Evaluation of a blood ammonia checker system
Hsu, W.S.; Chen, J.S.
Journal of the Formosan Medical Association 84(9): 1043-1050
1985
ISSN/ISBN: 0371-7682 PMID: 3867726 Document Number: 252627
A simple and rapid microdiffusion method for measuring blood ammonia concentration with the Blood Ammonia Checker (BAC) system was presented. Within-run and between-run precision were quite acceptable, with coefficients of variation (CV) between 3.5% to 6.4%. Although the mean recovery was only 86.0 .+-. 5.4%, the final concentration included not only the value recovered but also the ammonia generated in vitro. Eighty-four samples from hospitalized patients were analyzed by using both BAC system (y) and Ion Exchange Resin method (x). An excellent correlation was found between the 2 sets of results (r = 0.977, y = 0.880 x + 16.356). Comparative ammonia concentrations between the whole blood (x) and plasma (y) by using BAC system were investigated. The 2 sets of results correlated very well (r = 0.985, y = 0.953 x -2.374, n = 61). Therefore, if we were too busy to measure the ammonia concentration immediately, we could separate the plasma and freeze it at -20.degree. C till the time of measurement. Hemolysis caused significant over-estimation (7.4%) of the ammonia concentration. However, the extent of overestimation was so small that it would not seriously affect the interpretation of the ammonia concentration. The longer the reaction time, the higher the ammonia concentration. Hence the reaction time must be controlled accurately. When we used skin puncture for sampling, we could not sterilize with povidone iodine; even if alcohol was used for sterilization, we should wait for complete evaporation before sampling. Because povidone iodine could react with reagent and then liberate alkaline gas which could seriously overestimate the ammonia concentration; the white and opaque polypropylene film in the reagent plate was become transparent by alcohol, so the color for measuring could be affected. The precision of the BAC (reflectance meter) with CV 0.2% was very well. But there were variations among different BACs. Accordingly, when we choose one BAC, we should set up our own normal reference value. In this study our normal reference values were 21 .apprx. 73 N-.mu.g/dl. Our results indicate that BAC system really is a simple, convenient method with few interference and that it can be accepted for measuring blood ammonia concentration in a clinical laboratory.