Scholarlink

Background &amp; aims <p>Hepatitis E virus (HEV) represents a leading cause of acute viral hepatitis in China, yet large-scale studies characterizing its dynamic epidemiology and enabling early prediction of adverse outcomes remain scarce.</p> Methods <p>We performed a retrospective cohort study comprising 816 patients hospitalized with acute HEV infection (2019–2024). Predictors were identified through multivariable logistic regression, with&#xa0;Firth's penalized likelihood method applied to address potential small-sample bias, and bootstrap resampling (BCa 95% CI) used to validate estimate robustness. A prognostic nomogram was developed and externally validated in a prospective multicenter cohort, with performance evaluated using AUC, calibration curves, and decision curve analysis.</p> Results <p>Among 816 patients (69.6% male, median age 60&#xa0;years), HEV infection exhibited a winter–spring seasonal pattern in Shanghai, China. The overall 90&#xa0;day adverse outcome rate was 4.5% in the general population and 18.1% in cirrhotic patients (aOR = 4.37, 95% CI: 1.69–11.33; <i>p</i> = 0.002). No 90&#xa0;day adverse outcomes occurred in pregnant patients or those with HIV/AIDS. Chronic liver disease (CLD) (OR = 2.65, 1.27–5.70; Firth <i>p</i> = 0.008), MELD score (OR = 1.25, 1.18–1.32; Firth <i>p</i> &lt; 0.001), and neutrophil-to-lymphocyte ratio (NLR) (OR = 1.18, 1.09–1.27; Firth <i>p</i> &lt; 0.001) were independently associated with 90&#xa0;day adverse outcomes. The CLD–MELD–NLR nomogram achieved high accuracy in the general population (AUC = 0.94, 95% CI: 0.89–0.98), and the MELD–NLR nomogram also performed outstandingly in the CLD subgroup (AUC = 0.90, 95% CI: 0.84–0.96). Both models demonstrated good calibration and clinical utility.</p> Conclusions <p>The distinct seasonality and disease burden of HEV in Shanghai highlight the need for targeted local public health measures. The robustly validated&#xa0;CLD–MELD–NLR and MELD–NLR nomograms provide practical tools for early risk stratification and personalized management across diverse clinical settings.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Epidemiological characteristics in Shanghai and a prognostic nomogram for 90 day adverse outcomes in acute hepatitis E

  • 期刊论文

  • 发表日期：

    2026年2月12日

查看全文
  • Xue Mei,
  • Ying Zou,
  • Xin Zheng,
  • Hai Li,
  • Guohong Deng,
  • Jinjun Chen,
  • Yan Huang,
  • Zhongji Meng,
  • Yu Shi,
  • Xianbo Wang,
  • Yanhang Gao,
  • Feng Liu,
  • Liujuan Ji,
  • Yu Liu,
  • Hui Zhu,
  • Zhengguo Zhang,
  • Hongying Guo,
  • Wei Yuan,
  • Yuyi Zhang,
  • Zhiping Qian

摘要

Background & aims

Hepatitis E virus (HEV) represents a leading cause of acute viral hepatitis in China, yet large-scale studies characterizing its dynamic epidemiology and enabling early prediction of adverse outcomes remain scarce.

Methods

We performed a retrospective cohort study comprising 816 patients hospitalized with acute HEV infection (2019–2024). Predictors were identified through multivariable logistic regression, with Firth's penalized likelihood method applied to address potential small-sample bias, and bootstrap resampling (BCa 95% CI) used to validate estimate robustness. A prognostic nomogram was developed and externally validated in a prospective multicenter cohort, with performance evaluated using AUC, calibration curves, and decision curve analysis.

Results

Among 816 patients (69.6% male, median age 60 years), HEV infection exhibited a winter–spring seasonal pattern in Shanghai, China. The overall 90 day adverse outcome rate was 4.5% in the general population and 18.1% in cirrhotic patients (aOR = 4.37, 95% CI: 1.69–11.33; p = 0.002). No 90 day adverse outcomes occurred in pregnant patients or those with HIV/AIDS. Chronic liver disease (CLD) (OR = 2.65, 1.27–5.70; Firth p = 0.008), MELD score (OR = 1.25, 1.18–1.32; Firth p < 0.001), and neutrophil-to-lymphocyte ratio (NLR) (OR = 1.18, 1.09–1.27; Firth p < 0.001) were independently associated with 90 day adverse outcomes. The CLD–MELD–NLR nomogram achieved high accuracy in the general population (AUC = 0.94, 95% CI: 0.89–0.98), and the MELD–NLR nomogram also performed outstandingly in the CLD subgroup (AUC = 0.90, 95% CI: 0.84–0.96). Both models demonstrated good calibration and clinical utility.

Conclusions

The distinct seasonality and disease burden of HEV in Shanghai highlight the need for targeted local public health measures. The robustly validated CLD–MELD–NLR and MELD–NLR nomograms provide practical tools for early risk stratification and personalized management across diverse clinical settings.

Graphical Abstract