Effectiveness of dynamic epidemic intelligence networks for early detection and intervention in infectious diseases: a systematic review and meta-analysis
Abstract
BackgroundThe increasing frequency of global health emergencies highlights the critical need for surveillance systems that transcend the limitations of traditional laboratory confirmation. Dynamic epidemic intelligence systems and syndromic surveillance networks have emerged as vital tools for early detection, yet their operational effectiveness remains characterized by significant variability across geographic and clinical contexts.ObjectiveThis study aimed to provide a comprehensive meta-analytical evaluation of the effectiveness of dynamic epidemic intelligence systems, focusing on diagnostic classification accuracy, outbreak detection sensitivity, signal precision (Positive Predictive Value), and temporal forecasting reliability.MethodsA systematic review and meta-analysis were conducted, synthesizing data from original research articles searched across multiple databases (PubMed, Ovid/MEDLINE, and Embase) Methodological quality was appraised using structured six-domain quality assessment framework adapted from CASP principles and surveillance-system evaluation criteria. Diagnostic validity was synthesized using a bivariate random-effects (Reitsma) model to account for the correlation between sensitivity and specificity. Outbreak detection and alert precision were pooled using random-effects model, MSE was synthesized using inverse-variance random-effects meta-analysis on the log-transformed scale and back-transformed for interpretation. MAPE was summarized descriptively because only two estimates from a single study were available.ResultsThirty-four articles were included for analysis. The pooled diagnostic analysis demonstrated high specificity (0.97; 95% CI: 0.94–0.99) but moderate sensitivity (0.64; 95% CI: 0.37–0.85), with a Summary Receiver Operating Characteristic (SROC) Area Under the Curve of 0.96. While systems were highly effective at detecting established outbreaks (pooled rate: 0.91; 95% CI: 0.73–0.97), performance was markedly lower for rare pathogens such as Ebola (0.30) compared to international health threats (0.97). Alert precision was a significant challenge, with a pooled Positive Predictive Value of 0.36 (95% CI: 0.14–0.66), indicating that approximately 64% of system-triggered signals were “noise”.ConclusionsEpidemic intelligence systems are powerful instruments for excluding non-events and capturing large-scale outbreaks, yet their reliability for daily operational alerts is highly context dependent. To minimize investigative burden, future implementations must prioritize pathogen-specific optimization and the integration of diverse data streams to enhance sensitivity in humanitarian and low-resource settings.
Related
- Effectiveness of intervention using artificial intelligence for students with disabilities: systematic review and meta-analysis
- Artificial intelligence methods for infectious disease outbreak detection: a systematic review
- Effectiveness of artificial intelligence-assisted colonoscopy for colorectal lesion detection: a systematic review and meta-analysis of randomized controlled trials
- Effectiveness of an Artificial Intelligence Health Education Accurately Linking System (AI-HEALS) for the intervention of chronic kidney disease: a randomised controlled trial protocol
- Effectiveness and Safety Outcomes After Switching from Reference Biologics to Biosimilars for Treatment of Non-Infectious Uveitis: A Systematic Review and Meta-Analysis.
BibTeX
@article{zhang2026effectiveness,
title = {{Effectiveness of dynamic epidemic intelligence networks for early detection and intervention in infectious diseases: a systematic review and meta-analysis}},
author = {Jun Zhang},
journal = {Frontiers in Public Health},
year = {2026},
month = nov,
doi = {10.3389/fpubh.2026.1869243},
url = {https://www.frontiersin.org/articles/10.3389/fpubh.2026.1869243},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Effectiveness of dynamic epidemic intelligence networks for early detection and intervention in infectious diseases: a systematic review and meta-analysis
AU - Jun Zhang
PY - 2026
DA - 2026/11/02
JO - Frontiers in Public Health
DO - 10.3389/fpubh.2026.1869243
UR - https://www.frontiersin.org/articles/10.3389/fpubh.2026.1869243
AB - BackgroundThe increasing frequency of global health emergencies highlights the critical need for surveillance systems that transcend the limitations of traditional laboratory confirmation. Dynamic epidemic intelligence systems and syndromic surveillance networks have emerged as vital tools for early detection, yet their operational effectiveness remains characterized by significant variability across geographic and clinical contexts.ObjectiveThis study aimed to provide a comprehensive meta-analytical evaluation of the effectiveness of dynamic epidemic intelligence systems, focusing on diagnostic classification accuracy, outbreak detection sensitivity, signal precision (Positive Predictive Value), and temporal forecasting reliability.MethodsA systematic review and meta-analysis were conducted, synthesizing data from original research articles searched across multiple databases (PubMed, Ovid/MEDLINE, and Embase) Methodological quality was appraised using structured six-domain quality assessment framework adapted from CASP principles and surveillance-system evaluation criteria. Diagnostic validity was synthesized using a bivariate random-effects (Reitsma) model to account for the correlation between sensitivity and specificity. Outbreak detection and alert precision were pooled using random-effects model, MSE was synthesized using inverse-variance random-effects meta-analysis on the log-transformed scale and back-transformed for interpretation. MAPE was summarized descriptively because only two estimates from a single study were available.ResultsThirty-four articles were included for analysis. The pooled diagnostic analysis demonstrated high specificity (0.97; 95% CI: 0.94–0.99) but moderate sensitivity (0.64; 95% CI: 0.37–0.85), with a Summary Receiver Operating Characteristic (SROC) Area Under the Curve of 0.96. While systems were highly effective at detecting established outbreaks (pooled rate: 0.91; 95% CI: 0.73–0.97), performance was markedly lower for rare pathogens such as Ebola (0.30) compared to international health threats (0.97). Alert precision was a significant challenge, with a pooled Positive Predictive Value of 0.36 (95% CI: 0.14–0.66), indicating that approximately 64% of system-triggered signals were “noise”.ConclusionsEpidemic intelligence systems are powerful instruments for excluding non-events and capturing large-scale outbreaks, yet their reliability for daily operational alerts is highly context dependent. To minimize investigative burden, future implementations must prioritize pathogen-specific optimization and the integration of diverse data streams to enhance sensitivity in humanitarian and low-resource settings.
SN - 2296-2565
ER -