Bone-marrow transplantation is considered the treatment of choice for pediatric patients with recurring acute lymphoblastic leukemia, provided that a suitable donor is available. Many prognostic factors are known that help to predict the likely outcome of transplantation. We have implemented a system that applies probabilistic reasoning to the available data about individual patients to help determine the risk of recurrence and morbidity after transplantation, and to predict life expectancy. The resulting predictions can be used to decide whether marrow transplantation is the most desirable treatment modality for the patient.
