Diagnostic accuracy and short-term surgical outcomes in cases of suspected acute appendicitis

Objective — To test the hypothesis that, with modern diagnostic methods and antibiotics, more conservative use of surgery in cases of suspected appendicitis would not result in increased rates of short-term complications in confirmed cases.

Design — Population-based observational study using administrative data.

Setting — All Ontario hospitals in which primary appendectomy was performed from April 1, 1981, to March 31, 1992.

Patients — All 126,815 patients admitted to hospital for a primary appendectomy during the study period.

Outcome Measures — Diagnostic accuracy rate (acute appendicitis as the primary diagnosis), perforation rate, in-hospital death rate and length of stay.

Results — The diagnostic accuracy rate among the male patients was stable throughout the decade; among the female patients it rose significantly, from 71.7% in 1981 to 75.3% in 1991 (p < 0.01). The perforation rates increased significantly among both the female and male patients (p < 0.01), whereas the mean length of stay decreased (p < 0.05). Despite sex-related differences in the accuracy rates, the male and female patients had similar in-hospital death rates and mean lengths of stay. The institutional diagnostic accuracy rates, as determined from data for 1989-90 to 1991-92, ranged from 50.0% to 96.7%. Multivariate analyses of 27,189 confirmed cases of appendicitis at 175 hospitals revealed that perforation was a strong predictor of in-hospital death (odds ratio [OR] 2.46, 95% confidence interval [CI] 1.24 to 4.88), but comorbidity was the strongest predictor (OR 11.50, 95% CI 5.96 to 22.10). For each 10% increase in the diagnostic accuracy rate, the perforation rate increased 14% (OR 1.14, 95% CI 1.10 to 1.19), but the accuracy rate was not significantly related to the in-hospital death rate or the length of stay.

Conclusion — A higher diagnostic accuracy rate is associated with more perforated appendixes. Although perforation itself leads to adverse outcomes, a higher accuracy rate does not. This suggests that hospitals with higher accuracy rates incur more perforations, but, with close observation, timely laparotomy and the use of modern antibiotics, these patients have favourable outcomes. This contrasts with adverse effects of perforation among patients at high risk for perforation (especially very young children and elderly people) in centres at all accuracy levels. The variation in hospitals' diagnostic accuracy rates suggests that some proportion of appendectomies could be safely avoided.