Reliability of ultrasonography compared to radiography in revealing intestinal sand accumulations in horses.

Abstract

In order to investigate the reliability of ultrasonography in revealing intestinal sand accumulations, the cranioventral parts of the abdomen of 32 horses (age 3-20 years) with signs suggestive of gastrointestinal sand were examined radiographically and ultrasonographically. The amount of sand and its localisation, relative to the ventral abdominal wall, were evaluated radiographically and graded (scale 0-4). Motility of the intestine and whether it was lying against the ventral abdominal wall were evaluated ultrasonographically and the findings summed and graded (scale 0-3) to express the likelihood of the presence of sand. The ultrasonographic and radiographic grades were significantly associated. Out of 14 horses with a moderate or large ventral sand accumulation radiographically, 13 (92.9%) had a positive sand finding ultrasonographically. Out of 8 horses with no signs of sand radiographically, 7 were considered negative for sand ultrasonographically. The specificity of ultrasonography in detecting sand accumulations was 87.5% (7/8) and the sensitivity 87.5% (21/24). Small and more dorsally located accumulations were more difficult to detect ultrasonographically. Of the horses with a small or moderate amount of sand relatively ventrally or only a small part of sand close to the ventral abdominal wall, 70% (7/10) were considered suggestive or positive for sand ultrasonographically. The ventral aspect of sand accumulations was hyperechoic, causing varying acoustic shadowing and the intestine had decreased or absent motility. Ultrasonography revealed the length of the accumulations but gave very limited information of their height. Ultrasonography is a practical and reliable method for detecting sand accumulations but it cannot replace radiography.