We read with great interest the article by Odejide et al.1 Bone marrow aspiration and biopsy play a central role in the diagnosis and follow-up of many hematological and nonhematological disorders.2, 3 Obtaining adequate specimens for pathology analysis is crucial, because nondiagnostic samples often require repetition of the procedure, leading to patient discomfort, familial distress, and increased costs. The causes of bone marrow aspirate failure are diverse and occasionally complex or multifactorial. They may include factors related to the disease (eg, solid tumor infiltration, aplastic anemia, bone marrow fibrosis, or hypocellularity), the patient (eg, radiation therapy, anatomic variation, obesity, agitation, or anxiety), or the operator (eg, anxiety, improper needle location, or experience).4 In addition, it is important to keep in mind that the pathology team can also misclassify an adequate sample as nondiagnostic.4 We would like to contribute to the discussion by reporting the results of our quality-control process of morphologic assessment of bone marrow aspirates during a 6-month period between April 2013 and September 2013. At our institution, we have a team of 26 hematologists who are responsible for bone marrow aspirates and biopsies collected from children and adults in 5 outpatient units and 11 general hospitals in the São Paulo and Recife metropolitan areas of Brazil. Hematologists who performed the procedure also prepared the blood smears; they were assisted by a nurse technician in most cases. As a standard technique, we use 20-mL syringes for aspiration and obtain approximately 0.5 mL of bone marrow aspirate for the preparation of the smears for cytomorphological assessment. We also encourage operators to take advantage of a rapid panoptic stain in cases in which the macroscopic aspect of the smears is put into question. In accordance with our process, if the operator obtains an improper smear (eg, one that is devoid of bone marrow elements), and it is identified by the rapid panoptic stain, it is possible to perform a new procedure before the patient is dismissed. In addition, it is possible to discuss the addition of a bone marrow biopsy with the attending physician who had initially ordered only a bone marrow aspirate. Bone marrow smears were classified according the dilution in peripheral blood as adequate (good number of bone spicules with little dilution), suboptimal (moderate dilution), and nondiagnostic. The number of procedures per operator varied widely, with a median of 22 procedures (range, 3-110 procedures) performed over the period of the study. The majority of those aspirations included several other tests, such as immunophenotyping, karyotyping, fluorescence in situ hybridization analysis, molecular biology tests, and bone marrow trephine biopsy. The preferred site for the aspirate procedure was the posterior iliac crest at the lateral decubitus; however, puncture at the sternum, although discouraged at our institution, was performed in a few challenging cases. We performed a total of 1126 bone marrow aspirations for morphologic analysis over the 6-month period of the study. Patients had a median age of 48 years (range, birth to 96 years); 555 of the patients (49.2%) were female. We determined a rate of 16.1% suboptimal samples, and 1.3% nondiagnostic samples (samples were too diluted and/or devoid of bone marrow elements). In such circumstances, a new specimen was collected for analysis. In an effort to sustain low rates of suboptimal and nondiagnostic samples, we perform a monthly quality-control evaluation for which each team member completes a form with data regarding the tests collected and the specific conditions of the procedures that were performed. Data from those forms are audited by the leader of the hematology team and a monthly report is compiled detailing individual and team rates of suboptimal and nondiagnostic samples. For professionals with higher rates on those indicators, we offer an opportunity for recycling and one-on-one retraining. These monthly reports are sent to both the hematology and pathology teams, and we hold a regular forum for literature review and technical discussions. In conclusion, our results are similar to those reported by Odejide et al,1 with a trend toward a lower rate of nondiagnostic samples in our series. We believe that the introduction of regular quality audits and reports and a permanent forum for discussion can greatly benefit patients by reducing specimen recollection rates. The authors made no disclosures. Gustavo Loureiro, MD, PhD, Hematology Edgar G. Rizzatti, MD, PhD, Hematology Alex F. Sandes, MD, PhD, Hematology Maria de Lourdes Chauffaille, MD, PhD, Hematology Fleury Medicine and Health Sao Paulo, Brazil