Information and Information Systems in a Pandemic Influenza Emergency

• Introduction
• The World Has Changed
• What Will Happen in a Pandemic
• Coded States
• Information Needs
• Current State
• What Will Be Required
• Processes
• Implementation Considerations
• Conclusions
• References and Documents Reviewed

Introduction
Influenza is a highly contagious viral disease of the respiratory tract. Almost everyone will have had "a dose of the flu" at some time in their lives which is not life threatening and is commonly caused by a strain of influenza of low virulence or another upper respiratory tract pathogen. As a result, in developed countries the general public has little understanding of the potential seriousness of influenza in terms of morbidity and mortality. The virus has considerable ability to spread rapidly through populations. In New Zealand seasonal spikes of the disease typically occur in the winter months.

The influenza virus is remarkable for its ability to change through mutation or genetic re-assortment between individual viruses infecting the same cell. These processes are collectively referred to as "antigenic shift".

The virulence of any given strain of influenza is largely determined by two principal surface proteins or antigens. Haemagglutinin (H) is significant in the attachment of the virus to the respiratory tract and thus the virulence of the strain. Neuraminidase (N) is considered to be important for the propagation of the virus and thus the transmission of the infection within populations. There are 16 different H antigens and nine different N antigens. Influenza in humans has historically been caused by combinations of H1, H2, H3 along with N1 and N2.

Influenza is believed to be transmitted between humans primarily by airborne particles created when infected individuals cough or sneeze. In addition, the virus can survive on hard surfaces for up to 48 hours and on cloth or paper tissue for up to 12 hours. Groups living in close proximity, such as children in schools or aged people in rest homes can amplify transmission of the virus in a population.^{[ ]}

The H5N1 Avian Influenza ("Bird Flu") is a highly virulent strain, with a high mortality rate. However, the disease has a relatively low ability to be transmitted between individuals. There is a substantial risk that, over time, mutation of the Avian Influenza virus will create a strain that is both highly virulent and highly transmissible in the human population. In Public Health and Emergency Management circles, it is believed that this is an inevitable event, summarised by the commonly used phrase "not ‘if’ but ‘when’".^{[ ]}

Even without the threat of Bird Flu, Public Health authorities believe that an influenza pandemic is overdue. There were several such pandemics in the 20th century. The most severe of these in terms of morbidity and mortality was in 1918, when almost 40 million people died worldwide, and the most recent occurred in 1968.

In New Zealand, the government has initiated a number of measures to prepare for a pandemic. Businesses and other organisations are being encouraged to develop plans for responding to an outbreak, and Public Health preparedness exercises are being conducted periodically.

This paper discusses some of the information management issues that might be significant in a pandemic emergency.

The World Has Changed
The last influenza pandemic originated in the Asia–Pacific region in 1968–69 and became known as the Hong Kong flu. In the US, 50 million people were infected, and between 33,000 and 34,000 people died. This was considered to be a relatively minor (class 2) pandemic as the virus was highly transmissible, but not highly virulent.

It should be remembered that the world is now very different to what it was in 1968. It would be two years before Boeing flew the first 747 Jumbo Jet in 1970, ushering in the era of cheap global mass transportation. In 1968, computers were large expensive devices used only by the largest of corporations, and were not to be found in medical practices of any kind, let alone in primary care.

In the last 10 years, New Zealand has led the world in the adoption of information technology in primary care.^{[ ]} In the same period, public health surveillance systems have not progressed at the same pace, remaining largely paper based or, at best, utilising fax machines for reporting notifiable diseases. Dependence on electronic means for transmission of information regarding patients and their conditions is rapidly becoming the norm, and will certainly be complete in the next 5 to 10 years.

These two factors alone, a rapidly mobile World population and a highly computerised health system, are likely to make the next pandemic significantly different to what has gone before. Consideration should be given to the effect that these changes would have on the management of a pandemic emergency.

What Will Happen in a Pandemic
During a pandemic emergency, the focus is on limiting dissemination. People will be advised to stay at home and to minimise contact with one another in general as this is considered the only effective means to manage the spread of the disease. There is no known intervention that can eliminate or even reduce the rate of transmission of the virus, and thus the course of a pandemic has a certain inevitability about it. Forecasting models in the New Zealand population using incidence rates of 15% to 35% indicate up to 760,000 cases severe enough to require medical treatment and up to 3700 deaths might occur throughout the country. This is expected to place a huge strain on a health system that is already under severe pressure. ^{[ ]}

Standard Civil Defence measures will be activated in the event of a pandemic emergency. Affected centres will activate Incident Control Centres (ICCs) with responsibility for co-ordinating the response to the emergency. These will typically be located in pre-prepared secure facilities, such as those operated by ambulance services. Each ICC will have a Planning and Intelligence Team tasked with ensuring that the decision-makers are kept up to date with the latest available information.

The immediate effect of such a level of illness in the population in a small country such as New Zealand will be a significant impact on provision of services of all kinds. Providers of information technology infrastructure such as Internet service providers will struggle to maintain services as they lose staff capability through a combination of illness and voluntary or imposed social isolation to avoid infection.

Consequently, there is some risk that increasing dependency on information technology which will inevitably become less reliable will, if not properly planned for and managed, result in a further pressure on the system.

Coded States
The New Zealand National Health Emergency Plan (which includes provision for pandemic emergency) divides the country into four regions. Planning structure divides the stages of pandemic into four colour-coded states.

Code White is the term given to the current state. No pandemic exists, but it is expected to occur at some unknown time in the future. Emergency management activities in this phase are concentrated on planning and conducting exercises for the management of a pandemic.

Once the first cases of highly virulent and transmissible influenza have been identified, the emergency management state will be altered to Code Yellow. The first cases are almost certain to occur overseas, but it is conceivable that the first cases could occur within New Zealand borders. Information activities in Code Yellow concentrate on heightened monitoring and reporting, and keeping the virus out of the country as long as possible. This state is tagged "Keep it out" in the planning documents.

Once cases of pandemic influenza have been detected in New Zealand, a pandemic emergency will be declared and the emergency state shifted to Code Red. In this state, borders are likely to be closed and emergency management measures activated, including the establishment of emergency management centres – Incident Control Centres (ICCs) – across the four regions to co-ordinate health and civil defence activities ("Stamp it out"). These will deploy Planning and Intelligence Teams to monitor the flow of information in order to inform decisions on the co-ordination of health and civil defence resources.

Once the pandemic has run its course, Code Green will be declared. Activities will then focus on analysis and reporting of the impact of the pandemic, nature and spread of the disease, how well the emergency was managed and lessons to be learned for the future ("Recover from it")

Central government agencies will declare the change of status in each case using information provided by the World Health Organization and the internal public health surveillance systems. ^{[ ]}

During the Code Red state, primary treatment for pandemic influenza will be restricted to Community Based Assessment Centres (CBACs). Generally these will be located at existing major Accident and Medical centres

Information Needs
The information requirements during the different stages are broadly as follows:

Code White:
General information and advice; Establishment and maintenance of key standing data, such as population demographic data, and lists of practitioners and facilities (hospital, rest homes, general practices, etc); Key parameters for data to be understood and set; Inventories of emergency supplies.

Code Yellow:
Low-level reporting; Public health influenza surveillance data to be widely communicated; Establishment and testing of code red information systems.

Code Red:
Planning and Intelligence Teams fully activated; Key health providers to communicate with incident controllers on a regular basis (at least daily); Collection of data on individual cases, mortalities, etc, started.

Code Green:
Planning and Intelligence Teams stand down; Data collected during Code Red available for analysis and research.

Current State
The information systems currently deployed for surveillance of notifiable diseases are relatively unsophisticated, and lack automation and reliability features. Further, the national strategies for health information systems do not appear to foreshadow actions at the national level to improve this situation.^{[ ]} The Crown Research Institute ESR is said to have plans to enhance the surveillance systems using data extracted from practice management systems used by General Practitioners, but little progress has been reported to date.

Many of the information technology components required to effectively collect and disseminate key information in a pandemic emergency already exist. In particular, the vast majority of general practices use practice management systems to record information electronically at the time of patient consultations and are accustomed to referring data electronically to other agencies and groups. It should be relatively simple to co-opt features of these systems to capture the minimum dataset required for pandemic management.

There is an opportunity to integrate these various components with a view to enhancing the dissemination of information both during a pandemic emergency and for Public Health business-as-usual activities.

What Will Be Required
The information needs during a Code Red pandemic emergency are likely to be quite different to those in the preparatory stages and after the pandemic. At the height of the emergency, summarised information, rapidly transmitted, will be the order of the day. However, once the pandemic has run its course, there will be a need for epidemiologists and other researchers to examine data in a less pressured fashion. In practice, there will also be some need to examine the detailed information during Code Red, but this should not override the needs of the Planning and Intelligence Teams to have good information as rapidly and simply as possible.

Information systems requirements for pandemic planning should dovetail into other initiatives and not attempt to create something new and separate to that which exists already. Practicality is key; systems should not be so complicated or untested that they cannot be depended on to work properly when the time comes.

Planning and Intelligence Teams will be required to rapidly interpret a stream of data of considerable volume. In the past, measures such as marking maps with pins have been used. There are opportunities to explore electronic means to present the data in forms that make it possible to rapidly interpret what is happening. This is likely to require some form of real-time graphical presentation. There are some emerging technologies derived from those used to visualise the America’s Cup racing that could potentially be useful in this area. If such options were feasible in terms of availability at modest cost, it may be possible to provide the Planning and Intelligence Teams with a much improved means of providing high quality and timely information where needed.

It is necessary to allow for alternative means of data transmission and provide some redundancy to mitigate the risks of unreliable network infrastructure.

There will be a universal need for analysis of aggregated output, ie, information summarised from unit identifiable data which shows the detail of each individual and what has been happening with them. Although not all parties want to be able to see the detailed data, and it may not be necessary to view detailed data during the Code Red phase, it certainly will be important for post-pandemic review purposes.

The work undertaken to plan the information systems for a pandemic should be highly focussed, and consistent with the minimum requirements for good forward planning. There is a substantial risk of "scope creep" with the consequence that systems may be over-designed and under-tested to the point where they will not be practically useful, which is clearly not desirable.

Any work undertaken to prepare information systems for pandemic management should be consistent with a general improvement of public health surveillance systems in general.

Primary care data collection should utilise existing practice management systems and integrate with routine processes. Not all hospital systems will be as flexible as primary care practice management systems and consideration may need to be given to add-on systems in hospitals.

An absolute and total failure of Internet and other network services is considered unlikely, but systems may "flicker", ie, be periodically unreliable.

Fortunately, there is no genuine requirement for real time data. Updates every 24 hours to Planning and Intelligence Teams would be adequate for the purposes of monitoring the pandemic.

Processes
Data could be captured initially using systems that are already established in primary and secondary facilities (practice management systems and hospital information systems). Facilities would be expected to report their status periodically. The cycle of reporting would be determined from time to time by the Planning and Intelligence Teams.

Status reports could be extracted according to pre-defined criteria from facilities’ systems, packaged into a message format and transmitted simultaneously to a number of different destinations. Under normal circumstances, facility reports would include a "headline" report summarising the status of the facility, which would be combined with a minimal report for each case treated for pandemic influenza at the facility since the previous report.

A fallback option for submission of the headline facility status only could be provided by a simple static web page.

The suggested destinations for the status reports would be Incident Control Centres, the Regional Public Health Services and the DHBs within the region. Each of these destinations would have the same ability to receive and process the messages and store them in a database. The DHBs and regional public health services have their own networks which could be interlinked relatively easily for the purposes of accessing the data across DHBs. Incident Control Centres would have their own databases, which would have the advantage of redundancy, thereby mitigating risks due to network failures.

Details of patient consultations and treatments prescribed could be captured using a standard form based on a minimum dataset. This form would be used whenever a patient was seen in any clinical setting and their signs and symptoms appeared clinically compatible with pandemic influenza.

In primary care, the form could be deployed within the practice management system. Where hospital systems have the appropriate flexibility to quickly configure treatment capture forms, a similar mechanism could be deployed, otherwise there may be the opportunity to deploy an ancillary system, such as an installation of a primary care PMS or a custom-built database specifically for the purpose of capturing pandemic influenza data. Due to the mobile nature of hospital duty managers, deploying systems via mobile hardware such as tablet or laptop computers or personal digital assistants (PDAs) would have considerable advantages.

A defined, regular, reporting cycle would be established by the Planning and Intelligence Teams. This is likely to be 24 hours, initially, but may be need to be more frequent during a Code Red phase. At the end of the defined period, each facility would be expected to report its status to the relevant Planning and Intelligence Team. This would include hospitals, rest homes, general practices and CBACs.

In primary care facilities (including CBACs) reports would be extracted using a pre-defined query from the practice management system.

Data would be simultaneously forwarded to a number of defined target locations, most likely the DHBs, Regional Public Health Services and the Planning and Intelligence Teams. The primary mechanism would be the most commonly used mechanism for transmission between primary care facilities, and fallback mechanisms for transmission could include open email, copying to physical media and delivery by courier or fax/phone followed by manual data entry.

The daily reports would be loaded into a repository database at the destination facility. It is likely that copies of the repository database would be maintained at the Regional Public Health Services facilities and at the Incident Control Centres.

Hospitals could deploy some kind of mobile computer, such as a tablet, so that duty managers or their delegates could record hospital morbidity and mortality in a flexible manner. Tablet computers would also have

the advantage of having long battery life, therefore being capable of functioning in restricted power situations. Primary care settings should consider the deployment of uninterruptible power supplies (UPS) to ensure that desktop computers could continue to operate during power outages.

All facilities would require sufficient quantities of spare batteries for laptop and tablet computers.

It is envisaged that during Code White planning, the activities required to establish the processes outlined above would be implemented and trialled. Testing of the systems should be undertaken and refinements introduced as a result of the tests. Regular re-testing would ensure that the systems remain at an appropriate state of readiness.

In addition to the need to provide treatment, it is envisaged that there would be a need to capture data regarding deaths that occur in the community or otherwise outside a clinical care setting.

The regulations regarding who would have the authority to initially certify death in a pandemic emergency have yet to be established. However, it has been assumed that such certification will be limited to clinicians and possibly para-medical staff such as ambulance staff.

It is proposed that staff undertaking such certification would be equipped with mobile technology such as a laptop or tablet loaded with a practice management system or ancillary system used for the purpose of pandemic recording in hospitals, and would utilise similar methods for recording data.

Implementation Considerations
It is important that consideration be given to the implications of approaches used to ensure individual privacy protection.

It is considered that the exigencies of a pandemic emergency would be likely to override most concerns regarding protection of individual privacy. However, once Code Green is declared, privacy concerns would be likely to re-surface. It is worthwhile considering what measures could be taken in the planning and development stages to maintain public confidence at this phase.

There will be considerable long-term value in the dataset assembled during a pandemic emergency. Consideration should be given to creating a group charged with the long-term stewardship of this data.

Conclusions
New Zealand has an enviable reputation an innovator in health information systems. However, over-dependence on technology has risks in an emergency management scenario where services that can be taken for granted in normal circumstances may not be reliable.

This paper has discussed some pragmatic steps that could be undertaken immediately to mitigate these risks, while providing improvements to the management of public health surveillance overall.

It is not possible to predict when a pandemic will strike. Preparedness should focus on practical, prudent steps that can be undertaken now, rather than attempting to design a grand plan which may take years to realise and be half-complete when the emergency occurs.

References and Documents Reviewed

1. Auckland Regional Public Health Service. Highly Pathogenic Avian Influenza: Planning and Preparedness. 2006.
2. See for example the following:
   http://www.disasternews.net/news/article.php?articleid=3121
   http://www.occupationalhazards.com/News/Article/38330/ArticleDraw.aspx
   http://www.nehacert.org/catalog/index.php?main_page=product_info&cPath=5_6_8&products_id=23
   http://www.medscape.com/viewarticle/525642
   http://www.whitehouse.gov/homeland/pandemic-influenza.html
3. Didham, R., Martin, I., Wood, R., Harrison K. (2004). Information Technology systems in general practice medicine in New Zealand. Journal of the New Zealand Medical Association, 23-July-2004, Vol 117 No 1198. Available at http://www.nzma.org.nz/journal/117-1198/977/  
4. Wilson, N., Mansoor, O., Baker, M. (2005). Estimating the impact of the next influenza pandemic on population health and health sector capacity in New Zealand. Journal of the New Zealand Medical Association, 11-March-2005, Vol 118 No 1211 Available at http://www.nzma.org.nz/journal/118-1211/1346/  
5. Ministry of Health. 2004. National Health Emergency Plan: Infectious Diseases. Wellington: Ministry of Health. Available at: http://www.moh.govt.nz/moh.nsf/0/A7C725C02E537849CC256EE50004C1B2/$File/nationalhealthemergencyplan.pdf  
6. Health Information Strategy Steering Committee. 2005.Health Information Strategy for New Zealand. Wellington: Ministry of Health. Available at http://www.moh.govt.nz/moh.nsf/c43c7844c94e08cd4c2566d300838b43/edb64619d460f974cc2570430010fb71?OpenDocument