Abstract
Background
The aim of the study was to assess the incidence of undetected potassium EDTA contamination in routine blood samples submitted for biochemical analysis.
Methods
Over a seven-day period, serum EDTA concentrations were measured in all blood samples submitted for routine biochemical analysis.
Results
EDTA contamination was detected in 22 of 4789 samples submitted for analysis (0.46%), of which only seven (0.15%) would have been detected by staff using subjective methods. Detection of low but significant concentrations of EDTA was not necessarily associated with hyperkalaemia or hypocalcaemia. Haemolysis remains the leading cause of non-reporting of analyte results.
Conclusion
The presence of EDTA in serum does not always result in spurious hyperkalaemia or hypocalcaemia. Routine measurement of serum EDTA concentrations has the potential to reveal contamination which is not obvious using subjective measures and its routine measurement has the potential to identify minimally contaminated samples.
Introduction
Potassium EDTA is an anticoagulant commonly employed in the analysis of blood that acts by chelating divalent metal ions essential for the activation of the coagulation cascade system. However, contamination of serum samples by EDTA as a consequence of incorrect draw order can result in falsely elevated potassium, reduced alkaline phosphatase (ALP) activity and reduced concentrations of divalent ions, e.g. calcium and magnesium. To minimize the likelihood of EDTA contamination in serum samples, the draw order usually specifies that the serum sample should be taken prior to other anticoagulant containing tubes. However, this advice is not always followed, especially when patients are difficult to bleed, and this may result in serum contaminated with EDTA. Usually this manifests as a classical picture of extremely high potassium, low magnesium and calcium and, possibly, low ALP.
The aim of this study was to use a recently published method for the direct measurement of EDTA to reveal the extent of occult EDTA contamination.
Methods
Samples for routine biochemical analysis were, usually, collected into BDH Vacutainer tubes, although occasional samples were received in paediatric tubes or from other suppliers. Over a seven-day period, EDTA concentration was measured in all samples submitted for automated biochemical analysis.
EDTA was measured as described previously, 1 with modifications to run on the Beckman DxC. The method was found to be linear to at least 0.4 mmol/L; values over this concentration were recorded as >0.4 mmol/L. The lower limit of detection was found to be 0.03 mmol/L, but samples were considered to be contaminated with EDTA if the concentration exceeded 0.1 mmol/L. 1,2 The between-run precision was 5.4% and 4.2% at EDTA concentrations of 0.14 and 0.33 mmol/L, respectively. Potassium and calcium were measured by indirect ion-selective electrode. ALP and magnesium were measured using colourimetric methods supplied by Beckman-Coulter (High Wycombe, UK).
Results
A total of 4789 samples were available, of which 22 were found to be contaminated with EDTA (0.46%): four from general practice and the remaining 18 from hospital departments (3 paediatric, 3 casualty and the other samples from individual wards and departments). Of these samples, seven were identified by staff based on subjective assessment (high potassium, low calcium, magnesium and/or ALP with absence of delay or haemolysis). Trained phlebotomy staff were not associated with any instances of EDTA contamination in the study period.
The average EDTA concentration in the contaminated samples identified by staff was 0.36 mmol/L (range 0.15–>0.4 mmol/L) compared with 0.14 mmol/L (range 0.11–0.22 mmol/L) in those identified by the assay only. The range of analyte concentrations in these latter samples was potassium 3.8–7.2 mmol/L, corrected calcium 1.99–2.45 mmol/L and ALP 33–773 IU/L. The high potassium was seen in a patient with acute renal failure and no repeat sample was available. If this sample were removed, the range of potassium concentrations seen would be 3.8–4.8 mmol/L.
Over the study period, 10 samples had recorded potassium concentrations in excess of the telephone range (6.5 mmol/L and above). EDTA contamination may have been a contributing factor in a single patient as described previously (EDTA 0.11 mmol/L, potassium 7.2 mmol/L). The remaining hyperkalaemic samples had recorded EDTA concentrations ranging from <0.03 to 0.07 mmol/L. There were no cases where the recorded calcium or magnesium concentrations triggered the ‘telephone range’ (<1.5 and <0.5 mmol/L, respectively).
Haemolysis remains the leading reason for not reporting analyte results (Table 1). Measuring EDTA concentrations could increase the non-reporting rate by 0.21–0.46%.
Reasons for not reporting results
ALP, alkaline phosphatase
†Other reasons include incorrectly labelled samples and subjectively assessed EDTA contamination
Discussion
EDTA contamination is widely recognized as an artefactual cause of spurious results affecting a wide range of analytes. 3,4 The data in this study indicate that laboratory staff can identify gross EDTA contamination provided they have access to potassium, calcium and magnesium concentrations. However, this is a subjective opinion based on a typical biochemical pattern in the absence of obvious delay and haemolysis. The ability to measure EDTA concentration directly removes the need for a subjective judgement and provides an objective basis to confirm contamination and suppress results.
Sharratt et al. 2 have shown that EDTA contamination may be a contributing factor in a significant proportion of samples that trigger urgent laboratory action. This study did not reproduce this finding, but the number of samples was far smaller and, unlike the work by Sharratt et al. 2 did not rely on other analytes to trigger an EDTA analysis. Cornes et al. and Sharratt et al. highlight that EDTA contamination may mask significant hypokalaemia.
Haemolysis and delay in receipt remain the major causes of non-reporting of biochemical results. However, this study has demonstrated that EDTA contamination may affect up to 0.46% of routine samples submitted for automated analysis. Education of those involved in phlebotomy, especially those where this is not their primary role, is required with emphasis on the importance of correct draw order.
Introduction of routine EDTA analysis as part of the sample quality assessment may identify moderately contaminated samples that are missed by subjective appraisal.
DECLARATIONS