RESUMO
The aim of this study was to evaluate the transabdominal ultrasonography as a tool for predicting the success of abdominocentesis in horses. Patients were included in the study if a complete transabdominal ultrasonography examination and abdominocentesis were performed as part of the clinical work-up. Ultrasonographically, the amount of peritoneal fluid was assessed using a 4-point grading system, as well as the amount of peritoneal fluid collected. A χ(2) or Fischer exact test was performed to test for an association between the ultrasonography findings and fluid retrieved, and between the two grading scales. Interobserver and intraobserver agreement values were calculated using k statistics. Values of P<0.05 were considered significant; 109 horses met the inclusion criteria. Peritoneal fluid was identified ultrasonographically in 72 per cent of horses, and it was collected from 93 per cent of these cases. In horses with no peritoneal fluid identified at ultrasonography (28 per cent), fluid was collected in 70 per cent of cases. There is a significant association between transabdominal ultrasonography detection of peritoneal fluid and the likelihood to obtain a diagnostic amount of peritoneal fluid at abdominocentesis; however, even when peritoneal fluid is not detected during abdominal ultrasonography examination, an amount of peritoneal fluid useful for gross, clinicopathological and cytological evaluation can frequently be obtained.