Variation between batches in clotting factor assays.

Differences in factor V, VII and VIII potency between the standards used in the Northwick Park Heart Study (NPHS) are described in the previous paper. This paper discusses the variation from batch to batch within each standard. In every case there was significant variation between batches, the standard deviation of batch means being between a third and a half of that within batches. Some of this variation could be attributed to drifts over time, but most of the "batch effect" was not easily accounted for. It is relatively simple to correct for steady drift, but there is no satisfactory way of allowing for unexplained batch variation.