Patients with polycythaemia and normal controls have been studied to e
stablish and subsequently test non-conventional criteria for the diagn
osis of primary polycythaemia (primary proliferative polycythaemia, po
lycythaemia vera) as compared with conventional Polycythaemia Vera Stu
dy Group (PVSG) assessment. One criterion was erythroid colony formati
on from peripheral blood in a serum-free system, assayed alone and wit
h the addition of recombinant human erythropoietin (Epo), interleukin
3 (IL3), or alpha interferon (alpha-IFN) (Dudley et al. 1990). The rem
aining criteria were non-culture associated and comprised platelet dis
tribution width (PDW), platelet nucleotide ratio (ATP:ADP), serum eryt
hropoietin and clinical evidence of ischaemic vascular disease. The co
mbination of culture associated and non-culture associated variables,
by use of a simple additive scoring system, gave no false positive and
only 6% false negative results in distinguishing primary polycythaemi
a from other polycythaemias and normal controls in those (34 patients
Group A) used in its derivation. Testing the scoring system in a newly
presenting group (25 patients Group B) was highly satisfactory with n
o false positives and only a few false negative results (14%). Use of
these non-conventional criteria should allow more confident diagnosis
of primary polycythaemia, where conventional clinical and laboratory a
ssessment is inconclusive.