SESSION TYPE: Hematologic Problems in the ICU
PRESENTED ON: Monday, October 22, 2012 at 11:15 AM - 12:30 PM
PURPOSE: Diagnosis of Heparin Induced Thrombocytopenia (HIT) is clinical and can be further confirmed with laboratory testing. Serotonin Release Assay (SRA) is very specific but not available when clinical decisions are being made. This study will test the accuracy of two clinical scores and Heparin Dependent Antibody Assay (HDAA) in critically ill patients in both medical and surgical ICUs.
METHODS: Retrospective chart review of adult critically ill patients who had both HDAA and SRA testing. Based on the SRA as the gold standard we assessed the accuracy of two clinical scores; 4 T and HIT Expert Probability (HEP), and the Optical Density (OD) of HDAA. To obtain clinical scores, charts were reviewed up to the time of testing for HDAA and SRA, and then after that point for platelet trend and thrombosis.
RESULTS: 73 charts were reviewed, SRA was positive in 22%, 46% of which were surgical. Heparin was used for therapeutic indications more in HIT patients. High 4 T was more specific (90.5 %) than Hematology consult (40%), and HEP score in all, medical and surgical patients 45%, 55%, 23%, respectively. OD of 0.499 had 100% sensitivity, there was no difference in the accuracy of HDAA between medical and surgical patients, higher cut off OD had higher specificity; 21%, 64%, 81%, 94% for OD 0.499, 1.0, 1.5, 2.0 respectively. OD of 2.0, 1.5, and 1.0 were more specific than OD 0.499, OD 2.0 was more specific than OD 1.0, and OD 0.499 was more sensitive than OD 1.5. Thrombosis and upward platelet trend after stopping heparin together were more specific and less sensitive for HIT diagnosis compared to either.
CONCLUSIONS: High 4 T was more specific for HIT diagnosis than HEP score in medical and surgical patient. HDAA with OD 0.499 reliably excludes HIT, but higher cut off is needed for higher specificity.
CLINICAL IMPLICATIONS: high 4 T score and OD with cut off higher than 1.5 strongly suggest the diagnosis of HIT.
DISCLOSURE: The following authors have nothing to disclose: Majdi Hamarshi, Navin Bajaj, Michelle Haines
No Product/Research Disclosure InformationUMKC, Kansas City, MO