Personalised warfarin dosing in children after congenital heart surgery using the model-based approach
Oral anticoagulation with warfarin represents a major challenge to successful drug therapy in children. The aims of this study was to investigate the implementation in routine clinical practice, personalised warfarin dosing using a PK/PD model, in children after congenital heart surgery and to explore the experience of patients/parents and health care professionals with managing long-term warfarin treatment as well as their experience with the model-based dosing approach. The predictive performance of the PK/PD model was first validated using retrospectively collected data from a cohort of 60 children on long-term warfarin treatment. Seventy percent of the predicted doses were ideal with bias of -0.10 and precision of 0.19. A prospective interventional quantitative study was then conducted in two groups of children. Group 1 included 5 patients who started warfarin treatment for the first time after cardiac surgery. For the case subjects compared to the controls, the median time to achieve the first therapeutic INR values was longer (5 vs 2 days), the median time to stable anticoagulation was shorter (29.0 vs 96.5 days), the median time to over-anticoagulation was longer (15.0 vs 4.0 days), the median percentage of the INR observations within the target range (%ITR) was higher (70% vs 47.4%), the median percentage of time in therapeutic range (%TTR) was higher (83.4% vs 62.3%), the median frequency of INR measurements per month was comparable (5.0 vs 6.3) and the median frequency of dose alterations was also comparable (20.0 vs 21.0). Group 2 included 26 patients who were established on maintenance warfarin therapy. For the model-based dosing phase compared to the traditional dosing phase, the mean %ITR was 68.82% compared to 67.9% (p=0.84) and the mean %TTR was 85.47% compared to 80.2% (p=0.09). After excluding 5 patients who experienced medical issues during either phases of treatment, the mean %ITR was 71.28% compared to 65.51% (p=0.22) and the median %TTR was 91.8% compared to phase 77.3 % (p=0.03). The median frequency of INR measurements per month was 2.3 compared to 1.9 (p=0.08) and the median frequency of dose alteration was 6.5 compared to 2.5 (p=0.02). Patients with Fontan circulation had significantly higher %TTR during the model-based dosing phase than during the traditional dosing phase after excluding the 5 patients with medical issues (p=0.02). Semi-structured interviews were conducted with 3 doctors, 2 cardiac liaison nurses and four family representatives. Three thematic areas emerged from the doctors’ interviews; ‘medical and clinical knowledge’, ‘INR monitoring’ and ‘dose decision’. Four thematic areas emerged from the nurses’ interviews; ‘role of the cardiac liaison nurses in managing warfarin treatment’, ‘INR monitoring’, ‘dose decision’ and ‘adherence to the prescribed regimen’. Three thematic areas emerged from the families’ interviews; ‘managing warfarin treatment and the coping mechanism’, ‘warfarin dose decision’ and ‘adherence to warfarin treatment’. Both doctors and nurses found the new dosing approach useful and acceptable in patients with stable medical condition. Additionally, three of the families favoured that dosing be performed by a professional experienced with warfarin treatment regardless of the method used. This study has shown that model-based dosing can improve the anticoagulation control of warfarin and hence reduce its adverse events in children after congenital heart surgery. Further work is required to establish the clinical effectiveness and cost-effectiveness of the new dosing approach in this group of children.
- PhD