Materials and Methods

       The present prospective study was performed  during the period between March 2012 and December 2012. The Hospital’s research Ethics Committee approved the present study. Data on all patients referred to the Department of Radiology for emergency US during the present study period were collected. Emergency US was defined as an US requested on an emergency basis by he referring or primary physician. US for biopsy purposes or for emergency interventions, such as US-guided loculated pleural effusion drainage, were excluded.

       US was performed by experienced staff radiologists during the working hours (8.30 to 16.00 o’clock). Data collection forms were filled by experience staff radiologists. The final discharge diagnoses were obtained from medical charts and computerized records.

       Data collected for the present study included age, gender, admitting or provisional diagnosis, ultrasonographic findings and diagnosis, the final (discharge) diagnosis, time of US, other radiologic investigations, and therapeutic interventions. No attempt was made to elicit the opinion of referring physicians or  residents on whether  results of US “influenced” or “had an impact on” clinical decision making.

       All diagnoses were categorized according to organ-systems. A positive US result was defined as a clear abnormality detected by US. A negative US was defined as unclear findings or finding of no obvious abnormality. A negative final diagnosis was one that was vague, non-specific and not of clear clinical significance (for example, “dyspepsia”, “abdominal pain”, “muscle pain”, “ testicular pain” and “cramps”). The final diagnosis was assumed to represent the true status of the patient at discharge.

       The concordance or agreement between provisional diagnosis and US for a particular patient was defined as the same or practically the same diagnoses for both (for example, “acute appendicitis”, “appendicitis”, and “consistent with appendicitis” were considered practically the same diagnoses). US can “disagree” with or “exclude” the provisional diagnosis or were unclear. US could reveals an “alternative diagnosis” as well as demonstrated other, anatomically or pathologically unrelated but clinically important.

Results

       Three hundred sixty three patients underwent emergency US examinations at the Department of Radiology between March  and December 2012. Missing data due to incomplete record from medical charts and medical records, so only 285 patients had sufficient data to make the judgment.

       The vast majority of patients in the present study were adults (83%). The median age was 46 years. Male and female patients were equally represented (50% and 50% respectively). The most common anatomical region examined was abdominal (77%). Most of the examinations took place out of hours (67%).

       Table 1- 3 present data on characteristics of patients, anatomical regions suspected of harboring disease, and provisional ultrasound, and the final disease categories. Table 4 presents the main outcomes of the present study : agreement, impact and accuracy of US. Factors associated with the impact of US on clinical diagnosis and treatment are given in Table 5.

       According to the present study, age, gender, and timing of US had no relationship with the impact of US. There was a significant association between the anatomical region of US examination and impact: there was a greater chance of impact if the region examined was KUB (kidney, ureter and bladder) or if MSK US was performed, and a greater chance of no impact if the region examined was abdominal (upper and whole) (Table 5). US had a greater chance of having an impact if the provisional diagnosis was related to vascular or other (chest, cranial) systems, and less chance of impact for suspected appendicitis or intraabdominal infection and collection. There was also less chance of impact if the US diagnosis was intraabdominal infection or mass.

       The impact of US was greater when the final diagnosis was related to vascular problems, skin and soft tissue infections, or biliary obstruction and infection. US was least likely to impact on the final diagnoses of intraabdominal diseases, with the exception of gallstone disease.

       Agreement between provisional and US diagnoses had no association with the impact of US. It was to be expected, according to the present definition of impact, that further investigations were associated with less impact. Similarly, it was expected that the more accurate the US  examination, the more it would have an impact.

       The results of the check on the accuracy of US in diagnosing acute appendicitis and acue cholecystitis were as follows. For acute appendicitis, in 23 patients with available data, the sensitivity was 65.4% (95% CI: 45.6% to 82.1%) and the specificity was 95.4% (95% CI: 84.5% to 99.4%). For acute cholecystitis, in 66 patients with available data, the sensitivity was 86.7% (95% CI: 71.9%to 95.6%) and the specificity was 88.5% (95%  CI: 78.7% to 94.9%).

       The measures of impact on clinical management in the present study were consistent with the criteria of construct validity¹⁴. It seemed reasonable to expect that US should have a larger impact for problems specifically relate to vascular or other (chest, cranial) systems, compared to suspected abdominal abscess. Although the absolute measures of “impact” (94% impact) would very likely depend on the definitions used, qualitatively the findings in the present study seemed reasonable. For example, the apparently large impact of US on clinical practice was consistent with the ever-increasing use of US (otherwise, the use should decrease).