Comparison of Three Tests for Confirmation of the Diagnosis of Heparin Induced Thrombocytopenia.

In a recent two year period only 14% of platelet aggregation tests performed on 230 patients suspected of having heparin induced thrombocytopenia (HIT) were positive. The test (control platelet-rich plasma 0.14 mL, fresh heated patient serum 0.22 mL plus heparin 0.04 mL (1 Unit/mL final concentration) was performed using suitable controls on a commercial 4 channel aggregometer. 20% or greater increase in light transmission was a positive result. We then started a prospective study in which three different tests were done on the serum of each of 96 consecutive Hines Veterans Affairs Hospital patients who were clinically suspected of having HIT. The three tests were: platelet aggregation (PA) (same method as in the earlier study), C¹⁴ serotonin release (standard washed platelet method-SRA) and ELISA (GTI-PF4, Waukeshaw, Wisconsin), all performed on the same serum sample. The SRA and ELISA tests were performed on serum stored at −70°C. Blinded to the test results, a clinical score was determined from the record as HIGH (30% or greater fall in platelet count with recovery within 15 days off heparin), MEDIUM (same criteria but other cause present, or failure to recover platelet counts on continued heparin therapy) or LOW probability (failure to fall at least 30% or to recover from 30% or greater fall within 15 days observation). Results of the clinical scoring and laboratory tests were as follows: Clinical Probabilities: High (n=23), Medium (N=33), Low (n=40). Of the 56 who were High or Medium probability, 41% were High and 59% were Medium probability. Analysis of the results of the laboratory tests for sensitivity and specificity for each category revealed the following:

Conclusions: Sensitivity of the PA test was greater than either of the other two tests, with specificities for the PA and SRA being nearly equal. The ELISA test was relatively insensitive and less specific than then other two.