A €1.2 million European project, which concludes on 28th March, will help policymakers develop reliable plans for responding to flu epidemics, including pandemics. The world's first extensive cost-effectiveness analysis of epidemic responses has created models which will help governments avoid costly and ineffectual policies.

Annual epidemics are estimated to cause 3-5 million cases of severe illness and 250-500,000 deaths worldwide (WHO, fact sheet 211; March 2014). When an epidemic hits, the main factor limiting our response is cost. Governments must take tough decisions on how to allocate limited resources.

Data on epidemic responses across Europe has been collected using different approaches and success measures, making past responses very difficult to compare. This makes developing reliable preparedness plans extremely difficult.

This has resulted in a history of poor decisions, such as rolling out vaccination programmes through ad hoc vaccination centres without communicating them properly; or stockpiling anti-virals for mild outbreaks at great cost.

To address this problem and ensure more informed decisions are taken in future, the EC funded project, Fluresp, has developed specific cost-effectiveness models for flu responses.

This has enabled them to reliably compare potential public health responses and develop models that would help governments to select interventions according to the level of threat.

Helping epidemic planners make informed decisions

The project had three key outcomes for policymakers: a methodological approach to compare different interventions, a prioritisation tool, and recommendations to governments.

The proposed methodological approach allows healthcare planners to analyse large datasets on a wide range of public health responses. The key to the project was assessing all potential responses using the same measure: 'Cost per Success'. This defines all responses in terms of the financial cost to achieve a defined reduction in mortality (death) or morbidity (illness).

The world's first free public health response prioritisation tool allows governments to input their own data and compare different scenarios and costs. For example they could compare strategies for vaccinating a certain section of the population vs the entire population. This kind of multi-criteria analysis would have previously been extremely difficult and time-consuming, due to the quantity of data and the complexity of the analyses techniques.

The project also made overarching guidelines for governments based on the rigorous analysis of vast amounts of data. These include:

  • Vaccination programmes targeting the general population appear more cost-effective than targeting only at-risk people or health professionals.
  • Providing curative anti-virals to those with flu appears more cost-effective than distribution of prophylactics (intended to prevent flu) during a pandemic.
  • Using existing vaccination centres appears more cost-effective than creating ad hoc centres during a pandemic.
  • Screening at airport arrivals and promoting hand washing and mask wearing do not appear to be cost effective as standalone measures.
  • To improve future planning, governments must improve data collection (including on morbidity and mortality, costs of intervention and communication, and societal costs such as sick leave).
  • Healthcare planners must use a meaningful measure to assess responses to allow comparison across situations and countries. Cost per Success ratios are recommended, rather than complex metrics such as costs per Quality Adjusted Life Years (QALY) or per Disability Adjusted Life Years (DALY), which are the subject of an international methodological controversy. These are based on a number of poorly validated theoretical assumptions which can lead to dramatically divergent results using the same dataset, as demonstrated by another EU project (ECHOUTCOME).

Project Lead, Dr Ariel Beresniak, Paris-Descartes University, says: "Costs and ethics are the principle considerations in public health. Without robust information and a systematic approach to compare costs and effectiveness of different measures or combination of measures, governments cannot make the right choices. The Fluresp project provides models based on rigorous data analysis for governments to develop a tailored response plan, so when an outbreak or pandemic hits, they are ready to make an informed response."

This approach will now be used to encourage and assist governments to better manage seasonal flu but also to prepare for pandemics. This will help them to deal with outbreaks effectively without unnecessary expense to the taxpayer.

Beresniak concludes: "We aim to work with health policy makers to allow them to benefit from our findings and help them develop an appropriate approach for assessing the most efficient strategies for different epidemic scenarios".

Background to the research and supporting quotes

The project brought together experts in public health, influenza, health economics and computer sciences from ten European countries, who analysed large datasets on past responses.

The project looked at six levels of outbreak severity - from seasonal flu to a severe pandemic.

It examined 18 common public health responses including: sanitation (eg hand washing campaigns), controlling infection spread (eg closing schools), immunisation of different groups, antiviral distribution, and care for the sick.

Finally it looked at how effectively each reduced mortality and morbidity.

Each scenario was then given a 'Cost per Success' ratio. This provides a robust way to compare different scenarios in different situations and allows the development of reliable models.

The project was tested in four pilot European target countries (France, Italy, Poland and Romania).

Dr Pasi Penttinen, Head of disease programme, Influenza and other respiratory viruses, European Centre for Disease Prevention and Control: "We live under the constant threat of a pandemic and we need robust tools in place for proper preparedness. We have not yet appreciated the possibilities which large datasets can offer in healthcare planning, and the Fluresp project is the first step towards effectively harnessing them. Furthermore, its success may provide a systematic approach for developing new evidences in public health decision making."

Dr Sylvie Briand, Pandemic and Epidemic Disease department (PED) Director at the World Health Organisation, and collaborating partner on the project: "Governments must have a regularly updated plan for different scenarios that they can implement when a pandemic looms. Whilst they can't plan for everything, it is very important to define response scenarios in advance so they have a framework to work from."

Dr Jean-Paul Kress, CEO, Sanofi Pasteur MSD: "The Fluresp approach confirms that vaccination programmes against human influenza are effective not only in reducing flu mortality and morbidity, but are also cost effective when targeting the general population".