Predictors of Heparin-Induced Thrombocytopenia in Adult Cardiac Surgery Patients

Background: The 4Ts, HIT-Expert Probability (HEP), and Post-Cardiopulmonary Bypass (CPB) screening tools for heparin-induced thrombocytopenia (HIT) have not been validated in cardiac surgery patients. Evidence remains unclear regarding which screening tool most accurately predicts HIT in this population.

Methods: HIT-positive and HIT-negative patients who underwent on-pump cardiac surgery within a six-year period were matched 1:2 in a case-control design. Each patient was scored with the 4Ts, HEP, and CPB tools. Sensitivities and specificities of each tool were calculated using standard cut-offs. The Youden method was utilized to determine optimal cut-offs in receiver operating characteristic (ROC) curves of each score, then sensitivities and specificities were recalculated. A multivariable logistic regression was performed to determine the association of scoring tool components and relevant clinical characteristics with HIT.

Results: Using standard cut-offs for the scoring tools, sensitivities for the CPB, HEP, and 4Ts tools were 100%, 93.9%, and 69.4%, respectively. Specificities were 51%, 49%, and 71.4%, respectively. Using the Youden method-derived optimal cut-offs, sensitivity of the CPB score remained 100% with improved specificity to 88.9%. Sensitivity of the 4Ts score declined to 51% and specificity improved to 93.9%. Pattern of platelet decline, absence of clinically significant bleed, body mass index, coronary artery bypass graft surgery, and postoperative heparin duration were significantly associated with HIT.

Conclusions: The 4Ts score has limited utility in cardiac surgery patients, whereas the CPB and HEP scores with standard cut-offs demonstrated high sensitivity but low specificity. A cut-off of 3 points or higher on the CPB score could increase specificity while preserving high sensitivity.