Abstract: TH-PO1134

Pseudohypernatremia in Patients with Hemodialysis Catheters Locked with Trisodium Citrate

Session Information

Category: Fluid, Electrolytes, and Acid-Base

  • 704 Fluid, Electrolyte, Acid-Base Disorders

Authors

  • Brzosko, Szymon, DaVita, Bialystok, Poland
  • Drozdz, Maciej B., DaVita, Krakow, Poland
  • Rydzewska-Rosolowska, Alicja, Medical University of Bialystok, BIALYSTOK, Poland
  • Jacobson, Stefan H., Danderyd Hospital, Stockholm, Sweden
Background

International guidelines recommend arterio-venous fistula (AVF) as the preferred type of vascular access (VA) for hemodialysis (HD). Infections and patency-related complications of central venous catheters (CVC) pose a potential risk for patients. Trisodium citrate (TSC) locking solution is a promising alternative to unfractionated heparin (UFH) for prevention of CVC dysfunction. Pseudohypernatremia secondary to TSC contamination of blood specimens obtained from the CVC may potentially lead to profound but unnecessary diagnostic interventions. The aim of the study was to analyze pre-dialysis Na concentration and the prevalence of hypernatremia in relation to the type of locking solution (TSC vs UFH) and VA in a large cohort of HD patients (DaVita Poland).

Methods

The population studied consisted of 543 prevalent hemodialysis patients (45% female), mean age 67±14 years treated in a standard HD program (100% High Flux, 94% HD time>12h/week, 94% spKt/V>1.3, 18.5% CVC). We analyzed laboratory results from February & March 2017. Pre-dialysis Na concentration was compared between patients with CVC locked with 30%TSC, UFH (5000 IU/ml) and non-CVC types of VA (AVF, AVG). The proportion of patients with pre-dialysis Na ≥145 mEq/l (hypernatremia) was analyzed in both groups and from both months separately. Descriptive statistics was used together with ANOVA, t and X2 tests as appropriate.

Results

The mean plasma concentration of Na was significantly higher in patients with TSC locking solution (141±3 mEq/l n=26, 142±3 mEq/l n=35) when compared to UFH (138±3 mEq/l n=76, 138±4 mEq/l n=75) and non-CVC VA (138±2 mEq/l n=430, 138±3 mEq/l n=440; ANOVA p<0.001 for February & March respectively). The proportion of patients with Na≥145 mEq/l was significantly higher in patients with CVCs with TSC (n=11/61) as a locking solution than in patients with UFH (n=4/152; 18.0% vs 2.6%; X2 p<0.001). The highest plasma concentration of Na was 171 mEq/l, obtained from a patient with uneventful follow up. Repeated analysis from a peripheral vein showed a Na concentration of 138 mEq/l.

Conclusion

Pseudohypernatremia is may occur in HD patients when 30% TSC is used as a locking solution. Need for diagnostic laboratory investigations may be lowered by strictly following procedures for blood sample collection from CVC (including discarding proper volume of blood).