Comparison of diagnostic criteria for polycythaemia vera.

Three sets of diagnostic criteria for polycythaemia vera (PV); the Polycythaemia Vera Study Group (PVSG) criteria (1975), the British Committee for Standards in Haematology (BCSH) criteria (1996) and the World Health Organisation (WHO) criteria (2001) have been described. We compared the ability of each set of criteria to accurately diagnose PV and differentiate it from secondary, apparent and idiopathic erythrocytosis. A cohort was drawn from a clinical database of erythrocytosis patients currently attending the Belfast City Hospital and the relevant information from the time of diagnosis for each patient was assessed according to each set of criteria, with the BCSH criteria used as a comparator. Sufficient data was available on 71 patients: 46 PV, 8 idiopathic, 8 apparent and 9 secondary erythrocytosis. The BCSH criteria classified 34 of 46 patients (73.9%) as PV and the WHO criteria had a sensitivity and specificity of 100% for classifying PV. For idiopathic and apparent erythrocytosis, the specificity of the WHO criteria, compared to the BCSH criteria, declined to 66.7 and 87.5%, respectively. The PVSG criteria were limited by the unavailability of required data for some patients resulting in a sensitivity and specificity of 50% for PV and specificity of 100% for all other groups. The Janus kinase 2 (JAK2) V617F mutation was present in 34 (85.3%) PV, 2 (50%) IE, 1 (50%) apparent and no secondary erythrocytosis cases. We concluded that the BCSH criteria were the most accurate diagnostic criteria for PV as they had an acceptable level of sensitivity and could differentiate between PV and other erythrocytoses.