Multimodal Data Approaches for Examining the 2024-2025 Highly Pathogenic Avian Influenza Outbreak in the United States: Descriptive Study
Bibliographic record
Abstract
<sec> <title>BACKGROUND</title> Highly pathogenic avian influenza (HPAI) outbreaks have primarily affected wild and domesticated bird populations, with occasional human spillover. In 2024, the United States (US) reported the first known HPAI H5N1 infection in dairy cattle, which rapidly evolved into a multispecies outbreak among cattle and poultry with spillover into humans. Publicly available data remained siloed and fragmented across agencies, which has implications for timely response. Innovative multimodal surveillance methods present an opportunity to enhance early situational awareness through comprehensive, standardized data collection, integration, and visualization. </sec> <sec> <title>OBJECTIVE</title> This study aimed to describe observations from the application of enhanced surveillance methods that collect, integrate, and visualize multimodal data for real-time tracking of the 2024-2025 HPAI outbreak in the US as an innovative, transparent, repeatable, and scalable approach for open source public health surveillance for zoonotic or other emerging or re-emerging pathogens. </sec> <sec> <title>METHODS</title> The Global.health consortium conducted real-time, multimodal data collection on the US HPAI outbreak between February 1, 2024 and February 28, 2025 using publicly available data for human cases, animal outbreaks, wastewater surveillance, genomic data, research updates, policy actions, and response measures. This digital data stream of traditional and non-traditional sources was used to create outbreak resources—a line-list, event timeline, and interactive map—using a One Health framework to track emerging hotspots </sec> <sec> <title>RESULTS</title> Seventy human HPAI cases were confirmed across 13 US states, with exposure for nearly all (92.9%) cases associated with commercial agriculture and related operations. Only one human case of HPAI had ever been documented in the US prior to 2024, underscoring a sharp rise in incidence. We curated 682 Timeline entries across six distinct categories: human, cattle, response, birds, genome, wastewater, and mammals. California was identified as the outbreak epicenter with leading numbers in human cases (n= 38, 54.3%), cattle (n=748, 76.6%), and poultry infections (n=66, 20.3%) during the study period. Wastewater surveillance provided an early warning sign, identifying viral presence in California at least 81 days before the first dairy cattle case. </sec> <sec> <title>CONCLUSIONS</title> The integration of traditional and non-traditional public health surveillance data into a single view within a One Health framework improved contextual understanding and enhanced situational awareness during the 2024-2025 HPAI outbreak in the US. Wastewater detections identified early viral presence, marking a critical window for intervention, policy action, and response to curb spread. Access to an open source multimodal data platform - like that put forward by Global.health - in real-time can assist researchers, public health officials, and decision makers in understanding the origins, scope, and evolution of emerging zoonotic diseases that fragmented, more traditional surveillance systems may be unable to readily provide. Further research should be conducted to understand the full potential of multimodal data in real-time outbreak surveillance. </sec>
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.010 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".