Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Does leaving the tourniquet on during venepuncture affect serum electrolytes (in particular, serum potassium)?

Three Part Question

IN [adults without cardiovascular compromise undergoing venepuncture]
DOES [leaving the tourniquet on during blood collection]
AFFECT [serum electrolytes, in particular potassium]?

Clinical Scenario

The Trust introduced a new blood collection system that stated the tourniquet should be removed prior to blood collection during routine venepuncture. You conducted a quick survey of staff that revealed all clinicians (apart from one nurse who had recently attended her venepuncture training) leave tourniquets in place throughout the blood collection phase. You checked the current Trust policy, which also states that tourniquets should be removed prior to blood sampling. No rationale for this instruction is given and you wonder why tourniquet removal is recommended in the Trust policy. You carried out a quick Internet search that seemed to indicate serum electrolytes (potassium in particular) can be affected if the tourniquet is left on during blood collection.

Search Strategy

Medline 1946-07/2017 using OVID interface.
Embase 1974 - 07/2017 using OVID interface
Amed 1985-07/2017 using OVID interface
CINAHL 1937-07/2017 EBESCO host
({[exp. venipuncture OR venepuncture OR phlebotomy] AND [exp. tourniquet] AND [exp. serum electrolytes OR potassium]} LIMIT to [human and English language])

Search Outcome

4 papers were found, 3 of which were relevant to the 3-part question.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Wiederkehr, MR and Orson, WM
8 healthy volunteers (5 men, 3 women). Needles placed in antecubital veins of both arms and tourniquet placed on one arm. Total of 30 samples each from free-flowing and tourniquet blood draws. Tourniquet time < 1 min.Observational cohort study Small but statistically significant differences in potassium found with tourniquet application. Potassium (mEq/L)Free-flowing (3.7 +/- 0.05); with tourniquet (3.9 +/- 0.06) p <0.001. Values expressed as mean +/- SESmall sample size. Average difference between free-flowing and stasis tourniquet potassium levels 0.2 mEq/L (range 0.05-0.5 mEq/L); one volunteer had an increase of 0.5mEq/L. May not be clinically significant.
Lippi, G, Salvagno, GL, Montagnana, M, Brocco, G, and Guidi, GC
23 volunteer physicians (12 women, 11 men), mean age 32 yrs). Sequential venepunctures on 3 veins of the upper arms, alternating 1 arm with the other. 1st venepuncture = no tourniquet, 2nd and 3rd venepunctures = application of tourniquet for 1 min and 3 mins respectively. Blood collected by by single expert phlebotomist. Tourniquet pressure standardized at 60mmHg using sphygmomanometer. Observational cohort study.12 analytes in total: ALT, albumin, calcium, chloride, cholesterol, CK, creatinine, glucose, iron, potassium, sodium, urea. Observed a consistent decrease in concentrations of potassium, achieving statistical and clinical significance after 1- and 3-min stasis.Potassium: no stasis, 4.21 +/- SD 0.34; 1-min stasis, 4.09 +/- SD 0.27, p<0.001; 3-min stasis, 4.02 +/- SD 0.37 p=0.004Small sample size. Whilst standardized tourniquet pressure was used in the study, this does not reflect what happens in everyday clinical practice.
Cenzig, M, Ulker, P, Meiselman, HJ, and Baskurt, OK
10 healthy males aged between 25 and 30 yrs. Venepuncture was performed on right arm by anaesthetist. Blood pressure taken (left arm). Control blood samples obtained: venous stasis imposed for 30s by inflating cuff to 20mmHg above diastolic pressure. Cuff then deflated to mimic tourniquet removal and blood samples taken at 5, 30, 60, 90, 120, 150 and 180 seconds. This study imposed venous stasis for 30s and collected blood samples following deflation of the cuff (removal of tourniquet effect).Observational cohort study.Serum electrolytes (sodium, potassium, chloride, calcium, magnesium, nitrite and nitrate), blood gases and haematological parameters were measured. All analytes (including serum potassium) did not differ significantly at 30s intervals during the 3-min period following removal of tourniquet. However, haemorheological measures were affected. Potassium (mmol/L): control, 4.14 +/- 0.10; 5s, 4.10 +/- 0.11; 30s, 4.26 +/- 0.10; 60s, 4.20 +/- 0.12; 90s, 4.21 +/- 0.13; 120s, 4.21 +/- 0.10; 150s, 4.12 +/- 0.11; 180s, 4.07 +/- 0.10. All data presented as mean +/- SE.The pressure used for venous stasis was individualized to each patient and catheters were used for venepuncture (neither of which would happen in clinical practice). In addition, the rationale for applying this individualized tourniquet pressure (20mmHg above diastolic) was based on an unpublished pilot study of 1 patient.


There is some weak evidence to suggest that serum electrolytes are affected by the application of a tourniquet during blood collection. Our clinical question was focused on serum potassium and therefore limited in its scope. Further reviews in this field may be useful.

Clinical Bottom Line

The rationale underpinning the recommendation to remove the tourniquet during routine venepuncture should be included in Trust policy, and a local review of current venepuncture practice is required.


  1. Wiederkehr, MR and Orson, WM Factitious Hyperkalemia American Journal of Kidney Diseases 2000; Vol 36 (5): 1049-1053
  2. Lippi, G, Salvagno, GL, Montagnana, M, Brocco, G, and Guidi, GS Influence of short-term venous stasis on clinical chemistry testing. Clinical Chemistry and Laboratory Medicine 2005; 43 (8): 869-875
  3. Cenzig, M. Ulker, P, Meiselman, J, and Baskurt, OK Influence of tourniquet application on venous blood sampling for serum chemistry, hematological parameters, leukocyte activation and erythrocyte mechanical properties. Clinical Chemistry and Laboratory Medicine 2009; 47 (6): 769-776