| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
1 Department of Pathology and Laboratory Medicine Tan Tock Seng Hospital 11 Jalan Tan Tock Seng Singapore 308433
aFax 65-253-6507; e-mail Robert Hawkins{at}ttsh.com.sg.
To the Editor:
Because the increase in plasma potassium in hemolyzed specimens is linearly related to hemoglobin concentration (1), knowledge of the potassium/hemoglobin ratio can allow estimation of the magnitude of increase in plasma potassium. Previous estimates of the potassium/hemoglobin ratio have evaluated only a few apparently healthy volunteers (2)(3)(4). Interindividual differences, particularly in diseased populations, could produce variability in correction factor determination. This letter describes the interindividual variability in the potassium/hemoglobin ratio for in vitro hemolysis in 100 hospitalized patients.
Erythrocyte hemolysates were prepared by the procedure of Meites (5). Briefly, the erythrocytes from lithium-heparin-blood samples from 100 medical and surgical inpatients were separated by centrifugation at 2500g within 1 h of collection, washed three times with isotonic saline, and then lysed by the addition of an equal volume of distilled water. After freezing overnight and thawing, the hemolysates were centrifuged at 2500g for 30 min to remove the stroma.
For each of the 100 patient hemolysate samples, 100 µL of hemolysate was added to 2 mL of a common pooled serum to produce a severely hemolyzed sample. This was diluted in the common pooled serum to form a series of nine equispaced, linearly related samples. Each series of the diluted samples was analyzed in duplicate (same batch) for potassium (reported to two decimal places with an indirect ion-selective electrode) and hemolysis index (multiwavelength spectrophotometry with saline as the reagent) on a Roche Diagnostics 917 clinical chemistry analyzer. Pearson correlation coefficient calculation and Deming model linear regression analysis by Analyse-It add-on software were performed for each dilution series, and the slope of the potassium/hemoglobin relationship was noted.
The range of hemoglobin concentrations tested was 50–6000 mg/L. Within-batch CVs were as follows: potassium, 0.4%; hemoglobin, 2%. The range of correlation coefficients for the 100 dilution series was 0.99–1.00, confirming the linear relationship between potassium and hemoglobin. The median potassium/hemoglobin (mmol/g) ratio was 28.0, with a mean of 28.4, and the 95% population limits were 21.0 and 34.5 (Fig. 1
). Although the population range included published individual experimental values (1), the median and mean were lower than other individual estimates (1)(2)(3)(4)(6)(7). There was also a substantial difference between the 2.5 and 97.5 percentiles, which would create difficulty in the choice of ratio when correcting potassium values for hemolysis in a given case. Thus, because of the variability in the potassium/hemoglobin ratio demonstrated here, the correction of potassium increase in specimens exhibiting in vitro hemolysis is not advocated.
|
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  G. Dimeski Effects of Hemolysis on the Roche Ammonia Method for Hitachi Analyzers Clin. Chem., May 1, 2004; 50(5): 976 - 977. [Full Text] [PDF]
|
|
|
|
|
|
R. Hawkins Measurement of Whole-Blood Potassium--Is It Clinically Safe? Clin. Chem., December 1, 2003; 49(12): 2105 - 2106. [Full Text] [PDF]
|
|
Read all eLetters
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
