Accuracy of the Alvarado Score in Diagnosing Appendicitis Among Children Versus Adults.

BACKGROUND

The Alvarado score is widely used to risk-stratify suspected acute appendicitis; however, its diagnostic performance may vary across age groups.

AIM: To compare the diagnostic discrimination of the Alvarado score between adults (≥18 years) and children (<18 years) using receiver operating characteristic (ROC) analysis, with the primary outcome being the difference in area under the curve (AUC) between groups. Secondary analyses evaluated diagnostic performance at predefined clinical thresholds (>7 and ≥5).

METHODS

This retrospective cohort study included 743 consecutive patients evaluated for suspected appendicitis. The index test was the Alvarado score derived from documented presentation findings. The reference standard was histopathological confirmation of appendicitis. Patients were categorized into children and adults. Discrimination was assessed using ROC analysis, and diagnostic performance was evaluated at predefined thresholds (high >7; equivocal ≥5), reporting sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). ROC curves were compared between adults and children using the DeLong test.

RESULTS

The Alvarado score demonstrated moderate diagnostic discrimination in both groups. The area under the ROC curve (AUC) was 0.685 in adults and 0.732 in children, with no statistically significant difference between groups (DeLong p = 0.305). At the high threshold (>7), overall sensitivity was 50.5%, with 48.5% in adults and 55.7% in children, while specificity was 80.4% in adults and 78.7% in children (80.0% overall). At the equivocal threshold (≥5), sensitivity increased to 84.3% in adults, 91.4% in children, and 86.3% overall, with specificity decreasing to 25.7% in adults, 24.6% in children, and 25.4% overall. Positive predictive values ranged from 69.7% to 85.7%, while negative predictive values ranged from 43.5% to 55.6%.

CONCLUSIONS

The Alvarado score demonstrated moderate diagnostic discrimination in both adults and children, with no statistically significant difference in ROC performance between groups. At the traditional high threshold (>7), sensitivity was limited, whereas lower thresholds (≥5) increased sensitivity with reduced specificity. These findings support the use of the Alvarado score as a clinical risk stratification tool in both age groups.