Clinical relevance of computed tomography under emergency conditions. Diagnostic accuracy, therapeutical consequences

Purpose: To evaluate the diagnostic accuracy and therapeutic consequences of computed tomography performed on an emergency basis in a primary care hospital. Material and Methods: In 418 patients, 463 computed tomographies (thorax, abdomen, pelvis, spine, aorta, neck and extremities) were performed within 12 months, providing 999 diagnoses. The computed tomography diagnoses were retrospectively evaluated and correlated to surgery and discharge diagnoses. Therapeutical consequence were analyzed and allocated to a time period < 36 h (urgent) and ≥ 36-72 h (elective). Average age was 49 (1-94) years (41% female and 59% male). Discharge diagnosis was defined as gold standard, provided that it was supported by clinical, blood chemical, diagnostic and possible surgical data. Results: In 176 of 999 diagnoses (18%), the diagnoses were classified as ''noncorrelatable''. Of the 823 correlated diagnoses, 431 were true positive, 14 false positive, 66 false negative and 312 true negative. Sensitivity, specificity and diagnostic accuracy of computed tomography was 87,96 and 90%. Computed tomography had therapeutic consequences (surgery, drainage, puncture, reposition, thrombolytic therapy, chemotherapy, bronchoscopy, endoscopy, percutaneous transluminal angioplasty, coiling etc.) in 57% and no direct therapeutic interventions in 43%. Computed tomography excluded the suspected diagnosis in 36% and resulted in a conservative therapeutic regiment in 7%. Surgery was performed on 134 of the 418 patients (32%) who underwent computed tomography, with the surgery urgent in 71 (17%) and elective in 63 (15%) of the 418 patient. (orig.)