This study compares the clinical effectiveness of a new algorithm (model-based warfarin dosing) with standard practice (doctor's own judgement and intuition) designed to determine the most appropriate anticoagulant dose of warfarin in children after congenital heart surgery.

Warfarin is widely used in children after heart surgery in order to prevent complications that may arise due to blood clots forming in unwanted places - for example in the brain causing a stroke. If too little warfarin is given, there is a risk that clots may form. If the dose is too high, bleeding will result.

Traditionally, the approach to dosing warfarin in children has been to select a starting dose according to standard guidelines and to then adjust the doses for each child by monitoring the INR (which measures how quickly blood clots). However, a person's make up as well as external factors, such as age, weight, diet and genetics can affect warfarin dosing and this makes controlling the dose and hence controlling the INR, more difficult. Recently however, a more sophisticated dosing model has been developed by researchers which takes into account some of these factors. The model is designed to help doctors select the best dose of warfarin for children by individualising prescriptions.

The aim of the proposed research is, therefore, to compare the model based warfarin dosing with the traditional approach to warfarin dosing in children who have undergone congenital heart surgery. The research will involve children who are started on warfarin for the first time and also children who are already receiving ongoing warfarin therapy.

Inclusion Criteria: -Children from birth up to 18 years with congenital heart disease who have been treated or to be treated with warfarin after undergoing reconstructive heart surgery at Glenfield Hospital, Leicester, United Kingdom.

Exclusion Criteria:

1. Patients aged over 18 years who are treated as 'adults'.
2. Children who refuse assent, parents who refuse consent.
3. Any significant disease or disorder which, in the opinion of the direct care team, may either put the participant at risk because of study participation or adversely affect the participants' ability to participate in the study.