The diagnosis of heparin-induced thrombocytopenia (HIT) is problematic in the surgical ICU, as there are multiple potential explanations for thrombocytopenia. We conducted a study to assess the incidence, clinical presentation, and outcome of HIT in a cardiothoracic surgical ICU.
From January 2005 to December 2010, all patients with suspicion of HIT were prospectively identified, and data were collected retrospectively. Detection of anti-PF4/heparin antibodies and functional assays were systematically performed.
During the study period, 5,949 patients were admitted to the ICU (2,751 after cardiac surgery and 3,198 after thoracic surgery), of whom 101 were suspected to have HIT (1.7% [95% CI, 1.4%-2.0%]). Suspicion of HIT occurred at a median of 5 (4-9) days after ICU admission. Diagnosis was confirmed in 28 of 5,949 patients (0.47% [95% CI, 0.33%-0.68%]). Thrombosis was detected in 14 patients with HIT (50%) and in 12 patients without HIT (16%) (P = .0006). After receiver operating characteristic analysis (area under curve = 0.78 ± 0.06), a 4Ts score ≥ 5 had a sensitivity of 86% and a specificity of 70%. Course of platelet count was similar between the two groups. Six patients (21%) with HIT and 20 (27%) without died (P = .77).
Even with a prospective platelet monitoring protocol, suspicion for HIT arose in < 2% of patients in a cardiothoracic ICU. Most were found to have other causes of thrombocytopenia, with HIT confirmed in 28 of 101 suspected cases (0.47% of all patients in the ICU). The 4Ts score may have value by identifying patients who should have laboratory testing performed. The mortality of patients with HIT was not different from other very ill thrombocytopenic patients in the ICU.