View Abstract

A6042 - Simplified 2Ts Score for Heparin-Induced Thrombocytopenia Risk
Author Block: T. Gallo1, A. Padilla-Jones2, H. Bal3, S. Curry1, K. Ramos1, R. Raschke1; 1Division of Clinical Data Analytics and Decision Support, Department of Medicine, University of Arizona College of Medicine - Phoenix, Phoenix, AZ, United States, 2Banner Research Institute, Phoenix, AZ, United States, 3Banner - University Medical Center Phoenix, Phoenix, AZ, United States.
Rationale: Heparin-induced thrombocytopenia (HIT) is an uncommon, life-threatening reaction to heparin. HIT is characterized by the onset of thrombocytopenia in a timeframe consistent with the formation of anti-platelet factor 4/heparin antibodies. The four T’s (4Ts) score is a well-validated tool to assess the risk of HIT. However, implementation of the 4Ts score into computerized clinical decision support (CDS) is not feasible due to the need for assessment of subjective variables. The goal of this project was to compare the operating characteristics of a simplified, computer-based, “2Ts score” to a 4Ts score ≥4 which indicates an intermediate or high risk of HIT and a threshold at which laboratory testing is warranted. The 2Ts score is based on thrombocytopenia and timing, the first two components of the 4Ts score.
Methods: Platelet count data were collected for inpatients initiated on heparin or enoxaparin at Banner Health hospitals from July to September, 2017. The 2Ts scores were calculated for each patient using the point system in the table below. Two methods were used to determine the timing of the platelet fall; method 1 used the time required for a 30% decrease in platelets, while method 2 used the time of the greatest decrease in platelet count. Patients were categorized into four ordinal risk categories for HIT corresponding to 2Ts scores of 0-1, 2, 3 or 4. Forty charts were randomly selected for evaluation. Investigators performed 4Ts scoring via chart review and were blinded to the 2Ts scores. Area under the ROC curve (AUROC) was analyzed, comparing each level of 2Ts score to a 4Ts score of ≥4.
Results: The AUROC for the calculated 2Ts scores was 0.87 (95% CI: 0.76-0.99) for method 1 and 0.86 (95% CI: 0.73-0.98) for method 2. A 2Ts scoring cutoff of ≥3 for a positive result (4Ts score≥4) yielded the best operating characteristics with a sensitivity of 88.2% for method 1 and 76.5% for method 2. The specificity was 78.3% for both methods. Seventeen (43%) subjects had a 4Ts score ≥4.
Conclusions: A simplified computer-based HIT risk scoring method that evaluates thrombocytopenia and timing of platelet count fall can reasonably predict a clinical 4Ts score ≥4. Two different techniques to determine the timing of the platelet count fall sometimes yielded different 2Ts scores; both methods showed strong operating characteristics and no statistically significant difference between them. Computerized CDS could potentially use a simplified 2Ts score to evaluate HIT risk.