Prevalence and detection of citrate contamination in clinical laboratory.

OBJECTIVES

To study the prevalence of trisodium citrate (NaCitrate) contamination in hypernatraemic serum samples by direct measurement of citrate and to evaluate the performance of indirect markers for identification of NaCitrate contamination.

METHODS

Serum citrate was measured in all hypernatraemic serum samples (sodium ≥148 mmol/L) over a three-month period. The performance of serum chloride, sodium-chloride gap, indirect ion selective electrode (ISE)-direct ISE sodium disparity and osmolar gap in identification of NaCitrate contaminated samples was assessed against the 'gold-standard' direct citrate measurement.

RESULTS

In total, 27 NaCitrate contaminated samples were identified based on serum citrate concentration ≥1.5 mmol/L. The prevalence of citrate contamination was 3.1 % of hypernatraemic samples (n=875) and 0.017 % of all samples received for urea and electrolyte analysis (n=153,404). Most contaminated samples were from patients receiving haemodialysis (59.3 %), and the rest from inpatients. Cut-offs to give 100 % sensitivity were chloride ≤105 nmol/L (specificity 93.4 %), sodium-chloride gap ≥47 mmol/L (specificity 95.3 %), indirect ISE-direct ISE sodium disparity ≥3 mmol/L (specificity 81.9 %), and osmolar gap ≥39 mOsm/kg (specificity 2.8 %).

CONCLUSIONS

Trisodium citrate contamination is uncommon. Most contaminated samples were from patients receiving haemodialysis, likely because of contamination with citrate catheter locking solution. Screening with serum chloride or sodium-chloride gap can confidently exclude NaCitrate contamination in over 90 % of hypernatraemic samples, and in nearly all samples with sodium ≥155 mmol/L if metabolic alkalosis has been excluded. In the remaining samples, NaCitrate contamination can only be definitively confirmed or excluded by measurement of serum citrate. We propose algorithms to identify spurious hypernatraemia.