Inability of the activated partial thromboplastin time to predict heparin levels. Time to reassess guidelines for heparin assays

BACKGROUND: In treating venous thromboembolic disorders, patient outcomes appear to correlate with heparin levels. Due to pharmacokinetic and pharmacodynamic variations, a relationship between heparin dose and level cannot be reliably predicted in individual patients. Some patients have low heparin levels despite therapeutic activated partial thromboplastin times (aPTTs), which may increase their risk for recurrent thromboembolism. Patients with high heparin requirements appear to have fewer bleeding episodes with heparin level-guided therapy. The aPTT does not reliably correlate with heparin blood concentrations or antithrombotic effects. Consequently, heparin therapy monitored with heparin levels may be more effective and safer. OBJECTIVES: To prospectively determine whether (1) the aPTT therapeutic range adequately predicts heparin levels in 38 patients used to establish the therapeutic aPTT range as is currently recommended and (2) whether 3 paired sets of aPTT-antifactor Xa levels provide the basis for using aPTTs to predict subsequent heparin levels in individual patients (n = 27) receiving intravenous heparin for coronary artery disease or venous thromboembolic disease. RESULTS: In the therapeutic aPTT range established, the R2 value for the relationship was 0.4. Prediction intervals were wide. For an aPTT of 60 seconds, the 95% prediction interval estimates were heparin levels of 0.05 to 1.0 U/mL. In individual patients, the aPTT-antifactor Xa relationship had an average R2 value of 0.75. There was no consistent relationship between the aPTT and anti-factor Xa level in a significant number of patients. CONCLUSIONS: The aPTT does not appear to be a useful surrogate for heparin levels. These findings suggest that the current recommendations on the use of heparin levels should be expanded.