Automated spectrophotometric heparin assays. Comparison of methods

Three automated spectrophotometric heparin assays were investigated. The day-to-day reproducibilities in routine laboratory use were compared with two commercial manual kits for heparin determination. Regression analysis of the activated partial thromboplastin time (APTT) on results of any of the heparin assays shows that the heparin concentration cannot be deduced from the APTT values found in patients receiving heparin. The automated heparin assays that employ thrombin and Chromozym-Th or S-2238 were found to be most suitable for routine heparin determination. Heparin concentrations obtained from assays based on factor Xa inactivation were not significantly different from those employing thrombin (p less than 0.01), but revealed a wider standard deviation. The relationship between APTT and heparin level found was not related to the plasma antithrombin III concentration. The extra antithrombin III that is added in the assays had to be freed of heparin neutralising activity to obtain reliable estimates of the heparin concentration in the low range (0-200 U/l).