Eight Years Experience in External Quality Assessment Scheme (EQAS) for Peripheral Blood Morphology

Introduction: The Asociación Española de Hematología y Hemoterapia (AEHH) organizes a National Haematology EQAS and the Hospital Clinic of the University of Barcelona (HCB) serves as the EQAS Coordination Centre (CC). We report results from the eight year long study and the evaluation of laboratory performance for blood smear interpretation by the Spanish haematology EQAS.

Objectives: To evaluate:

1. The overall performance of different laboratories in identifying pathological blood cells and

2. If their performance improves for following samples.

Material and methods: During the period 2000–2008, replicates of 30 different stained blood smears were sent to 604 EQAS participants for cell morphology evaluation. The samples included blood films from patients with the following diagnoses: acute myeloid leukemia (AML: 8), acute lymphoblastic leukemia (ALL: 3), myeloproliferative disease (MPD: 7), Sézary disease: 3, B cell neoplasm: 4, Plasmodium falciparum infection: 3, and sickle cell disease: 2. Patient details corresponding to the samples were disclosed, such us sex, age, haemoglobin value and white blood cell count. The participants were asked to select up to four significant morphology features using the coding list provided by the EQAS CC. At the end of the run the participants received a brief clinical history of the patient, the final diagnosis and the reference results. For each survey, individual results were assessed against the morphological reference results established by the Cytology group of the AEHH. We divided the participants into two groups according their services to hospitalized (HPL) or non-hospitalized patients (NHPL) in order to compare the diagnostic accuracy in these different laboratory types.

Results: HPL showed higher participation ratios and better performance levels than NHPL. For the blood films corresponding to AML and MPD, the presence of blast cells was reported by a mean of 93 % of HPL and 87 % of the NHPL. Only 74% and 65% of the HPL and NHPL respectively reported blast cells in ALL blood films. Nevertheless, the performance of the participants in the diagnosis of ALL showed an improvement in the successive surveys. In the T-neoplasm smears, a mean of 64 % of HPL and 44 % of NHPL reported either Sézary cells or atypical lymphocytes. In addition, not improvement was noted in successive surveys of Sézary disease. Morphological reports for B cell neoplasm slides showed that atypical lymphocyte cells were detected by 56 % and 34 % of HPL and NHPL respectively. Mean values of correct answers in HPL and NHPL for Plasmodium slides were 90 and 78 % respectively, without improvement of the performance in the successive surveys. In sickle cell anaemia samples, a mean of 72 % and 69 % of HPL and NHPL reported abnormal red cells, showing better performances for following samples.

Conclusions: Pathological lymphoid cells were the most difficult to identify and our results confirmed the importance of the immunophenotyping in the diagnosis of lymphoproliferative disorders. Our experience of EQAS for peripheral blood morphology showed higher participation and better performance level in HPL with respect to HPL and this observation probably is in relation with the higher degree of haematology specialization in HPL. Since blood smear observation is a crucial tool in the diagnosis of the haematological disorders, specific educational programs aimed to improve peripheral blood morphology analysis should be considered.