Scholarlink

Background and aim <p>Recurrent spontaneous bacterial peritonitis (SBP) is a major concern for cirrhotic patients with ascites. This study seeks to identify predictors of recurrent SBP using clinical factors, inflammatory markers, and machine learning models.</p> Patients and methods <p>The study involved 347 patients with cirrhotic ascites and SBP. Receiver Operating Characteristic (ROC) curve analysis assessed the predictive ability of biomarkers. A composite score was created to evaluate the risk stratification model. Different machine learning models were compared for predictive accuracy.</p> Results <p>Eighty-three patients (23.9%) experienced recurrent SBP. Independent predictors of recurrence in multivariable analysis included acute kidney injury (AKI), elevated C-reactive protein (CRP) levels, higher serum bilirubin levels, a higher model for end-stage liver disease (MELD) score, proton-pump inhibitor (PPI) use, and lack of β-blocker use. A composite 10-point score (including AKI, CRP &gt; 50&#xa0;mg/L, low albumin levels &lt; 2.5&#xa0;g/dL, ascitic protein &lt; 1.0&#xa0;g/dL, albumin/ascitic ratio &lt; 2.5 [2 points], MELD ≥ 15, diabetes, multidrug-resistant organism [MDRO] infection, and non-use of β-blockers) stratified the risk of recurrence into low (0–3: 15%), moderate (4–6: 45%), and high (7–10: 80%) categories. Machine learning models outperformed supervised machine logistic regression in predicting recurrence. Logistic regression achieved 70% accuracy, 65% sensitivity, and 68% specificity. The decision tree model improved accuracy to 75%, sensitivity to 72%, and specificity to 71%. The random forest model showed the best performance with 78% accuracy, 77% sensitivity, and 76% specificity.</p> Conclusion <p>A composite score, combined with machine-learning models like random forest, enhances risk assessment for SBP recurrence. Clinical predictors such as AKI, CRP, bilirubin, MELD, PPI use, and β-blockers non-use help in targeted prevention.</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

Risk stratification of spontaneous bacterial peritonitis recurrence: integrating acute kidney injury, biomarkers, composite scores, and machine learning models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月7日

查看全文
  • Nasser Mousa,
  • Alaa Elmetwalli,
  • Mostafa Abdelsalam,
  • Mohamed Wahba,
  • Mohamed Selim,
  • Dina Nour,
  • Eman Abdelkader,
  • Ahmed El-Eraky,
  • Amany Hasson,
  • Ahmed E Taha,
  • Eman Mousa,
  • Adel El-Assmy,
  • Ali El-Assmy,
  • Sherif Shiha,
  • Muhammad Diasty,
  • Mohammed Abdelaziz,
  • Shereen A. Mourad,
  • Nader Elmalky,
  • Marwa Mansour

摘要

Background and aim

Recurrent spontaneous bacterial peritonitis (SBP) is a major concern for cirrhotic patients with ascites. This study seeks to identify predictors of recurrent SBP using clinical factors, inflammatory markers, and machine learning models.

Patients and methods

The study involved 347 patients with cirrhotic ascites and SBP. Receiver Operating Characteristic (ROC) curve analysis assessed the predictive ability of biomarkers. A composite score was created to evaluate the risk stratification model. Different machine learning models were compared for predictive accuracy.

Results

Eighty-three patients (23.9%) experienced recurrent SBP. Independent predictors of recurrence in multivariable analysis included acute kidney injury (AKI), elevated C-reactive protein (CRP) levels, higher serum bilirubin levels, a higher model for end-stage liver disease (MELD) score, proton-pump inhibitor (PPI) use, and lack of β-blocker use. A composite 10-point score (including AKI, CRP > 50 mg/L, low albumin levels < 2.5 g/dL, ascitic protein < 1.0 g/dL, albumin/ascitic ratio < 2.5 [2 points], MELD ≥ 15, diabetes, multidrug-resistant organism [MDRO] infection, and non-use of β-blockers) stratified the risk of recurrence into low (0–3: 15%), moderate (4–6: 45%), and high (7–10: 80%) categories. Machine learning models outperformed supervised machine logistic regression in predicting recurrence. Logistic regression achieved 70% accuracy, 65% sensitivity, and 68% specificity. The decision tree model improved accuracy to 75%, sensitivity to 72%, and specificity to 71%. The random forest model showed the best performance with 78% accuracy, 77% sensitivity, and 76% specificity.

Conclusion

A composite score, combined with machine-learning models like random forest, enhances risk assessment for SBP recurrence. Clinical predictors such as AKI, CRP, bilirubin, MELD, PPI use, and β-blockers non-use help in targeted prevention.