Molecular assays for the diagnosis of sepsis in neonates: a diagnostic test accuracy review.

Summary

Across 68 studies (n=14,309), neonatal molecular assays showed pooled sensitivity of 0.91 and specificity of 0.88 against culture, but with notable heterogeneity and low to very low certainty. The authors recommend randomized trials to evaluate clinical utility and cost-effectiveness as add-on tests.

Key Findings

• Pooled sensitivity 0.91 (95% CI 0.85–0.95) and specificity 0.88 (95% CI 0.83–0.92) versus culture for neonatal sepsis molecular assays.
• Substantial heterogeneity persisted across test type, gestational age, and onset type, with overall low-to-very low certainty.
• Sensitivity analyses restricted to higher-quality studies and single-sample designs yielded similar performance estimates.

Clinical Implications

Molecular assays can complement cultures to speed decision-making and may reduce unnecessary antibiotics in suspected neonatal sepsis, but adoption should be coupled with stewardship and evaluated via pragmatic RCTs.

Limitations

• High heterogeneity with low to very low certainty; culture as imperfect reference standard may bias estimates.
• Lack of randomized trials assessing clinical utility, outcomes, and cost-effectiveness of assay implementation.

Future Directions

Conduct pragmatic RCTs comparing add-on molecular testing versus standard care in suspected neonatal sepsis, focusing on antibiotic exposure, time to targeted therapy, clinical outcomes, and cost-effectiveness.