Idiopathic thrombocytopenic purpura is a relatively common pediatric disorder. It tends to follow a self-limited course, but treatment with gamma globulin or corticosteroids is often instituted to prevent bleeding complications. It has been standard practice to perform a bone marrow aspiration to rule out leukemia as part of the initial work-up of a child with an abnormally low platelet count. Several older studies questioned the need for this procedure but did not provide convincing evidence because of small sample size or methodology. Calpin and associates performed a retrospective study to determine whether any diagnoses of leukemia were discovered in a large series of bone marrow aspiration samples that were collected from children diagnosed with acute idiopathic thrombocytopenic purpura who had typical hematologic features of this disease.

The reports of all bone marrow aspirations performed at the Hospital for Sick Children in Toronto from January 1, 1984, until May 31, 1996, were reviewed. This cohort included both inpatient and outpatient reports on bone marrow aspirations that were performed to confirm a provisional diagnosis of idiopathic thrombocytopenic purpura. The peripheral blood smear and complete blood count performed nearest to the time of the bone marrow aspiration were also reviewed to see if the patient had typical features of idiopathic thrombocytopenic purpura. “Typical features” included a normal hemoglobin, a normal total white blood cell and neutrophil count, and a platelet count of less than 50,000 × 10³ per μL (50,000 × 10⁹ per L). The researchers then analyzed the outcomes of all bone marrow aspirations performed in children to confirm idiopathic thrombocytopenic purpura with typical laboratory features to see if any other diagnosis was found. In a similar fashion, the final diagnoses of all children who underwent bone marrow aspiration to confirm idiopathic thrombocytopenic purpura but who had abnormal laboratory features were also obtained.

During the time reviewed, 7,612 bone marrow aspirations were performed. Of these, 484 (6.4 percent) were performed to confirm a provisional diagnosis of idiopathic thrombocytopenic purpura. During the same 12-year period, 672 children were admitted to the institution with a primary diagnosis of acute idiopathic thrombocytopenic purpura. The study estimate indicated that 72 percent of children with idiopathic thrombocytopenic purpura underwent a bone marrow aspiration. Among 332 children with a platelet count of less than 50,000 × 10³ per μL (50,000 × 10⁹ per L) but otherwise normal or typical hematologic parameters, no cases of leukemia were diagnosed. In the group of 135 children with atypical hematologic features for idiopathic thrombocytopenic purpura, three children were diagnosed with leukemia.All three of these children had at least two atypical hematologic values in the complete blood cell count, and all had a clinical presentation not consistent with idiopathic thrombocytopenic purpura. Eight additional children in the group with atypical laboratory results were found to have aplastic anemia. The authors conclude that in children who present with thrombocytopenia and otherwise typical hematologic features of idiopathic thrombocytopenia—platelet count less than 50,000 × 10³ per μL (50,000 × 10⁹ per L) with a normal hemoglobin, white blood cell count and neutrophil count—the risk of missing a diagnosis of leukemia is less than 1 percent. They also cite the significant cost of this procedure ($700 to $800) and associated morbidity, including pain, bleeding and infection.