Introduction

Delays in diagnostic paracentesis for suspected spontaneous bacterial peritonitis (SBP) contribute to preventable mortality in cirrhosis. This study aimed to develop and internally validate a pragmatic bedside score using routinely available admission variables to stratify the risk of SBP at hospital presentation.

Material and methods

In this retrospective single-center cohort study, 1,312 consecutive adults with cirrhosis and ascites admitted between January 2020 and January 2025 were screened. A total of 1,000 unique patients met the eligibility criteria and were randomly divided into derivation (n = 700) and validation (n = 300) cohorts. SBP was defined as an ascitic polymorphonuclear leukocyte count ≥ 250/mm3. Multivariable logistic regression was used to identify independent predictors, which were incorporated into a four-component bedside score (SCAN: serum sodium, C-reactive protein, age, and neutrophil-to-lymphocyte ratio). Model discrimination and calibration were assessed.

Results

Spontaneous bacterial peritonitis was diagnosed in 185 patients (18.5%). In the validation cohort, the SCAN score demonstrated good discriminatory performance (AUROC 0.91, 95% CI: 0.87-0.94) and acceptable calibration (slope 0.96, Brier score 0.06). The probability of SBP increased progressively from 2.2% at a score of 0 to 75.0% at scores of 5-6. A prespecified cut-off of ≥ 3 points yielded a sensitivity of 88.1% and a specificity of 91.7%.

Conclusions

The SCAN score is a simple bedside tool that stratifies the risk of spontaneous bacterial peritonitis using routine admission variables. It may assist clinicians in prioritizing timely diagnostic evaluation, complementing guideline-recommended management. Prospective multicenter validation studies are warranted.