Intra- and inter-observer reliability of combined segmental measurement techniques for predicting immediate post-deployment intraluminal tracheal stent length in dogs.

Abstract

This study evaluated segmental measurement techniques for predicting immediate post-deployment intraluminal tracheal stent length in dogs with naturally occurring tracheal collapse. Radiographs of 12 client-owned dogs that underwent intraluminal tracheal stent placement were retrospectively reviewed. Tracheal lengths were divided into 1, 2, 3, or 4 equal segments. Stent lengths were predicted using the widest dorsoventral height of each segment, with and without the addition of 10%, and an accompanying foreshortening chart. Techniques were compared for intra- and inter-observer reliability, and post-deployment stent length predictability. There was good to high intra- and inter-observer reliability for all segmental measurements; median intra-class correlation coefficients were 0.98 and 0.92, respectively. Measuring 2 segments without the addition of 10% to the widths was significantly more accurate in predicting immediate post-deployment stent length in terms of percent (P = 0.03) and absolute difference (P = 0.02). Segmental measuring techniques are repeatable amongst observers and may help guide stent selection.