WHO. REPORT OF THE WHO CONSULTATION ON SURVEILLANCE FOR PANDEMIC INFLUENZA - Geneva, Switzerland 10-12 December 2007

GLOBAL INFLUENZA PROGRAMME

REPORT OF THE WHO CONSULTATION ON SURVEILLANCE FOR PANDEMIC INFLUENZA - Geneva, Switzerland 10-12 December 2007

[Full PDF Document is available at this LINK. EDITED. Excertps.]


1. Background

In the 40 years since the 1968?1969 H3N2 influenza pandemic, substantial societal and public health changes have occurred. Many of these changes are expected to affect the emergence, spread and control of the next pandemic. For example, people travel more widely, more frequently and to more remote areas using increasingly faster modes of transportation. This increased global ?connectivity? and interdependency of societal systems will most likely contribute to the accelerated global spread of a new pandemic virus. The rapidity with which a novel virus can circumnavigate the globe was amply demonstrated by the Severe Acute Respiratory Syndrome (SARS) experience of 2003; on the other hand, SARS also demonstrated that it is possible to use modern technology to mount a complex public health response.

Today, the global public health community has new and improved tools to help prepare for and respond to a pandemic that were not available in 1968?69. These include antiviral drugs and nascent technologies to speed the development of pandemic vaccines; improved molecular and genetic techniques to analyse and track the evolution of influenza viruses; and mathematical methods to model the evolution and spread of a pandemic virus, estimate incidence and prevalence and assess the impact of pharmaceutical and nonpharmaceutical measures on disease
transmission and associated morbidity and mortality.

The intervening decade since the first emergence in 1997 of avian influenza A (H5N1) has given the world an unprecedented opportunity to prepare for a possible pandemic. The World Health Organization (WHO) has developed a protocol outlining how the first emergence of a pandemic virus might be rapidly contained before it has spread widely. Most countries have developed pandemic preparedness plans; some national plans include the integration of government and non-government sectors. Many countries are considering how best to implement various public health measures during a pandemic and some have strategic stockpiles of antibiotics, antivirals, human H5N1 influenza vaccine and personal protective equipment.

Importantly, the International Health Regulations (2005) have come into force. They include key provisions for surveillance and notification to WHO of ?events which may constitute a public health emergency of international concern?,1 such as cases of a new subtype of influenza.

Epidemiological surveillance at the global level can help countries anticipate a pandemic?s impact and guide their response as the pandemic evolves. Systems exist at global and national levels to monitor seasonal influenza and detect the emergence of influenza viruses with pandemic potential. These systems, however, have not been designed to cope with a pandemic situation where tens of millions of people are infected over a very short period of time.

The heterogeneous nature of influenza surveillance is an additional challenge.

In some countries, infectious disease and/or influenza surveillance systems barely exist (if at all), while at the other end of the spectrum are countries with multiple and sophisticated systems.

Systems can be laboratory-, diseaseor syndrome-based; some are integrated, some generic and others are influenza-specific. Electronic tools for data collection and transmittal are increasingly being used, although many systems remain relatively set by a variety of factors with limited flexibility.

Against this backdrop, policy-makers, the media and the public have high expectations that public health organizations will respond swiftly, efficiently and effectively once a pandemic begins.

Surveillance data will be essential to inform the public health response at local, national and global levels.


2. The consultation

With a view to addressing the above expectations, WHO convened a technical consultation on surveillance for pandemic influenza from 10 to 12 December 2007. The consultation, attended by 97 experts and key stakeholders from 25 countries, considered what information would be needed during a pandemic, whether existing surveillance systems would be capable of collecting this information, and ways to analyse and disseminate key information during a pandemic.

The programme and list of participants of the consultation can be found in Annexes 1 and 2 respectively. The outcome of the consultation will inform a working group that will develop guidelines for pandemic influenza surveillance at the global level. The consultation was part of a series of interrelated consultations held in 2007 and 2008 to address various aspects of pandemic preparedness, including an update and revision of the 2005 WHO global influenza preparedness plan and a consultation on disease control measures during a pandemic.

The objectives of the consultation were:

1. to determine the core information that should be gathered by WHO to help Member States monitor and manage a pandemic situation at national level;
2. to identify approaches and tools for WHO to receive data from Member States, and to analyse and disseminate the information in a timely manner;
3. to identify essential next steps to produce the guidelines and improve global disease surveillance during a pandemic.


3. Lessons and opportunities identified from relevant experiences

Summary points:
- Advance planning is critical to specify what information should be collected during a pandemic and how data will be managed, analysed and shared at the global level.
- Some information will be used in real time and some will be analysed after the pandemic.
- Information will come from multiple sources, and analysis and interpretation may not be straightforward. Globallevel working groups with relevant expertise should be used to strengthen the
analysis process.

There are few examples of public health responses that compare to the scale and magnitude of a pandemic. Experience from previous pandemics highlight gaps in knowledge and the need to develop more systematic approaches to data collection. The global response to SARS included new approaches to surveillance and strategic partnerships that can be adapted for an influenza pandemic.

Previous influenza pandemics:
Information about previous pandemics can be obtained from international media reports published at the time, public health reports written soon after the event and retrospective analyses.
Using these sources, it is possible to derive information on etiology and virological subtype, transmission, attack rates, case fatality rates, serial intervals, effective reproductive numbers (Ro),
epidemic curves, and who and which community functions were affected. A consistent finding, as illustrated by data from the United Kingdom (UK), was local variation in mortality rates during a pandemic.

Little of this information was known even during the last pandemic in 1968?69. Some of the key information gaps that remain today are an understanding of the dynamics of the disease, how it spreads and the effectiveness of various interventions such as masks, school closures, movement restrictions and border closures.

SARS:
The global public health response to SARS in 2003, including surveillance and reporting of cases, is probably the closest experience to an influenza pandemic in modern times. During the outbreak, WHO provided guidance on surveillance for the new disease and collected and disseminated global surveillance data. A minimum global dataset was specified to support monitoring, risk assessment, decisionmaking and evaluation needs.

Practical aspects of surveillance included provision of a data dictionary and clear instructions about the frequency and methods of reporting. WHO and some countries found it helpful to identify separate operational and decision-support teams. The former (called Team A) was immersed in the day-to-day operations to collate, analyse and track the epidemic. The latter (called Team B) functioned at ?arm?s length? to review and reflect on surveillance data and other information, pose questions to facilitate evaluation and lend support for decision-making.

The rapidly evolving situation, complicated by multiple stakeholders with varying information needs, required open and transparent access to data. However, this raised complex ethical and data ownership issues for both WHO (as custodian of the global data) and countries that were not easily resolved. Development of a well-nuanced data sharing plan in advance of a pandemic is strongly advised to specify what level of data will be shared and how it can be used.

During the SARS outbreak, WHO formed several ?virtual? working groups including the SARS Epidemiology Working Group. This Group of laboratory experts met regularly by teleconference to
exchange information and data in real time, review guidance documents drafted by WHO and national public health agencies and serve as a problem solving resource for WHO.

Their teleconferences and meeting minutes were also shared with the Clinical Working Group. The Epidemiology Working Group also prepared a consensus document on the epidemiology of SARS.

Communication within the Group was facilitated by a secure web site; in the future collaborative electronic work spaces may serve as an additional resource to share preliminary information, assist in rumour alert and verification and host on-line discussions.


4. Expectations and perspectives of countries and the media

Summary points:
Pandemic influenza surveillance at the global level must consider:

- what data to collect in order to assess and monitor key parameters over time;
- different ways to collect data such as special studies, routine surveillance and the use of sentinel sites;
- processes to ensure data quality and reliability;
- how to provide the best information available in a timely and easily understood format;

- how frequently information should be updated; and
- changing information needs over the course of the pandemic.

Member States and the media are among the key partners and stakeholders whose perspectives can help inform the development of a plan for global pandemic influenza surveillance.

Member Sates:
Irrespective of their country?s level of resources and development, consultation participants consistently voiced high expectations about what information WHO should provide at the global level during a pandemic (see annexes 3 and 4). Participants indicated that the following information would help them address national concerns such as morbidity, mortality and social disruption as well as anticipate needs in health care and other sectors in the allocation of scarce resources:

- epidemiological, clinical and virological parameters of the pandemic virus;
- pharmaceutical (antivirals and vaccines) and nonpharmaceutical interventions;
- details on the global spread of the disease;
- surveillance and case definitions; and
- triggers to start and stop interventions.

It was agreed that priorities related to data collection and analysis will change during the pandemic and this needs to be taken into consideration when planning for pandemic surveillance at national and global levels. While the pandemic is under way, information will be urgently needed for immediate action, e.g. on the severity of the disease and how many additional treatment centres are needed. Other information collected during the pandemic could be analysed after termination of the pandemic, e.g. the number of patients who received a complete course of
antiviral therapy. These changing information needs (Figure 1) will require different types of data collection approaches, including detailed case investigations or special studies linked to outbreaks early in the pandemic, and traditional disease surveillance for ?core? data during the pandemic. The discussion raised questions about how information would be collected, shared with WHO and aggregated at the global level. It was also noted that the quality of data was likely to vary and such limitations needed to be acknowledged.

To cope with the overwhelming amount of information that a pandemic would generate, participants stressed the importance of establishing a process whereby data could be carefully sifted and synthesized to best inform and guide public health action.

Practical suggestions for WHO included daily or twice daily briefings (in view of different global time zones) and pre-notification to Member States before official announcements to other Member States and/or public. If possible. Information sharing should be open and transparent. Challenges for pandemic surveillance in resource-poor countries are particularly acute and include a low
level of commitment by policy-makers and partners; little capacity for collection, collation, analysis and timely reporting of surveillance data; inadequate laboratory resources, notably basic equipment, supplies, facilities and trained staff; lack of systems for seasonal influenza surveillance; and inadequate logistics support and training.

The media:
An influenza pandemic will generate considerable demand for accurate and timely information from many parties including WHO, countries, politicians, the media, scientists, public health officials, modellers, the business community and the general public. Although there will be no shortage of information, much of it will be confusing and at times inaccurate depending on the source and quality of the data, as well as the expertise and knowledge of the spokesperson. The media will certainly access global, regional and national surveillance data. However, other types of information will be sought such as the impact of the pandemic on the economy, schools, transit, travel and health care, and the availability of antivirals, personal protective equipment and other supplies.

Important work can be undertaken in advance of the pandemic such as simulation exercises. In addition, all stakeholders should be advised to expect uncertainty and incomplete information, especially during the early stages of a pandemic. WHO reports should serve as a credible, stabilizing source of information and guidance.


5. Review of existing surveillance systems: options for adaptation during a pandemic

Summary points:

- Surveillance systems span a spectrum of complexity, methods and resources. Modifying existing systems will be necessary to meet the demands of a pandemic.
- Pandemic surveillance will not be done in a uniform way and global pandemic surveillance will integrate data from heterogeneous national, regional and global systems.
- The challenge is to define both the minimum data to be shared at global level on a routine basis and the more detailed data to be collected by a subset of countries.
- Advance planning, practical guidance and simple systems that accommodate varying levels of infrastructure and capacity can maximize the number of countries participating in a global
surveillance system.


5.1 Examples of global-level systems

Two WHO systems collect influenza surveillance data at the global level:

- the Global Influenza Surveillance Network (GISN) with FluNet as data reporting system and an early warning system.

The GISN is a network of laboratories involved in virological surveillance. It plays a significant role in supporting WHO?s recommendations on influenza vaccine composition. The GISN comprises 118 National Influenza Centres (NICs) in 89 countries, four WHO Collaborating Centres (CC) for Reference and Research on Influenza, three key national reference laboratories involved in vaccine virus selection and development, and a WHO Collaborating Centre for Studies on the Ecology of Influenza in Animals (FiG. 2). NICs collect specimens in their country and perform primary virus
isolation and preliminary antigenic characterization. They ship newly isolated strains to WHO CCs for more sophisticated antigenic and genetic analysis, the results of which form the basis of WHO recommendations on the composition of influenza vaccine for the northern and southern hemispheres each year. In addition, the GISN updates seasonal influenza diagnostic reagents and monitors antiviral susceptibility. As at 2008, only minimal epidemiological information was reported to the GISN.

Created in 1995 with data from 78 countries, FluNet is a web-based interactive data reporting, query and mapping system for support and coordination of national and global influenza surveillance. Currently, 83 NICs contribute to FluNet. It facilitates real-time monitoring of influenza activity around the world and includes information about influenza-like illness (ILI) activity and the number of isolates and specimens processed.

The GISN also has an important role in monitoring and assessing the potential risk that influenza A (H5N1) and other viruses pose for a pandemic. GISN laboratories perform detailed molecular and antigenic analyses of H5N1 isolates, identify H5N1 human viruses suitable for vaccine development, develop and update diagnostic reagents and protocols for human infections, provide confirmatory diagnostic services, assist in outbreak investigations and monitor antiviral susceptibility.

The second WHO global surveillance system is for early warnings.

This system screens daily reports from the media using tools as Global Public Health Intelligence Network (GPHIN) and reports from formal and informal networks on events of potential global
public health importance such as H5N1 infections or other influenza viruses with pandemic potential. Efforts to verify unsubstantiated reports and rumours can include an epidemiological investigation if required. WHO?s Global Alert and Response Network (GOARN), comprising 114 international organizations and partners, can provide expertise to carry out epidemiological investigations and support for rapid containment of an emerging pandemic.

WHO has long-established systems for virological surveillance of seasonal and novel influenza; however, it does not have a comparable system to collect epidemiological and virological data during a pandemic. The early warning system will be instrumental at the beginning of a pandemic but may well not be adequate for sustained surveillance until the termination of a pandemic.


5.2 Examples of regional-level systems

At the regional level, two interesting examples illustrate the challenges of integrating national sources of data. These are the European Influenza Surveillance Scheme (EISS), a surveillance system with both virological and epidemiological components, and the Integrated Disease Surveillance (IDS) system, an early warning system focused on rapid detection of epidemic diseases in Africa.

EISS comprises 35 member countries and a population base of 498 million persons. Both virological and epidemiological data are collected and reported using a web-based platform. Virological data are supplied by national reference laboratories that receive specimens from general practitioners (GPs), hospitals and other sources. EISS hosts the Community Network of Reference Laboratories for Human Influenza in Europe (CNRL). Epidemiological data, including age-specific community data, are collected by a network of 25 750 sentinel physicians who also supply
community-based specimens for analysis. Each week data are processed over three days and published in an electronic bulletin. 2

The system requires six to seven persons to operate but could function with fewer. It offers several advantages including an enhanced database and harmonized key surveillance indicators such as age groups and laboratory activities despite differing health systems across Europe. Areas for improvement are further harmonization (e.g. a common definition for reporting of ILI), inclusion of cases that do not seek care from GPs and a mechanism to share regional data with modellers. EISS will operate out of the European Centre for Disease Prevention and Control (ECDC) as from September 2008 and will serve as the base for pandemic surveillance in Europe.

IDS is a regional strategy to provide timely data for decision-making and public health interventions to control 19 priority communicable diseases. Recently it has evolved to include events specified under the IHR (2005). The system relies on the use of simple case definitions, collection of minimal information, integration of reporting forms and provision of feedback. Reporting frequency ranges from realtime case-based/line listing information to weekly, monthly and quarterly aggregated data. Casebased information includes name, age, sex, address, location (urban/rural), date of onset/admission, laboratory result, diagnosis and outcome. Aggregated data (i.e. number of cases and deaths) are age-stratified as < 5 years and ≥ 5 years. Although IDS
has helped improve epidemic detection and response capacities, progress has been slow; it required 10 years for 43 countries to implement the system. Influenza-specific surveillance is not carried out in many countries in the African region and there are only nine NICs in eight countries in this region. Enhanced surveillance for H5N1 within IDS includes ?zero reporting? carried out
currently by 23 countries, coordination with animal surveillance, and rumour/event verification. Some countries in the region are planning to implement severe acute respiratory illness (SARI)/ILI surveillance through bilateral cooperative agreements with the US Centers for Disease Control and Prevention (CDC).

Pandemic influenza surveillance in the African region could theoretically follow the IDS model, i.e. collection of case-based data for the first 100?200 cases per country (as is done for meningococcal disease) followed by a move to reporting aggregated data (cases/deaths) and the number of countries/provinces/districts affected on a daily or weekly basis.

Nonetheless, many challenges can be anticipated for IDS-based pandemic influenza surveillance, for instance the timeliness of reporting (related in part to the time required for information to
flow from the Ministry of Health to WHO Country and Regional Offices and then Headquarters), incomplete information, data quality issues, communication failures and lack of resources.

During the discussion it was noted that the WHO Regional Office for the Americas and CDC have developed a generic protocol for influenza surveillance that integrates epidemiological and laboratory components. The protocol outlines a two-pronged approach: a sentinel surveillance system for ILI in outpatients and SARI and SARIrelated mortality in hospital patients coupled with an enhanced nationwide notifiable disease surveillance system for unusual occurrences of acute respiratory infections. PAHO plans to implement the protocol on a pilot basis in 2008. The difficulty of introducing standardized methodology in areas with pre-existing systems was noted.


5.3 Examples of country-level systems

To participate in global surveillance for pandemic influenza, countries need a surveillance system that is able to gather the required data. Three types of country-level surveillance systems
were reviewed:

- Long-standing seasonal influenza surveillance systems with both virological and disease components. During a pandemic, such systems can serve as a foundation upon which new components can be added or existing components restructured.
- Sentinel surveillance or early warning systems in community and hospital settings to detect human cases of infection with H5N1 or other novel influenza virus strains that may signal the start of a pandemic.
- Surveillance systems for diseases such as polio that could be used during a pandemic with some adaptation.

Countries without pre-existing surveillance systems when a pandemic starts may be able to implement an ad hoc system modelled on approaches used during ?chaotic situations.?

Long-standing influenza surveillance systems:
Canada, France, Japan and the UK described their current seasonal influenza programmes. All four countries rely on a multi-component strategy that typically includes (i) assessment of morbidity through sentinel physicians and hospital networks; (ii) virological surveillance; and (iii) mortality data (with the exception of Canada which is piloting population-based severity/mortality surveillance during 2007?08). Data are reported as influenza rates per population. Other country-specific components include school absenteeism (Japan), ?cold/flu/fever? calls to a 24/7 nurse-led clinical helpline (UK), hospital bed activity (UK), provincial-level qualitative influenza ?activity? (Canada), outbreak reporting (UK) and surveillance for hospitalized severe cases (Canada
and France [paediatric only]).

Canada, Japan and the UK described plans to adapt their seasonal systems to a pandemic situation. During the pre-pandemic period, or at the initial occurrence of pandemic influenza, each of the three countries plans to collect more detailed case- or clusterbased data such as clinical characteristics and course, transmission patterns, response to antivirals, virus characterization and antiviral susceptibility testing. The Netherlands plans to undertake a multidisciplinary systematic follow-up of early cases and their contacts using a unified database, and to conduct seroprevalence surveys in existing serodatabases.

The ECDC is exploring the possibility of using the UK?s enhanced web-based reporting system for detailed data collection throughout Europe. This enhanced data collection would continue for a
limited period of time (e.g. covering the first few hundred cases as in the UK).

Over the course of the pandemic Japan plans to restructure its seasonal ILI surveillance and continue its virtual clinicians? network, pneumonia, eventbased, mortality, virological and vaccine adverse event surveillance. In the UK, pandemic influenza data will be reported using a recently developed web-based informatics portal. In addition to the usual seasonal surveillance elements, new schemes will be added such as data on persons receiving antiviral treatment at dedicated distribution facilities, antimicrobial susceptibility testing, molecular diagnostics testing and real-time modelling. Work is ongoing to increase capacity for both real-time modelling and more frequent reporting by influenza networks. Daily reporting of modelled and observed data (e.g. rates of ILI/100 000 population) is planned and will be folded into a larger daily report for government ministers that provides information about hospital bed availability and local resilience and
events (e.g. school closures, ambulance services, etc.).

Canada plans to reduce and streamline reporting data during the pandemic/mitigation phases and not report actual numbers of cases. Instead surveillance will focus on core monitoring activities including virological surveillance, provincial level influenza activity assessments, some form of all-cause mortality monitoring (options for piloting are under way), antiviral adverse events and vaccine safety. Laboratory testing will be prioritized using population-based sampling and/or targeted investigations. Challenges include the unknown impact that anticipated changes in health-care service delivery (e.g. centralized ?flu assessment clinics?) will have on collection of surveillance data and the variations in data collection/management systems across public health institutions and laboratories at the provincial and territorial levels.

During the plenary discussion, the United States of America indicated that it planned to move from reporting ?case counts? during a pandemic to reporting rates of hospitalization for respiratory illness as an example of one of several indicators that are being developed.

Sentinel systems developed since H5N1:
Cambodia and Turkey have reported both human and animal cases of H5N1 influenza. They described how surveillance is currently undertaken amidst the continued threat of H5N1. Disease surveillance in Cambodia is both passive and active and comprises several components. Passive surveillance for 12 diseases is undertaken at all government health facilities using a syndromic approach, although an evaluation in 2005 documented that completeness of reporting and regular, timely analyses were lacking. Two additional passive approaches are village-based event
surveillance in humans for which 29 ,000 village volunteers have been trained and hotlines established for human disease and elevated poultry deaths. Active sentinel surveillance occurs at three levels: outpatient, inpatient and community. However, participation in these schemes is limited: four outpatient clinics undertake active ILI surveillance, two hospitals survey for acute lower respiratory illness (ALRI) and 25 convenience sample villages in one province (~20,000 population) are visited weekly to detect febrile illness and collect specimens for testing. The Pasteur Institute of Cambodia has received funding from CDC to study the epidemiology of seasonal influenza including an assessment of seasonality, incidence and hospitalization rates, risk factors for
severe outcomes and an estimation of influenza-related deaths. Although an existing network of active sentinel sites, coupled with in-country virological capacity, offers the potential for surveillance during a pandemic, many challenges can be anticipated related to the overwhelming burden on health-care systems and the uncertainty of adequate staffing.

Since the appearance of H5N1 human cases in Turkey in early 2006, the country has established surveillance for upper respiratory infection symptoms and ?flu like? symptoms in 10 health centres in each of its 14 provinces. ?Special vigilance? occurs among schoolchildren and other risk groups. Specimens are obtained and tested for influenza A and B viruses. The current system could be used during a pandemic; during interpandemic periods, consideration is being given to integrating other emerging infections.

(...)
-
1 International health regulations (2005), 2nd ed. Geneva, World Health Organization, 2008.
2 http://www.eiss.org/cgifiles/bulleti...gi?season=2008 accessed on 19 December 2008
-
-----