Evaluation of calibration methods for size estimation in the pediatric cardiac catheterization laboratory.

Limited data are available on the accuracy of various calibration methods used to estimate the size of cardiovascular structures during interventional pediatric cardiac catheterizations. The aim of this study was to evaluate, in an experimental model, several of these commonly used calibration methods. Thoracic box models were constructed in 4 sizes to simulate average chest sizes of infants, children, adolescents, and adults. Four anatomic positions were studied: aortic valve in posteroanterior projection, pulmonary valve in lateral projection, and isthmus in posteroanterior and lateral projections. The following calibration methods were examined: (1) a 5Fr pigtail catheter or a 1-cm radio-opaque marker catheter, placed at the level of the anatomic structure, (2) a body surface marker in multiple positions, and (3) a 1 x 1-cm radio-opaque rectilinear grid placed either at the center of the thorax or in an anatomic position. Three independent observers made size estimations. The most accurate method for size estimation, in all anatomic positions and for all box sizes, was the 1-cm radio-opaque marker catheter (mean error </=4%). The surface marker was moderately accurate in the smallest box (mean error 3% to 6%), but had a mean error as high as 33% in the adult-sized box. The 5Fr pigtail angiographic catheter overestimated the size of the structure by 3% to 13% (mean 7%). Intra- and interobserver differences in size estimations were not statistically significant. Among the size calibration methods used during pediatric cardiac catheterizations, catheters with radio-opaque markers within the structure of interest were the most accurate. The use of surface markers in larger patients may introduce significant error, risking a potentially inadequate or adverse outcome from the intervention.