Review of Australia’s Health Sector Response to Pandemic (H1N1) 2009: Lessons Identified

3.3 Key Issues and Lessons Identified

Page last updated: October 2011

Accurate and rapidly available pandemic surveillance data is of paramount importance in understanding the nature of the disease, its clinical picture and its burden in the community. The challenge is to balance the level of detail needed to inform decision makers and for public information purposes, and the expectations of public health officials, ministers, the media and the public. More work is needed in order to agree on expectations about the level and sustainability of data collection from the outset of a pandemic response. There is a need to balance the tension between public information needs, or interests, and the associated opportunity costs of diverting limited public health resources, as well as the impact of this on the timeliness of reporting and analysis of surveillance data. The information needed to inform public health policy is not necessarily the information needed to inform the public about the disease.

The collection and reporting of data is very resource intensive. Skilled epidemiologists are a limited resource needed for the analysis and interpretation of data, especially in the early stages of a pandemic. In 2009, epidemiologists were engaged in routine reporting driven by the need for frequent updates of numbers for the Situation Reports and the required reporting to WHO, which affected their capacity to analyse Australian data as required for the development of the first Australian Influenza Surveillance Summary Report. Planning needs to emphasise the importance of early analysis, interpretation and reporting of detailed data, and of a skilled workforce to support this work. Analysed data provide evidence for decision makers on disease transmission and severity to inform policy, such as when to relax public health measures at a time when vulnerable people have not been vaccinated.

Communication to the public needs to include a summary of the situation, acknowledgement of what remains uncertain about the outbreak, and plans for what will be undertaken next. Ministers were under pressure to report the latest numbers, and the need for frequent and resource-intensive updating of case, hospitalisation, ICU and death counts (three times per day in the first months) to meet public expectations proved a strain on surveillance systems and staffing resources. This also affected the capacity to analyse further the available surveillance data.

The collection and reporting of all requested data continued throughout the pandemic rather than changing with the phases as anticipated. Standards on data collection need to be agreed on, including guidelines on when to de-escalate or cease surveillance activities once they are no longer needed to inform decision making. The burden on Commonwealth and jurisdictional resources resulted in some jurisdictions independently making decisions to stop collecting and reporting certain data. This affected the completeness and reliability of the national data. An earlier, coordinated cessation of enhanced data collection would have resulted in a smaller but higher-quality data set. In some instances where ongoing data were required, sentinel systems or data from sentinel states or territories could have sufficed. Consideration could also be given to other areas that might provide ongoing evidence, such as pharmacy records of antiviral medications dispensed, ambulance 000 records or case data, and school absenteeism rates.

Maintaining NetEpi required enormous effort and over time became difficult as the number of cases increased. Double handling of some data was required as some jurisdictions were also required to enter cases into local systems. As a result of this added burden, some states made independent decisions to stop using NetEpi, in particular once virus transmission was widespread. The routine automated NNDSS reporting process or jurisdiction-based systems were then used. This resulted in a loss of information nationally, since NNDSS is capable of receiving only core data on cases and not the enhanced data that were being collected during the 2009 pandemic. Earlier implementation of interfaces between NetEpi and jurisdictional systems could have resulted in the collected data being more complete and in the process of extraction and analysis at national level being much less complex. While it may be difficult to combine information from various databases, each can contribute to building a clinical and epidemiological picture.

Top of Page

Jurisdictional reports of the number of people hospitalised per day and the proportion of those requiring additional care (high dependent or intensive care) proved to be a very useful severity marker. These data are extremely difficult to access or interpret, as jurisdictional surveillance systems rely on hospitalisation and death data which are collected for purposes other than surveillance. Few states are able to collect these data effectively and within the time frame needed for surveillance, so most had to rely on resource-intensive, manual follow-up and reporting of hospitalisations and deaths. It is important to continue the work that commenced in the years before the pandemic to move towards automated extracts from hospital surveillance systems.

There is no national system in Australia to collect real-time death data during a public health emergency.10 While jurisdictions had been working towards achieving this prior to the pandemic, New South Wales was the only jurisdiction that had an almost real-time system available. Although these data represented only one state, they provided very useful information indicating that the number of deaths from pandemic (H1N1) 2009 was not high. Manual collection of mortality data was resource intensive and added significant strain on public health capacity.

Adapting systems and collection mechanisms as case definitions changed over time proved to be difficult. Robust data dictionaries are critical to the uniform understanding of data fields and protocols, such as those for cross-border notifications. The ASPREN system and state-based ILI presentations to healthcare providers were good indicators of ILI in the community, but there were limitations with the representativeness of the ASPREN network. ASPREN was augmented to incorporate sentinel laboratory testing, and the implementation of this system should continue.

Continued development of routine seasonal influenza surveillance to include standard indicators of severity, including emergency department (ED) presentations, hospitalisations, ICU presentations, use of ECMO and deaths, would enable easy escalation during a pandemic.

While all levels of government in Australia worked collaboratively to ensure that a national surveillance picture was available to decision makers and to keep the public informed, completion of a surveillance plan for the collection, analysis and reporting of data at national level would enhance national capability. Protocols for sharing jurisdictional surveillance data with researchers and modellers need to be agreed on, as these data were not readily available and the potential benefits of disease modelling were not realised. In addition, while the NHMRC assisted by providing funds for urgent research, the research was not able to be implemented in time, or made available, to assist in reviewing the assumptions about the pandemic virus to inform the public health response.

Recommendation 8:
Complete a surveillance plan for the collection, analysis and reporting of data at national level.


Top of Page

10 Collecting mortality data can be delayed by coronial investigations, delays in clerical coding of the cause of death, and legislative constraints to accessing the records.



Document download

This publication is available as a downloadable document.

Review of Australia’s Health Sector Response to Pandemic (H1N1) 2009: Lessons Identified(PDF 1023 KB)