Use and evaluation of national and international surveillance
This section covers:
- Purpose of Surveillance
- Principles of surveillance
- Categories of surveillance
- Sources of surveillance data
- Sources of specific surveillance data
- Evaluation of a Surveillance System
Surveillance is 'the continued watchfulness over the distribution and trends in the incidence of disease through the systematic collection, consolidation and evaluation of morbidity and mortality reports and other relevant data'
(Alexander Langmur, founder of Centers for Disease Control and Prevention, Atlanta, USA 1960)
Communicable disease surveillance systems’ primary purpose is to allow the rapid control of outbreaks by taking early preventive action (though surveillance does not include taking that action itself). This requires: early detection though collection, collation and analysis of data on disease incidence and spread, and dissemination of this information leading to investigation, and implementation of control measures; speed is important (Giesecke J. Modern Infectious Disease Epidemiology). For these reasons the following definition of surveillance is also favoured:
Surveillance is the 'on-going scrutiny, continuous analysis, interpretation and feedback of systematically collected data, generally using methods distinguished by their practicability, uniformity and frequently, their rapidity, rather than complete accuracy'.
(Last JM. A Dictionary of Epidemiology. New York: Oxford University Press, 1998.)
The main source of incidence data comes from clinician notifications of infectious diseases, which is often a statutory obligation (when the diseases are known as ‘notifiable diseases’). Increasingly, other sources of data – such as semi-automated electronic reporting of microbiological data on laboratory-confirmed infections - are also being used.
- Allows individual cases of infection to be notified and collated
- Measures incidence of infectious disease, with changes potentially indicating an outbreak
- Tracks trends in occurrence and risk factors of an infectious disease allowing targeted interventions
- Enables priority setting and planning of control measures e.g. to a particular region
- Evaluation of existing control measures
- Syndromic surveillance may detect emergence of new infections of public health importance.
(Adapted from Hawker et al. Communicable Disease Control and Health Protection Handbook, Third Edition. 2012, Blackwell Publishing)
Principles of surveillance
- A good surveillance system requires a case definition which includes clinical and / or microbiological criteria
- The case definition must be sensitive enough to detect cases, but also specific enough to prevent too many ‘false positives’. Different levels of case definition are often used to describe a case with increasing specificity e.g. ‘possible’, ‘probable’ and ‘confirmed’, dependent on confirmatory clinical and microbiological data.
- Several surveillance systems with differing levels of sensitivity and specificity may be used in combination to provide multi-faceted information aspects such as burden, high risk groups and laboratory-confirmed characteristics such as antimicrobial susceptibility profile
- Systematic collection of data: systematic, regular and uniform
- Analysis of data to produce statistics: by time, place and person
- Interpretation of statistics to provide information: but care needed with timeliness, completeness, representation
- Distribution of this information to those who require it for action, for instance via national bulletins. For example, in national communicable disease epidemiology reports such as PHE’s Health Protection Report, or the CDC Morbidity and Mortality Weekly Report (MMWR). Information is increasingly distributed online.
- Continuing surveillance to evaluate actions taken.
Categories of surveillance
- Special effort to collect data and confirm diagnoses to ensure more complete reports, such as surveys and outbreak investigations
- Encompasses formal and informal communications (such as phone calls or internet searches to seek information).
- Use of routine sources of information such as notifications, laboratory reports, Hospital Episode Statistics (HES).
- Traditional public health surveillance relies on clinicians notifying public health authorities about diseases, which can lead to time delay from the reporting clinician or awaiting diagnostics. By collation and analysis of real-time health data (such as presentations to primary care with a particular symptom e.g. influenza-like illness, rather than clinician notified diagnoses) which may indicate an important public health threat, ‘Syndromic surveillance’ allows for the earlier detection of outbreaks or health threats than would normally occur using traditional notification systems, and therefore earlier targeted action (CDC MMWR 2004).
- These systems facilitate syndromic and routine surveillance; rather than collecting data from all notifying clinicians, only a sample are asked to report data at a regional or national level, on a regular (e.g. weekly/monthly) basis (Giesecke J. Modern Infectious Disease Epidemiology).
- Enhanced surveillance is the collection of data above that collected for routine surveillance, generally at a patient level and often to design or evaluate an intervention or to aid control of more important health hazards. For example, gaining epidemiological as well as microbiological information about MRSA bacteraemia cases in order to target and evaluate the effect of healthcare interventions on the incidence of the infection.
Sources of surveillance data
- Statutory notifications
- Laboratory reports
- Serological surveys
- Sentinel reporting systems Routine Primary and Secondary Care utilisation data, e.g. in England, Hospital Episode Statistics (HES) data
- Death certificates (limited use in industrialised countries because few infectious diseases lead to death)
- Enhanced Surveillance for infections of public health importance to combine epidemiological and microbiological data, e.g. meningococcal disease, TB
- Vaccine use (COVER statistics in England)
- Sickness absence
- Epidemic reports, e.g. respiratory illness outbreaks in care homes are monitored as part of influenza surveillance
- Media reports
- Social media activity
- Animal reservoir and vector studies
- Special systems: See below for examples.
Sources of specific surveillance data (in England)
- Data collated by Respiratory Diseases Department (RDD) of the PHE Centre for Infectious Disease Surveillance and Control (CIDSC)
- RDD also carries out surveillance for novel respiratory viruses such as MERS Coronavirus.
- Primary Care sentinel surveillance weekly returns service of consultation rates for influenza like illness (principle measure of flu activity)
- Other syndromic surveillance data - Emergency Department attendances and NHS 111 (health advice helpline for the general public) cold and flu calls
- Flusurvey.org.uk – online reporting of flu symptoms by the general public
- Medical Officers of Schools Association (MOSA) scheme
- Mortality data from Office for National Statistics (ONS) weekly death reports
- Laboratory reports – Datamart surveillance scheme of positive influenza virology samples from selected laboratories; RCGP sentinel scheme – selected practices post nose and throat swabs from patients presenting with ILI for further characterisation by PCR
- Hospital surveillance through the mandatory UK Severe Influenza Surveillance Scheme (USISS), monitoring critical care admissions of confirmed flu cases
- Also, a USISS voluntary sentinel surveillance scheme for other hospitalised patients with confirmed influenza.
- A statutory notification
- Data collated and analysed by TB section at PHE CIDSC
- Enhanced Tuberculosis Surveillance system - ETS (since 1999)
- Collects data on demographic, clinical and risk factor details, treatment outcome, drug sensitivity and species typing
- Death certificates
- TB incident and outbreak surveillance.
Healthcare Associated Infections (HCAI)
- PHE has surveillance programs covering
- Staphylococcus aureus (Methicillin resistant Staphylococcus aureus (MRSA) and Methicillin sensitive Staphylococcus aureus (MSSA))
- Escherichia coli bacteraemia
- Clostridium difficile infection
- Enhanced surveillance of the first 4 infections is a mandatory requirement in regulated healthcare organisations in England and requires sign-off by the reporting organisation’s CEO (Mandatory Enhanced MRSA, MSSA and E. coli bacteraemia, and C. difficile surveillance, Public Health England 2016), see
- Utilises the HCAI Data Capture System (DCS)
- For MRSA bacteraemias, all positive cases must also have a Post Infection Review (PIR) completed, identifying how the case occurred and what actions could be taken to prevent a reoccurrence
- Also allows capture of voluntary information such as the source/primary focus of MRSA bacteraemia
- Surgical Site Infection (SSI) Surveillance service also requires mandatory reporting of SSI in 4 categories of orthopaedics and voluntary surveillance in 13 categories of surgical procedures
- Patient Administration Systems (PAS) can also contribute data.
STIs and HIV
- Genitourinary medicine clinical activity dataset (GUMCAD) – patient level data on diagnoses made and services provided from GUM and sexual health services
- Chlamydia Testing Activity Dataset (CTAD) – patient level data on testing and diagnoses
- HIV and AIDS new diagnosis database (HANDD) – new HIV diagnoses, AIDS and deaths
- Survey of prevalent HIV infections diagnosed (SOPHID) – people seen for HIV care
- HIV and AIDS Reporting system (HARS) – intended to replace HANDD and SOPHID
- CD4 Surveillance Scheme – Reports of CD4 cell count measures from microbiology laboratories
- National Survey of Sexual Attitudes and Lifestyles (NATSAL) sexual health behaviour survey
- Hepatitis B is a notifiable disease
- Antenatal screening for infectious diseases in pregnancy also provides data on HIV, hepatitis B, and syphilis (as well as Rubella susceptibility)
- Blood, tissue and organ donors screening – NHS Blood Transfusion/PHE surveillance scheme includes hepatitis B and HIV.
Evaluation of a Surveillance System
- similar considerations are also used when establishing a surveillance system
Steps to be taken when evaluating a surveillance system:
1. Describe the public health importance of the health event under surveillance; consider:
- total number of cases, incidence and prevalence
- indices of severity (death rate, case fatality ratio)
- can it be prevented?
2. Describe the surveillance system to be evaluated:
List its objectives, for example:
- detecting and monitoring outbreaks
- detecting and monitoring trends
- setting priorities and allocating resources
- describe the health event under surveillance
- state the case definition
- state the population and choice of denominator
- draw a flow chart of the system
- describe the components and operation of the system
- what information is collected?
- who collects the data?
- time to collect data?
- mode of data transfer and storage?
- who analyses data?
- how is the data disseminated?
- who is it disseminated to?
3. Discuss validity and repeatability of case definitions in various types of surveillance data:
4. Indicate the level of usefulness by describing the actions taken as a result of data collected by the surveillance system. (e.g. useful in prevention and control of adverse event):
- what actions are taken because of the data?
- who uses data to initiate action?
- list anticipated uses of data.
5. Describe cost / resources needed to run surveillance:
- direct: data collection/analysis/interpretation and dissemination.
6. Evaluate the quality of the surveillance system by assessing its attributes
Key attributes of a surveillance system:
Other attributes of a surveillance system:
7. List conclusions and recommendations:
- is the system meeting its objectives?
- address the need to continue/modify surveillance
- NB: no system perfect; trade-offs must always be made.
Ways to improve the system:
- improve awareness of disease by increasing education
- simplify reporting
- frequent feedback
- use multiple sources and methods
- active surveillance
- sentinel surveillance
- complex modelling of surveillance data in combination with data from other sources, for example to aid forecasting
© Sarah Anderson, Gayatri Manikkavasagan 2008, David Roberts and Kiran Attridge 2016