Steroid therapy is increasingly being used as initial therapy for ITP because of concern regarding the safety of blood products. Bone Marrow Aspiration (BMA) is routinely performed prior to starting steroid therapy in children with idiopathic thrombocytopenia, primarily to rule out leukemia. We constructed a decision tree for the initial management of a child presenting with classic ITP. The three strategies are: 1. All patients receive an initial BMA; 2. Only patients at high risk for leukemia receive a BMA; or 3. No patients receive a BMA. High risk criteria include any of: platelet count>50×109/L; Haemoglobin <100g/L (6m-12m) or <110g/L(>lyr); white blood cell count <5 (6m-6yr) or <4×109/L(>6yr); or absolute neutrophil count <1.5 (6m-6yr) or<2×109/L (>6yr). The analysis considers the short-term morbidity of BMA versus the anxiety of not having a BMA, as well as the long-term implications of inappropriately giving steroids to a patient with undiagnosed leukemia.

The base case results were as follows: 1. BMA all - 64.179 quality of life adjusted years (QALYs); 2. High Risk - 64.180 QALYs; 3. BMA none - 64.130 QALYs. These results indicate a toss-up between strategy 1 and 2, as there is less than a one day difference in QALYs. Both options are slightly preferred over strategy 3, with a 18 day difference. The choice of preferred strategy is sensitive to: the time and utility of anxiety over the uncertainty of the diagnosis; the sensitivity of the leukemia risk assessment; the increased chance of dying from leukemia after inappropriately receiving steroids; and the time of BMA disutility.

Performing a BMA only in children with high risk laboratory features for leukemia appears to be a clinically valid alternative to performing a BMA on all children presenting with features of ITP. If economic considerations are taken into account, the risk strategy appears to be the preferred initial management strategy. More reliable data on all of the sensitive varaibles is needed before a clear recommendation can be made.