Beyond Acute Jaundice: refining case definitions for suspected hepatitis E in South Sudan and Bangladesh

Résumé fourni par la source

Summary The true burden of hepatitis E is poorly understood due to a lack of routine diagnostic testing and nonspecific symptoms. We evaluated the sensitivity and specificity of international case definitions for suspected hepatitis E and explored whether alternative case definitions could more accurately identify those with true hepatitis E, thus improving our ability to interpret surveillance data. We used data from acute jaundice surveillance in South Sudan (March-December 2022) and Bangladesh (December 2014-September 2017), two settings where hepatitis E genotype 1 circulates. Individuals seeking care with acute jaundice syndrome (AJS) were asked about signs/symptoms of hepatitis E and underwent testing with anti-hepatitis E virus (HEV) IgM ELISA. We used an ensemble of classification models to assess how well signs/symptoms could distinguish between AJS with and without HEV. To explore alternative case definitions, we estimated the sensitivity and specificity of all combinations of signs/symptoms. Among patients with AJS, 20% in South Sudan and 38% in Bangladesh had IgM antibodies. International case definitions for suspected hepatitis E had variable sensitivity (53-96%) and specificity (6-60%) across study populations. Alternative case definitions had poor discrimination in both populations (AUC SouthSudan =0.64; 95% Confidence Interval (CI): 0.57, 0.71; AUC Bangladesh = 0.60; 95% CI: 0.57, 0.63). No alternative case definitions had both sensitivity and specificity above 60% in both study populations. Acute jaundice surveillance in two distinct populations revealed that signs/symptoms alone are insufficient to distinguish HEV-related cases from other causes of acute jaundice. Accurately characterizing the burden of hepatitis E and assessing the cost effectiveness of vaccines will require expanded use of diagnostic testing.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.