Health and cost benefits of monitoring infectious diseases using novel statistical methods.
Submitting InstitutionUniversity of Strathclyde
Unit of AssessmentMathematical Sciences
Summary Impact TypeTechnological
Research Subject Area(s)
Mathematical Sciences: Statistics
Medical and Health Sciences: Public Health and Health Services
Summary of the impact
Research on novel statistical methods for disease surveillance and
influenza vaccine effectiveness has led to the development of a suite of
automatic systems for detecting outbreaks of infectious diseases at Health
Protection Scotland (HPS). This work has improved the public health
response and helped to reduce costs in Scotland and also in the wider UK
and EU by providing real-time early warning of disease outbreaks and
timely estimates of the effectiveness of the influenza vaccine. This
research, commissioned by the Scottish Government, through HPS, and also
the UK National Institute for Health Research (NIHR) and the European
Centres for Disease Control (ECDC), but used in a wider context by many
others, formed the basis for the HPS response to the H1N1 Influenza
Pandemic and monitoring of the effects of Influenza Vaccines.
Context: Health Protection Scotland is the NHS agency concerned
with protecting the population of Scotland from the effects of infectious
diseases and outbreaks that are a major public health problem. It is also
responsible for assessing the impact of vaccination programmes on health.
The seasonal influenza vaccination programme is estimated to cost in
excess of £15 million annually (the cost was £14M in 2008) and it is
necessary to establish that the vaccine used is effective. Diminishing
resources and a large number of diseases means that little time can be
allocated to any one disease. Consequently, research into the use of
sophisticated statistical techniques to process automatically large
volumes of data was carried out in close collaboration with consultants
and epidemiologists at HPS.
Key Findings: Statistical models based upon Poisson regression
techniques were developed for statistical exception reporting . These
models were extensions of research critically reviewed in  and led to
the development of a syndromic surveillance system for Scotland based upon
daily data provided by NHS24, the out of hours telephone service for
medical advice . Methods for modelling these data were developed and
existing algorithms for signalling an exception were modified and compared
using real and simulated data. The initial research was carried out in
2006- 07, based upon earlier work in the PhD thesis of G. McCabe. Further
research has considered both spatial and temporal effects , which
permits HPS to provide epidemiological guidance at sub national (Health
Board) level and thus increase the specificity of the HPS response. The
research demonstrated that these surveillance systems detect outbreaks
early and are important in monitoring disease activity.
Competitive research grants funded by ECDC (EpiConcept), the Chief
Scientist's Office (CSO), NIHR, and HPS have permitted the utilisation of
HPS data to develop statistical methods for the estimation of vaccine
effect, particularly for seasonal and pandemic influenza, and the Human
Papilloma Virus (HPV) vaccine. Information on cohorts of patients was
assembled and linked with other health related details available in
Scotland, such as laboratory and hospital data and cause of death
statements. Statistical models for the estimation of vaccine effectiveness
were produced and compared. Over and above the novelty of the
epidemiological research, the novel statistical aspects of this research
are in the application of state-of-the-art statistical models and methods
for the control of confounding, such as instrumental variables, propensity
scores and negative confounders imbedded within time dependent Cox models,
with a correlated error structure to account for the clustering of
patients within GP practices in the cohorts. The research provided timely
estimates of the effect of the influenza vaccine in Scotland, as well as
an understanding of the effect of the influence of the methodology. [4, 5]
Key Researchers: Research was, carried out by the Epidemiology and
Population Modelling group at Strathclyde:
Chris Robertson (Professor, 2002- ) supported by D Greenhalgh (Reader,
1990- ) and G Gettinby (Professor, 1978 - ), undertook the initial
research into the surveillance systems.
C Robertson undertook the research on statistical methods for estimating
vaccine effectiveness in association with Dr J McMenamin (Health
Protection Scotland), Dr R Pebody and Dr N Andrews at Public Health
England, Dr D Fleming at Royal College of General Practitioners,
Birmingham, Dr C Simpson, Dr N Lone and Prof A Sheik (Edinburgh
University), and Prof L Ritchie (Aberdeen University). The key researchers
in this project were Prof C Robertson, Dr McMenamin and Dr Lone who
provided the medical background, and Dr Simpson who was PI on 2 NIHR
grants and is an expert on data systems.
References to the research
References 1, 3 and 5 best exemplify the quality of the body of
research. References1, 4 and 5 are part of the REF2014 submission
1. Kavanagh. K., Robertson, C., Murdoch, H., Crooks, G., McMenamin, J.
Syndromic surveillance of influenza-like illness in Scotland during the
Influenza A H1N1v pandemic and beyond. Journal of the Royal Statistical
Society Series A: Statistics in Society, 2012, 175 (4) pp.939-958
2. Unkel, S., Farrington, C., Garthwaite, P. H., Robertson, C., &
Andrews, N. Statistical methods for the prospective detection of
infectious disease outbreaks: a review. Journal of the Royal Statistical
Society: Series A (Statistics in Society), 2012, 175(1), 49-82.
4. Kavanagh K., Robertson C., McMenamin J. Assessment of the Variability
in Influenza A(H1N1) Vaccine Effectiveness Estimates Dependent on Outcome
and Methodological Approach. PLoS ONE, 2011, 6(12): e28743.
5. Simpson C.R., Ritchie L.D., Robertson C., Sheikh A., McMenamin J.
Effectiveness of H1N1 vaccine for the prevention of pandemic influenza in
Scotland, UK: a retrospective observational cohort study. Lancet Infect.
Dis., 2012, Sept, 12(9),696-702. Epub 2012 Jun 26.
Other evidence for quality of research:
18 grants with a total value of over £2.4M were awarded by HPS, HPA, ECDC,
CSO, NIHR, SEHD, EPSRC since 2003. Examples are
1. Simpson C.R., Lone N., Ritchie L.D., Robertson C., Sheikh A.,
McMenamin J. Early estimation of pandemic influenza Antiviral and Vaccine
Effectiveness (EAVE) - use of a unique community and laboratory national
linked dataset. 2012-15, NIHR, £215,000.
2. Simpson C.R., Lone N., Ritchie L.D., Robertson C., Sheikh A.,
McMenamin J. Seasonal Influenza Vaccine Effectiveness (SIVE), 2010-12,
3. Cruickshank M., Campbell C., Choi Y., Cubie H., Cushieri K., Donaghy
M., Imrie J., Robertson C., Sullivan F., Weller D. The Scottish Cervical
Cancer Prevention Programme: Assessing And Modelling The Impact Of HPV
16/18 Immunisation, 2010-15, CSO, £450,000.
Details of the impact
Process from research to impact:
C Robertson has a joint appointment with Health Protection Scotland and
this close working relationship between the researchers based in
Mathematics and Statistics at Strathclyde and those at Health Protection
Scotland enables the timely identification and investigation of important
health research areas. Implementation of the research outputs comes
through the contribution of university staff to core HPS strategy and
planning groups, such as HPV Surveillance in Scotland, the MRSA Screening
Programme Board, Pandemic Influenza, and the Olympic and Commonwealth
Games Surveillance Group (Source1). Wider dissemination takes place
through reports to Government and HPS study reports (Source 2).
Consequently, the results of Strathclyde's research are often transferred
to and implemented by HPS before publication in journals.
Preparatory research at the university led to the establishment of a
national surveillance system for the weekly "organism report" around the
time of the G8 summit in Gleneagles in July 2006. At this time, HPS had no
facility for processing large amounts of varied, daily produced data
within a unified surveillance system and the university research led
directly to the establishment of such systems. A review of the
surveillance strategy by HPS in the aftermath of the G8 conference
concluded that statistical methods for the automatic scanning of a large
amount of daily data were crucial to its operational effectiveness. In the
period 2008-13, at least 5 separate data analysis systems have been
developed and tested, based on adoption of Strathclyde's research.
Types of Impact
Monitoring Influenza Pandemic:
During the H1N1v Influenza Pandemic, HPS systems based on Strathclyde's
research were instrumental in successfully monitoring the evolution of the
pandemic, estimation of the transmission dynamics of the virus, monitoring
the numbers of people dying in relation to those expected to succumb; and
development of a national system for surveillance of GP consultations,
including the use of NHS24. HPS placed, and continues to place, a huge
reliance on these systems so that it can function effectively (Source 3).
Outputs derived from the university research, were regularly used by
government, health officials, and health protection agencies. Reports were
prepared for the both the UK Government's Scientific Advisory Group for
Emergencies (SAGE) and the Scientific Pandemic Influenza Subgroup on
Modelling (SPI-M),which monitored the course of the pandemic(Source 4).
Without the university research, HPS would not have had advanced syndromic
surveillance systems, nor would it have had as sophisticated a response to
pandemic influenza surveillance (Source 5). The benefits to HPS, and
consequently the general population, have been apparent throughout the
period 2008-13, with greatest impact in the period April 2009-June 2010.
The Clinical Director at HPS during this period (Source 1) states "This
research had a great impact at HPS in helping to mitigate the increased
workload and pressures on the organisation associated with managing a
national outbreak and reporting, each day, on a wide range of issues".
Vaccination is the main way of protecting the population against some
infectious diseases and HPS monitors the impact of the national
vaccination programmes. The university's research on vaccine effectiveness
is carried out in conjunction with colleagues at HPS (Dr Jim McMenamin) ,
Edinburgh University (Dr Colin Simpson, Dr Nazir Lone, Prof Aziz Shiekh)
and Aberdeen University (Prof Lewis Ritchie). A successful HTA grant
during the H1N1v pandemic yielded one of the first publications on the
effectiveness of the pandemic vaccine, demonstrating that the vaccination
programme had prevented infection, hospitalisations and death. This group
has also worked on Seasonal Influenza Vaccine Effectiveness (HTA grant
2011-12) and preparatory work to develop data systems for pandemic
preparedness (EAVE, HTA 2013-2015). As a result of this research HPS has
published mid-season and end-of-season estimates of the effect of the
influenza vaccines since 2009. These results are communicated to (a)
Scottish Government through reports and meetings, (b) Public Health
England and, hence, the UK Government, and (c) the I-Move network through
meetings and reports, where the results have assisted in guiding public
vaccination policy at national and European level.
Improved effectiveness of Health Protection Scotland:
Strathclyde's research working in partnership with HPS, has permitted the
demonstration of a positive health impact of the influenza vaccination
programme on the health of the UK population, with the added benefits of
improving the effectiveness of HPS in carrying out its responsibilities.
The surveillance systems provide HPS with the ability to monitor disease
levels in relation to expected levels and to detect outbreaks of
infectious diseases in a timely fashion. When combined with appropriate
interventions, a reduction in the magnitude of the outbreaks can be
achieved, with a reduction in the time in which they can be controlled.
Furthermore, there is a workforce impact as epidemiologists are freed from
routine surveillance, leading to a more efficient deployment of staff
within HPS. A conservative estimate of the savings is one staff per year
since 2009 at £50,000 per annum (Source 6).
Reach and Significance: The results of the research on influenza
vaccine effectiveness have been presented annually at a European network
on vaccine effect, I-Move, which reports to ECDC and WHO (Source 7). This
regular reporting has influenced other national programmes demonstrating a
global reach. Early estimates of the effect of the current seasonal flu
vaccine are crucial for the planning of the next year's flu vaccine. As a
result of the university research, HPS, in conjunction with Public Health
England (PHE), are better able to provide timely estimates for the UK
(Source 8) placing HPS/Strathclyde University amongst the leaders in the
early estimation of influenza vaccine effect in Europe.
A consultant epidemiologist at HPS has confirmed "This research is
carried out at Strathclyde University but has a national, European and
international impact. Since 2009 we have developed a system of linked
datasets which, with complex statistical modelling, can be used to provide
timely estimates of the effectiveness of the seasonal influenza vaccine.
The results of this research have an impact at national level, where they
are communicated to Scottish Government, through reports and meetings and
to Public Health England and thence the UK Department of Health.
Furthermore, the impact is wider through the contribution of HPS in a
European network of researchers and public health officials from European
Centre for Disease Control. This data allows in-season estimation of the
severity of influenza as it impacts the public health and the success or
otherwise of public health measures to limit its impact. The network
regularly contributes to the World Health Organisation influenza vaccine
strain selection meeting where the composition of the seasonal influenza
vaccine for the next season is decided" (Source 6).
Sources to corroborate the impact
- Statement from Clinical Director, HPS will corroborate that the
university research is crucial to workings of HPS and that university
researchers are a vital part of HPS working groups.
will corroborate that university research is published in reports by
Health Protection Scotland and which are then submitted to Scottish
and Elliott, A. J., N. Singh, P. Loveridge, S. Harcourt, S. Smith, R.
Pnaiser, K. Kavanagh et al. "Syndromic surveillance to assess the
potential public health impact of the Icelandic volcanic ash plume
across the United Kingdom, April 2010." Eurosurveillance 15, no. 23
(2010).will support the claim that the results of the university
research continue to have an impact
- HPS Report to SAGE Committee. HPS_School_Closures_Scotland 23 July
2009 and HPS Report to SAGE Committee. SAGE paper impact of schools in
Scotland Opening after summer break Sept 2009 will support the claim
that University research contributed to SAGE committee during pandemic
will support the claim that the Scottish response to H1N1v was
- Statement from Consultant Epidemiologist at HPS will affirm that
university research is crucial for influenza surveillance in Scotland
and improves efficiency at HPS
- Valenciano M, Ciancio BC. I-MOVE: a European network to measure the
effectiveness of influenza vaccines. Euro Surveill.
2012;17(39):pii=20281. Available online:
demonstrates influenza surveillance in Scotland and vaccine
effectiveness estimates from UK are timely and are used to inform the
development of next season's vaccine.
- Pebody, Richard, P. Hardelid, D. Fleming, J. McMenamin, N. Andrews, C
Robertson, D. Thomas et al. "Effectiveness of seasonal 2010/11 and
pandemic influenza A (H1N1) 2009 vaccines in preventing influenza
infection in the United Kingdom: mid-season analysis 2010/11." Euro
Surveill 16, no. 6 (2011): 19791. Supports the claim that university
research is crucial for influenza surveillance in Scotland and that this
research is timely and is used to inform the development of next