This issue of the Journal includes an article1 in the realm of clioepidemiology—epidemiology using historical data (after Clio, the muse of history).2 The article examines Japanese experience with pandemic influenza during 1918–19 and is notable for studying both morbidity and mortality data. It demonstrates that the issues are complex and that just because rural areas may have sometimes been found to experience lower pandemic-related mortality rates, this may not necessarily correspond with lower rural infection and morbidity rates.

Another notable feature of this article is the possibility that the lower urban morbidity rates were attributable to public health (e.g. social distancing) interventions and/or individual level interventions (e.g. mask use) in this part of Japan. Other recent evidence from this pandemic has also indicated that another island nation (Iceland) successfully used public health measures to reduce disease spread.3 Maritime quarantine used by mainland Australia, Tasmania, and some Pacific Islands also prevented or delayed arrival of this pandemic.4

Various recent studies in the United States provide additional evidence that quarantine and “protective sequestration” were effective for some communities for the 1918 pandemic.5 Similarly, social distancing interventions (e.g. school and workplace closures) were sometimes successful when implemented early and if sustained.6–8

Social distancing interventions appear to also have helped when the pandemic eventually reached Sydney.9 But the Australia experience also warns of the problems of people circumventing the quarantine blockade at state borders and refusing to wear masks.10

These general successes with pandemic prevention and mitigation contrast with most other countries for 1918, including New Zealand. Our country hardly attempted serious public health measures to control the pandemic11 and there were only a few isolated examples of local control successes.12 Furthermore, errors by New Zealand officials resulted in the spread of the pandemic to Samoa with catastrophic results, a failure for which there has been an official New Zealand Government apology.13

More recently New Zealand has invested substantially in pandemic influenza planning14,15 including extensive simulation exercises.16 Furthermore, a new revision (“version 17”) of the New Zealand pandemic plan is currently underway. Yet the New Zealand health sector still lacks a defined agenda for pandemic influenza research.

The need for such a research agenda has been articulated elsewhere (e.g. for the USA17) and actively developed and funded in Australia by their National Health and Medical Research Council (NHMRC).18

To start the discussion for New Zealand we provide some initial thoughts on potential pandemic influenza research priorities:

Clioepidemiology—Much remains to be learnt about New Zealand’s 1918 pandemic experience. Further analysis of individual level data from citizens and military personnel could be rewarding. For example, to understand why some communities appeared to manage the epidemic much better than others and to investigate differential impacts by socioeconomic position, ethnicity and for healthcare workers.

Modelling spread and containment—Expansion of previous modelling work done in New Zealand19 to utilise freely available modelling software could be performed (e.g. using InfluSim20). If detailed travel and time-use data were collated then this could even be fed into one of the supercomputer models available (as used in the US21). Such work may clarify the scope for travel restrictions to control spread (e.g. between the North and South Island), and inform when best to institute and lift school and other closures.

Seasonal influenza epidemiology—New Zealand has a comprehensive surveillance system for seasonal influenza that includes primary care, laboratory, hospitalisation, mortality, and immunisation coverage data. Findings from this system are already providing insights into the potential impact of influenza immunisation at a population level.22–24

Internationally, such surveillance has provided information on the emergence of adamantane resistance in influenza A viruses25 and the global spread and seasonality of influenza.26 With some further refinements, the New Zealand surveillance system could give additional insights into the real burden of disease from seasonal influenza, seasonality,27 and the spread within the country—which could have implications for a pandemic control. It may also be able to clarify the impact of interventions such as routine school holidays on influenza transmission (as per a study in France28).

Population vulnerability to influenza—Seasonal influenza can also be used to investigate patterns of population vulnerability to influenza according to demographic, socioeconomic, and environmental factors. Research in New Zealand is currently using a large cohort-study of social housing applicants and tenants to investigate the effects of housing conditions on hospitalisation with seasonal influenza and pneumonia.

Initial findings suggest that while household crowding has a modest effect on the risk of hospitalisation, having young children living in a household is a key risk factor for adult hospitalisation.29 A limitation of all work based on seasonal influenza is that pandemic influenza may behave somewhat differently.

Border screening for influenza—New Zealand is currently researching screening instruments for arriving airline passengers to evaluate the performance of methods such as declaration cards, temperature testing, throat/nasal swab collection, and post-arrival follow-up.30 Given the critical importance of the “keep it out” component of New Zealand’s pandemic influenza plan, this research area seems particularly urgent.

Current social distancing phenomena—Research could be done on studying the lessons arising from home detention for prisoners in terms of informing New Zealand responses to social distancing non-pharmaceutical interventions (NPIs). Utilising this “natural experiment” would be much less expensive than paying a random selection of people to adopt such social distancing measures for a trial period. Similarly, there could be work to explore the impact on New Zealand families with children that arise from sudden school closures—e.g. associated with norovirus outbreaks.

Experiments to assess potential interventions—The Centers for Disease Control in the US is funding experiments to investigate the effectiveness of NPIs—e.g. on mask use and hand hygiene.31 In New Zealand, there is also a Health Research Council funded study on using hand sanitisers to reduce illness absences in primary school children.32

Other health initiatives such as “the Sneeze Safe programme” (www.sneezesafe.co.nz) promoting respiratory hygiene among pre-school and primary school children, based on early intervention studies in Antarctica,33 provide a basis for further systematic research. Such studies are expensive, but the interventions being investigated may have wide benefits in reducing the morbidity from a wide range of infectious diseases.

Stockpile management—As some of the New Zealand stockpiles of antibiotics, antiviral medication, and of H5N1 vaccine start to expire, there will increasingly be a need for testing to determine if some of these supplies may still be worth retaining for emergency use (i.e. while efficacy persists over a certain minimal level).

Public attitudes to NPIs—Research on public attitudes to NPIs among New Zealanders could help assess acceptability and likely uptake. This research could use surveys (e.g. as per overseas studies34,35), focus groups, hui, and citizens’ panels or citizens’ juries. There is some experience with the latter in New Zealand.36

It might even be reasonable to plan on conducting short-duration and rapid turn-around public surveys during a pandemic to help guide the acceptability of public health recommendations and to identify problems with various social distancing NPIs. Indeed, this idea has been suggested in the US for pandemic influenza.35

Studying virtual worlds—Virtual worlds may partially reflect real-world human behaviour and are therefore attracting the interest of epidemiologists and social scientists.37–39 Therefore, researchers could study a typical New Zealand “virtual hospital” or emergency department with avatars controlled by real hospital staff for a day. This process could help inform how the health system responds to various pandemic scenarios.

Identifying co-benefits of influenza pandemic planning—Research could detail the spin-off benefits from pandemic planning for improvements in seasonal influenza control, improvements in other infectious disease control (e.g. if basic hygiene standards are improved) and for improving response to other natural disasters.

In conclusion, this brief preliminary research agenda is far from complete but it does suggest that there is plenty of scope for further worthwhile research in this country. Many of these research outputs will also be of value to New Zealand society even if the next influenza pandemic is far in the future.

Author information: Nick Wilson, Senior Lecturer, Michael G Baker, Associate Professor, University of Otago, Wellington; Lance C Jennings, Virologist, Canterbury Health Laboratories, Christchurch

Acknowledgements: Our thinking in this area has been stimulated by work for the Centers for Disease Control and Prevention (USA) (via grant: 1 U01 CI000445-01) and by work for the New Zealand Ministry of Health.

Correspondence: Dr Nick Wilson, Department of Public Health, University of Otago Wellington, PO Box 7343 Wellington South, New Zealand. Email: nick.wilson@otago.ac.nz