Objective To determine the prevalence and nature of new clinically significant abdominal x-ray (AXR) findings and the proportion of patients receiving additional imaging in the emergency department (ED).

Methods This was a retrospective audit of consecutive adult patients, who presented to a tertiary referral ED (annual census 70 000) between September and December 2008. Data were extracted from radiologist reports and the medical records of patients with new significant AXR findings. The electronic radiology record was further interrogated to determine which patients received additional imaging and whether this correlated with the original AXR findings.

Results Of 997 cases that met the inclusion criteria, 121 (12.1%, 95% CI 10.2 to 14.4) and 43 (4.3%, 95% CI 3.2 to 5.8) had new clinically significant and insignificant AXR findings, respectively. Among the significant findings, the predominant diagnoses were bowel obstruction (72.7%), renal calculi (14.0%) and sigmoid volvulus (5.0%). Patient signs and symptoms were poorly associated with significant AXR findings. In all, 334 (33.5%, 95% CI 30.6 to 36.5) cases had additional imaging after the AXR. However, more patients with clinically significant AXR findings went on to have additional imaging (difference in proportions 23.0%, 95% CI 13.2 to 32.9, p<0.001)

Conclusions The yield for clinically useful information from the AXR is low and this investigation may be overused. Positive findings are associated mostly with bowel obstruction. As the proportion of patients ordered additional imaging was considerable, the utility of the preliminary AXR is questionable, especially in cases where the diagnosis is clear. Guidelines for AXR imaging are recommended to assist clinicians with investigation ordering.