Emerging zoonotic diseases frequently originate in wildlife and are driven by human activities such as land-use change and urbanization. The authors note that a large proportion of emerging infectious disease events are zoonotic and often originate from wildlife. Effective surveillance and prompt response require collaborative, multidisciplinary approaches that integrate human, animal, and environmental health perspectives under the One Health concept.
The paper presents an index-based alert system to enhance early detection of potential zoonotic outbreaks by exploiting real-time wildlife health data collected through an existing national platform.
The Brazilian SISS-Geo (Wildlife Health Information System) is a free digital platform established in 2014 to enable real-time, georeferenced collection of wildlife health and environmental observations via mobile devices. SISS-Geo supports contributions from health professionals, researchers, environmental managers, and the general public. The platform integrates occurrence registration, expert diagnoses, visualization tools, and real-time alerts for wildlife-related health events.
SISS-Geo has been applied in operational support during major wildlife health events in Brazil, including responses to Yellow Fever (YF) outbreaks. The platform is cited in international One Health planning documents as an example of an integrated surveillance tool that can support early warning and response systems by collecting and integrating wildlife health data for multisectoral action.
To improve SISS-Geo’s fixed-rule alerting, the authors developed a dynamic alert approach based on a Multi-Attribute Zoonotic Alert Index (Z-Alert). The Z-Alert assigns a numerical score to spatially and temporally clustered records from SISS-Geo, with the goal of quantifying each cluster’s level of attention for surveillance teams.
The system architecture comprises two principal stages: a clustering stage that groups related records, and a scoring stage that computes the Z-Alert index for each cluster using optimally weighted cluster attributes.
In the clustering stage, SISS-Geo records—representing animals reported as alive, dead, or sick—are grouped when they are spatiotemporally related. This grouping facilitates identification of patterns relevant to epidemiological surveillance that might not be evident from individual reports.
For the scoring stage, the authors used a multi-objective optimization approach to derive optimal weights for the attributes that compose the Z-Alert index. The optimization balances multiple objectives so that the index reflects the relative importance of different cluster features when prioritizing surveillance actions.
The paper reports that this combined use of clustering techniques and multi-objective optimization enhances the alert system’s capacity to flag clusters that merit investigation.
Each identified cluster receives a numeric Z-Alert score computed from weighted attributes. Representative attributes listed include the percentage of dead animals, the temporal interval among records, and the geographical spread of the cluster. These attributes are used because they are epidemiologically informative when assessing potential outbreaks.
The system allows health managers to set custom alert thresholds, enabling adaptation of sensitivity and specificity of alerts to local operational needs and risk tolerances. This customization supports context-specific prioritization of prevention and investigation actions.
To validate the model, the authors used historical Yellow Fever data provided by Brazil’s Ministry of Health. The validation assessed whether the index-based alerts from grouped SISS-Geo records could support surveillance teams in prioritizing actions, improving risk assessment, and optimizing resource allocation.
Details on validation procedures, performance metrics, and quantitative results are reported in the full article. The source indicates that historical YF data were instrumental in demonstrating the system’s utility for supporting outbreak detection and response.
The SISS-Geo platform offers different levels of data access: registered users receive limited general datasets, while sensitive geolocated confirmed disease cases and specific animal health conditions are restricted to authorized institutional users. Because of legal and privacy constraints, the complete dataset used in the study cannot be publicly shared.
However, a minimal dataset sufficient to reproduce the analyses and the author-generated code are publicly available on GitHub (repository referenced in the source). The Brazilian Ministry of Health also provides anonymized municipality-level data on confirmed human cases and non-human primate epizootics of Yellow Fever via open data channels.
Readers wishing to access full SISS-Geo institutional data must follow the platform’s institutional data access procedures and contact the coordination team as detailed on the SISS-Geo website.
The index-based Z-Alert system is presented as a practical tool to support early warning and response by enabling surveillance teams to prioritize clusters for investigation. Specific implications described include improved risk assessment, more efficient allocation of limited resources, and promotion of cost-effective strategies for outbreak control.
Because SISS-Geo collects geographically precise reports from diverse contributors, the enhanced alert system can accelerate information dissemination and simplify response efforts, including in remote regions where traditional surveillance is challenged.
The authors acknowledge data access and privacy constraints: sensitive geolocated records and confirmed case details are restricted, limiting public sharing of the complete study dataset. Institutional procedures govern access to these data for authorized users.
The source emphasizes that the alert system is configurable and intended to complement, not replace, existing surveillance workflows and expert judgment. The Z-Alert score provides a prioritization metric to guide investigations but requires integration with local epidemiological knowledge and operational capacity.
The study introduces a Multi-Attribute Zoonotic Alert Index (Z-Alert) integrated with the SISS-Geo platform to improve early detection and prioritization of wildlife-associated events that may signal zoonotic outbreaks. By combining spatiotemporal clustering of reports with multi-objective optimization to weight informative attributes, the system produces customizable alert scores to support surveillance and response. Validation using historical Yellow Fever data illustrates the approach’s potential to assist health managers in risk assessment and resource prioritization while contributing to One Health early warning efforts.
For reproducibility, a minimal dataset and the code used for analyses are available on the authors’ GitHub repository; access to full SISS-Geo institutional data is governed by platform procedures and privacy regulations.