Abstract
BACKGROUND: Spontaneous bacterial peritonitis (SBP) is a frequent and life-threatening complication of cirrhosis, contributing to considerable morbidity and mortality. METHODS: A cross-sectional derivation study was conducted to assess the diagnostic accuracy of two sepsis-related calcitonin peptide family biomarkers, mid-regional pro-adrenomedullin (MR-pro-ADM) and procalcitonin, in ascitic fluid for identifying bacteriologically confirmed SBP (BC-SBP). In a subsequent validation study, the diagnostic performance of the 'SBP score' was evaluated in an independent patient cohort using an absolute polymorphonuclear (PMN) leukocyte count threshold of ≥250 cells/mm(3) as the diagnostic benchmark for diagnosing SBP. RESULTS: In the derivation study, the concentration of MR-pro-ADM in ascitic fluid was significantly higher in patients with BC-SBP compared to those without BC-SBP (3.14 nmol/L [IQR, 2.39-6.74] vs. 1.91 nmol/L [IQR, 1.33-2.80]; p = .0002). Bayesian ANOVA indicated that MR-pro-ADM was highly discriminative for diagnosing BC-SBP, with a substantial Bayes factor (BFM = 2505), whereas procalcitonin exhibited poor discriminatory performance. Receiver-operating characteristic (ROC) analysis identified an optimal MR-pro-ADM cut-off of ≥2.50 nmol/L for diagnosing BC-SBP, with an area under the ROC curve (AUROC) of 0.746 (95% CI, 0.685-0.801; p < .0001). Multivariable logistic regression identified three independent predictors of BC-SBP, which were subsequently incorporated into the 'SBP score' (MR-pro-ADM ≥2.5 nmol/L, absolute PMN count ≥250 cells/mm(3) and Child-Pugh score). In the validation study, the 'SBP score' demonstrated an AUROC of 0.993 (95% CI, 0.929-1.000; p < .0001) for diagnosing SBP. CONCLUSION: MR-pro-ADM in ascitic fluid emerges as a promising biomarker for SBP diagnosis. Combining MR-pro-ADM with absolute PMN count and Child-Pugh score in the 'SBP score' greatly improves the diagnostic accuracy of SBP.