Article, Biochemistry

Agreement of serum potassium measured by blood gas and biochemistry analyzer in patients with moderate to severe hyperkalemia

a b s t r a c t

Purpose: Several studies investigated the agreement between central laboratory biochemistry analyzers and Blood gas analyzers for potassium measurements. However, data are scarce when the potassium level is moder- ate to severely high. We aimed to evaluate the agreement between central laboratory biochemistry analyzers and blood gas analyzer in terms of Serum potassium level measurement because differences in potassium at this level translate into very different clinical actions.

Basic procedures: This was a retrospective medical record review study in which patients who presented to the emergency department and had serum potassium levels >= 6 mmol/L were included. Patients who did not have simultaneous potassium measurement by blood gas analyzer were excluded. We included all patients meeting potassium criteria irrespective of their underlying disease or comorbidities. We evaluated agreement between the measurement methods with Pearson correlation, Bland-Altman plot, and Sign test.

Main findings: A total of 118 blood sample pairs were included. The mean serum potassium level measured by biochemistry analyzer was 6.78 +- 0.79 mmol/L, whereas it was 6.16 +- 0.86 mmol/L by blood gas analyzer (P b .001, Sign test). There was a strong correlation (P b .001, r = 0.864) between the 2 methods, but agreement was relatively poor. Blood gas analyzer tended to measure potassium significantly lower than measured by bio- chemistry analyzer. The mean difference between the methods was 0.62 +- 0.43 mmol/L. Principal conclusions: In patients with moderate to Severe hyperkalemia, blood gas analyzer and biochemistry an- alyzer gives significantly different serum potassium results which may be clinically important.

(C) 2016

Introduction

Hyperkalemia is a life-threatening Electrolyte disorder and com- monly seen in patients who present with acute kidney injury or chronic kidney disease [1]. Because exact serum potassium level is crucial in de- termining the treatment approach in hyperkalemia, prompt measure- ment is of clinical importance. To this end, blood gas analyzers are frequently used in emergency departments in patients with suspected acute kidney injury while central biochemistry results are pending. A number of studies in the literature examined the agreement of blood gas analyzers and central laboratory biochemistry analyzers in mea- surement of serum potassium levels, with conflicting results [2-5].

* Corresponding author at: SakaryaUniversitesi, EgitimveArastirmaHastanesi, 54100, Sakarya, Turkey. Tel.: +90 5058899885.

E-mail addresses: [email protected] (S.B. Acikgoz), [email protected] (A.B. Genc), [email protected] (S. Sipahi), [email protected]

(M. Yildirim), [email protected] (B. Cinemre), [email protected] (A. Tamer), [email protected] (Y. Solak).

Moreover, most of these studies examined mild to moderate hyperkalemia in various clinical settings. Severe hyperkalemia is a po- tentially fatal electrolyte imbalance, and the agreement of these 2 methods should be clearly established to use these methods inter- changeably in a safe manner. Thus, we studied the agreement of potas- sium measurements in blood gas analyzer and central biochemistry laboratory analyzer at our hospital in moderate to severe hyperkalemia (serum potassium N 6 mmol/L).

Methods

This was a retrospective study. We screened the patients who pre- sented to the emergency department of our hospital with serum potas- sium levels >= 6 mmol/L (measured by biochemistry analyzer) and also had potassium levels measured by a blood gas analyzer device between October 2014 and January 2015. Adult patients with moderate to severe hyperkalemia were included in the study irrespective of their presenta- tion concern and underlying diseases. Patients who did not have a blood

http://dx.doi.org/10.1016/j.ajem.2016.01.003

0735-6757/(C) 2016

S.B. Acikgoz et al. / American Journal of Emergency Medicine 34 (2016) 794797 795

gas analyzer serum potassium measurement were excluded from the study. We exerted great care to make sure that blood samples were drawn simultaneously or temporally very close to each other for measure- ments of potassium in both devices. Venous blood serum potassium levels were measured by using both central laboratory biochemistry analyzer (ICT [ISE] Module of ARCHITECT c16000; Abbott Laboratories. Abbott Park, IL) and blood gas analyzer (ABL 700; Radiometer, Copenhagen, Denmark).

Statistical analysis

Statistical analyses were performed via IBM SPSS Statistics version 17. Serum potassium levels measured by biochemistry analyzer and blood gas analyzer were not distributed normally; thus, we performed Sign test to compare the medians of the 2 measurement methods. We performed Bland-Altman plot to analyze the agreement between the 2 measurement methods. Pearson correlation analysis was used to calculate the correlation of the 2 potassium measurement methods. We also constructed a line graphic to show correlation and level of agreement of the biochemistry and blood gas analysis. A P value b .05 was accepted as significant.

Results

A total of 118 blood sample pairs from 118 adult patients were in- cluded in the study. Mean age of the patients was 70 +- 13 years. Men comprised the 44% of the whole group.

The mean serum potassium level measured by biochemistry analyz- er was 6.78 +- 0.79 mEq/dL (median, 6.5 mmol/L; range, 6-9.4). The mean potassium value measured by blood gas analyzer was 6.16 +-

0.86 mEq/dL (median, 5.9 mmol/L; range, 4.5-9.3). Blood gas analyzer tended to measure potassium significantly lower than measured by bio- chemistry analyzer. Of 118 potassium pairs, blood gas analyzer mea- sured serum potassium lower than biochemistry analyzer in 114 measurements (96.6%). The difference between the 2 measurement methods was statistically significant (Sign test, P b .001).

There was a strong correlation between the 2 methods, but agree-

ment was relatively poor (Fig. 1). Fig. 2 depicts the Bland-Altman plot showing agreement between biochemistry analyzer- and blood gas

analyzer-measured serum potassium levels. The mean difference be- tween the 2 methods was 0.62 +- 0.43 mEq/dL. The difference between the 2 devices ranged from 0.19 to 1.05 mEq/dL. Fig. 3 demonstrates the histogram showing frequencies of difference in the potassium values attained by biochemistry analyzer and blood gas analyzer.

Discussion

The salient finding of this study was that there was a poor agreement between serum potassium measurements by blood gas analyzer and biochemistry analyzer in patients who had moderate to severe hyperkalemia. At the higher end of the serum potassium spectrum as in this study, difference up to 1 mmol/L may mean significant differ- ences in the management of patients. In our study, blood gas analyzer tended to measure serum potassium significantly lower than measured by biochemistry analyzer. Many physicians also in our institution rely on serum electrolyte levels attained rapidly by blood gas analysis. Al- though much faster than central laboratory biochemistry analyzer re- sults, serum potassium values attained by blood gas analyzers may mislead physicians to the belief that the patient does not have a life- threatening hyperkalemia.

A substantial portion of clinicians rely on point-of-care blood gas an- alyzer results for potassium levels to base their Treatment decisions. Jose and Preller [6] assessed the attitude of physicians with a survey. The au- thors evaluated at the same time the agreement between results of blood gas analyzer and central laboratory biochemistry analyzer mea- surements of potassium in intensive care unit patients. A total of 48.4% of the surveyed clinicians reported that they would base clinical deci- sions on results obtained from the blood gas analyzer, whereas the rest of the clinicians preferred to wait for the results from the biochem- istry. The mean difference of the potassium between the 2 methods was found to be 0.03 mmol/L. However, the main serum potassium level measured by biochemistry analyzer in this study was 3.96 +- 0.605 mmol/L and was much lower compared with our results.

Several studies assessed the agreement of blood gas analyzer and biochemistry analyzer in terms of serum potassium values; however, to the best of our knowledge, none of them evaluated the higher end

Fig. 1. Correlation and agreement between potassium values measured by biochemistry analyzer and blood gas analyzer (red line: line of agreement, black line: correlation line). Note the strong correlation (P b .001, r = 0.864) but poor agreement of the 2 measurement methods.

796 S.B. Acikgoz et al. / American Journal of Emergency Medicine 34 (2016) 794797

Fig. 2. Bland-Altman plot showing agreement between biochemistry analyzer- and blood gas analyzer-measured serum potassium levels. The mean difference between the 2 methods was 0.62 +- 0.43 mmol/L. (P b .001, t test of mean difference, 95% confidence intervals: 1.462 and -0.222).

of the potassium spectrum, that is, serum potassium N 6 mmol/L [2-5,7,8]. Moreover, whereas some studies demonstrated that there is no difference between the 2 measurement methods, others revealed significant difference. Bloom et al [4], similar to our results, found that blood gas potassium was on average 0.46 mmol/L lower than laboratory levels (maximum 1.03 mmol/L). The authors concluded that this level of difference may be clinically significant and that blood gas analyzer

results in isolation should not be relied upon. However, the range of po- tassium levels for patients included in their study (measured by bio- chemistry analyzer) was 3-5.6 mmol/L and was lower compared with our patients. Quinn and colleagues [3] assessed the agreement of blood gas analyzer and central biochemistry laboratory analyzer in measurement of electrolytes. The authors reported that serum potassi- um showed no significant difference within the physiological range.

Fig. 3. Histogram showing frequencies of difference in the potassium values attained by biochemistry analyzer and blood gas analyzer.

S.B. Acikgoz et al. / American Journal of Emergency Medicine 34 (2016) 794797 797

On the other hand, at higher concentrations of potassium (N 5 mmol/L), compared with standard venous estimates, blood gas potassium read- ings varied significantly (mean difference was 0.44 mmol/L).

In a recent and large retrospective study involving 11 000 paired

samples from 3 hospitals that evaluated the agreement between sodi- um, potassium, and calcium results from blood gas and biochemistry laboratory analyzers [2], the authors concluded that there was sufficient agreement between the serum potassium levels to be used in the clini- cal decision making. However, the mean level of potassium in the whole group was 4.2 +- 0.63 mmol/L measured by biochemistry laboratory an- alyzer. This was a fairly large study with enough power to detect differ- ences between the 2 methods, but again, serum potassium levels were within physiologic limits in almost all patients.

Our sample size is relatively small but we included patients with moderate to severe hyperkalemia who, in our opinion, constitute an im- portant cohort in which precise measurement of serum potassium mat- ters. Because our design is retrospective, though our endeavor to avoid we cannot guarantee, some potassium pairs might have some time lapse between them. Despite these shortcomings, this study is the first to report a significant difference between blood gas analyzer and bio- chemistry analyzer for serum potassium measurements.

In conclusion, our results show for the first time that, at higher spec- trum of serum potassium levels, blood gas analyzer and central labora- tory biochemistry analyzer differ significantly in the measurement of serum potassium. This difference at this level of serum potassium may

translate into significant clinical consequences when blood gas analyzer results are accepted in isolation without biochemistry results. Further larger sample-sized studies are needed to establish the difference of the performance characteristics of these 2 methods in the measurement of serum potassium levels.

References

  1. Kovesdy CP. Management of hyperkalaemia in chronic kidney disease. Nat Rev Nephrol 2014;10:653-62.
  2. Mirzazadeh M, Morovat A, James T, Smith I, Kirby J, Shine B. Point-of-care testing of electrolytes and calcium using blood gas analysers: it is time we trusted the results. Emerg Med J 2015.
  3. Quinn LM, Hamnett N, Wilkin R, Sheikh A. Arterial blood gas analysers: accuracy in determining haemoglobin, glucose and electrolyte concentrations in critically ill adult patients. Br J Biomed Sci 2013;70:97-100.
  4. Bloom BM, Connor H, Benton S, Harris T. A comparison of measurements of sodium, potassium, haemoglobin and creatinine between an emergency department-based point-of-care machine and the hospital laboratory. Eur J Emerg Med 2014;21:310-3.
  5. Jain A, Subhan I, Joshi M. Comparison of the point-of-care blood gas analyzer versus the laboratory auto-analyzer for the measurement of electrolytes. Int J Emerg Ment Med 2009;2:117-20.
  6. Jose RJ, Preller J. Near-patient testing of potassium levels using arterial blood gas analysers: can we trust these results? Emerg Med J 2008;25:510-3.
  7. Johnston HL, Murphy R. Agreement between an arterial blood gas analyser and a ve- nous blood analyser in the measurement of potassium in patients in cardiac arrest. Emerg Med J 2005;22:269-71.
  8. Budak YU, Huysal K, Polat M. Use of a blood gas analyzer and a laboratory autoanalyzer in routine practice to measure electrolytes in intensive care unit pa- tients. BMC Anesthesiol 2012;12:17.