Background: CT and ultrasound (US) are increasingly recommended to establish the diagnosis of appendicitis, but population-based rates of misdiagnosis (negative appendectomy [NA]) have not improved over time. The objective of this study was to determine the relationship between CT/US and NA in common practice.
Study design: Using data from the Group Health Cooperative on all patients undergoing appendectomy between 1980 and 1999, a longitudinal study was conducted to determine the frequency of NA over time and a case-control, medical record-based study of a subset of patients from the 1990s was conducted to determine the accuracy of CT/US.
Results: Of 4,058 patients undergoing appendectomy (mean age 31 +/- 18.6 [SD] years, 49.6% women), 631 (15.5%) had an NA. The overall incidence of NA remained stable over time at 1.5/10,000 patient-years, as did the age and gender adjusted rate (incident rate ratio 0.95, 95% CI 0.97, 1.01). In 1999, nearly 40% of patients had either CT or US. The aggregate sensitivity of these tests was only 74.2% (95% CI 65.7, 83.7), with a positive predictive value of 95.1% (95% CI 91.5, 96.8). CT scans were 88.3% sensitive, with 97.2% positive predictive value (95% CI 92.9, 100). Ultrasounds were 69.5% sensitive, with a positive predictive value of 94.1% (95% CI 89.6, 96.4). More than one in five tests obtained in patients with NA were positive for appendicitis (21.7% for CT and 20.8% for US).
Conclusions: The rate of NA was unchanged over time despite the introduction and use of CT/US, and this appeared to be related to the inconsistent performance characteristics of the tests. This study cautions against overreliance on CT/US in diagnosing appendicitis and emphasizes the need for test benchmarking in routine practice before establishing protocols for presumed appendicitis.