Auswertung von Ringversuchen zur zytologischen Knochenmark-Diagnostik der Gesellschaft zur Förderung der Qualitätssicherung in medizinischen Laboratorien e.V. (INSTAND)

Interlaboratory comparison („round robin“) of diagnostic quality in haematological cytomorphology from the society for the Advancement for quality assurance in medical laboratories (INSTAND-Düsseldorf) Introduction: The cytomorphology of haematological diseases is an important diagnostic tool. The focus of this interlaboratory comparison (“round robin” test) is diagnostic quality. 1772 single bone marrow specimens were evaluated. Results were categorized according to the nature of the mistake and the severity of the disease. When evaluating the results, everyday demands in the field of cytomorphology were taken into account, as well as realistic clinical valuation standards in diagnosis. These results were discussed while taking into consideration the WHO-classification and the legal framework in the area of quality assurance. Methods: The selection and evaluation of the specimens was conducted by a cytologist from a reference laboratory. The participants were to submit a cytomorphological diagnosis with the help of a questionnaire and little additional information. In conclusion, the participants received a detailed error analysis and explanatory video prints of the specimens. This analysis was restricted to laboratories achieving more than 70% of the total score. Results: The error rates were spread over a severity scale of 4 levels (range between 0,0 to 35,1 %). Error rates were high for non-pathological specimens (16,1 %) and myeloproliferative neoplasias (10,0 %), the latter despite straightforward clinical information such as a thrombocytosis over an extended period of time. False diagnoses in hemolytic anemia were 13,9 %, in megaloblastic anemia 15,1 % within which most cases were confounded with myelodysplasia. Myelodysplastic cases showed error rates up to 23,5 %. MDS-subtypes error rates were RARS (11,5 %), RAEB (35,1 %) and 5q-minus-syndrome (34,6 %). In acute myeloid leukemia errors appeared in 5,6 %. The Remission status in acute myeloid leukemia (7,9 %) and acute lymphoblastic leukemia (10,6 %) is a challenge, especially in AML M4Eo. In this particular case only 17,2 % of the participants diagnosed an early relapse. From a different point of view when participants finally diagnosed a case, such as MDS type RAEB, results were correct in 87,9 % of cases. In contrast, in the case of reactive changes the result was correct in only 10.6 %. Conclusion: In addition to offering extensive training and continued education programs, a panel of experts, such as in the ongoing study group trials should be available to determine reference diagnoses. Modern digital media can simplify the necessary work routines.