Key result
Clinical scoring systems (AUC 0.67-0.72) performed equally well as physician gestalt (AUC 0.72) in predicting appendicitis, and do not obviate the need for CT imaging.
Why the study?
Do clinical scoring systems or physician gestalt obviate the need for CT in patients with possible appendicitis?
Observational (n=287)
No
Do clinical scoring systems or physician gestalt obviate the need for CT in patients with possible appendicitis?
Clinical scoring systems perform equally well as physician gestalt in predicting appendicitis, but neither obviates the need for CT imaging when deemed necessary by a physician.
Scoring systems match gestalt but do not safely obviate CT; leaves open whether refined thresholds could reduce imaging in select patients.
OBJECTIVE: To determine whether clinical scoring systems or physician gestalt can obviate the need for computed tomography (CT) in patients with possible appendicitis. METHODS: Prospective, observational study of patients with abdominal pain at an academic emergency department (ED) from February 2012 to February 2014. Patients over 11 years old who had a CT ordered for possible appendicitis were eligible. All parameters needed to calculate the scores were recorded on standardised forms prior to CT. Physicians also estimated the likelihood of appendicitis. Test characteristics were calculated using clinical follow-up as the reference standard. Receiver operating characteristic curves were drawn. RESULTS: Of the 287 patients (mean age (range), 31 (12-88) years; 60% women), the prevalence of appendicitis was 33%. The Alvarado score had a positive likelihood ratio (LR(+)) (95% CI) of 2.2 (1.7 to 3) and a negative likelihood ratio (LR(-)) of 0.6 (0.4 to 0.7). The modified Alvarado score (MAS) had LR(+) 2.4 (1.6 to 3.4) and LR(-) 0.7 (0.6 to 0.8). The Raja Isteri Pengiran Anak Saleha Appendicitis (RIPASA) score had LR(+) 1.3 (1.1 to 1.5) and LR(-) 0.5 (0.4 to 0.8). Physician-determined likelihood of appendicitis had LR(+) 1.3 (1.2 to 1.5) and LR(-) 0.3 (0.2 to 0.6). When combined with physician likelihoods, LR(+) and LR(-) was 3.67 and 0.48 (Alvarado), 2.33 and 0.45 (RIPASA), and 3.87 and 0.47 (MAS). The area under the curve was highest for physician-determined likelihood (0.72), but was not statistically significantly different from the clinical scores (RIPASA 0.67, Alvarado 0.72, MAS 0.7). CONCLUSIONS: Clinical scoring systems performed equally well as physician gestalt in predicting appendicitis. These scores do not obviate the need for imaging for possible appendicitis when a physician deems it necessary.
No takes yet. Share an insight, caveat, or question.
Golden et al. (2016) conducted an observational in Possible appendicitis (n=287). Clinical scoring systems (Alvarado, MAS, RIPASA) vs. Physician-determined likelihood was evaluated on Prediction of appendicitis (Area under the curve). Clinical scoring systems (AUC 0.67-0.72) performed equally well as physician gestalt (AUC 0.72) in predicting appendicitis, and do not obviate the need for CT imaging.