NT-proBNP shows 78% sensitivity and BNP shows 75% sensitivity for detecting paroxysmal atrial fibrillationBlood tests may help identify heart rhythm issues after stroke

INTRODUCTION: Paroxysmal atrial fibrillation (PAF) is a common cause of cryptogenic stroke (CS). It is often asymptomatic, and prolonged electrocardiographic monitoring is not widely available. Natriuretic peptides (NPs) have emerged as potentially adjunctive biomarkers to support risk stratification. METHODS: A systematic search was conducted in PubMed, Embase, CENTRAL, Scopus, Web of Science, and LILACS up to June 2025. Studies in adults with CS that measured NT-proBNP, BNP, or MR-proANP, and that used electrocardiographic monitoring as a reference standard, were included. Risk of bias was assessed using the QUADAS-2 tool, and data were meta-analyzed using a bivariate model and a hierarchical summary receiver operating characteristic model. Certainty of the evidence was assessed in accordance with the GRADE methodology. RESULTS: Eight observational studies involving 1616 patients were included, with an overall PAF prevalence of 19%. For NT-proBNP and BNP, respectively, pooled sensitivity was 78% (95% CI, 63%-87%) and 75% (95% CI, 56%-87%); specificity, 64% (95% CI, 53%-73%) and 67% (95% CI, 45%-84%); and diagnostic odds ratio, 6.18 (95% CI, 3.94-9.70) and 6.10 (95% CI, 4.42-8.42). The certainty of the evidence was low for NT-proBNP and moderate for BNP. Heterogeneity was associated with selected cutoff values, analytical platform, and reference-standard monitoring. CONCLUSION: NPs are potentially useful adjunctive biomarkers to support PAF risk stratification in patients with CS. However, the certainty of evidence for NT-proBNP was low, and for BNP moderate. Incorporating them into stepwise strategies may help stratify patients, but their incorporation requires prospective validation. NPs alone should neither guide anticoagulation nor replace rhythm monitoring.