Feed

Enterovirus D68 (EV-D68) has been implicated in clusters of acute flaccid myelitis (AFM) and severe respiratory illness; however, the magnitude and consistency of association across settings and over time remain uncertain. We quantified the association between EV-D68 detection and acute neurologic outcomes (particularly AFM) and explored design- and time-related heterogeneity. Following PRISMA 2020 and MOOSE guidance, we systematically identified observational studies reporting associations between EV-D68 detection and neurologic outcomes. Random effects meta-analysis was performed using REML with Knapp-Hartung adjustment; heterogeneity was summarised with τ² and I². Prespecified moderators were examined with meta-regression. Small-study effects were evaluated using contour-enhanced funnel plots, Egger and Peters tests, and PET-PEESE bias-adjusted models. PROSPERO registration: 1152300. Across 98 studies, the pooled odds ratio (OR) was 1.39 (95% CI 1.14-1.69), with high heterogeneity (I² = 98.9%; prediction interval 0.24-8.15). By design, respiratory surveillance studies showed stronger association (OR 1.59, 95% CI 1.35-1.86, k = 78) whereas AFM case-control studies did not (OR 0.86, 95% CI 0.40-1.86, k = 20). In multivariable meta-regression, study design and calendar year explained about 47% of inter-study variance. Predicted ORs declined from 2014 to 2022 across regions. Funnel asymmetry and small-study effects were suggested; PET-PEESE bias-adjusted estimates remained above the null. The EV-D68 outcome association is context-dependent and time-varying. Surveillance datasets enriched during outbreak waves drive the pooled signal, while AFM case-control designs yield attenuated estimates after adjustment. Standardising diagnostics and integrating design-specific surveillance will improve risk estimation and AFM preparedness.