Anticoagulation in thrombocytopenic patients with hematological malignancy: A multinational clinical vignette-based experiment

Background: Thrombocytopenia in cancer patients with an indication for anticoagulation poses a unique clinical challenge. There are guidelines for the setting of venous thromboembolism but not atrial fibrillation (AF). Evidence is lacking and current practice is unclear. Objective: To identify patient and physician characteristics associated with anticoagulation management in hematological malignancy and thrombocytopenia. Methods: A clinical vignette-based experiment was designed. Eleven hematologists were interviewed, identifying 5 relevant variable categories with 2–5 options each. Thirty hypothetical vignettes were generated. Each physician received 5 vignettes and selected a management strategy (hold anticoagulation; no change; transfuse platelets; modify type/dose). The survey was distributed to hematologists and thrombosis specialists in 3 countries. Poisson regression models with cluster robust variance estimates were used to calculate relative risks for using one management option over the other, for each variable in comparison to a reference variable. Results: 168 physicians answered 774 cases and reported continuing anticoagulation for venous thromboembolism or AF in 607 (78%) cases, usually with dose reduction or platelet transfusion support. Overall, management was affected by platelet count, anticoagulation indication, time since indication, type of hematological disease and treatment, and prior major bleeding, as well as physician demographics and practice setting. The CHA2DS2-VASc score and time since AF diagnosis affected anticoagulation management in AF. Conclusion: This study indicates what the widely accepted management strategies are. These strategies, and possibly others, should be assessed prospectively to ascertain effectiveness. The decision process is intricate and compatible with current venous thromboembolism guidelines.