Predicting and evaluating the effect of bivalirudin in cardiac surgical patients

Bivalirudin, used in patients with heparin-induced thrombocytopenia, is a direct thrombin inhibitor. Since it is a rarely used drug, clinical experience with its dosing is sparse. We develop two approaches to predict the Partial Thromboplastin Time (PTT) based on bivalirudin infusion rates. The first approach is model free and utilizes regularized regression. It is flexible enough to be used as predictors bivalirudin infusion rates measured over several time instances before the time at which a PTT prediction is sought. The second approach is model based and proposes a specific model for obtaining PTT which uses a shorter history of the past measurements. We learn population-wide model parameters by solving a nonlinear optimization problem. We also devise an adaptive algorithm based on the extended Kalman filter that can adapt model parameters to individual patients. The latter adaptive model emerges as the most promising as it yields reduced mean error compared to the model-free approach. The model accuracy we demonstrate on actual patient measurements is sufficient to be useful in guiding the optimal therapy.