|Year : 2020 | Volume
| Issue : 2 | Page : 89-93
Comparison of two methods for the measurement of serum chloride
M Keerthi Thej, Aparna R Bitla
Department of Biochemistry, Sri Venkateswara Institute of Medical Sciences, Tirupati, Andhra Pradesh, India
|Date of Submission||18-Jan-2020|
|Date of Acceptance||02-Mar-2020|
|Date of Web Publication||4-Aug-2020|
Aparna R Bitla
Professor, Department of Biochemistry, Sri Venkateswara Institute of Medical Sciences, Tirupati 517 507, Andhra Pradesh
Source of Support: None, Conflict of Interest: None
Background: Discrepancies in electrolyte concentration, namely sodium and potassium using the direct and indirect ion selective electrode (ISE) methods have been described. The present study was taken up to compare serum chloride values obtained using the direct and indirect ISE methods.
Methods: Serum chloride measurement was done in residual samples received in the clinical laboratory using direct ISE method and indirect ISE method. Imprecision, accuracy and comparability of the methods were studied as per the Clinical and Laboratory Standards Institute EPA guidelines. The American Clinical Laboratory Improvement Amendments (CLIA) guideline of acceptability of target ±5% was considered as the quality goal.
Results: Both the methods had acceptable imprecision and inaccuracy of <5%. Good agreement between the two methods was seen as assessed using Bland–Altman plot and Passing–Bablok regression analysis with a correlation coefficient of 0.9838. A significant difference in serum chloride levels measured in patient samples covering the entire analytical range was observed (P < 0.001), with the results of serum chloride by indirect ISE being lower compared to the direct ISE method. The difference between the two methods was below the target ±5% and thus met the CLIA guidelines.
Conclusions: The findings of the present study show that serum chloride estimation using direct ISE and indirect ISE methods are comparable.
Keywords: Chloride, direct ion-selective electrode, indirect ion-selective electrode
|How to cite this article:|
Thej M K, Bitla AR. Comparison of two methods for the measurement of serum chloride. J Clin Sci Res 2020;9:89-93
| Introduction|| |
Chloride (Cl−) is the most common anion in the extracellular fluid. It represents 70% of the total anion content. It is involved in many physiological functions including the maintenance of osmotic pressure, electron neutrality and muscular activity. Normal concentration of serum chloride (Cl−) in humans ranges between 98 and 110 mmol/L.
Serum or plasma chloride concentrations are mainly measured by ion-selective electrode (ISE)-based potentiometric methods. Measurement of electrolytes by this method is simple, quick, inexpensive and accurate.
Two methods have been described for the measurement of serum electrolytes by ISE: direct ion-selective electrode method and indirect ion-selective electrode method. In the direct ion-selective electrode method, the specimen is bought to electrolyte surface without any dilution and measures the electrolyte activity in the activity in the plasma water (mmol/kg H2O), which in turn, is converted to the readout concentration by a fixed (ion-specific) multiplier. This method is the principle behind the working of point of care analysers for rapid determination of electrolytes. In the indirect ISE method, samples undergo a 20–30-fold dilution with high ionic activity buffers as per the analytical system used. The high ionic activity buffers ensure that the activity coefficient virtually remains constant for different samples. Autoanalysers principally use the indirect ISE method since it requires very low sample volume.
A few studies have shown discrepancies in electrolyte concentration, namely sodium and potassium using the direct and indirect ISE methods.,, There is paucity of data on chloride levels measured using these two methods. The present study was thus taken up to compare serum chloride values obtained using direct and indirect ISE methods.
| Material and Methods|| |
The study was conducted in the Department of Biochemistry, Sri Venkateswara Institute of Medical Sciences, Tirupati, Andhra Pradesh, India. The study was conducted after obtaining approval from the Institutional Ethics Committee. Thirty samples received in the clinical laboratory for biochemical analysis were used after analysis of the requested parameters. Blood samples were allowed to stand for 30 min for clotting. Serum was separated by centrifugation at 2000 rpm for 15 minutes. The separated serum was used for the measurement of serum chloride using the two methods on the same day.
Serum chloride measurement using direct ion-selective electrode method was done using ExDs analyser (Kanagawa, Japan) and indirect ion-selective electrode method in Beckman Coulter AU480 analyser (California, USA).
In the direct ion-selective electrode method, that is, ExDs analyser; the specimen is bought to electrolyte surface without any dilution, and the activity of relative ion is measured in the given sample. While in the indirect ISE methods, the sample is diluted with high ionic activity buffers in ratios of 1:31 (sample 20 μL, reagent 618 μL).
The method evaluation was conducted as per the Clinical and Laboratory Standards Institute EPA guidelines.
Two-level assayed human sera control material from Randox laboratories (Randox, BT29 4QY, United Kingdom) was used to test the between day imprecision (CVbd) and for calculation of inaccuracy (bias) for both the methods. Serum chloride was measured in these control material daily for 20 days using both the methods. The mean, standard deviation (SD) and coefficient of variation (%) were calculated for each of the decision levels measured. CVbd is considered as a measure of random analytical error, while bias (inaccuracy) is considered as a measure of systematic error. Bias was calculated using the following formula:
Bias (%) = ([mean value obtained − target value)/target value) × 100%
Target value was obtained from the product insert provided by the manufacturers (Randox, BT29 4QY, UK).
To measure within-day imprecision (CVwd), two different pooled sera were collected from the samples remaining after analysis of the requested parameters at two decision levels, 95 and 110 mmol/L. Each pooled sample was aliqouted into 20 analyser sample cups and used for the measurement of serum chloride using both the methods. The mean, SD and coefficient of variation (%) were calculated for each of the decision levels measured.
Serum chloride measurement done using direct ion-selective electrode method on ExDs analyser (Kanagawa, Japan) was compared with serum chloride measurement done using indirect ion-selective electrode method on Beckman Coulter AU480 analyser (California, USA). Agreement between the methods was analysed using appropriate statistical methods. The American Clinical Laboratory Improvement Amendments (CLIA) guideline of acceptability of target ±5% was considered as the quality goal.
Continuous variables are expressed as mean ± SD. Paired measurements using the two methods were assessed using paired samples t-test. Correlation between the methods was expressed using Pearson's correlation analysis. Agreement between the two estimations was tested using Bland–Altman plots. Passing–Bablok regression analysis was performed to obtain the regression equation for the two methods and 95% confidence intervals (CI) for the intercept and slope. Methods are considered to be comparable if the 95% CI for intercept includes 0 and 95% CI for slope includes 1. Lack of zero in 95% CI for intercept is an indicator of constant systematic error, while lack of one in 95%CI for slope is an indicator of proportional systematic error. All statistical analyses were performed by using statistical software MedCalc (Version 12.1, Ostend, Belgium) and Microsoft Excel (Microsoft, Redmond, WA, USA). A P < 0.05 was considered statistically significant.
| Results|| |
The performance characteristics of the two methods are shown in [Table 1]. The imprecision for both the methods (both between day and within day) was below 5%. Bias of the two methods at the two decision limits studied using commercial control material was also found to be <5%.
|Table 1: Performance of direct and indirect ion.selective electrode methods|
Click here to view
[Figure 1] shows the scatter plots representing paired measurements of serum chloride in patient samples using direct ion-selective electrode method and indirect ion-selective electrode method. A significant difference in serum chloride levels measured in patient samples covering the entire analytical range was observed (P < 0.001). Mean values for the samples measured using indirect ISE method were lower (104.3 ± 7.35 mmol/L) compared to paired measurement using the direct ISE method (108.87 ± 7.65 mmol/L). The mean difference between the two methods was 4.57 lower for indirect ISE method with 95% CI of − 5.0826 to −4.0507.
|Figure 1: Scatter plot representing serum chloride measured using direct ISE method and indirect ISE method (n = 30). ISE = Ion-selective electrode|
Click here to view
A good correlation was observed between the two methods with a correlation coefficient of 0.9838. Bland–Altman analysis for indirect ISE method result minus the direct ISE result had limits of agreement of −7.3 to −1.9 mmol/L ([Figure 2]). All the difference fell within the 1.96 SD of the mean value showing good agreement between the two methods.
|Figure 2: Bland–Altman plot showing the agreement between the direct ISE and indirect ISE method for serum chloride levels in patient sample (n = 30); the mean of the two methods (X-axis) is plotted against the difference between the two serum chloride measurements using the two methods (Y-axis). The solid horizontal line represents the overall mean of the differences (bias = -4.6 mmol/L), and the dashed lines show the range containing the mean of the differences ±1.96 standard deviations, which represent the limits of agreement (-7.3 mmol/L and - 1.9 mmol/L). ISE = Ion-selective electrode|
Click here to view
Passing–Bablok regression analysis is shown in [Figure 3]. The regression equation showed an intercept of − 5.00 with 95% CI intervals for intercept being − 5.0000 to 5.0909. The slope was 1.00 with 95% CI intervals for slope being 0.9091–1.0000. This also indicates good agreement between the methods since 95% CI for regression line intercept includes the value 0 and 95% CI for slope includes value 1.
|Figure 3: Passing–Bablok regression of the direct and indirect ISE methods for serum chloride measurement in patient samples, n = 30; concentration range = 88–118 mmol/L. The solid line represents the regression equation (y = ax + b; where A-regression line’s intercept. B-regression line’s slope) with its 95% confidence intervals (dotted line). The thin solid line represents the identity line consistent with correlation between the two methods (Correlation coefficient = 0.981; P < 0.001). ISE = Ion-selective electrode|
Click here to view
Bias was also calculated from the regression equation at three levels: 95.0 mmol/L, 100 mmol/L and 110 mmol/L. [Table 2] shows the bias at these three levels. The bias was more than 5% at 90 mmol/L and 100 mmol/L and <5% at 110 mmol/L.
|Table 2: Bias estimation at decision levels from Passing-Bablok regression analysis|
Click here to view
| Discussion|| |
The performance of both the methods was found to be acceptable using the CLIA recommendations of target ± 5%. Both the methods were also found to be comparable in the present study as evident from the good correlation coefficient and results of the Bland–Altman plots as well as Passing–Bablok regression analysis.
The analytical goals differ across countries with the different professional bodies recommending their own quality goal. For chloride, the Royal College of Pathologists of Australasia Quality Assurance Programs, Australia, recommends a quality goal of 3 mmol/L at chloride concentration of ± 100 mmol/L and an analytical goal of 3% for concentrations over 100 mmol/L. Richtlinien der Bundesärztekammer (RiliBAK) guidelines of the German Medical Association for the quality assurance of laboratory medical examinations recommend a quality goal of 8%. Hence, the criteria is stricter when the Australian guideline is implemented. The widely used CLIA guidelines were used in the present study.
The findings of the present study are in line with the previous studies, which have shown lower electrolyte levels using the indirect ISE method compared to the direct ISE method.,, In an earlier study a similar significant difference in plasma chloride as well as plasma sodium level measurements from the central laboratory and the intensive care unit (ICU) blood gas analyser based on the indirect ISE and direct ISE, was reported. The results obtained using these two methods were found to be significantly different, as was observed in the present study. The difference was attributed to the lower albumin levels observed in ICU patients. Similarly, lower sodium and chloride levels were reported using the indirect ISE method in samples with high glucose concentration.
Results of indirect ISE method are known to be dependent on the solid content in the sample and assume plasma water content of 93% with the remaining being the solid content, that is, proteins and lipids. Hence, in conditions with higher solid content, the ion concentration measured can be lower. This effect is known as 'electrolyte exclusion.' However, the effect of these on serum chloride levels in the present study was not evaluated.
The observed bias between the two methods although acceptable can be crucial in decision-making. Hence, the caution is recommended while interpreting patient results, especially in the critical care setting. Laboratories having analysers working on both these methods should exert caution when reporting patient values. Serial measurements for a patient should preferably be performed using the same equipment so as to avoid confusion in interpreting the results obtained using the two methods.
Thus, the findings of the present study show that serum chloride estimation using direct ISE and indirect ISE methods are comparable. However, the caution is recommended in the interpretation of serum chloride levels in the acute care setting.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Berend K, van Hulsteijn LH, Gans RO. Chloride: The queen of electrolytes? Eur J Intern Med 2012;23:203-11.
Tietz NW, Pruden EL, Siggaard-Andersen O. Electrolytes. In: Burtis CA, Ashwood ER, editors. Tietz Fundamental of Clinical chemistry. 4th
ed. Philadelphia: WB Saunders; 2006. p. 95.
Levy GB. Determination of sodium with ion-selective electrodes. Clin Chem 1981;27:1435-8.
Stove V, Slabbinck A, Vanoverschelde L, Hoste E, De Paepe P, Delanghe J. How to solve the underestimated problem of overestimated sodium results in the hypoproteinemic patient. Crit Care Med 2016;44:e83-8.
Story DA, Morimatsu H, Egi M, Bellomo R. The effect of albumin concentration on plasma sodium and chloride measurements in critically ill patients. Anesth Analg 2007;104:893-7.
Clinical and Laboratory Standards Institute. Method Procedure Comparison and Bias Estimation Using Patient Samples; Approved Guideline. CLSI Document EP9-A3. 3rd
ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2013.
Westgard JO, Carey RN, Wold S. Criteria for judging precision and accuracy in method development and evaluation. Clin Chem 1974;20:825-33.
Bilić-Zulle L. Comparison of methods: Passing and Bablok regression. Biochem Med (Zagreb) 2011;21:49-52.
Al-Musheifri A, Jones GR. Glucose interference in direct ion-sensitive electrode sodium measurements. Ann Clin Biochem 2008;45:530-2.
Spasovski G, Vanholder R, Allolio B, Annane D, Ball S, Bichet D, et al
. Clinical practice guideline on diagnosis and treatment of hyponatraemia. Eur J Endocrinol 2014;170:G1-47.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]