P. H. J. van der Voort

Direct sodium measurement prevents underestimation of hyponatremia in critically ill patients

Dear Editor, Maintaining an optimal fluid and electrolyte balance is an important task for ICU professionals. Clinical awareness of pseudohyponatremia is commonly known and integrated into guidelines [1]. However, the opposite condition, i.e., pseudohypernatremia, is currently underestimated and warrants attention [2]. Most chemistry analyzers measure electrolyte concentration after obtaining a 6 to 46 times dilution. For the final calculation of the electrolyte concentration it is assumed that 7 % of the volume consists of fats and proteins, the remaining 93 % is the water fraction in which electrolytes are found. Therefore, in critically ill patients with often a low total protein/albumin concentration, sodium concentration is overestimated when using an indirect method [3]. When direct sodium measurement with blood gas analyzers is used this problem is not present. In this study we compare the difference between direct and indirect sodium measurement with the total protein and albumin concentration. Included were consecutive patients who agreed that their blood samples may be used for research. The results from 2,336 patients were reviewed in retrospect. Indirect measurement of sodium, albumin, and total protein concentration was performed on a Cobas System (Roche, the Netherlands). For the indirect sodium measurement a 1:31 dilution is used. Direct measurement of sodium was performed on a blood gas analyzer (Radiometer, the Netherlands). Direct and indirect sodium measurements were performed within a 1-h time frame. We arbitrarily defined 4 mmol/l as the clinically relevant difference. In 198 of 2,336 patients (8.5 %) indirect sodium results were more than 4 mmol/l higher than results obtained with the direct method. In 80/198 patients (40 %) this resulted in a pseudohypernatremia ([145 mmol/l, 21 patients) or pseudonormonatremia (135–145 mmol/l, 59 patients) based on the indirect method whereas direct measurement revealed a normoor hyponatremia respectively. In 407 patients the total protein concentration was measured, showing a strong significant correlation (linear regression analysis) with the difference between direct and indirect sodium measurement (Fig. 1). Similar results were found for the albumin concentration (data not shown). In previous papers pseudohypernatremia (or pseudonormonatremia) is mainly described in critically ill patients in ICU departments [2, 3]. However, in our research 21 % of samples in which a difference of more than 4 mmol/l was found derived form the ICU department, the remaining 79 % were from the emergency department (43 %) and the general wards (36 %). This emphasizes that pseudohypernatremia is not only found in ICUs, but also in the general hospital population. Thus, measurements of electrolytes with a blood gas analyzer is more reliable when a pseudohypernatremia is suspected. However, in contrast to chemistry analyzers, pre-analytic errors such as high potassium due to hemolysis or underestimation of positively charged electrolytes due to heparin excess can be missed with blood gas analysis [4]. In critical ill patients a catabolic state and consequently low total protein concentration are commonly found. Tight regulation of fluid

Uncommon complications during chest tube placement: a potential role of tube material.

mained unchanged. A left-sided thoracotomy revealed the chest tube in the left pulmonary artery. The tube was removed with reconstruction of the pulmonary artery. The testicular mass appeared to be a pneumococcal abscess. After several weeks the patient died because of irreversible multiple organ failure. Chest tube placement is frequently associated with complications [4]. Pulmonary artery cannulation is extremely rare but was recently reported by Jaillard et al. [1]. Surgical removal and repair are obligatory. However, our report shows that in a stable condition surgical intervention can be postponed to allow necessary preparations to be made. The second point that we want to address is the tube material. We changed chest tubes (from polyvinylchloride to polyurethane) and experienced two major complications shortly thereafter, whereas no complications had been observed in previous years. One complication is described here; the other complication was chest tube placement in the lung itself leading to intrapulmonary bleeding. In both patients the tube was inserted without trocar. This suggests that the chest tube material and its characteristics, such as stiffness and design of the tip, may play a role. Figure 4 (Electronic Supplementary Material) shows that the polyvinylchloride tube is more flexible than the polyurethane tube. In conclusion, we confirm the pulmonary artery cannulation as a complication of chest tube placement. However, emergency operation is not always necessary. Furthermore, the chest tube material may be an additional risk factor for a fausse route even when a trocar is not used.