A set of training materials for professionals working in intervention epidemiology, public health microbiology and infection control and hospital hygiene.
Need help with your investigation or report writing? Ask the Expert. Free advice from the professional community.
You can't make decissions on this page's approval status because you have not the owner or an admin on this page's Group.
The overall aim of surveillance is to provide information that is required to inform public health action. Although 'information for action' is often used as a meaningful synonym for surveillance, it does not help to define the purpose of surveillance any further unless we can answer the questions of 'what action?', 'what information?' and 'whose action?'.
The action may range from the management of the contacts of a single case of infection (e.g. meningococcal meningitis), to the investigation and control of an outbreak, to the development and implementation of national policy e.g. a new immunisation programme or a change in national guidance on antimicrobial therapy. Actions may also include decisions to stop particular interventions, on the basis that surveillance has demonstrated a lack of effect or has provided evidence of the elimination of the problem. Surveillance might also identify risk factors for disease that can be the focus of research studies into disease aetiology or might identify cases for inclusion in research studies.
Because the range of public health actions that might be taken is so diverse, it is helpful to define a subset of aims that link to specific areas or types of public health action. These aims, and the associated actions, include:
One of the first questions to be addressed in surveillance system design is that of where the data to be used should be drawn from? The answer will depend on the target population about which the surveillance outputs will be used to inform actions and on the relative ease and cost of capture of data from the potential sources available. Where the objectives of surveillance include estimating the distribution of cases within the population, and changes in that distribution, then the choice of target population, and of the mechanism for ascertaining cases, should be guided by considerations of the representativeness of the resulting information. Many surveillance systems are based on the collection of data from health services, which may not have an easily defined catchment population, and which may include a biased subset of all cases of a particular condition (see figure 1). Many surveillance systems based on universal reporting of cases (e.g. national notification systems or national laboratory reporting schemes) are subject to significant ascertainment bias, with under-representation of asymptomatic and less severe cases, and also relative over-representation of cases at the upper and lower most ends of the age spectrum. Those responsible for surveillance systems, and particularly for the interpretation of surveillance data, must therefore consider potential biases that will arise from the choice of data source, and ensure that recipients of the outputs are aware of those biases.
In some situations surveillance, particularly when it's primary purpose is early warning of outbreaks or epidemics, does not necessarily need to provide measures of disease occurrence that are representative of the general population. In the case of early warning surveillance, it is often more efficient, and sensitive, to focus data collection on a suitable sentinel group e.g. surveillance in the early stages of the emergence of HIV infection in many countries focused on high risk groups such as men who have sex with men and injecting drug users. Nevertheless, the representativeness of data reported from potential data sources needs to be considered not only when developing new surveillance systems, but also when using well established systems.
The purpose of surveillance is to provide information for action, and as such the design of a surveillance system should be shaped by the information requirements (surveillance system outputs) of those responsible for taking the control and prevention action that is to be informed by the system. Typically surveillance outputs will be in the form of tables or graphs showing counts or rates of cases/events, or proportional morbidity or mortality, categorised by demographic, geographic or other risk factors. These may be accompanied by statistical measures of trend or exception, and increasingly geographical information systems are being used to analyse and present surveillance outputs. Surveillance outputs may also take the form of listings of cases requiring public health follow up, or be data files that can be subject to additional analyses or used in record linkage by the recipients.
It is not always possible, usually for reasons of practicability or cost, to provide exactly the information that policy makers, service planners or health service providers might require, and often the choice of outputs is a balance of what it is possible to provide within the constraints of resources and data availability and what is actually required (see box for example).
Those responsible for the control of HIV/AIDS would ideally like surveillance systems to provide information on the incidence of infection. However, the mild and non-specific nature of HIV seroconversion illness means that true 'incident' cases are rarely detected, and so direct measures of incidence from surveillance based on clinical service testing are not possible (although this could be achieved through long term study of defined cohorts, with regular testing). Surveillance systems have therefore been developed that can provide proxy measures of incidence, or at least provide information on how recently detected infections are likely to have been acquired. This has largely been done by making use of advances in laboratory techniques, such as the development of routine assays of CD4 cell counts (higher counts at the time of first HIV positive test indicate more recent infection), and 'detuned' HIV assays.
The frequency and timeliness of outputs is a particular issue in communicable disease surveillance, where early identification of outbreaks or epidemics is required in order to mount effective measures to interrupt transmission or to prepare other responses. This need for frequent and timely outputs has important consequences for the operation of surveillance systems, such as the need for continuous data capture systems, rapid data transfer and analysis, and the human and information technology resources required to support those processes.
You need to be logged in to post comments.
You can log in here. You can register here if you haven't done so yet.