Pseudo-hyperkalaemia in ambulatory samples: the never-ending story?

Hyperkalaemia is a major electrolyte disturbance, which can be life-threatening [1, 2]. Pseudo-hyperkalaemia has been described since 1955 but remains a frequent problem for clinicians, and is mainly linked to blood collection [3]. The pre-analytic phase is of particular relevance to avoid pseudo-hyperkalaemia generating useless treatments and hospitalizations.

We conducted a study to determine the difference in hyperkalaemia prevalence in outpatient primary care between blood sampling performed at the laboratory versus at the patient's home.

We conducted a retrospective analysis of all the blood samples with kalaemia measurement from January to December 2016 in a laboratory in the southeast of France (Draguignan). Samples taken at home were either delivered directly by the nurse to the laboratory for those taken in town, or to laboratory relay stations. The daily average temperature (the average of the 24-h temperatures) was provided by the national meteorological research centre.

The main outcome was to describe kalaemia levels and the proportion of hyperkalaemia (kalaemia ≥5.5 mmol/L) over 1 year, according to the sampling location (at the laboratory or at home with immediate delivery, at home sampling with relay station) and the impact of ambient temperature. Kalaemia was presented as mean and compared using the analysis of variance. Hyperkalaemia was presented as a number and compared using the Chi-squared test. Multivariable linear regression models were built for each of the three sampling conditions. An overall logistic regression model was built to assess the independent impact of sampling conditions, temperature, age and kidney function on the hyperkalaemia risk. All analyses were performed using R software version 4.0.1 [4].

Between January and December 2016, there were 87 868 requests for biological analyses recorded at the medical biology laboratory. Among them, 6025 hospitalized or institutionalized patients were excluded. Of the 81 843 remaining requests, 28 137 (34.4%) included a potassium measurement. A total of 1823 haemolytic samples or those without haemolytic evaluation were excluded, and 762 additional pre-analytic non-conforming samples were also excluded. Of the 24 552 remaining blood samples, 55.2% were obtained at patients’ homes. The mean age of patients was 59.8 years. A total of 3.29% of patients had at least one episode of hyperkalaemia ≥5.5 mmol/L and 0.78% had at least one episode of hyperkalaemia ≥6 mmol/L. Mean kalaemia was significantly higher for home sampling, especially for those samples that transited through a relay station, compared with laboratory sampling (4.66 ± 0.54 versus 4.48 ± 0.38 mmol/L, P < 0.0001). Prevalence of hyperkalaemia was significantly higher for home sampling, 3.5% (P < 0.0001) with immediate delivery and 7.5% for home sampling with relay station, compared with 1.3% for laboratory sampling (Table 1).

The mean daily kalaemia increased during the winter when the blood sample was performed at home compared with the laboratory (Figure 1). We found an increase of 0.032 mmol/L of kalaemia for a diminution of 1°C in minimal temperature [beta coefficient = −0.032 (0.034; 0.030), P < 0.001] for home sampling with relay station (Table 2). Concerning hyperkalaemia, the odds ratio (OR) was 5.67 [95% confidence interval (CI) 4.65–6.93] for home sampling with relay station (Table 3). The temperature increase is associated with a lower prevalence of hyperkalaemia [for 1°C increased, OR 0.92 (95% CI) 0.91–0.93] (Table 3).

FIGURE 1:

Daily average kalaemia [home sampling (A) and laboratory sampling (B)] according to the daily minimal temperature during 2016.

Open in new tabDownload slide

Table 3.

Open in new tab

OR for the risk of hyperkalaemia (K >5.5 mmol/L) in all seasons

References are laboratory sampling.

Table 3.

Open in new tab

OR for the risk of hyperkalaemia (K >5.5 mmol/L) in all seasons

References are laboratory sampling.

In our study, we highlighted a relation between hyperkalaemia and blood samples done at home. It can be explained by factors such as blood sample conservation methods and type of transport, which impact kalaemia results in vitro. These two important factors are never reported by the laboratory, while they could help clinicians with kalaemia interpretation. When samples are preserved at +4°C, we observed a significant elevation of kalaemia by inhibition of NA/K adenosine triphosphate (ATP) ase = NA/L ATPase (NaK-ATPase) in <2 h [5]. A study showed a higher prevalence of hyperkalaemia in winter (15%) compared with summer (6%) in ambulatory samples. This contrasted with a stable rate in hospitalized patients [6]. We observed similar results in our cohort. It is therefore important not to forget the impact of transport duration to the laboratory, which is probably more important in samples going via a relay station. It is known that kalaemia increases if the delay before centrifugation is too long. The extracellular leak of potassium is no longer compensated by the NaK-ATPase [7]. This effect can be attenuated by a controlled ambient temperature [8]. The temperature could be improved by using a controlled ambient temperature (18–25°C). Physicians never know where blood samples are collected, nor whether the official recommendations for blood collection and storage temperatures have been applied. Thus, it is impossible to have an idea of the risk of pseudo-hyperkalaemia as opposed to real hyperkalaemia on a blood analysis result.

Many incidents of hyperkalaemia could be in fact pseudo-hyperkalaemia. It is important to keep in mind which pre-analytic factors can influence kalaemia in vitro. The temperature of sample conservation and lead time to the laboratory should more often be considered by laboratories and clinicians.

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES

Author notes