Development and independent regional validation of a dynamic ultrasound appendicitis score for suspected acute...

Acute appendicitis is the most common surgical emergency. Although CT is highly accurate, radiation exposure is a major concern in children, young adults, and pregnant women. Ultrasound is radiation-free but operator-dependent, and current scores rely on static weights without accounting for important patient-finding interactions.

To develop and independently validate a Dynamic Ultrasound Appendicitis Score (Dynamic UAS) with interaction effects to improve risk stratification and potentially avoid unnecessary CT examinations.

A prospective multicenter study collected development (n = 802) and independent regional external validation (n = 262) data. The Dynamic UAS was derived using multivariable logistic regression with interaction terms and internally validated using bootstrapping (1000 resamples).

A web-based calculator was developed. A sensitivity analysis addressed limitations of the reference standard, non-visualized appendix, and missing data. Appendicitis prevalence was 33.3% (development) and 32.8% (validation). The Dynamic UAS performed excellently in diagnosing acute appendicitis (AUC 0.96 and 0.94) and outperformed known scores (AUC 0.81 to 0.91). At the ≥ 6 level, sensitivity was 91.7%/90.7%, and specificity was 97.1%/96.0%. Calibration was good (slope 0.89, Brier 0.098). In non-visualized appendix cases, the AUC was 0.89 (95% CI: 0.84 to 0.93) in the external validation cohort and 0.90 (95% CI: 0.85 to 0.94) in the development cohort. The model is not very useful for distinguishing complicated from uncomplicated appendicitis (AUC 0.72).

The Dynamic UAS is a novel, independently validated tool that may reduce unnecessary CT use in radiation-sensitive populations and resource-limited settings. Future studies in different international settings are needed.