Premium
Identification of Sources of Inter‐Laboratory Variation in Factor VIII Assay
Author(s) -
Kirkwood T. B. L.,
Rizza C. R.,
Snape T. J.,
Rhymes I. L.,
Austen D. E. G.
Publication year - 1977
Publication title -
british journal of haematology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.907
H-Index - 186
eISSN - 1365-2141
pISSN - 0007-1048
DOI - 10.1111/j.1365-2141.1977.tb01029.x
Subject(s) - chromatography , third stage , coefficient of variation , stage (stratigraphy) , haemophilia a , chemistry , medicine , haemophilia , biology , surgery , physics , training (meteorology) , paleontology , meteorology
S ummary . A repeated finding of national and international collaborative studies of standard factor VIII preparations has been that systematic differences exist between laboratories in their measurement of the relative activities of the same pairs of factor VIII preparations. A workshop meeting was held at the Oxford Haemophilia Centre during 23–26 November 1976 to investigate which of the possible sources of variation between laboratories were responsible. Participants from 16 British laboratories (nine using one‐stage assays and seven using two‐stage assays) performed a total of 273 valid assays using three freeze‐dried preparations of differing purity (a plasma, an intermediate and a high purity concentrate). The results of assay with each participant using their normal system established that, if the participants were a representative cross‐section, approximately one‐third of one‐stage laboratories would show a systematic difference from the overall mean of at least 16% with a corresponding figure for the two‐stage laboratories of 9%. Various features of the assay systems were then modified in a controlled series of experiments. The results showed conclusively that (i) differences between reagents accounted for most of the variation between laboratories, and (ii) the two‐stage assays were, on average, detecting relatively more activity in the more purified preparations than the one‐stage assays. The results also suggested that the use of buffer or citrate‐saline as opposed to haemophilic plasma for the initial dilution of concentrates did not affect the assay results.