Original Article
Brief Report
Pseudohyperkalemia Caused by Fist Clenching during Phlebotomy
N Engl J Med 1990; 322:1290-1292May 3, 1990
- Article
CLINICIANS are occasionally confronted with the finding of an elevated serum or plasma potassium level in an otherwise healthy person. Such an abnormality may herald the presence of occult mineralocorticoid deficiency or a defect in renal tubular transport.1 Alternatively, it may represent so-called pseudohyperkalemia, caused by the release of potassium from formed elements in the blood in patients with leukocytosis or thrombocytosis.2 , 3 We report here on a patient in whom pseudohyperkalemia resulted from the common practice of repeatedly clenching and unclenching a fist during venipuncture. This maneuver, which has been passed on from generation to generation of house staff, medical students, and other practitioners of phlebotomy as a way to make the veins more prominent, may in fact increase the plasma potassium concentration by 1 to 2 mmol per liter. Although the phenomenon was recognized almost four decades ago,4 it continues to be overlooked, as the following case illustrates.
Case Report
A 55-year-old university professor was referred to the General Clinical Research Center at San Francisco General Hospital for the evaluation of hyperkalemia. Six months earlier a serum potassium concentration of 5.9 mmol per liter was noted on routine examination. A serum potassium level of 6.9 mmol per liter was measured three days later, and he was admitted to the local university medical center for further evaluation. His renal function was normal (serum creatinine, 97 μmol per liter), as were his plasma levels of aldosterone (166 pmol per liter), cortisol (304 nmol per liter), and renin activity (0.44 ng per liter per second). While he was in the hospital, his serum and plasma potassium concentrations were normal (3.9 to 4.6 mmol per liter) at every measurement (n = 5). During subsequent visits to the outpatient clinic, however, he was again found to have high serum potassium concentrations (5.4 to 6.6 mmol per liter). His white-cell and platelet counts were normal. He exercised regularly, but his serum potassium level was elevated even on days when he had not exercised, and it remained elevated during treatment with hydrochlorothiazide (50 mg per day). He was then referred for consultation.
He reported no medical problems and was an avid jogger who regularly ran between 13 and 16 km (8 and 10 miles) per day. His only medication was colestipol hydrochloride (20 g per day), which had been prescribed for the treatment of hypercholesterolemia three years earlier. Because of the apparent discrepancy between the serum potassium values obtained in the hospital and at the outpatient clinic, he was questioned further about the means used to obtain the blood samples. He stated that in the clinic, samples had been obtained with use of a tourniquet-and-fist-clenching technique, whereas the samples that were drawn in the hospital had been obtained from an indwelling catheter with no tourniquet or fist clenching.
His blood pressure was 146/80 mm Hg, and his pulse rate 54 beats per minute. He was a lean, well-developed man in excellent health, and his physical examination was entirely normal. His hemoglobin concentration was 2.23 mmol per liter (14.4 g per deciliter), hematocrit 44.3 percent, white-cell count 5.6×109 per liter, and platelet count 361×109 per liter. The plasma level of sodium was 142 mmol per liter, chloride 107 mmol per liter, and total carbon dioxide 27 mmol per liter. The serum creatinine level was 88 μmol per liter (1.0 mg per deciliter), the serum calcium, phosphorus, and glucose concentrations were normal, and the serum cholesterol level was 6.2 mmol per liter (238 mg per deciliter). The results of urinalysis, thyroid-function tests, and chest radiography were normal. The plasma potassium concentration in a sample of blood obtained from an indwelling catheter without stasis or fist clenching was 4.1 mmol per liter, and it increased to 5.1 mmol per liter within one minute after the application of a tourniquet and repeated fist clenching. Because of the discrepancy in potassium levels obtained with different methods of phlebotomy, we undertook four studies.
Methods
The studies were approved by the Committee on Human Research of the University of California at San Francisco, and informed consent was obtained from the study subjects — the patient and the four healthy men from 31 to 59 years old who served as controls. Four different protocols were employed. An 18-gauge angiographic catheter (Deseret, Sandy, Utah) was placed in a large vein in one or both antecubital fossae. The patency of the catheters was maintained with a heparin lock. The subjects remained in a recumbent position for 30 minutes before and throughout the study protocols.
Protocols
Effect of a Tourniquet with and without Fist Clenching
In each subject, three base-line blood samples were obtained from each arm at 30-second intervals. A tourniquet was then applied snugly to the mid-biceps region of one arm (the test arm) for one minute. With the tourniquet still in place, the subjects then repeatedly clenched and unclenched the fist of that arm for one minute, after which the tourniquet was removed and the fist clenching stopped for an additional minute. As a time control, in a separate experiment the tourniquet was left in place for two minutes with no fist clenching. Blood samples were obtained at 30-second intervals from both the test and the contralateral (control) arm for plasma potassium measurements throughout these maneuvers.
Effect of Handgrip without a Tourniquet
Two blood samples were obtained from each arm five minutes apart before the handgrip maneuver was performed. Thereafter, using a dynamometer, the subjects maintained a constant handgrip for two minutes at 30 percent of their predetermined maximal gripping strength, then relaxed the hand. Blood samples were obtained from both the test and contralateral arms at one-minute intervals during the handgrip maneuver and for three minutes thereafter.
Effect of Graded Increases in Occlusion Pressure
We used a sphygmomanometer cuff placed on the mid-biceps region to increase occlusion pressure from 0 to 150 mm Hg in 30-mm increments at 30-second intervals in the four control subjects. Before inflation of the cuff and at the end of each 30-second period, blood samples were collected from an antecubital vein distal to the cuff for plasma potassium measurements.
Effect of a Tourniquet on Local Venous pH and Lactate Levels
In a separate study involving only the patient, we evaluated the effect of a tourniquet with and without fist clenching on venous pH and plasma lactate levels. For this study, two base-line blood samples were obtained from each arm at two-minute intervals before a tourniquet was applied. The tourniquet was then applied to one arm for two minutes, after which — with the tourniquet in place —the patient repeatedly clenched his fist for one minute. Thereafter, the tourniquet was removed and the fist clenching stopped. Blood samples for measurements of plasma potassium and lactate concentrations, venous-blood pH, and total carbon dioxide content were obtained from both arms at one-minute intervals throughout the study.
Laboratory Methods
Plasma potassium concentrations were measured with use of a flame photometer. The interassay coefficients of variation were 1.8, 0.6, and 0.7 percent at potassium concentrations of 2.8, 5.0, and 7.0 mmol per liter, respectively. Venous pH was measured with use of an MK2 analyzer (Radiometer Microsystem 3, Cleveland) and total carbon dioxide with a carbon dioxide analyzer (Corning 960, Medfield, Mass.). The plasma bicarbonate concentration and partial pressure of carbon dioxide were calculated from these measurements. Plasma lactate concentrations were determined by spectrophotometric measurement of the NADH formed in the oxidation of lactate.
Results
Alone, the application of a tourniquet had no effect on plasma potassium concentrations, whereas the addition of clenching the fist increased potassium levels in both the patient (increase, 1.0 mmol per liter) and the controls (mean [±SEM] increase, 1.0±0.1 mmol per liter; P<0.001) (Fig. 1Figure 1
Effects of the Application of a Tourniquet plus Fist Clenching (Upper Panel) and Tourniquet Alone (Lower Panel) on Plasma Potassium Concentrations.). Handgrip exercise also raised the plasma potassium concentration in the patient (increase, 1.4 mmol per liter) and the control subjects (mean increase, 1.0±0.2 mmol per liter; P = 0.002) (Fig. 2Figure 2
Effect of Handgrip Exercise on Plasma Potassium Concentrations.). There was no change in plasma potassium concentrations in the samples obtained from the contralateral arm during any of these maneuvers. In the four control subjects, there was no change in the mean plasma potassium concentration when arm occlusion pressure was increased stepwise from 0 to 150 mm Hg.In the study performed in the patient alone, the effect of a tourniquet with or without fist clenching on plasma potassium and acid–base indexes was studied (Table 1Table 1
Effect of Stasis and Forearm Exercise on Local Venous Plasma Potassium, Bicarbonate, and Lactate Levels, Blood pH, and Partial Pressure of Carbon Dioxide (pCO2).). As demonstrated earlier, the application of a tourniquet alone had no effect on the plasma potassium concentration, but the addition of fist clenching increased it 1.1 mmol per liter. Plasma lactate levels did not change with the tourniquet alone, but did increase when the fist was clenched as well. The plasma lactate concentration continued to increase after the tourniquet had been removed and the fist relaxed, whereas the plasma potassium concentration returned to base-line levels. Venous pH did not change with the tourniquet alone or with fist clenching as well, but it did begin to decrease during the recovery phase. The decrease in pH could be accounted for by an increase in the partial pressure of carbon dioxide.Discussion
The discrepancies in serum and plasma potassium levels measured at various times in the patient resulted from the methods of phlebotomy used. Both repeated fist clenching and isometric handgrip increased the plasma potassium level by as much as 1.6 mmol per liter in the venous effluent from the contracting arm. In blood samples obtained from the arm that was not exercised, there were no changes in plasma potassium concentrations. In studies performed in the control subjects, similar increases in plasma potassium concentrations were found in the venous effluent from the exercising arm. Thus, fist clenching during phlebotomy leads to an increase in the local potassium concentration that could mislead clinicians about the true plasma or serum potassium concentration in their patients.
The elevation in potassium concentration that results from fist clenching is due to the local release of potassium caused by the contraction of forearm muscles.4 Potassium is released during exercise because intracellular electronegativity declines during the depolarization of muscle cells, and this favors the release of potassium rather than its uptake.5 This local hyperkalemia has been implicated as a factor in the increase in blood flow to an exercising muscle.6 , 7 The increased concentration of potassium in venous blood during exercise provides an underestimation of the level in the interstitial fluid of the contracting muscle. Potassium in the interstitial fluid had been demonstrated to increase to as much as four times the level in the inflowing arterial blood.7 This marked rise in the interstitial potassium level during exercise may increase the consumption of oxygen and lactate and the production of heat.8 , 9 In our study, however, the changes in venous pH and lactate levels in the forearm that are induced by exercise did not appear to be responsible for the elevations in the level of potassium. When the fist was clenched, venous pH remained unchanged, and the potassium concentration increased; when fist clenching stopped, venous pH decreased, plasma lactate levels increased, and plasma potassium returned to base-line levels (Table 1). This is in accord with the observations that marked hyperkalemia is unusual with either lactic10 , 11 or respiratory12 acidosis and that changes in blood pH correlate poorly with plasma potassium levels.13 , 14
It has been well known since the 1930s15 that an exercising muscle loses potassium, but the potential for error in measuring levels of potassium was first recognized in 1961 by Skinner, who posed the question, "Could the ritual of forearm exercise so often encouraged during venepuncture [sic] affect the local potassium level?"16 He proceeded to demonstrate that forearm exercise raised plasma potassium concentrations by as much as 2.7 mmol per liter in healthy subjects. In addition, he investigated the extent to which forearm exercise was part of venipuncture procedures by observing the phlebotomy technique of house staff. All the observed house staff, who had no knowledge of the study, encouraged forearm exercise during phlebotomy. The serum potassium concentrations in the samples obtained by the house staff averaged 0.4 mmol per liter higher and were as much as 0.8 mmol per liter higher than those obtained without exercise in the other arm five minutes later.
There are obvious consequences when high plasma or serum potassium levels are found, so it is important that the results be valid. Otherwise, as in our patient, such a finding may result in unwarranted and costly studies or inappropriate treatment. In another example, Brown et al.17 noted that in patients with aldosteronism, forearm exercise during phlebotomy raised serum potassium from subnormal to normal levels. Thus, the technique may obscure a diagnosis of aldosteronism. In view of these examples, it would seem advisable to avoid fist clenching altogether and to rely on venous stasis alone, if needed, as an aid in performing phlebotomy.
Supported by grants (HL-11046, DK-39964, RR-79, and RR-83) from the National Institutes of Health. Dr. Don is a Clinical Associate Physician of the General Clinic Research Center (RR-83) at San Francisco General Hospital.
Source Information
From the Medical Service, San Francisco General Hospital Medical Center, and the Department of Medicine, University of California, both in San Francisco.
- References
References
1
Sebastian
Web of Science | Medline2
Hartmann
CrossRef | Web of Science | Medline3
Branson
Full Text | Web of Science | Medline4
Farber
Web of Science | Medline5
Clausen
CrossRef | Web of Science | Medline6
Knochel
CrossRef | Web of Science | Medline7
Kjellmer
Medline8
Kaye
CrossRef | Web of Science | Medline9
Knochel JP. Potassium gradients and neuromuscular excitability. In: Seldin DW, Giebisch G, eds. The kidney: physiology and pathophysiology. Vol. 2. New York: Raven Press, 1985:1207–21.
10
Oster
Web of Science | Medline11
Perez
CrossRef | Medline12
Schwartz
CrossRef | Web of Science | Medline13
Sterns
CrossRef | Web of Science | Medline14
Adrogue
CrossRef | Web of Science | Medline15
Fenn 16
Skinner
CrossRef | Web of Science17
Brown
CrossRef | Web of Science | Medline
- Citing Articles (23)
Citing Articles
1
Qing H. Meng, John Krahn. (2011) Reverse pseudohyperkalemia in heparin plasma samples from a patient with chronic lymphocytic leukemia. Clinical Biochemistry 44:8-9, 728-730
CrossRef2
R.A. Sulaiman, P.J. Twomey, R. Gama. (2011) Mitigation and detection of spurious potassium and sodium results. Clinica Chimica Acta 412:1-2, 1-6
CrossRef3
Masanori Seimiya, Toshihiko Yoshida, Yuji Sawabe, Kazuyuki Sogawa, Hiroshi Umemura, Kazuyuki Matsushita, Fumio Nomura. (2010) Reducing the Incidence of Pseudohyperkalemia by Avoiding Making a Fist During Phlebotomy: A Quality Improvement Report. American Journal of Kidney Diseases 56:4, 686-692
CrossRef4
Johannes Steinfurt, Markus C. Müller, Anke Seidel, Richard Salm, Andreas Ochs. (2010) Paresen, Myalgien und massive CK-Erhöhung: eine schwere neurologische Erkrankung?. Medizinische Klinik 105:7, 496-500
CrossRef5
Goce Dimeski, Tony Badrick, Andrew St John. (2010) Ion Selective Electrodes (ISEs) and interferences—A review. Clinica Chimica Acta 411:5-6, 309-317
CrossRef6
Namrata Setia, James H. Nichols. 2009. Preanalytical Variation in Clinical Laboratory Testing. .
CrossRef7
Neha R. Chawla, Joel Shapiro, Ronald L. Sham. (2009) Pneumatic tube “pseudo tumor lysis syndrome” in chronic lymphocytic leukemia. American Journal of Hematology 84:9, 613-614
CrossRef8
Toshiro Sugimoto, Nobuhiro Ogawa, Atsunori Kashiwagi. (2007) Pseudohyperkalemia in a patient with rapidly progressive glomerulonephritis. European Journal of Internal Medicine 18:3, 258
CrossRef9
Domenic A. Sica. (2006) Antihypertensive Therapy and Its Effects on Potassium Homeostasis. The Journal of Clinical Hypertension 8:1, 67-73
CrossRef10
Paul S. Kellerman, James M. Thornbery. (2005) Pseudohyperkalemia Due to Pneumatic Tube Transport in a Leukemic Patient. American Journal of Kidney Diseases 46:4, 746-748
CrossRef11
H A. Tran. (2005) Extreme Hyperkalemia. Southern Medical Journal 98:7, 729-732
CrossRef12
Michael Stowasser, Richard D. Gordon. (2004) The Aldosterone???Renin Ratio in Screening for Primary Aldosteronism. The Endocrinologist 14:5, 267-276
CrossRef13
J Ian S Robertson. (2004) Aldosterone excess in hypertension. Journal of Hypertension 22:4, 861
CrossRef14
Donald Schwartz, Neil Roy Connelly, P. Manikantan, J. H. Nichols. (2003) Hyperkalemia and Pyloric Stenosis. Anesthesia & Analgesia 97:2, 355-357
CrossRef15
Noha Abou-Madi, Elliott R. Jacobson. (2003) Effects of Blood Processing Techniques on Sodium and Potassium Values: A Comparison Between Aldabra Tortoises ( Geochelone gigantea ) and Burmese Mountain Tortoises ( Manouria emys ). Veterinary Clinical Pathology 32:2, 61-66
CrossRef16
H. Wisser, Th. Bertsch, D. Wisser. (2002) Preanalytical Prerequisites for the Quality of Samples/Präanalytische Vorbedingungen der Probenqualität. LaboratoriumsMedizin 26:5-6, 284-290
CrossRef17
S. Narayanan, W. G. Guder. (2002) Präanalytische Variable und ihr Einfluß auf die Qualität von Laboratoriumsbefunden/Pre-Analytical Variables and their Impact on the Quality of Laboratory Results. LaboratoriumsMedizin 26:5-6, 263-266
CrossRef18
M KAPOOR, G CHAN. (2001) FLUID AND ELECTROLYTE ABNORMALITIES. Critical Care Clinics 17:3, 503-529
CrossRef19
Sheshadri Narayanan, PhD. (2000) The Preanalytic Phase: An Important Component of Laboratory Medicine. American Journal of Clinical Pathology 113:3, 429-452
CrossRef20
Giuseppe Famularo, Fabio Massimo Corsi, Manlio Giacanelli. (1999) Iatrogenic Worsening of Hypokalemia and Neuromuscular Paralysis Associated with the Use of Glucose Solutions for Potassium Replacement in a Young Woman with Licorice Intoxication and Furosemide Abuse. Academic Emergency Medicine 6:9, 960-964
CrossRef21
(1999) Treatment of Hypokalemia. New England Journal of Medicine 340:2, 154-155
Full Text22
K Mäkelä, V. Kairisto, O. Peltola, T. Hollmén, A. Virtanen, K. Pulkki, V. Näntö. (1995) Effect of platelet count on serum and plasma potassium: evaluation using database information from two hospitals. Scandinavian Journal of Clinical & Laboratory Investigation 55:s222, 95-100
CrossRef23
(1991) Pseudohyperkalemia and Platelet Counts. New England Journal of Medicine 325:15, 1107-1107
Full Text
- Letters
