Evaluation of a Reflex Testing Algorithm for Suspected Heparin-Induced Thrombocytopenia.

Abstract

We implemented a policy of reflex serotonin-release assay (SRA) testing for all patients with a positive heparin-induced thrombocytopenia (HIT) immunoassay. We identified all patients who had SRA testing sent as a consequence of a positive HIT immunoassay test. We reviewed charts of patients to calculate the 4Ts clinical score, determined the effect of testing on clinical management, and documented the change in utilization of direct thrombin inhibitors (DTIs). The likelihood of a positive SRA varied with the optical density (OD) of the immunoassay. The performance of the immunoglobulin G (IgG)-specific and polytypic enzyme-linked immunosorbent assay was not statistically different. Both OD and 4Ts score correlated with the likelihood of a positive SRA but demonstrated poor specificity. Discontinuation of DTIs in patients with negative SRAs resulted in decreased drug utilization. The IgG-specific HIT immunoassay OD correlates with the likelihood of a positive SRA but does not achieve high specificity. The reflex testing algorithm allows for definitive classification of patients, and the cost of such a reflex testing program may be offset by decreased utilization of DTIs.