Syndromic Surveillance in Veterinary Practice

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-11

Key Takeaways

• Syndromic surveillance in veterinary medicine systematically monitors pre-diagnostic health signals, such as abortion storms in ruminants or neurological signs in equids, to detect emerging disease events earlier than traditional laboratory-confirmed case reporting.
• Core data sources include clinical practitioner records, laboratory test requests, mortality reports, and production data, each with distinct properties and limitations regarding standardization, timeliness, and representativeness of true disease incidence.
• The primary analytical challenge involves distinguishing true population-level aberrations from background noise, seasonal variations, and reporting artefacts, necessitating robust baseline establishment and multivariate monitoring approaches.
• Veterinary syndromic surveillance holds significant value for the "One Health" paradigm by serving as an early warning system for zoonotic diseases, such as arboviruses, where animal populations can act as sentinels for human health risks.
• Implementation success hinges on clearly defining the health event of interest in syndromic terms, identifying existing data streams, establishing reliable baselines, and pre-defining response protocols to ensure timely investigation and action.
• Recognized complications include reporting fatigue, alert fatigue among analysts, data misclassification due to lack of standardized coding, and temporal artefacts from operational changes, all of which require specific troubleshooting and corrective actions.

Syndromic surveillance in veterinary medicine is the systematic, continuous monitoring of health-related data that precede definitive diagnosis, with the objective of detecting unusual patterns that may signal an emerging disease event. Unlike laboratory-based surveillance, which depends on confirmed pathogen identification, syndromic surveillance operates on clinical signs, production indicators, and other proxy measures that can be captured in near real time. This article examines the conceptual foundations, data sources, analytical methods, and operational considerations of veterinary syndromic surveillance for a research-oriented readership. It addresses how surveillance systems can be designed to detect aberrations in animal populations before laboratory confirmation is possible, and it evaluates the evidence base for current approaches.

The clinical question at the center of this field is straightforward: can systematic monitoring of non-specific health signals detect an outbreak earlier than traditional case reporting? The answer matters because emerging zoonoses, vector-borne diseases, and production-limiting conditions often produce non-specific clinical presentations in the early phase of an epizootic. Abortion storms, neurological signs in horses, and unusual mortality in birds are examples of syndromic presentations that may precede laboratory confirmation of a specific aetiology. For the veterinary researcher, understanding the design logic, data requirements, and analytical constraints of syndromic surveillance is prerequisite to evaluating existing systems or developing new ones.

This article draws on the European Triple-S project inventory, systematic reviews of veterinary syndromic surveillance initiatives, and international standards from the World Organization for Animal Health (WOAH). It does not cover laboratory-based surveillance, confirmatory diagnostic testing, or molecular epidemiology, except where those topics intersect with the interpretation of syndromic signals.

At a Glance

Conceptual Foundations

Syndromic surveillance rests on a simple epidemiological premise: clinical signs precede diagnosis, and diagnosis precedes official reporting. By capturing data at the earliest of these stages, surveillance systems can compress the interval between an emerging health event and its detection. The approach originated in human public health, where automated monitoring of emergency department visits, pharmacy sales, and school absenteeism was developed to detect bioterrorism events and emerging respiratory pathogens. Veterinary applications have adapted these methods to animal populations, with the additional complexity of multiple species, varied production systems, and heterogeneous data capture mechanisms.

The Triple-S project, a European Commission co-financed initiative that ran from 2010 to 2013, conducted an inventory of veterinary syndromic surveillance systems across Europe. The inventory identified 27 systems in 12 countries, at development stages ranging from project phase to fully operational. The survey revealed substantial diversity in objectives, target populations, and data providers, and it highlighted the relative novelty of the field. The authors noted that the inventory itself was an active process, requiring direct contact with animal health experts to identify systems that were often not documented in peer-reviewed literature.

A systematic review published in 2011 examined veterinary syndromic surveillance initiatives and found that development efforts were concentrated in two areas: animal health monitoring and the use of animals as sentinels for human public health. The review identified clinical data from practitioners and laboratory data as the principal information sources, with a range of additional sources under exploration. The authors observed that early development focused on data collection strategies and historical data analysis, in part because of limitations inherent in how animal health data are captured.

The Surveillance Logic

Syndromic surveillance operates on the principle that many emerging disease events produce detectable changes in population-level health indicators before individual cases are diagnosed. The signal may be subtle, such as a small increase in abortion submissions to diagnostic laboratories, or dramatic, such as a cluster of neurological cases in horses. The surveillance system's task is to detect these changes against a background of normal variation, seasonal patterns, and reporting artefacts.

The choice of syndrome definition is the central design decision. A syndrome definition specifies which clinical signs, laboratory test requests, or other indicators will be monitored. Definitions must be broad enough to capture relevant cases but specific enough to avoid excessive false alarms. The Triple-S inventory found that systems varied considerably in their syndrome definitions, with some monitoring single clinical presentations and others aggregating multiple indicators into composite syndromes.

Data timeliness is a critical constraint. Syndromic surveillance adds value only when it detects events earlier than traditional surveillance would. This requires data streams that are captured and transmitted frequently, ideally daily or weekly. The 2011 review noted that diagnostic laboratories appeared to provide the most readily available data sources for veterinary syndromic surveillance, because laboratory test request data are already digitised and transmitted regularly. However, laboratory data capture only those cases where testing is pursued, introducing selection bias that must be accounted for in interpretation.

Data Sources and Their Properties

Clinical data from veterinary practitioners remain the most direct source of syndromic information. Practice management software can record presenting complaints, clinical signs, and preliminary diagnoses in structured formats that are amenable to automated extraction. The 2016 systematic review of animal health syndromic surveillance confirmed that clinical data from practitioners and laboratory data remain the main data sources, while noting that production data, mortality data, abattoir data, and internet search activity have been the subject of increasing methodological development.

Each data source has distinct properties that affect its utility for surveillance. Practitioner clinical data are clinically rich but may be incomplete, non-standardized, and subject to practice-level variation in coding behavior. Laboratory data are more standardized but capture only a fraction of clinical cases and reflect testing patterns that vary with economic conditions, practitioner preferences, and diagnostic capacity. Production data, such as milk yield, feed consumption, or egg production, are collected continuously on many farms and can detect health events that affect productivity before clinical signs are recognized. Mortality data from rendering plants or farm records can detect unusual death patterns, particularly in poultry and swine operations.

The 2016 review identified the lack of classification standards as a persistent barrier to automated syndromic analysis. Without standardized coding of clinical signs across practices and regions, aggregating data from multiple sources requires substantial data cleaning and harmonisation effort. The review also noted that reporting sustainability remains a challenge, with systems based on passive notification or voluntary data transfers struggling to maintain participation over time.

One Health and Zoonotic Applications

Veterinary syndromic surveillance has particular value in the detection of zoonotic disease events, where animals may serve as early warning indicators for human health threats. Arboviruses provide a clear example. Rift Valley fever and Wesselsbron virus are associated with abortion storms in livestock, while West Nile virus, Shuni virus, and Middelburg virus cause neurological disease outbreaks in horses. Avian mortality may signal Bagaza virus and Usutu virus activity. A 2018 review of African arboviruses with zoonotic potential noted that syndromic surveillance in animals may serve as an early warning system for detecting zoonotic arbovirus outbreaks before human cases are recognized.

The One Health rationale extends beyond arboviruses. Climate change is expected to alter the geographic distribution of vector-borne diseases, and integrated surveillance systems that monitor animal, human, and environmental signals may detect emerging threats earlier than sector-specific systems. A 2018 One Health review argued that the cost of emerging vector-borne zoonotic pathogen outbreaks may be substantially lower if detected early in vectors or livestock instead of later in humans, and it called for regional and global integrated syndromic surveillance and response systems.

The operational implication is that veterinary syndromic surveillance systems should be designed with data sharing and interoperability in mind. A system that monitors abortion storms in livestock may have public health value if the data can be shared with human health authorities in a timely manner. This requires attention to data governance, privacy protections, and cross-sectoral communication protocols, topics addressed in the later sections of this article.

Implementation Pathways

The gap between syndromic surveillance theory and operational practice is wide. The Triple-S inventory of European veterinary syndromic surveillance initiatives identified 27 systems across 12 countries, at stages ranging from project phase to fully active operation. That heterogeneity is instructive. Implementation succeeds when the system is matched to the data that already exist, the signal that matters for the target population, and the decision that the system is meant to inform.

The Assessment Sequence

Before any data stream is selected, the surveillance question must be specified. The sequence runs as follows.

First, define the health event of interest in syndromic terms. A syndrome is a cluster of clinical signs, not a diagnosis. For early detection of Rift Valley fever, the syndrome of interest is abortion storms in ruminants. For West Nile virus, it is acute neurological disease in equids. The case definition must be simple enough for a data provider to apply without laboratory confirmation, because the entire point is to capture events before the laboratory result exists.

Second, identify the data that are already being generated. The systematic review of animal health syndromic surveillance from 2011 to 2016 found that clinical data from practitioners and laboratory data remain the principal sources, but production data, mortality data, abattoir data, and internet search activity have all been evaluated as candidate indicators. The correct choice depends on what is recorded, how reliably it is recorded, and how quickly it reaches the analyst.

Third, establish the baseline. Syndromic surveillance is inherently statistical. A signal is an aberration from an expected pattern, so the expected pattern must be characterized before the system goes live. This requires historical data spanning at least several years to account for seasonality, particularly in production species where reproductive cycles and management calendars drive predictable variation.

Fourth, define the response protocol. A syndromic alarm is not a diagnosis. The response is a structured investigation: confirm the excess is real, rule out artefact, and escalate to laboratory investigation if the pattern is plausible. Without a pre-agreed response pathway, alarms accumulate and the system loses credibility with the clinicians who supply the data.

Decision Points and What Changes Them

The first decision is whether syndromic surveillance is the right tool at all. For diseases with pathognomonic lesions or rapid laboratory confirmation, traditional case reporting is more efficient. Syndromic surveillance earns its complexity when the threat is emerging, the clinical presentation is nonspecific, or the laboratory infrastructure is limited. The cost of outbreaks of emerging vector-borne zoonotic pathogens may be substantially lower if they are detected early in livestock instead of later in humans, which argues for syndromic approaches in exactly those settings.

The second decision is data source selection. This is where species and production system matter most. In intensive pig and poultry operations, mortality data are recorded daily and are highly standardized, making them an excellent syndromic stream. In extensive cattle systems, abortion reports may be the only consistently recorded reproductive event. In companion animal practice, electronic medical records contain free-text clinical notes that require natural language processing to extract syndrome categories, a substantial technical investment. The review of veterinary syndromic surveillance initiatives noted that diagnostic laboratories provide the most readily available data sources, because laboratory submissions are already digitised, standardized, and time-stamped.

The third decision is the analytical method. Univariate monitoring of a single data stream is simpler but produces more false alarms, because any transient disturbance in that stream looks like an outbreak. Multivariate systems that monitor several streams concurrently are inferentially more accurate, and Bayesian methodologies are particularly suited to discovering the interplay among multiple syndromic data sources. The trade-off is complexity. A system monitoring abortion reports, mortality, and clinical consultations simultaneously must model the correlation structure among those streams, which requires statistical expertise that many practices and even regional authorities lack.

Monitoring Parameters and Their Interpretation

Each parameter has a characteriztic failure mode. Abortion rates are exquisitely sensitive to reporting fatigue. Mortality proportions in poultry are confounded by management changes such as thinning or feed transitions. Neurological case counts in horses are so small that a single cluster can trigger an alarm while a genuine outbreak in a neighbouring region produces no signal at all. The analyst must know which failure mode applies to each stream and must communicate that uncertainty to the response team.

A Case Study in Implementation

Consider a regional veterinary authority responsible for livestock health across a mixed farming area with dairy cattle, sheep, and a small horse population. The authority decides to implement syndromic surveillance for arboviral disease after a neighbouring region experiences a West Nile virus outbreak.

The system is built on three streams. The first is abortion reports from veterinary practitioners, submitted through an existing notifiable disease portal. The second is daily mortality data from the three largest dairy operations, transmitted automatically from farm management software. The third is neurological case reports from the two equine referral hospitals in the region.

The case definition for the equine stream is any horse presenting with acute onset ataxia, paresis, or cranial nerve deficits, regardless of vaccination status. The definition for the abortion stream is any abortion in a ruminant during the vector season, defined locally as May through October.

Baseline data are assembled from the preceding five years. The abortion stream shows a strong seasonal peak in late winter, which reflects nutritional and management factors unrelated to arboviral disease. The equine stream shows no seasonal pattern but has a high variance because the annual case count is only 15 to 25. The mortality stream is stable except for a predictable spike in July, when the dairy operations cull cows.

The analytical approach is a multivariate Bayesian model that monitors all three streams jointly. The model is calibrated to raise an alert when the posterior probability of an aberration exceeds a threshold set by simulation to produce no more than one false alarm per season.

In the third year of operation, the equine stream produces two neurological cases in a single week, both from the same county. The model does not alarm, because the posterior probability remains below threshold. The authority investigates anyway, because the spatial clustering is unusual. Both horses are seronegative for West Nile virus but positive for Shuni virus, a less well known arbovirus with zoonotic potential. The abortion stream shows no aberration, and the mortality stream is flat.

The investigation concludes that the cluster represents a genuine but locally contained outbreak. The authority issues an advisory to equine practitioners in the county, recommending vector control and diagnostic testing for arboviral disease in any febrile neurological case. No further cases are reported.

This case illustrates the central lesson of syndromic surveillance implementation. The statistical system did not fire, but the investigation protocol did. The value of the system was not the alarm, which was correctly suppressed by the multivariate model, but the structured attention that the monitoring process brought to a rare event. The system worked because the response protocol was specified in advance, because the data streams were already being collected, and because the investigating clinician knew which differential diagnoses to pursue when the syndrome appeared.

Data Source Catalogue

The following data sources have been used or evaluated for veterinary syndromic surveillance, with their operational properties.

Practice management software. Clinical records from first-opinion practices contain consultation dates, species, and free-text notes. The advantage is timeliness and coverage of the entire presenting population. The disadvantage is the absence of standardized coding, which forces reliance on text mining. This source is most practical in companion animal practice, where electronic records are near-universal, and least practical in mixed or large animal practice, where recording is more variable.

Diagnostic laboratory accession data. Laboratory submissions are digitised, standardized, and time-stamped. They capture changes in clinician suspicion, which can precede confirmed diagnosis by days. The limitation is that submission behavior is influenced by cost, outreach activity, and regulatory requirements, so the data reflect testing patterns as much as disease incidence.

Mortality data. Daily mortality logs from intensive production operations are highly standardized and often transmitted automatically. They are the most reliable syndromic stream in poultry and swine. In cattle and sheep, mortality recording is less consistent, and the data may be confounded by management events.

Abattoir data. Condemnation records and gross pathology findings at slaughter capture subclinical disease that never reaches a veterinarian. The limitation is the delay between disease occurrence and slaughter, which can be weeks in cattle and months in breeding stock.

Production data. Milk yield, feed intake, egg production, and weight gain are recorded continuously in modern production systems. Deviations from expected curves can signal disease before clinical signs appear. The challenge is that production data are influenced by management decisions, nutrition, and environment, making the attribution of a deviation to disease difficult.

Internet search activity. Search engine queries for veterinary terms have been evaluated as a proxy for disease activity. The advantage is complete automation and zero burden on data providers. The disadvantage is that search behavior is influenced by media coverage and public concern, also by true disease incidence.

Wildlife mortality reports. Die-offs in wild bird populations have signalled arboviral activity, including Bagaza virus and Usutu virus. Wildlife data are opportunistic and geographically biased, but they can provide early warning in regions where domestic animal surveillance is sparse.

The selection of data sources must be guided by the surveillance objective, the species at risk, and the existing recording infrastructure. A system designed for early detection of zoonotic arboviruses in a region with intensive livestock production will reasonably combine mortality data, abortion reports, and equine neurological case counts. A system designed for the same purpose in a region with extensive pastoral production will rely more heavily on community-based reporting and wildlife mortality, because the structured data streams do not exist. The WOAH animal health surveillance standards provide the international framework for integrating such systems with notifiable disease reporting obligations.

Recognized Complications and Failure Modes

Syndromic surveillance systems fail in characteriztic patterns, and most failures trace to data flow instead of analytic method. Reporting fatigue is the most frequently documented complication. Data providers, initially motivated, reduce submission frequency as novelty fades, and the system degrades silently because completeness metrics are rarely monitored in real time. Detection of this failure requires tracking submission counts per provider per week and comparing them against expected baselines, also checking whether the database grows.

Alert fatigue operates at the analyst level. Systems tuned for sensitivity generate frequent alarms, and analysts begin triaging them with decreasing rigour. The discriminating check is the alarm-to-action ratio: if more than a small fraction of alarms trigger no investigation, threshold settings or case definitions need revision. The systematic review of animal health syndromic surveillance progress notes that reporting sustainability remains a persistent barrier to automated analysis and interpretation.

Data misclassification produces a subtler failure. Clinical signs recorded in free text or non-standard codes resist aggregation, and the resulting noise masks true signals. Detection requires periodic audits comparing a sample of submitted records against the original clinical notes. The inventory of European veterinary syndromic surveillance initiatives identified the lack of classification standards as a defining characteriztic of the field, with systems varying substantially in how they define and record syndromic events.

Temporal artefacts constitute a third failure class. Holiday closures, weekend staffing changes, and seasonal practice patterns create periodic fluctuations that algorithms may interpret as outbreaks. The discriminating check is calendar overlay: plot the signal against known service disruptions before investigating further.

Common Errors and Corrective Actions

Less experienced analysts often mistake statistical significance for epidemiological significance. A cluster that exceeds the threshold may reflect a coding change, a new staff member, or a local referral pattern shift. The corrective action is to require a minimum cluster size and duration before initiating investigation, and to verify data provenance first.

A second error involves over-interpreting single-stream signals. A rise in respiratory signs in one clinic may be noise, whereas the same rise distributed across multiple clinics in a region carries meaning. The review of veterinary syndromic surveillance initiatives emphasizes that multivariate monitoring, which concurrently examines several data streams, is inferentially more accurate than univariate approaches. Analysts should therefore resist acting on single-source alarms until corroborating streams are checked.

A third error is neglecting the denominator. Increases in raw counts may reflect increased patient volume instead of increased disease incidence. Corrective action is to normalize against practice-level patient counts or population estimates before interpreting trends.

Limitations of Current Evidence

The evidence base for veterinary syndromic surveillance remains thin in several areas. Most published systems are pilot projects or retrospective validations instead of prospectively evaluated programs. The systematic literature review covering 2011 to 2016 found that production data, mortality data, abattoir data, and internet search metrics have been studied mainly for indicator development and validation, not integrated into running systems. Consequently, the operational performance characteriztics of these data sources under real-world conditions are largely unknown.

Expert opinion differs on the optimal balance between sensitivity and specificity in system design. Some argue for maximally sensitive systems that generate many alarms, relying on human triage to filter false positives. Others advocate for stricter thresholds to preserve analyst attention for credible signals. Neither position has strong prospective evidence, and the choice remains context-dependent.

The WOAH animal health surveillance standards describe general surveillance principles but do not prescribe specific syndromic methods, reflecting the absence of consensus on best practice. Similarly, the CDC principles of epidemiology provide foundational surveillance concepts developed for human health, and their direct transferability to veterinary contexts with different data structures and reporting incentives is not fully established.

Escalation and Referral Criteria

Clinicians encountering unusual clusters of clinical signs should escalate through defined pathways. The threshold for regulatory reporting is not a specific case count but the presence of a pattern consistent with a notifiable disease, particularly one with trade implications. The WOAH terrestrial animal health code specifies reporting obligations for listed diseases, and national veterinary authorities maintain their own notifiable disease lists that take precedence locally.

Referral for specialist consultation is warranted when the syndromic pattern suggests a disease outside the practitioner's diagnostic comfort zone, when human health implications are possible, or when the cluster persists despite negative routine diagnostics. Laboratory involvement is indicated for confirmation of index cases, strain characterization, and antimicrobial susceptibility testing, even when the syndromic signal itself is clear.

Zoonotic potential changes the urgency calculus. The assessment of zoonotic arboviruses of African origin notes that abortion storms in livestock and neurological disease outbreaks in horses may signal arbovirus activity with human health implications, and that syndromic surveillance in animals can serve as early warning for zoonotic outbreaks. When such patterns appear, public health authorities should be notified in parallel with veterinary authorities.

Troubleshooting Reference

Frequently Asked Questions

How Much Does a Veterinary Syndromic Surveillance System Cost to Establish and Maintain?

Costs vary widely with data source and scale. Systems built on existing electronic practice records or laboratory data have lower startup costs than those requiring new data collection infrastructure. The European inventory of veterinary syndromic surveillance initiatives found systems operating at different levels of development, from pilot projects to fully active programs, reflecting substantial variation in resource commitment. Ongoing costs are dominated by personnel time for data curation, statistical analysis, and interpretation instead of software licensing. Sustainability of reporting is a recognized challenge, as data providers may lose motivation without visible feedback. Start with a single data stream already collected routinely, then expand only after demonstrating value.

What Can Be Done When Ideal Data Infrastructure Is Unavailable?

When automated data feeds are not feasible, begin with manual, structured data capture using standardized forms. Paper-based recording of presenting signs can be digitised later. Diagnostic laboratories often provide the most readily available data sources for syndromic surveillance in animal health, as noted in the review of veterinary syndromic surveillance initiatives. Abattoir data and mortality records are also accessible in many regions without new technology. Use simple spreadsheet-based monitoring with weekly counts and basic threshold alerts. The statistical methods matter less than consistent case definitions and reliable reporting. A modest system that runs consistently outperforms an elaborate one that collapses from reporting fatigue.

How Do Surveillance Approaches Differ Between Livestock, Companion Animals, and Wildlife?

Livestock systems benefit from defined populations, production records, and established reporting channels. Abortion storms and neurological signs are well-recognized indicators, particularly for arboviral diseases such as Rift Valley fever and West Nile virus, where syndromic surveillance in animals may serve as an early warning system for zoonotic outbreaks. Companion animal data are fragmented across many small practices, requiring aggregation platforms. Wildlife surveillance relies on mortality events and sentinel species, with less denominator information. The systematic review of animal health syndromic surveillance confirms that clinical and laboratory data remain the main sources across species, but production and mortality data are more developed for livestock. Species-specific case definitions are essential because the same syndrome carries different predictive value in different populations.

What Records Should Be Kept to Support Syndromic Surveillance?

Maintain the minimum dataset: unique animal identifier, species, breed, age, sex, date of presentation, postcode or farm location, and the syndromic classification assigned. Record the case definition version used, since definitions evolve. Preserve raw clinical notes alongside coded data to allow retrospective reclassification. Document denominator data, such as total practice consultations or herd size, because counts without denominators are difficult to interpret. The WOAH animal health surveillance standards emphasize that surveillance data must be verifiable and traceable. Store data in a format that supports time series analysis, with consistent date fields and no silent corrections. Audit trails for data entry errors are critical, as reporting sustainability and classification standards remain ongoing limitations in the field.

How Should I Explain Syndromic Surveillance Findings to a Client or Supervisor?

Frame the explanation around the distinction between individual case management and population-level signal. A single unusual case is not an outbreak, but a cluster of similar presentations may be. Explain that syndromic surveillance detects patterns of clinical signs before a specific diagnosis is confirmed, which is particularly valuable for emerging zoonotic diseases. The One Health perspective on climate change notes that detecting outbreaks early in livestock instead of later in humans substantially reduces costs. Use concrete language: "We have seen three neurological cases this week where we would expect one." Avoid alarming language and emphasize that most alerts are investigated and resolved without finding a serious cause. Provide the supervisor with a written summary of counts, expected baselines, and the threshold that triggered the alert.

How Do I Decide Whether an Alert Requires Immediate Action or Continued Observation?

Apply a structured triage based on cluster size, severity, and known differentials. Compare current counts against the historical baseline for that week and region, accounting for seasonality. Assess whether the syndrome has known notifiable causes in your jurisdiction. Consider whether the affected population has shared risk factors, such as common feed, water source, or vector exposure. The CDC principles of epidemiology describe outbreak investigation steps that begin with verifying the diagnosis and confirming the excess over expected counts. If the cluster involves severe disease, high case fatality, or a zoonotic agent, escalate immediately. If the cluster is mild and self-limited, continue daily monitoring for 48 to 72 hours before deciding whether investigation is warranted. Document the rationale for either decision.

Related Clinical & Scientific Guides

• Evaluating Veterinary Surveillance System Attributes
• Network Analysis for Infectious Disease Spread in Animal Populations
• Randomized Controlled Trials in Veterinary Field Settings

References and Further Reading

• Inventory of veterinary syndromic surveillance initiatives in Europe (Triple-S project): current situation and perspectives.. 2013.
• Veterinary syndromic surveillance: Current initiatives and potential for development.. 2011.
• Assessing the zoonotic potential of arboviruses of African origin.. 2018.
• Climate change and One Health.. 2018.
• Public health surveillance and infectious disease detection.. 2012.
• Animal health syndromic surveillance: a systematic literature review of the progress in the last 5 years (2011-2016).. 2016.
• WOAH Animal Health Surveillance Standards. WOAH.
• CDC Principles of Epidemiology in Public Health Practice. CDC.
• MSD Veterinary Manual, Professional Edition. MSD Veterinary Manual.

Related Articles

• Risk-Based Surveillance in Animal Health
• Time Series Analysis for Veterinary Disease Surveillance
• Data Quality Assurance in Veterinary Surveillance Systems
• Designing Case Definitions for Veterinary Surveillance
• Designing and Implementing Animal Disease Surveillance Systems

This article is educational professional reference material for veterinary audiences. It is not a substitute for veterinary diagnosis, individual clinical judgment, current product labeling, or applicable regulatory requirements.