Hypothesis / aims of study
Nocturia affects more than 60% of adults above 65 years old and the prevalence only increases further with age with a significant impact on the patient’s quality of life (1). As people grow older, nocturnal polyuria (NP), becomes the main cause of nocturia and is by itself caused by a loss of normal circadian rhythm of water and salt handling in the healthy aging kidney (2).
Desmopressin is the only type of medication with an evidence based effect on bothersome NP, but clinical implication remains limited in older people. Though rare, one of the most feared dose-dependent side effects is hyponatremia (<135 mmol/L), but even with the new low dose formulation, regular serum sodium monitoring remains necessary when initializing desmopressin in adults older than 65 years old (3). Therefore, we want to make desmopressin a safe treatment option for (older) patients at risk for hyponatremia, by introducing a new way of sodium monitoring. The goal is to reduce the risk of hyponatremia, enhance patient safety in desmopressin use and ultimately introduce self-monitoring of sodium levels. The first step in the aforementioned is to validate capillary sodium.
Study design, materials and methods
All adult patients (M/V/X 18-99 years old) admitted to the urology ward that were planned to receive a standard venous blood collection were eligible to participate. There were no exclusion criteria. 100 capillary blood samples were collected through a single finger prick with a BD Microtainer® contact-activated lancet, in order to collect a couple of capillary blood drops (250µL) in a lithium-heparin tube. Venous blood was collected through regular venipuncture in a lithium-heparin tube. Venous and capillary plasma sodium were analyzed by indirect ion selective electrode (ISE) method.
Statistical analysis was performed using MedCalc version 20.015. Each patient acted as its own control for the capillary and venous blood sample. The primary outcome was the agreement between capillary and venous sodium, measured by the intraclass correlation coefficient (ICC). Inter-method differences were analysed by the method of Bland and Altman and a Passing-Bablock regression. Sodium results are reported in mmol/L and shown as mean, standard deviation (S.D.) and range (min, max).
Results
Of the 100 paired blood samples, 3 samples were excluded due to insufficient capillary blood volume and hence the inability to analyse sodium levels. After sensitivity analysis of the difference between venous and capillary sodium, 1 outlier was identified and excluded from further data analysis. A total of 96 venous and capillary samples were statistically analyzed.
The median age of patients was 63 years old, ranging from 18-87 years, of which 71% were men.
The mean (S.D.) venous plasma value was 138,3 mmol/L (±3.3), ranging from 126 to 144 mmol/L, compared to a mean (S.D.) capillary plasma value of 138.5 mmol/L (±3.5), ranging from 127 to 145 mmol/L. There was no significant statistical difference observed between venous and capillary sodium (mean difference -0.23 mmol/L (0.2), p=0.374). The ICC for single measures between capillary and venous sodium was 0.82 (95% CI 0.75-0.88). The Bland-Altman plot visualizes the agreement between the two methods (Fig.1), indicating a mean difference of -0,23 mmol/L with a lower limit of agreement of -4.1 mmol/L and an upper limit of 3.4 mmol/L. The correlation between both groups was plotted by a Passing-Bablok regression (Fig.2) showing no systematic (intercept = 0.00), nor proportional (slope = 1.00) difference between venous and capillary sodium measurements.
Interpretation of results
With this research we show that there is no significant difference between venous and capillary indirect ISE sodium measurements in a real life clinical setting. It is however important to bear in mind as a clinician that there is inter- and intravariability (test-retest differences) between chemical analyses. In clinical practice, these minor differences become of particular interest in case of borderline hypo- or hypernatremia. However, in case of the latter and specifically for hyponatremia, clinical assessment will overrule borderline lab results and precocious measures will be taken to either prevent or treat hyponatremia.
This study is the first step in developing a patient centered method for preventing hyponatremia in elderly patients and other patients at risk taking desmopressin. Capillary blood collection has the advantage of being less invasive, using smaller blood volumes, having no need to gain venous access and having no hematoma formation compared to a standard venous blood collection, which is particularly of interest for older patients. In the future part of the study, we do recommend using more appropriate techniques for capillary blood collection. These methods have the benefit of being more patient friendly in use and less painful, leaving less margin for sampling error and requiring no difficult hand-eye coordination, so the patient can easily provide its own capillary blood collection without the need of another (trained) person. Patients participating actively in their own treatment and health will have a better outcome in therapy adherence. We hypothesize that patients being more involved in their treatment and having access to an easy, fast and painless method of sodium monitoring will develop less frequently hyponatremia and/or experience fewer hospital admissions, when taking desmopressin.