2 Public Health Surveillance
Surveillance is the collection of data related to health events to guide public health action.
It is the systematic, ongoing collection, management and analysis of data and the
timely sharing of information so that action can be taken.
Systematic
An organised process that has rules about how it operates.
Ongoing
Happening all the time- there is no end date to surveillance.
Public health surveillance data can be collected for all types of health events. These include communicable and non- communicable diseases, food or drug safety, environmental factors, and more. Health events should be under surveillance if they are important and the data collected can be used for action. There are many reasons why a health event may be important and importance can differ by location. Some reasons include that the health event is common (high incidence), has severe outcomes (e.g. disability or death), spreads easily between people, had a large economic impact or is important to the population.
Types of action that surveillance data can be used for include disease prevention activities (e.g. vaccination programs), control activities (e.g. outbreak control) and increasing knowledge (e.g. research to develop new prevention tools).
Surveillance cycle
There are four key components of the surveillance cycle (pictured below):
Component |
Description |
|---|---|
| Data collection | Relevant health data are collected about people who meet a case definition. |
| Data management and analysis | The health data are combined and organised and regularly analysed to see if there is something happening that may need public health action (e.g. if there are more cases of a disease than normal it may indicate the start of an outbreak). |
| Data dissemination | Regularly reporting the findings of data analyses back to the data collectors, the community, and those responsible for policy or public health action. |
| Public health action | The data are used as an evidence base to take action to control or prevent the spread of disease.
|
Surveillance system objective
Every surveillance system must have clear objectives that explain the reasons for collecting data and the aims that the system is trying to achieve. The system should also have a clear description of how data are collected, managed, and analysed and how the the data will be used to prevent or control disease. A system will often have more than one objective. The surveillance system objectives will determine the most important system attributes to incorporate into system design, monitoring, and evaluation.
Examples:
- An objective of a tuberculosis surveillance system may be to identify people with active tuberculosis so that they can receive treatment.
- An objective of a sexually transmitted infection surveillance system might be to inform testing and communication strategies to prevent the spread of disease.
Active surveillance
Active surveillance collects data by actively trying to find new cases. This means the pubic health authorities contact health providers, call or visit health facilities, and/or review records. It can also involve health workers going into the community to try to find cases by going house to house.
Active surveillance is often used when trying to eliminate a disease (e.g., poliomyelitis), or during the early stages of outbreak investigations.
Example:
The public health authorities may call hospitals every day to find out if they have treated any new patients with a particular disease.
Passive surveillance
Passive surveillance involves health care providers (e.g. doctors, nurses, laboratory staff) reporting cases of a disease using a documented surveillance process. When patients who meet a case definition go to health facilities, passive surveillance relies on this information to be entered into the system. The reporting site then sends the information to the higher health system level. There is no active search for cases.
Example:
A staff member at a health facility reports the number of cases who attended the clinic with diarrhoea in a surveillance report form. The report forms are sent to the next administrative level in the health system according to the surveillance process.
Indicator-based surveillance
Indicator-based surveillance is routine, structured reporting of cases of a disease or syndrome, typically from a health facility. Reporting is regular and ongoing, following an agreed process.
Example:
National syndromic surveillance system, where data are reported from health facilities on a regular basis, according to a reporting schedule. The data are entered into a surveillance database and analysed regularly.
Event-based surveillance
Event-based surveillance is unstructured reporting of unusual events that may impact public health, such as unusual disease patterns, animal disease or die-offs, chemical spills or environmental contamination. These reports may come from formal channels, such as from healthcare workers at facilities, or from informal channels, such as local media and community members reporting to facilities or to dedicated hotlines. Events detected by event-based surveillance require urgent verification to check if there is a real public health threat that requires action.
Examples:
- A farmer calls a dedicated hotline to report that 70 of their 80 chickens died suddenly and unexpectedly.
- A doctor calls the health department directly to alert them that they have seen six community members that morning from the same village with signs of neurotoxicity, and normally they would have none.
Population-based surveillance
Population-based surveillance is a type of surveillance that covers the entire population, where all diagnosed cases are reported to the system.
Example:
Notifiable disease surveillance systems require health care providers to report to the health department all cases diagnosed with specific diseases or presenting with defined clinical signs and symptoms (syndromes). Measles and tuberculosis are notifiable diseases in many countries.
Sentinel surveillance
Sentinel surveillance is when only selected sites are required to report to the surveillance system. This type of surveillance is used when it is not practical or necessary to identify all cases in a population. Sites may be chosen due to geographical spread, capacity to diagnose or test, or for financial reasons (i.e., the most cost-effective sites are included). Due to being a sample of sites, the data are not representative of the entire population, but are an indicator of what is happening with particular diseases or syndromes in those select areas.
Example:
Having laboratory staff at every facility in a country may not be possible. Instead, hospitals with suitably trained laboratory staff may be a sentinel site for malaria surveillance. This sentinel surveillance system will not be able to count all malaria cases in the country, but it can identify trends within the specific catchment areas those hospitals serve.
Syndromic surveillance
Syndromic surveillance uses case definitions that count a person as a case if they have clinical signs and symptoms (syndromes) rather than a laboratory diagnosis. This means that identifying cases can be fast because it is based on clinical signs and symptoms that are seen at a clinic, rather than needing to wait for a laboratory confirmation. It is useful for early detection of outbreaks and clusters where there may not be laboratory capacity to diagnose cases. This enables a rapid response to initiate control measures.
Example:
Using a syndromic case definition of ‘Acute fever ≥38°C and non-vesicular rash’ enables healthcare workers to quickly identify potential measles cases if there is limited laboratory capacity to test. While a laboratory test is subsequently recommended to confirm the diagnosis, by using the syndromic case definition, control measures can begin quickly as if it is a true measles case to prevent onward transmission.
Surveillance system attributes
Characteristics of a surveillance system that enable it to meet its objectives and gather data that can be used to inform public health action. Depending on the objectives of a surveillance system, some attributes will be more important than others.
Surveillance systems can be described by some of the following attributes: acceptability, data quality, flexibility, simplicity, stability, representativeness, and timeliness. The below descriptions of the attributes are sourced from the Centers for Disease Control and Prevention guidance on evaluating surveillance systems.
A surveillance system’s objectives frame the way in which we monitor and evaluate a system’s performance over time. For example, when timeliness is a critical attribute for a surveillance system, it is important to monitor the time taken between each step in the system on an ongoing basis, and to evaluate regularly.
The usefulness of a surveillance system is determined by how well the system meets it’s objectives. It is important to demonstrate how the data and information from the system are being used to improve public health. Usefulness can be influenced by the attributes of the system.
Acceptability
Willingness of people (often health staff) and organisations to contribute data to the surveillance system. It is related to whether participants believe that the information collected is useful and that what they are required to do is reasonable for meeting the surveillance system’s goal.
Some factors influencing the acceptability of a particular system include:
- The importance of the event to public health (if an issue is considered to be important to public health, people may be more accepting of the system)
- The system acknowledges the contribution of health staff (people want to feel that their effort is meaningful and valuable)
- Dissemination of data back to reporting sources and stakeholders (contributors can understand how the data are used for decision-making)
- How much time and effort are required to make the system function (efficient systems are more acceptable)
- Privacy and confidentiality of patient and participant information increases acceptability
- Cooperation of different people participating in the system
Data quality
Data quality refers to whether data in the system are complete and valid.
- Data completeness is related to whether there are missing records or data fields. It is often measured by the number or proportion of unknown or blank responses to variables in the system. The term “zero reporting” refers to reporting when there are no known cases. When calculating data completeness it is important to recognise a true zero case, rather than a “blank” response.
- Data validity means that the data in the system are accurate and true. It is often related to how many errors can be found in the data in the system.
Flexibility
The surveillance system can adapt and continue to meet its objectives when surveillance needs or the conditions in which the system operates change. Ideally, this should be able to be done without adding many new resources to the system (e.g. staff or funding). Changes that a surveillance system might need to adapt to include the emergence of a new, important health condition, a change to the case definition or method of diagnosing the condition, or changes in the availability of funding or data providers.
Example:
Adapting existing respiratory surveillance (e.g., for influenza-like illness or severe acute respiratory infections) to include COVID-19 surveillance shows systems were flexible to respond quickly to a new public health threat.
Simplicity
Simplicity refers to how easy the system is to understand and operate. A simple system makes it easy for people to participate, reduces the chance of errors in the system and makes it easy for data and analysis from the system to be used for public health action. Sometimes some complexity may be needed to achieve the objectives of the system, but it should be minimised as much as possible. Systems should be as simple as possible while still meeting their objectives.
Simplicity may be related to:
- The information needed to decide if a person has the condition under surveillance (e.g., may be a simple syndromic case definition or rapid test, or more complex clinical and laboratory diagnosis)
- The data that are collected on cases (e.g., demographic data are simpler than needing many items of risk factor data, such as occupation and travel history)
- How the data are collected (e.g., collecting data from one source is simpler than needing to collate data from multiple sources)
- The method to store and collate data (e.g., whether paper-based or electronic databases are used and whether they are easy to navigate)
- How the data are analysed and interpreted (e.g., it is easier to have a few standard analyses rather than many complicated analyses)
Stability
The system operates consistently without error or failure and is able to be accessed by participants in the system when needed.
An unstable system could be a system where there are insufficient staff to consistently be able to detect cases or analyse data. Another example might be a system that uses an electronic database but regular power outages prevent users from accessing the database when needed.
Representativeness
The data collected by the system represents the true nature of the health event in the population under surveillance. This means that analysis of data in the system gives and accurate understanding of when and where the health event is occurring and who is affected by it (time, place and person), and this provides good evidence for deciding if and what public health action should be taken.
Surveillance data from all health facilities across the country may be more representative than data from two sentinel sites, for example. However, data from sentinel sites can still be representative if the sites are chosen carefully to provide an accurate picture of what is occurring in the general population.
Another aspect of representativeness is the completeness of reporting. When a surveillance system has a set number of reporting sites, it is possible to measure the completeness of sites reporting to the system. Sites reporting late or not reporting at all affects the representativeness of the surveillance system during that time period.
Timeliness
The measure of time it takes for each step of the surveillance system to occur and how quickly information is available to enable public health action to occur. The importance of timeliness between each step in the surveillance system depends on the system’s objectives and the health event under surveillance.
Examples:
- For a surveillance system with a primary objective of detecting outbreaks, timeliness is crucial as it is important to implement control efforts quickly. There needs to be a short time between detecting a case or an outbreak and public health response.
- For a surveillance system with a primary objective of monitoring incidence of disease over time for an annual report, timeliness is less important and there is likely to be more time between steps in the surveillance cycle.
