Lack of Value of Clinical Predictors of Heparin Induced Thrombocytopenia (HIT) in Complex Patients.

HIT is a devastating complication of heparin use that can result in thrombosis and death. However, it can be difficult to diagnose. Although the clinical hallmark of HIT is a drop in platelet counts in patients receiving heparin, in complex patients there may be many confounding reasons for thrombocytopenia and thrombosis. One proposed diagnostic tool for HIT is to quantitate the percent drop in platelet count from the highest level. Also Warkentin has proposed the “4T’s” scoring system where 4 clinical indicators (% drop in platelet count, timing of heparin dosing, presence of thrombosis, and lack of other causes of thrombocytopenia) are used to derive a score to indicate High, Intermediate, or Low probability of HIT.(Curr Hem Rep, 2, 148–157) We sought to verify the predictive value of these clinical systems. Over a 30 month period, we reviewed the charts of 544 patients who were suspected of having HIT. 61% of these patients were in intensive care units, 47% were peri-operative, and 22% were cardiac patients. HIT was diagnosed by a combination of review of the clinical course plus results of platelet activation assay, and with at least 2 faculty hematologist agreeing on the diagnosis. Data was abstracted concerning platelet count, timing and use of heparin, presence of thrombosis, and other causes of thrombocytopenia including infection and major surgery. 18 out of 544 patients (3.3%) had heparin induced thrombocytopenia. 16 of 18 patients received therapeutic standard heparin and 2 patients received only prophylactic doses of LMWH. There was no difference in nadir platelet count nor percent drop in HIT (75±79,000/Ul, 65±26%) vs those without HIT (90±71,000/uL, 53±24%). Two major causes of thrombocytopenia in this population were infections and surgery. 175 patients (32%) had positive blood cultures with a mean platelet drop of 55±23% preceding the positive cultures. 47% had major surgery with a mean platelet drop afterwards of 60±22%. A drop in platelets of over 50% had a 82% sensitivity for HIT but only a 43% specificity. Using different cut-offs did not reveal a value with better predictability. Interestingly, only 2 (11%) of the HIT patients scored on the 4T’s as High probability with 39% being Intermediate and 50% as Low. This was due to presence of other causes of thrombocytopenia in patients and the fact that 30% of the HIT patients were on heparin longer that 10 days. 4% of the non-HIT group scored as High and 28% as Intermediate but close review of their subsequent clinic course did not reveal evidence for HIT. Simple clinical tools for diagnosing HIT are of limited utility in ill and complex patients due to other competing causes of thrombocytopenia and the potential for long duration of heparin exposure.