Error estimation in the quantification of alkaline phosphatase isoenzymes by selective inhibition methods

A method for calculating, by selective inhibition, the activities in serum of isoenzymes of alkaline phosphatase (EC 3.1.3.1) originating from bone, liver, intestine, and placenta produced results for a sample of patients for which the imprecision was five- to 25-fold higher than that of the alkaline phosphatase method used--too imprecise for routine clinical use. Error analysis by direct calculation and by Monte Carlo estimation revealed that the algorithm used in the method completely accounted for the increase in imprecision of the isoenzyme estimations. I recommend that all methods involving such algorithms or the principle of multicomponent analysis should undergo a thorough error analysis by use of the techniques described here, to obtain an estimate of the increase in imprecision that is ascribable to the particular numerical technique used.