initial alcohol measurement, each sample was aliquoted into two separate clean tubes (1 mL). One of the aliquoted tubes was stoppered immediately after aliquoting and remained stoppered during the experiment; while the other two tubes (original sample tube and the second aliquoted tube) remained open. During the experiment all three tubes were kept at room temperature. Alcohol concentration was measured at 30 minutes, 1, 2 and 3 hours after the initial measurement in all 3 tubes. The differences between the time intervals for each test tubes were examined using repeated measures Anova or Friedman test. The deviation from the initial concentration was calculated for all three test tubes for each time interval. The calculated deviations were compared with desirable imprecision specifications (DI) according to the RiliBÄK (DI b 9%). Results: We found a statistically significant difference between the initial concentration and the concentration in unstoppered tubes for all the investigated time intervals; however, the DI was exceeded only in the original tube and in the tube B, 3 hours after the initial measurement (−9.2% and −12.6%, respectively). Conclusions: Alcohol concentration can be accurately measured in the unstoppered samples within two hours upon decapping the tube, when stored at room temperature. Longer storage time (N 2 hours) in the unstoppered samples introduces significant bias in alcohol concentration. © 2013 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Alcohol concentration measurement is a frequently requested laboratory analysis for diagnostics, therapeutic monitoring or other purposes [1]. In order to assure an accurate total testing process, besides having a standardized analytical procedure, it is important to know the pre-analytical variables that might influence the results, such as sample type or storage conditions. So far several studies have investigated the stability of alcohol concentration in samples from living human subjects within days from blood collection. In 1983 Winek and Paul [2] analyzed the stability of alcohol in whole blood samples stored in sealed tubes at room temperature (22–29 °C) or in a refrigerator (0–3 °C) for 14 days and found that the stability of ethanol did not change significantly during their experiment. In addition, Penetar and colleagues [3] also investigated the stability of ethanol concentrations. They used plasma samples (EDTA and potassium oxalate/sodium fluoride), serum samples (plain serum and serum with sodium fluoride as additive) and whole blood ⁎ Corresponding author at: University Department of Chemistry, University Hospital Centre “Sestre Milosrdnice”, Vinogradska 29, 10 000 Zagreb, Croatia. Fax: +385 1 37 68 280. E-mail addresses: [email protected] (A. Saracevic), [email protected] (A.M. Simundic), [email protected] (L. Dukic).
samples (EDTA and potassium oxalate/sodium fluoride) that were aliquoted into polystyrene tubes and stored for 10 days at room temperature (25 °C) or in a refrigerator (4 °C). Similar to the results of Winek and Paul, their findings have shown that alcohol concentration does not change significantly regardless of sample type or storage conditions. A different study explored the stability of alcohol concentration within 35 days from the collection and showed that alcohol concentration may decrease in whole blood if stored in a closed tube on 26.7, 32.2 or 37.8 °C, but not in serum or plasma samples when kept in the same conditions [4]. Alcohol concentration decreases in unstoppered samples within the first few hours from the collection and such samples are not suitable for alcohol analysis. Although, to the best of our knowledge, the evidence for the evaporation rate of alcohol from uncapped samples has not yet been provided, it is recommended to keep the tubes stoppered to avoid possible evaporation of ethanol [5]. Therefore, the aim of this study was to investigate the stability of alcohol concentration in unstoppered tubes during the first three hours after measuring the ethanol concentration and removing the stopper from the tube. Materials and methods The study was carried out in the Clinical Institute of Chemistry, University Hospital Center “SESTRE MILOSRDNICE” during February
0009-9120/$ – see front matter © 2013 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clinbiochem.2013.11.006
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95 Table 1 The results of repeated measures Anova analysis. Tube
30 minutes
1 hour
2 hours
3 hours
Original bias; P −1.06 ± 2.17; −1.52 ± 1.95; −2.60 ± 3.04; −2.82 ± 3.04; 0.0275 b0.001 b0.001 b0.001 A bias; P −0.81 ± 1.94; −0.02 ± 1.95; −0.53 ± 2.39; −0.43 ± 1.95; 0.155 1.000 1.000 1.000
93
relationship was examined using Pearson's correlation coefficient if the data were normally distributed, or using Spearman's correlation coefficient in case of non-Gaussian distributions. The statistical analysis was done using the MedCalc® programme, version 11.5.1.0 (F. Schoonjans, Belgium). P values b 0.05 were considered significant. The deviation from the initial concentration for all test tubes for each time interval was calculated using Microsoft Excel (Microsoft office 2003, Microsoft, Redmond, Washington, USA).
and March 2013. Samples with alcohol concentration N 2.17 mmol/L (N0.1 g/L) were included in the study (N = 44). Original sample tubes used in the experiment were Vacuette sample tubes (Greiner Bio-One, Kremsmünster, Austria) either with serum clot activator or lithium heparin, as additives. All alcohol measurements were carried out on Vitros 250 analyzer (Ortho Clinical Diagnostics, Johnson & Johnson Medical S.p.A., Milano, Italy) with original Vitros reagents. According to the manufacturer's instructions ethanol concentration remains stable if the tubes are kept tightly closed at room temperature for 3 hours, at 2–8 °C for one week or frozen for 6 months [6]. After the initial alcohol measurement, each sample was aliquoted into two separate clean tubes (1 mL). One of the aliquoted tubes (tube A) was stoppered immediately after aliquoting and remained stoppered during the experiment; while the other two tubes (original sample tube (tube O) and the second aliquoted tube (tube B)) remained open. Tubes made of polystyrene (PS) used for sample storage during the experiment were manufactured by Starstedt (Nümbrecht, Germany). During the experiment all three tubes were kept at room temperature. The average room temperature in our laboratory is 23.5 °C (min 22.1 °C– max 25.1 °C). The average ambient humidity is 55%. Alcohol concentration was measured at 30 minutes, 1, 2 and 3 hours after the initial measurement in all 3 tubes.
Results
Statistical analysis
Discussion
All data sets have been tested for normality using Kolmogorov– Smirnov test. The differences between the time intervals for each test tubes were examined using repeated measures Anova if the data were normally distributed, or using Friedman test in case of non-Gaussian distributions. Additionally, the deviation from the initial concentration was calculated for all three test tubes for each time interval according to the formula [(Cx − Ci)/Ci)] × 100; where Ci represents the mean value for the initial measurement and Cx the mean value of each tube and each time interval. The calculated deviations were compared with the desirable imprecision (DI) according to the RiliBÄK (desirable imprecision b 9%) [7]. In order to investigate whether the stability shows concentration dependence we examined the relationship between the alcohol concentrations in each tube for each time interval and the deviation from the initial concentration for each time interval and each tube. The
Current recommendations state that samples for alcohol measurements should be tightly closed in order to prevent evaporation into the atmosphere [5]. Despite the fact that there was a statistically significant difference between the initial concentration and the concentration in unstoppered tubes for all the investigated time intervals, the desirable bias was not exceeded during the first two hours of the experiment. A few studies have so far investigated the stability of alcohol in samples from living humans and found that alcohol concentration is stable within days if properly stored. For example, back in 1983 Winek and Paul [2] examined the stability of ethanol in whole blood samples that were taken from participants in uncontrolled conditions (variable additives and anticoagulants in sample tubes) and in controlled conditions (additives: sodium heparin or potassium oxalate with sodium fluoride). The samples were stoppered and stored at room temperature (22–29 °C; number of samples in the three studied groups taken under uncontrolled and controlled conditions with
The mean initial alcohol concentration was 32.47 mmol/L (range: 20.60–84.20 mmol/L), which is equal to 1.49 g/L (range: 0.95–3.88 g/L). The results of repeated measures Anova test are presented in Table 1. There was no statistically significant difference between the initial concentration and the concentrations in the closed tube A for the investigated time intervals. For the unstoppered tubes, a statistically significant difference was observed between the initial concentration and the measured concentrations during the entire experiment. Deviations from the initial concentration exceeded the desirable imprecision in the original tube and in the tube B, 3 hours after the initial measurement. The observed biases were − 9.2% and − 12.6% (biases found for the original tube and the tube B, respectively). Mean alcohol concentrations in the original tube and in the tube B 3 hours after the initial measurement were 28.93 and 27.63 mmol/L (1.33 and 1.27 g/L), respectively. The deviation of alcohol concentration was within the acceptable criteria for the closed tubes during the entire experiment. The decrease of alcohol concentration with time is presented in Fig. 1. The results of the correlation analysis are presented in Table 2. Overall, the stability showed poor or no concentration dependency.
Decrease of alcohol concentration with time
initial value (%)
Percentage of the
105% 100% 95%
original tube
90%
tube A
85%
tube B 80% 0
30
60
120
180
DSI criteria
Time (min) Fig. 1. Decrease of alcohol concentration with time. DSI criteria — desirable specification for imprecision according to the RiliBÄK (desirable imprecision b 9%).
94
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95
Table 2 Results of correlation analysis to test whether alcohol stability shows concentration dependence. Tube
Original
Tube A
Tube B
Time interval
30 minutes 1 hour 2 hours 3 hours 30 minutes 1 hour 2 hours 3 hours 30 minutes 1 hour 2 hours 3 hours
X
Y
Alcohol concentration (mean ± SD)/mmol/L
Bias (mean ± SD)/%
31.47 30.81 29.90 28.86 31.68 32.33 31.89 30.38 31.25 31.03 29.73 27.56
−3.1 −4.7 −7.0 −9.3 −1.7 −0.4 −1.3 −1.0 −2.8 −3.8 −7.5 −12.6
± ± ± ± ± ± ± ± ± ± ± ±
23.00 22.57 21.70 21.48 23.00 24.30 23.65 22.79 22.57 22.57 21.70 20.83
± ± ± ± ± ± ± ± ± ± ± ±
6.0 3.9 4.8 6.1 3.7 3.9 4.1 4.1 4.3 5.1 5.3 7.1
correlation coefficient, r
P
−0.131 0.197 −0.192 0.144 −0.248 0.197 −0.005 −0.085 −0.312 −0.186 −0.105 0.094
0.398⁎⁎ 0.201⁎ 0.212⁎ 0.356⁎ 0.104⁎ 0.201⁎ 0.976⁎ 0.594⁎ 0.039⁎ 0.220⁎ 0.496⁎ 0.554⁎⁎
⁎ Pearson's coefficient. ⁎⁎ Spearman's coefficient.
heparin and oxalate/fluoride as additives were 42, 5 and 5, respectively) or in a refrigerator (0–3 °C; number of samples in the three studied groups taken under uncontrolled and controlled conditions with heparin and oxalate/fluoride as additives were 30, 5 and 5, respectively) for 14 days, whereas ethanol concentration was measured on days 1, 2, 5, 7 and 14. The authors found no difference in alcohol concentrations regardless of sample type or storage conditions. About a decade later, Winek and his colleagues [4] investigated the stability of alcohol concentration in whole blood taken from individuals outside their laboratory (the authors had no control over collection or storage conditions prior sample submission) and pooled human serum samples spiked with absolute ethanol to obtain concentrations of 1.50, 2.05 and 3.20 g/L, which is equal to 32.55, 44.49 and 69.44 mmol/L, respectively (additives added to the pools: potassium oxalate or potassium oxalate with sodium fluoride) stored with stoppers during 35 days on 26.7, 32.2 and 37.8 °C. The alcohol concentration in whole blood decreased by 10–19% after 35 days of storage, while there was no significant change in serum alcohol concentration. Recently, another study, published in 2007, investigated the stability of alcohol in whole blood samples collected into tubes with EDTA or fluoride/oxalate, and urine samples [8]. The researchers have divided the samples into 5 groups (15 samples each) depending on storage conditions and the timing of alcohol measurements, as follows: group 1 — samples stored at room temperature with measurements done on days 1, 7 and 14; group 2 — samples stored at 4 °C and alcohol measured after 1–3 months of storage; group 3 — alcohol measured after more than 3 months of storage at 4 °C; group 4 — samples stored at − 20 °C and alcohol measured after 1–3 months of storage; group 5 — alcohol measured after more than 6 months of storage at −20 °C. Additional two groups of samples (one group of samples taken into EDTA tubes and one group with fluoride/oxalate tubes) were stored at 4 °C up to 3 months to compare changes in alcohol concentration between them. The authors found a significant decrease in alcohol concentration in all groups except in group 4 and no difference between whole blood samples taken with EDTA or fluoride/oxalate if stored up to 3 months at 4 °C. Interestingly, the decrease of 22.4% after storing the samples for 14 days is contrary to what Winek and Paul found back in 1983. The reason for such discrepancy in the results could be the sample number used in these studies or the methods used for ethanol measurements (gas chromatography used by Winek and Paul; Roche commercial test with alcohol dehydrogenase used by Mandic-Radic and colleagues). In addition, a year after Mandic-Radic and co-workers, Penetar, together with his colleagues [3] also examined the effect of storage conditions on alcohol concentration. They measured ethanol concentration in serum (serum without additives and with added sodium fluoride), plasma (EDTA or potassium oxalate/sodium fluoride) and whole blood samples (taken with EDTA or potassium oxalate/sodium
fluoride as additives) stored for 10 days at room temperature (25 °C) or in a refrigerator (4 °C) and found, similarly to Winek and Paul, that neither sample type or storage condition influenced alcohol concentration significantly. In a clinical setting it is important to provide accurate results for alcohol concentration especially when there is a doubt in severe alcohol consumption without clinical symptoms of abuse [9]. Nowadays analytical errors have been reduced due to technical and analytical improvements and the most of erroneous results occur from errors in the extra-analytical phase [10,11]. Since additional requests for analysis are not uncommon in a clinical laboratory testing it is important to know the stability of certain analytes in blood, plasma or serum samples [12,13]. To the best of our knowledge, the evidence for the evaporation rate of alcohol and the exact time frame within which ethanol can be reliably measured in unstoppered tubes from uncapped samples, has not yet been provided. Our study, for the first time, explores whether and for how long, unstoppered tubes, already tested for certain biochemical parameters, are acceptable for alcohol analysis. This study was conducted in a Clinical laboratory within a University Hospital Centre where alcohol is measured for diagnostics or alcohol abstinence monitoring and as such is limited to samples used for medical purposes only. Conclusion Finally, we conclude that samples are acceptable for alcohol determination within 2 hours after removing the stopper, when stored at room temperature. Beyond 2 hours there is an unacceptable bias in alcohol measurement from unstoppered aliquoted samples. Such samples should be rejected for analysis and a new sample should be requested. Acknowledgement This work was supported by the Ministry of Science, Education, and Sports, Republic of Croatia (project number: 134-1340227-0200). References [1] Frajola WJ. Blood alcohol testing in the clinical laboratory: problems and suggested remedies. Clin Chem 1993;39:377–9. [2] Winek CL, Paul LJ. Effect of short-term storage conditions on alcohol concentrations in blood from living human subjects. Clin Chem 1983;29:1959–60. [3] Penetar DM, McNeil JF, Ryan ET, Lukas SE. Comparison among plasma, serum, and whole blood ethanol concentrations: impact of storage conditions and collection tubes. J Anal Toxicol 2008;32:505–10. [4] Winek T, Winek CL, Wahba WW. The effect of storage at various temperatures on blood alcohol concentration. Forensic Sci Int 1996;78:179–85.
A. Saracevic et al. / Clinical Biochemistry 47 (2014) 92–95 [5] Porter WH, Moyer TP. Clinical toxicology. In: Burtis CA, Ashwood ER, editors. Tietz fundamentals of clinical chemistry. St. Louis, Missouri: Saunders Elsevier; 2001. p. 642. [6] Instructions for use; Vitros ALC slides. Version 1.0, Pub. No. MP2-110 Ortho-Clinical Diagnostics, Inc.; 2002. [7] http://www.westgard.com/rilibak.htm/ . [accessed May 10, 2013]. [8] Mandic-Radic S, Dzingalasevic G, Lukovic N. Stability of ethanol in blood and urine samples. J Med Biochem 2007;26:241–4. [9] Okruhlica L, Slezakova S. Clinical signs of alcohol intoxication and importance of blood alcohol concentration testing in alcohol dependence. Bratisl Lek Listy 2013;114:136–9.
95
[10] Simundic AM, Lippi G. Preanalytical phase — a continuous challenge for laboratory professionals. Biochem Med 2012;22:145–9. [11] Lippi G, Becan-McBride K, Behúlová D, Bowen RA, Church S, Delanghe J, et al. Preanalytical quality improvement: in quality we trust. Clin Chem Lab Med 2013;51:229–41. [12] Cuhadar S, Koseoglu M, Atay A, Dirican A. The effect of storage time and freeze–thaw cycles on the stability of serum samples. Biochem Med 2013;23:70–7. [13] Ho B, Ho E. The most common nonconformities encountered during the assessments of medical laboratories in Hong Kong using ISO 15189 as accreditation criteria. Biochem Med 2012;22:247–57.
