Comparison of clinical judgment and diagnostic ultrasonography in the diagnosis of acute appendicitis: experience with a score-aided diagnosis.

OBJECTIVE To evaluate the diagnostic accuracy of clinical judgment and diagnostic ultrasonography (US) used routinely and to create a scoring system to aid diagnosis. DESIGN Prospective, double-blind study. SETTING University hospital, Denmark. SUBJECTS 222 Consecutive patients suspected of having acute appendicitis admitted between 0800 and midnight from June 1990 to June 1992. INTERVENTIONS 148 Patients (67%) underwent appendicectomy and the remaining 74 patients were observed. 193 Patients (87%) had a diagnostic US examination. 21 Predictive variables were collected prospectively to create a scoring system. MAIN OUTCOME MEASURES Results of surgical pathological findings, clinical outcome (observed group), diagnostic US, and values of diagnostic score. RESULTS The decision to operate was made by a junior surgeon solely on the clinical examination, which yielded a diagnostic accuracy of 76%, specificity of 58%, and negative appendicectomy rate of 36%. 193 Patients underwent diagnostic US conducted by the radiologist on call of whom 123 were operated on, 78 for histologically proven appendicitis. US had a diagnostic accuracy of 72%, sensitivity of 49%, and specificity of 88%. Of the 21 predictive factors for acute appendicitis 11 were significant (p < 0.05): total white cell count (WCC) (>10 x 10[9]/1), migration of pain to the right lower quadrant, gradual onset of pain, increasing intensity of pain, pain aggravated by movement, pain aggravated by coughing, anorexia, vomiting, indirect tenderness (Rovsing's sign), muscle spasm, and sex. These 11 predictors were assigned an appropriate weight, based on the likelihood ratio, and used to create a scoring system. The score performed poorly if it was used to separate patients for observation and those for appendicectomy. However, if the score was used with two cut-off points resulting in three test zones (low, intermediate, and high risk of having acute appendicitis), some diagnostic benefit was seen for those patients within the zones of high and low probability. CONCLUSION The clinical judgment of a junior surgeon was disappointing, and diagnostic aids are desirable to reduce the negative appendicectomy rate. Diagnostic US performed poorly as a routine procedure. Application of an up to date scoring system might be of some help to patients with a high or low probability of acute appendicitis, but any conclusion about its clinical application cannot be drawn from this study.