Diagnostic Accuracy of the Modified Vesikari Score for Predicting Severe Acute Gastroenteritis in Children: A Prospective Diagnostic Accuracy Study

Technical Summary: Diagnostic Accuracy of the Modified Vesikari Score for Predicting Severe Acute Gastroenteritis in Children

Problem Statement
Acute gastroenteritis (AGE) remains a leading cause of morbidity and mortality in children globally, particularly in low- and middle-income countries. While the clinical severity of AGE varies from self-limiting illness to life-threatening dehydration and shock, early identification of severe cases is critical for optimizing fluid management and determining the need for hospitalization. Although the Modified Vesikari Score (MVS) is a standardized tool used in research and vaccine trials, its diagnostic accuracy in routine clinical practice—specifically within resource-limited primary referral settings like Type D hospitals—remains under-evaluated. This study addresses the gap in evidence regarding the MVS's ability to predict severe AGE in such environments.

Methodology
The authors conducted a prospective diagnostic accuracy study at Candi Umbul Regional General Hospital, a Type D facility in Magelang Regency, Indonesia.

• Study Population: 70 consecutive children aged 1–59 months presenting with acute gastroenteritis (defined as ≥3 loose/watery stools in 24 hours for <14 days).
• Index Test: The Modified Vesikari Score (MVS) was calculated at presentation based on seven parameters: duration and frequency of diarrhea, duration and frequency of vomiting, fever, dehydration status, and treatment requirements. A score ≥11 was classified as "severe" for the purpose of the index test.
• Reference Standard: An independent pediatrician, blinded to the final MVS where possible, performed a clinical assessment based on established management guidelines. Severe AGE was defined by the presence of moderate-to-severe dehydration, need for intravenous rehydration, hospitalization, or signs of significant clinical deterioration.
• Analysis: Diagnostic performance was evaluated using Receiver Operating Characteristic (ROC) curve analysis to determine the Area Under the Curve (AUC). Sensitivity, specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV), and Likelihood Ratios (LR+ and LR−) were calculated.

Key Results

• Sample Distribution: Of the 70 participants, 30 were classified as severe by the MVS (≥11), while 25 were classified as severe by the pediatrician's clinical assessment (reference standard).
• Discriminative Ability: The ROC analysis revealed excellent discriminative performance with an AUC of 0.975 (95% CI: 0.927–1.000).
• Diagnostic Metrics:
  • Sensitivity: 96.0% (indicating a high ability to correctly identify children with severe AGE).
  • Specificity: 86.7% (indicating a high ability to correctly identify non-severe cases).
  • Predictive Values: PPV was 80.0%, and NPV was 97.5%.
  • Likelihood Ratios: The Positive Likelihood Ratio (LR+) was 7.20, and the Negative Likelihood Ratio (LR−) was 0.05.
• Confusion Matrix: Among the 25 children confirmed as severe by the reference standard, 24 were correctly identified by the MVS. Among the 45 non-severe cases, 39 were correctly identified by the MVS.

Significance and Claims
The paper claims that the Modified Vesikari Score demonstrates excellent diagnostic accuracy for predicting severe acute gastroenteritis in children within a Type D hospital setting. The authors highlight that the score's high sensitivity (96.0%) and very low negative likelihood ratio (0.05) make it a highly reliable tool for ruling out severe disease in low-risk patients, which is crucial for preventing complications like hypovolemic shock. Conversely, the high positive likelihood ratio (7.20) supports its utility as a "rule-in" tool for identifying cases requiring intensive intervention.

The authors assert that the MVS serves as a practical, rapid, and cost-effective bedside instrument that can assist clinicians in resource-limited settings to prioritize patients for intravenous therapy or hospitalization. However, they maintain a modest stance, emphasizing that the score should be used complementarily with clinical judgment and not as a standalone diagnostic tool. The study concludes that while the MVS shows outstanding performance, further multicenter studies with larger sample sizes and microbiological confirmation are recommended to validate these findings across diverse healthcare settings.

Limitations Noted by Authors
The authors acknowledge several constraints: the reliance on clinical judgment as the reference standard (introducing potential subjectivity), the single-center design with a relatively small sample size (limiting generalizability), and the lack of routine microbiological confirmation for viral or bacterial etiologies.