Undetected spurious hypernatraemia wastes health-care resources

A recent test on a 63 year old woman showed marked hypernatraemia and moderate hypokalaemia (Table 1). These results were confirmed by reanalysis and telephoned to the requesting primary care physician, who arranged for the patient to be admitted to the hospital as a matter of urgency. However, repeat electrolytes in this otherwise well patient were normal (Table 1). The patient recalled that there had been difficulty obtaining blood and that samples had been transferred between tubes. We considered the possibility of sodium citrate contamination as the patient had also had blood collected into a sample tube for erythrocyte sedimentation rate (ESR) measurement, which requires a tube containing trisodium citrate as an anticoagulant.

Table 1

Influence of contaminants on electrolyte results

	Patient		Control 1			Control 2			Control 3
	Initial	Repeat	Baseline	Coagulation	ESR	Baseline	Coagulation	ESR	Baseline	Coagulation	ESR
Sodium (mmol/L)	170	137	141	167	182	141	165	174	143	168	180
AVL:sodium (mmol/L)	145	135	139	152	152	138	150	148	142	151	149
Potassium (mmol/L)	2.7	4.0	4.4	3.6	3.0	4.2	4.0	3.7	3.8	2.8	2.6
AVL:potassium (mmol/L)	2.5	4.0	4.6	3.3	2.6	4.3	3.7	3.2	3.8	2.8	2.3
Urea (mmol/L)	5.5	8.3	5.2	5.0	4.1	5.1	4.6	4.2	6.6	5.6	5.3
Creatinine (μmol/L)	46	52	86	55	55	4.6	38	32	101	82	72
Chloride (mmol/L)	64	103	99	81	54	104	87	66	107	89	66
Glucose (mmol/L)	4.8	5.3	5.5	–	4.1	4.9	4.5	4.1	4.7	4.1	3.8
Osmolality (mmol/L)	275	291	294	296	271	295	300	284	297	297	280
Calculated osmolality (mmol/L)	350	295	292	343*	372	292	339	356	297	346	369
Osmolar gap (mmol/L)	−75	−4	+2	−47	−99	+3	−39	−72	0	−49	−89

We were unable to find published reports of sample cross-contamination with sodium citrate. We, therefore, collected blood samples from volunteers (MC, CF and RG) into Sarstedt serum/z4 gel tubes, sedivette 4NC/3.5 (ESR) and coagulation 9NC/3 tubes using the Sarstedt Safety Monovette^® system (Sarstedt Safety Monovette, Aktiengesellschaft & Co, Nümbrecht, Germany). Blood from the sedivette 4NC/3.5 and coagulation 9NC/3 tubes was then decanted into serum Z/4 sample tubes and the separated plasma analysed for creatinine and electrolytes on the Roche MODULAR^® analyser (Roche Diagnostics GmbH, Mannheim, Germany). Electrolytes were also measured using direct ion-specific electrode (ISE) on the Roche AVL9181 electrolyte analyser (Roche Diagnostics) and osmolality on the Advanced Micro Osmometer™ (Advanced Instruments Inc, Nowood, MA, USA). These results were then compared with the uncontaminated baseline serum results (Table 1).

The data indicate that pseudohypernatraemia due to sodium citrate contamination is characterized by hypernatraemia with disproportionately low serum chloride and a negative osmolar gap (measured osmolality − calculated osmolality) similar to that observed with contamination with Citra-Lock™ (a sodium citrate ‘catheter locking’ anticoagulant). ^1–3 In addition, for the first time, we report a lower sodium concentration measured using a direct ISE compared with indirect ISE as a feature of pseudohypernatraemia due to sodium citrate contamination. The molar anion:cation dissociation ratios for trisodium citrate and sodium chloride are 3:1 and 1:1, which explains the negative osmolar gap and the lower than expected serum chloride concentration.

Pseudohypernatraemia due to hypoproteinaemia is also characterized by lower direct compared with indirect ISE sodium measurement and a negative osmolar gap, ⁴ but the serum chloride concentration is appropriate for serum sodium. Serum chloride will therefore discriminate between pseudohypernatraemia due to hypoproteinaemia and sodium citrate contamination. Further analysis of the initial patient sample was consistent with decanting of blood from a sedivette (ESR) sample tube into a serum/z4 gel tube (Table 1).

In summary, we report unrecognized pseudohypernatraemia because of sodium citrate sample contamination leading to inappropriate hospital attendance and investigation. Training and education regarding correct blood collection is essential in preventing spurious electrolyte results. ⁵ As in this case, clinicians should suspect results are spurious if they are inconsistent with the patient's medical condition and medication and request a further blood sample. Laboratories must, however, take some responsibility for identifying spurious sodium results. We suggest that hypernatraemia on indirect ISE measurement be confirmed using a direct ISE or, if unavailable, serum osmolality measured. Discordant results indicate pseudohypernatraemia. Serum chloride measurement will then differentiate between pseudohypernatraemia because of hypoproteinaemia and sodium citrate contamination.

We wish to emphasize the importance of detecting spurious hypernatraemia to prevent wasting limited health-care resources and causing unnecessary patient anxiety.

DECLARATIONS

Competing interests: None.

Funding: None.

Ethical approval: N/A.

Guarantor: RG.

Contributorship: Laboratory work was carried out by MC. Ideas and discussion were equally contributed to by all authors as was the writing and editing process.