| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Institut für Laboratoriumsmedizin, Deutsches Herzzentrum München, Klinik an der TU München, Munich, Germany
aAddress correspondence to this author at: Institut für Laboratoriumsmedizin, Deutsches Herzzentrum München, Klinik an der TU München, Lazarettstrasse 36, D-80636 Munich, Germany. Fax 49-89-1218-1013; e-mail braun{at}dhm.mhn.de.
To the Editor:
Electrolyte measurements predominantly use ion-selective electrodes (ISEs) for sodium, potassium, and chloride. In contrast to the reliability of the sodium and potassium electrodes, erroneous chloride ISE values have occasionally been reported (1)(2)(3). We recently observed an increasing number of falsely high chloride concentrations in patient plasma samples. These results appeared at unpredictable intervals after replacement of the electrode on a Cobas Integra analyzer (Roche Diagnostics). Repeat analyses with a coulometric method yielded significantly lower results.
The falsely high values appeared not to have been caused by a general low selectivity of the chloride electrode, which responds to other ions such as iodide, thiocyanate, nitrate, and bromide (1)(2)(3). More likely was a time-dependent deterioration of the chloride electrode, for which the manufacturer claims a lifetime of 3 months but which needed to be changed more frequently. To attain reliable chloride ISE results, electrodes could be changed on a more frequent schedule (e.g., weekly), but that approach is expensive. Hence we aimed to introduce a reliable and inexpensive indicator of deterioration of the electrode.
Routine blood specimen were collected in tubes containing ammonium heparin (Sarstedt) and immediately centrifuged after arrival in our laboratory. Selectivity of the ISE was checked by comparison with chloride concentrations measured by coulometric titration on a Model 925 chloride analyzer (Corning). For quality control we used DuotrolTM abnormal (Biomed; target value, 127 mmol/L; control A) and PrecinormTM U (Roche Diagnostics; target value, 111 mmol/L; control B). As a selectivity control we used ISE solution 1 (Roche Diagnostics) supplemented by an aqueous KSCN solution. Briefly, we added 25 µL of 5 mmol/L aqueous KSCN (Merck), corresponding to 12.5 µmoles of KSCN, to 10 mL of ISE solution 1, which contains 115 mmol/L NaCl. Because of the high sensitivity of the chloride electrode for KSCN, this solution mimics a chloride concentration of 
120 mmol/L. A loss of selectivity should lead to markedly increased values.
After placement of a new electrode, results for the selectivity control solution gradually increased with a final peak at 2–6 weeks (Fig. 1
). After replacement of the electrode, selectivity control solution results returned to the expected concentration range (Fig. 1
). Results for control A showed this effect less well and could not demonstrate unambiguously deterioration of the electrode (Fig. 1
). Control B varied statistically around the target value (Fig. 1
) and therefore was not a suitable indicator.
|
The time dependence of the deterioration was consistent with an "aging" of the electrode, which could be the reason for the emerging loss of selectivity. One possible explanation for this effect could be the coating of the membrane with protein, particularly fibrinogen. Daily preventive maintenance performed according to the manufacturer’s instructions did not improve our results sufficiently. After such a procedure the selectivity of a worsening electrode was reconstituted for some hours only. We believe that this short-lived reconstitution was attributable to partial peeling of the electrode surface. After the maintenance procedure, falsely increased results for patient samples were detected more frequently than in the first hours after reconstitution. Repeat analyses with the coulometric reference method revealed significantly lower chloride concentrations [mean (SD) difference, 3 (3.7) mmol/L] in most cases. In 3% of those patient samples, the differences were >10 mmol/L.
Replacing the chloride electrode whenever the result for selectivity control solution exceeded its 3 SD limit reduced the rate of falsely high (difference for ISE minus coulometry >10 mmol/L) results to 0.3%. The mean difference between ISE and coulometry has been halved [mean (SD) difference, 1.7 (1.75) mmol/L]. Thus the introduction of this selectivity control solution offers a novel and inexpensive way to achieve a maximum useful lifetime of the chloride electrode.
Acknowledgments
Roche Diagnostics supported our work by substitution of defective electrodes free of charge.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and control B (
)] and chloride equivalents in the selectivity control solution (•).