One core laboratory at two international sites, is that feasible? An inter-core laboratory and intra-observer variability study.

To assess the magnitude of differences in QCA outcomes between two cooperating core laboratories in a single trial, we have carried out an inter-core laboratory variability study. Two QCA experts at the Montreal Heart Institute and Heart Core Leiden both analyzed 32 lesions (pre- and post-intervention) in accordance with previously agreed upon standard operating procedures. One of the experts analyzed the whole image set twice to determine the intraobserver variability. The inter-core laboratory differences in the acute gain (n = 31 pairs) are non-significant. The systematic errors of the individual measurements (n = 63 analyses) show an excellent intraclass correlation coefficient of reliability (>75%), except for the stent length (67.7%). The corresponding random errors are small. In general, the intra-observer systematic and random errors are both slightly smaller than those for the inter-core laboratory study. QCA analyses in clinical trials can be carried out in core laboratories at two different locations if and only if highly standardized conditions are maintained. Copyright 2002 Wiley-Liss, Inc.