Background: We present a study aimed to define the analytical goals for the determination of hemoglobin A 2 , a minor hemoglobin present in human blood normally accounting from 2.5 % to 3.3 % of total hemoglobin, and typically increased up to 6 % -7 % in subjects carriers of β -thalassemia. Methods: The analytical goals have been derived using two approaches, the first one based on biologic variation, and the second one based on the opinion of experts. Results: The data obtained by studying 17 adult non-carrier healthy subjects, from whom we took blood samples every 2 weeks for 2.0 months, indicated a small intraindividual biologic variation (CV I of 0.7 % ), with respect to a larger between-subject variation (CV G of 7.7 % ). The minimum levels for imprecision, bias and total error derived from the analysis of these data were: 0.5 % , 2.9 % and 4.5 % , respectively. The limits derived from the opinion of experts were based on a questionnaire with three clinical cases, which was circulated among two teams of international experts, and on a discussion about the clinical needs. The average total error derived from such surveys ranged between 7.0 % and 9.5 % .
Conclusions:The various methods to derive analytical performance goals gave different limits, thus indicating the need for an increased communication between clinicians and laboratory professionals on this matter.