New Zealand's Experience of the 1918–19 in Uenza Pandemic

ARTICLE

New Zealand’s experience of the 1918–19 in uenza pandemic: a systematic review a er 100 years Jennifer A Summers, Michael G Baker, Nick Wilson

ABSTRACT BACKGROUND: The 1918–1919 influenza pandemic has been New Zealand’s most severe disaster event (around 9,000 deaths). We aimed to review the literature related to this pandemic in New Zealand and among New Zealanders overseas, to identify any remaining research gaps (given ongoing risks of future influenza pandemics and from new pathogens, eg, synthetic bioweapons). METHODS: Systematic literature searches and comparisons with international findings for this pandemic to facilitate identification of research gaps. RESULTS: A total of 61 relevant publications were identified. The epidemiological patterns reported were largely consistent with the international literature for this pandemic. These features included the w-shaped age-distribution for mortality, and the much higher mortality rates for indigenous people (ie, seven-fold for Māori vs New Zealand European). But some novel risk factors were identified (eg, large chest size as a risk factor for death in military personnel), and there was an extremely high mortality troop ship outbreak (probably related to crowding). In contrast to some international work, there was an apparent lack of a socio- economic gradient in mortality rates in two studies using modern analytical methods. New Zealand work has clearly shown how the pandemic spread via the rail network and internal shipping routes and the rarity of successful measures to prevent spread in contrast to some other jurisdictions. It has also found a marked lack of memorials to the pandemic (in contrast to war memorials). Nevertheless, some research gaps remain, including on the apparent marked reduction in birth rates in 1918–1919 and the reasons for no socio-economic gradient despite other New Zealand evidence for occupational class variation in lifespan at this time. CONCLUSIONS: This is a relatively well-studied disaster event but there remain important research questions relating to this pandemic in New Zealand. Filling these gaps may contribute to improved planning for managing future pandemics.

he 1918–19 infl uenza pandemic documented using both historical and epide- was one of the most severe natural miological methods. The knowledge of the Tdisasters in recorded human history. events leading up to the pandemic, and its It occurred during the fi nal stages of the long-term effects, are also increasingly well First World War (WW1) and this partic- documented. However, a century later, it ularly lethal H1N1 viral pandemic strain is prudent to re-examine all the evidence of infl uenza is estimated to have killed from a range of sources to gain a clearer 50–100 million people worldwide.1 Very few perspective of this pandemic and its impact localities (other than a few remote islands) on New Zealand. Therefore, this current escaped the deadly impact of this pandem- paper aimed to review the literature related ic, and for the peoples of New Zealand, the to the 1918–19 infl uenza pandemic in New pandemic would be described as “the worst Zealand and among its peoples (including single human disaster in recorded New Zea- New Zealand military personnel overseas) land history”.2 and to identify any research gaps. The Various aspects of New Zealand’s latter is of importance for this country’s experience of the pandemic have been preparedness for inevitable infl uenza and other pandemics in the future.

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 54 www.nzma.org.nz/journal ARTICLE

primary focus and provide understating of Methods the pandemic (with further details on each We aimed to systematically identify all publication in Appendix 4). In total, we published literature related to the 1918–19 identifi ed fi ve books, two case-series, nine infl uenza pandemic either in New Zealand discussion/commentary pieces, 15 epidemi- or among New Zealanders outside of New ological studies, 24 reviews and six theses. Zealand. The searches were conducted These publications use a variety of methods using the Cochrane Library, Google Scholar, and spanned different academic fi elds, Embase, PubMed, various University thesis such as social history, epidemiology, public repositories and grey literature for all health, health services and biostatistics. studies published from 1914 to 2018 (to The earliest identifi ed publication was a take into account evidence of pre-pandemic case series in 1918 of New Zealand military waves3,4). Search strategies used in data- personnel in the New Zealand Expeditionary bases and other archival/thesis catalogues Force (NZEF), which described post-mortem are detailed in the Appendix (Appendix 1, 2 details of cases.5 In contrast, the most recent and 3). We also examined the bibliographies publication was published in 2018 and it of all the identifi ed literature for additional reviewed mortality patterns within the publications. Pacifi c region during the pandemic.6 There has been steady increase in the Results number of publications on the pandemic The experience of the pandemic in New and New Zealand, with the vast majority of Zealand and among New Zealanders has studies being published in the last decade. been evaluated both during and since the This pattern may refl ect the increasing pandemic occurred in 1918, from a variety availability of records from 1918 such as of primary sources from the period (for the digitisation of military fi les for online example military records or personal archives and online newspaper records accounts) and secondary historical/epide- (Figure 1). But there has also been a growth miological publications. A total of 61 in all medical research, including for publications were identifi ed which relate “seasonal infl uenza” in the New Zealand to the 1918–19 infl uenza pandemic in New context (eg, 63 Medline-indexed publications Zealand or among New Zealanders as a as of late 2018).

Figure 1: Cumulative frequency of publications related to New Zealand’s experience of the 1918–19 inﬂ uenza pandemic.

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 55 www.nzma.org.nz/journal ARTICLE

Figure 2: Pandemic-related mortality per week among the civilian populations in the North and South Islands of New Zealand and NZEF personnel who died in New Zealand.4,12

Note: some of the civilian data include discharged NZEF personnel.

The main impact of the 1918–19 inﬂ uenza the military personnel. These reports have pandemic occurred in New Zealand during been used in several epidemiological studies November 1918 for both civilian and of the outbreak as they provide detailed military populations situated within New description of the individuals on board.4,10,11 Zealand (Figure 2). While for New Zealand Those on board the Tahiti experienced one military personnel situated either at sea (on of the worst ship outbreaks worldwide board the troopship HMNZT Tahiti) or in during the pandemic, and the personnel in the Northern Hemisphere, the experience of the Northern Hemisphere experienced two the pandemic was rather different. Oﬃ cial mortality waves, in November 1918 and in reports of the troopship outbreak are held late February 1919 (much like other popula- at Archives New Zealand7–9 and detail the tions in Europe) (Figure 3). course of the outbreak and the outcomes for

Figure 3: Pandemic-related mortality among New Zealand military personnel by location during the 1918–19 in the deﬁ ned pandemic periods.4

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 56 www.nzma.org.nz/journal ARTICLE

Research domains identiﬁ ed in a previous psychology.2 Given the signiﬁ cant increase review article for New Zealand as areas in publications, many of these domains have which needed further development included been developed (Table 1), along with several epidemiology, Māori history, social history, studies examining the health system in New military history and disaster sociology/ Zealand during the pandemic.

Table 1: Identiﬁ ed literature on the 1918–19 inﬂ uenza pandemic in New Zealand categorised by major research domains.

Research Comment domain Epidemiology Various epidemiological methods have been employed to examine the pandemic. For example, one (including risk study gave estimates of reproduction numbers (range 1.3 to 3.1) and a seven-day lag period from diag- factors for nosis until pandemic death indicating secondary pneumonia as cause of death13 while another study mortality) estimated a generation time (days) of between 2–4 days.14 Other studies have used case-control and cohort designs with individual level patient data to explore risk factors for mortality among various populations and environments.10,15–27 There was a ‘w’-shaped distribution pattern for mortality rates, with young adults (in particular men) dying at higher than expected rates, which was consistent with international findings.4,10,12,22,24,28–33 The New Zealand work has clearly shown how the pandemic spread via the rail network and internal shipping routes12 and rarity of successful measures to prevent spread (ie, there were only some successful moves in a small town26 and a few institutions12).

Impact on An estimate of Māori mortality during the pandemic is at 2,160 (registered deaths), but as Rice has Māori argued, the true figure is likely to have been much higher.34 Both civilian (7.3x higher) and military death rates (2.3x higher) were disproportionately higher among Māori populations compared with European ones.18 Several authors have theorised that these di erences were possibly due to factors such as: less protective exposure to a milder first wave (associated with rural residence), higher rates of chronic disease (eg, tuberculosis) and poorer access to healthcare/social support during the pandemic.2,6,12,18,27,32,34–39

Social history In the social history domain, two recent theses concentrate on the impact and role of children during (also with the pandemic, of which they both describe how older children were called upon to provide relief in some overlap both their own homes and in the wider community.40,41 The impact of housing on mortality outcomes with the has also been discussed, with particular evaluation of housing/crowding in Wellington,42 Christ- epidemiology) church,42,43 Auckland,44,45 and also the poor housing among the Māori population.35,37 Local outbreaks have also been evaluated such as in the town of Nightcaps (which experienced one of the highest Euro- pean mortality rates during the pandemic),46 Nelson,47 Temuka48 and the Coromandel Peninsula.49 The probable impact of crowds during Armistice celebrations was also described in terms of their impact in spreading the pandemic both within cities and to more rural areas.12,34 and comparative work with New Zealand compared to Australia,50 Japan51 and Iceland.22

Specific The relationship of the pandemic to WW1 and the military environment have been relatively well impacts explored for the New Zealand military beginning in 1918 with Eyre and Lowe’s case series and on military epidemiological analysis of the pandemic among military personnel.5,15 A majority of the publications personnel have been of epidemiological studies using modern-day statistical techniques conducted in the last 10 years.2,4,6,10,11,17,19,21,23–25 The medical response to the pandemic within the military has been described, including a vaccination study (using a very broad-spectrum vaccine) and outbreaks within military camps.4,35,52–54 The detrimental e ect of crowding was explored, which has implications for today’s pandemic planning.10,55 Higher mortality rates were found among Māori/Paciﬁ c Island military personnel,21,27,35 ‘fresh recruits’,4,19 those with larger chest sizes,4,23 those aged around 25 to 29 years,4,21,23 those with earlier WW1 deployment4 and those with prior hospital admissions for respiratory conditions.4 The experience on troopships has also been evaluated, in particular the very severe outbreak on the HMNZT Tahiti.4,10,11,56

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 57 www.nzma.org.nz/journal ARTICLE

Table 1: Identiﬁ ed literature on the 1918–19 inﬂ uenza pandemic in New Zealand categorised by major research domains (continued).

Research Comment domain New Zealand’s The role of the New Zealand ship the SS Talune in spreading the pandemic throughout the Pacific role in Islands has been described.57 This included Western Samoa (modern day ‘Samoa’) which resulted in pandemic approximately 25% of the adult population dying in the pandemic.6,58,59 The Talune also spread the spread in the pandemic to Fiji and Tonga.57 South Pacific

Disaster There has been little research on the sociological and psychological aspects of the pandemic. One sociology/ study commented on the possible e ect of WW1 involvement (psychological and physical stress) on psychology immune system response during the pandemic among New Zealand military personnel.23 Another study identified the relative lack of memorials to the pandemic (especially compared with war memorials) and the implications of this absence for public education concerning future pandemic disasters.60

Health Several authors have examined the impact of the pandemic on the health system in New Zealand, the Systems use of temporary hospitals during the pandemic, the limitations of the New Zealand Health Depart- ment’s response and the impact of the Minister of Public Health in 1918, Mr GW Russell, having multiple ministerial portfolios during war time.9,12,22,34,44,61–64 One author describes the pandemic as a worst case, with substantial health sta absent due to the war e ort and remaining New Zealand-based sta succumbing to the pandemic themselves.63,65 Authors describe the legislative changes and impact upon the New Zealand health system following the 1919 Royal Commission Report on the pandemic9 such as the 1920 Health Act.9,44,66

Pathology and Two studies describe the pathology and post-mortem results among New Zealanders (either in Duned- virology in, New Zealand67 or among military personnel located in the UK).28 The thesis by Champtaloup (at the University of Edinburgh) is a relatively new finding from this review and as it includes detailed drawing plates and photomicrographs of the NZEF pandemic cases of which examples are included in this review (Figures 4 and 5). Given that there are no known existing specimens of the pandemic in New Zealand and indeed few samples known worldwide, these plates of post-mortem findings are a particularly notable resource demonstrating the care taken in trying to understand the nature of the 1918–19 pandemic. An earlier study by Eyre and Lowe does not specify the post-mortem details as being from New Zealanders.5 In contrast to some other countries, there are no known existing pathological specimens relating to the pandemic held in New Zealand.

Other areas There are several sources that cover multiple domains such as epidemiology and social history, such (including as Rice’s three books,12,33,34 and Bryder’s two publications which span a whole city.44,45 Known resources multiple (not listed in Appendix 4) used in some of the above references in this table include WW1 Casualty and domains) Roll-of-Honour,68,69 military reports7–9 and other New Zealand Government reports.9,59,70–72

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 58 www.nzma.org.nz/journal ARTICLE

Figure 4: Plate 1 (Coloured) Lung Case 680: ‘haemorrhagic oedema type’.

Figure 5: Plate 2 (Coloured) Trachea & Bronchi: ‘along the trachea and bronchi appear swollen lymphatic glands dotted with haemorrhages’.

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 59 www.nzma.org.nz/journal ARTICLE

indigenous peoples from this pandemic.6 Discussion Similarly, risk factors for death identifi ed Main findings and interpretation in a New Zealand case-control study23 were A century has passed since the 1918–19 often the same as those reported elsewhere infl uenza pandemic reached the shores of (eg, co-existing chronic disease or being New Zealand and among the New Zealand a recent military recruit). The spread of military personnel serving in the Northern the pandemic via railway and shipping Hemisphere as part of WW1. The infl uenza networks was also part of the international pandemic occurred during the fi nal stages pattern—though rarely has it been shown of WW1 and no doubt spread further in such detail as in the work of Rice for New during Armistice celebrations resulting in Zealand.12 Our informal observations are the pandemic having the grim record as that memorials to this pandemic are rare New Zealand’s worst human disaster. The internationally—which is also the case for works of Rice12,34 and one of the current New Zealand.60 4 author’s thesis, has identifi ed 8,831 deaths Unusual patterns with the New as a results of the 1918–19 pandemic among New Zealanders. Rice’s34 more recent book Zealand data There were unusual patterns for pandemic suggests a more accurate total may be impacts and response in New Zealand which over 9,000 deaths if undocumented deaths contrast with other populations around the among the Māori population are taken world, such as the following: into account. The majority of these deaths occurred in New Zealand during a six-week • Some novel risk factors for pandem- period in November and December 2018. ic-related mortality were identifi ed However, the fi rst substantial numbers of in a New Zealand case-control study; New Zealanders dying during the pandemic with males with larger chest size occurred eight weeks earlier in September having an increased mortality risk 1918 on board the troopship HMNZT 107 (possibly related to the increased Tahiti on route to the UK carrying New lung capacity of soldiers which might Zealand military reinforcements. ‘increase the chance of a cytokine storm’).23 Studies in the last decade have increasingly focused on the pandemic’s impact • There was an apparent lack of a on military populations; due to access to socio-economic gradient in mortality rich and detailed medical records for large rates for the New Zealand European volumes of individuals. As roughly 40% of population10,12,23 with no apparent eligible male New Zealanders served in the variation by different housing districts NZEF during WW1, estimates of morbidity in Auckland.45 This fi nding (albeit and mortality can be generalised to the with only the two military studies wider population, such as the identifi cation using modern analytic methods), is in of increased pandemic mortality risks asso- contrast to some of the international ciated with being a young adult, being Māori literature. For example, a socio-eco- or Pacifi c, or having had pre-pandemic nomic gradient has been described respiratory admissions to hospital. for Norway,73 Sweden74 and Chicago, US.75 The fi nding of no gradient in Patterns consistent with the New Zealand is also in contrast with international data evidence for lifespan differences In general, the epidemiology of the by male occupational class in New pandemic in New Zealand was similar to Zealand at this time.76 that of other countries. A notable simi- • The similar male vs female mortality larity was the w-shaped age distribution rates among Māori were also for mortality rates (albeit with some more somewhat unusual—in contrast to the in-depth analysis that may relate to the relatively higher mortality in males vs role of the 1890s infl uenza pandemic in females in the New Zealand European this pattern).24 The higher mortality rate population.31 One of the possible for Māori and Pacifi c soldiers in the New reasons is Māori being exposed to Zealand military (see Table 1), is also greater crowding (more mixing of consistent with the higher burden for

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 60 www.nzma.org.nz/journal ARTICLE

men and women), but possibly also But other apparent gaps are as follows: the greater similarities in chronic • There is scope for better quantifying respiratory disease burdens by sex in the extent of undocumented Māori Māori (eg, due to greater similarities deaths with studies that compare in smoking prevalence). However, this registered deaths with names on area requires further research. memorials in urupā (Māori burial • New Zealand was like most settings sites). This type of research by a where there were only a few (or no) Māori researcher would also have successful acts of border control or the benefi t of further building Māori ‘protective sequestration’ (ie, there research capacity. were only some successful moves in a • Modern spatial analysis could be 26 12 small town and a few institutions. applied to data on the spread of the In contrast, some other jurisdictions pandemic (as detailed in maps on the successfully used maritime quar- rail and shipping networks compiled 58 22 antine, road closures and measures by Rice).12 This analysis could then such as school closures and public estimate the speed of pandemic 77 gathering bans. spread (eg, in kilometres per day and Strengths and limitations of this week). review • The relationship of the natality A strength of this review is that it is the shocks of 1918 and 1919 in New fi rst systematic review of this pandemic Zealand to the pandemic have not in the New Zealand context in the journal been thoroughly analysed, although literature. It has also encompassed a wide Pool provides a brief assessment variety of academic fi elds. However, this of decreased fertility among Māori review also has its limitations. For example women.35 It is known that infl uenza it has focused on secondary sources mostly infection is associated with fetal as they provide summary accounts/data. loss and stillbirth (as documented in Furthermore, there was little review of Dunedin, New Zealand, during the fi rst-hand accounts, such as newspaper pandemic)67 and it is likely to have articles, letters or diaries (other than been a factor in the sudden decline military sources, eg, Summers et al11). It in New Zealand’s annual birth rate is also conceivable that there are journal in 1918 by 9% and in 1919 by 17% articles or theses that were published in the (relative to 1917) [Data calculated early part of last century which have not yet from the 1924 Yearbook70 and based been included in the online medical data- on changes in rates]. This issue only bases which we searched. appears to have been studied in a few international settings, eg, US and Scan- Possible implications for future dinavia,73,78 Taiwan,79 and Sri Lanka.80 research This topic has some contemporary A fi rst step is probably to consider further relevance given suboptimal vacci- research around those unusual aspects of nation of pregnant woman against the pandemic in New Zealand that differed infl uenza.81 from the international literature as detailed Given the importance of learning from above (ie, the large chest size as a risk factor pandemics, there is perhaps a case for for death, the apparent lack of a socio-eco- the Ministry of Health to allocate funds nomic gradient in mortality rates, and to the Health Research Council to specif- the similarity of male and female Māori ically support research to address these mortality rates). For example, the use of knowledge gaps. Once this work was done, modern biostatistical methods for additional an updated review of the potential lessons analysis of other data relevant to socio-eco- relevant to modern pandemic planning nomic gradients, eg, by using the suburb could be conducted, eg, in terms of external data in Rice’s Christchurch research43 or the border control, internal border control, Auckland-based work by Bryder.45 health system preparedness and provision

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 61 www.nzma.org.nz/journal ARTICLE

of voluntary sector nursing care to neigh- use the centennial of the 1918 pandemic bours, etc. as an opportunity to refl ect on our read- Several studies evaluated the effect of the iness for the kinds of pandemics that we pandemic on the New Zealand health system may face over the next 100 years and plan both during and post-pandemic. There accordingly. One example is the need for appears to be consensus that the health rapid border closure to prevent entry of system was under immense strain, with particularly severe pandemic diseases. Such 83,84 limited resources/staffi ng and a crippled measures could be highly cost-effective. administration that was unable to provide clear leadership in the face of an over- Conclusions whelming infl uenza pandemic. If anything In the centenary year of the 1918–19 can be learned from this experience it is that infl uenza pandemic, now is the time for pre-pandemic planning is essential for New New Zealand to refl ect on its impact on Zealand in order to face inevitable future New Zealand in terms of its mortality, pandemics. Such planning needs to include its long-term effects on the physical and scenarios of similar intensity to the 1918 psychological health of survivors, and infl uenza pandemic and potentially more indeed its role in shaping New Zealand’s extreme events. health system and society. Most importantly A greatly improved Health Act in 1920 was for today’s New Zealanders, the pandemic one of the responses to the 1918 infl uenza provides insight into the nature of infl uenza pandemic, which aimed to address the pandemics and how societies response to limitations of the previous legislation such events. Despite the large body of work in terms of administrative responsibil- to date, there remain important knowledge ities/duties by restructuring the health gaps relating to this pandemic in New department and allowing special measures Zealand. Filling these gaps may contribute during periods of infectious disease to improved planning for managing future outbreaks.44,66,82 Perhaps New Zealand could pandemics.

Appendix 1: Literature search criteria Inclusion criteria Disease Influenza (specifically the 1918-19 Influenza Pandemic)

Population • Cases located in New Zealand during the pandemic period. • New Zealanders who contracted influenza during the pandemic period. • Cases of pandemic influenza as a result of the viral strain in New Zealand. Population groups of interest (based on inclusion criteria): • Māori/Pacific Island peoples • New Zealand Expeditionary Forces

Search terms See Appendix 2.

Language restrictions Foreign language papers if translation available.

Search dates 1914 to present day.

Exclusion criteria Study design Book reviews

Search dates Pre-World War One (ie, before 1914)

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 62 www.nzma.org.nz/journal ARTICLE

Appendix 2: Database search strategies Cochrane Libraries Search date: 22nd August 2018

ID Search Hits #1 influenza:ti,ab,kw 5,397

#2 Zealand 12,877

#3 1918:ti,ab,kw 108

#4 1918 334

#5 #1 and #2 and #4 1

#6 #3 and #5 1

#7 pandemic 712

#8 #7 and #2 21

#9 *flu* 183,023

#10 #9 and #2 2,769

#11 #10 and #4 20

#12 #10 and spanish 134

#13 #11 or #12 151 Google Scholar Search date: 22nd August 2018

ID Search Hits #1 In article: influenza zealand 72,800

#2 In article: 1918 zealand 199,000

#3 In article: 1918 influenza zealand 9,560

#4 In article: Zealand 1918 spanish 30,900

#5 In article: Zealand influenza spanish 10,200

#6 All in title: influenza zealand 209

#7 All in title: 1918 zealand 128

#8 All in title: influenza zealand 1918 35 Embase Search date: 22nd August 2018 Embase 1974 to 2018 Week 34 Embase Classic 1947 to 1973 Global Health 1973 to 2018 Week 32 HMIC Health Management Information Consortium 1979 to May 2018 Ovid MEDLINE® Epub Ahead of Print, In-Process & Other Non_indexed Citations, Ovid MEDLINE® Daily, Ovid MedLINE and Versions®

NZMJ 14 December 2018, Vol 131 No 1487 ISSN 1175-8716 © NZMA 63 www.nzma.org.nz/journal ARTICLE

ID Search Hits #1 (1918 influenza pandemic and Zealand).af. 18

#2 (pandemic and Zealand).af. 7,660

#3 (pandemic and influenza and Zealand).af. 586

#4 (“1918” and Zealand and influenza).af. 89

#5 (“1918” and Zealand and pandemic).af. 71

#6 (“1918” and Zealand).af. 193

#7 2 or 3 760

#8 1 or 4 or 5 or 6 193

#9 (Spanish and Zealand and influenza).af. 31 PubMed Search date: 22nd August 2018

ID Search Hits #1 (1918) AND influenza 1,239

#2 (Spanish) AND influenza 1,756

#3 ((1918) AND influenza) AND zealand 34

#4 ((spanish) AND influenza) AND zealand 7

#5 #1 OR #2 OR #3 OR #4 2,767

#6 (1918) OR spanish 411,227

#7 #6 AND influenza 2,767

#8 #7 AND zealand 40 Grey literature Search date: 22nd August 2018 Grey literature – searched key words such as “1918”, “inﬂ uenza”, Spanish”, and “Zealand”. www.greylit.org/ 0 www.opengrey.eu/ 0 http://oaister.worldcat.org/ 20 ntrl.ntis.gov/NTRL/ 465 www.archives.govt.nz – searched “1918 influenza” or “Spanish influenza” 85 27 Wellcome Trust 86 http://wellcomelibrary.org/ 1 University Library catalogues: 6 New Zealand: http://catalogue.library.auckland.ac.nz http://ourarchive.otago.ac.nz http://library.victoria.ac.nz, http://researcharchive.vuw.ac.nz/ http://mro.massey.ac.nz http://nzresearch.org.nz/ United Kingdom/International: http://ethos.bl.uk/ http://www.era.lib.ed.ac.uk/ http://www.openthesis.org/

Other sources 3

Appendix 3: PRISMA 2009 Flow Diagram87

Appendix 4: Additional detail on the publications (secondary sources) included in this systematic review Available through external link: http://www.nzma.org.nz/__data/assets/pdf_ﬁ le/0003/86655/Appendix-4-FINAL.pdf

Competing interests: Nil. Acknowledgements: We take this opportunity to thank Professor Geoffrey Rice for his long history of scholarship regarding this pandemic in New Zealand. We also thank the various New Zealand Government agencies which have made archival material available to ourselves and other researchers (particularly the digitalisation of military ﬁ les). Author information: Jennifer A Summers, Postdoctoral Research Fellow in Medical Statistics, School of Population Health and Environmental Sciences, Faculty of Life Sciences and Medicine, King’s College London, London, United Kingdom; Michael Baker, Professor of Public Health, Department of Public Health, University of Otago, Wellington; Nick Wilson, Professor of Public Health, Department of Public Health, University of Otago, Wellington. Corresponding author: Dr Jennifer A Summers, 4th Floor Addison House, Guy’s Campus, Faculty of Life Sciences and Medicine, King’s College London, London, United Kingdom. [email protected] URL: https://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no- 148714-december-2018/7769

REFERENCES: 1. Johnson N, Mueller J. 6. Shanks GD, Wilson N, 1918. Emerg Infect Dis. Updating the Accounts: Kippen R, Brundage JF. 2010; 16(12):1931–1937. GLobal Mortality of the The unusually diverse 11. Summers J. Pandemic infl u- 1918–1920 ‘Spanish’ mortality patterns in the enza outbreak on a troop Infl uenza Pandemic. Bull Pacifi c regions during ship - diary of a soldier in Hist Med. 2002; 76:105–115. the 1918–21 infl uenza 1918. Emerg Infect Dis. 2. Wilson N, Baker M. pandemic: refl ections at 2012; 18(11):1900–1903. the pandemic’s centenary. Ninety Years on: what we 12. Rice GW. Black Novem- Lancet Infect Dis., 2018: p. still need to learn from ber: The 1918 infl uenza pii: S1473-3099(18)30178-6. the “Black November” pandemic in New Zealand. 1918 about pandemic 7. NZ Government. “Health 2005, Christchurch: Canter- infl uenza. N Z Med J, 40th Reinforcements, bury University Press. 2008; 121(1285):136–138. Medical Report on 13. Sertsou G, Wilson N, Baker “Tahiti” Epidemic 1918” 3. Oxford JS, Sefton A, Jackson M, et al. Key transmission WA.1. Box 1/3/25. Record: R, et al. Early Herald Wave parameters of an institu- 10/79. Wellington: Outbreaks of Infl uenza in tional outbreak during the Archives New Zealand, 1916 prior to the Pandemic 1918 infl uenza pandemic National Offi ce. 1918. of 1918. Intl Congr Ser. estimated by mathematical 2001; 1219:155–161. 8. O’Neill CEJ. Court of modelling. Theor Biol 4. Summers JA. The Burden Enquiry regarding Med Model. 2006; 3(38). HMNZT Transport and Risk Factors for 14. Nishiura H, Wilson N. “Tahiti”. 1918: London. Death from the 1918-19 Transmission dynamics Infl uenza Pandemic 9. Denniston SJE, Mitchelson of the 1918 infl uenza amongst the New Zealand E, McLaren D. Report of pandemic in New Zealand: Military Forces of World the Infl uenza Epidemic analyses of national War One [PhD thesis]. Commission. 1919, and city data. N Z Med J. 2013, University of Otago, Appendix to the Journals 2009; 122(1296):81–86. Dunedin, New Zealand. of the House of Represen- 15. Eyre JWH, Lowe EC. Report tatives of New Zealand. 5. Eyre JWH, Lowe EC. upon the Autumn Infl u- Prophylactic Vaccina- 10. Summers J, Wilson N, enza Epidemic (1918) as it tions against Catarrhal Baker MG, Shanks G. affected the N.Z.E.F. in the affections of the respira- Mortality Risk Factors for United Kingdom. Lancet. tory tract. Lancet. 1918; Pandemic Infl uenza on 1919; 193(4988):553–560. 192(4963):484–487. New Zealand Troop Ship,

16. McSweeny K, Coleman Infl uenza Other Respir 1918 Infl uenza Pandemic A, Fancourt N, et al. Was Viruses. 2014; 8(3):329–338. in New Zealand. N Z Med rurality protective in the 24. Wilson N, Oliver J, Rice G, J. 2018; 131(1481):6–8. 1918 infl uenza pandemic et al. Age-specifi c mortal- 33. Rice GW. Black Novem- in New Zealand? N Z Med ity during the 1918–19 ber: The 1918 Infl uenza J. 2007; 120(1256):U2579. infl uenza pandemic and Epidemic in New Zealand. 17. Shanks GD, MacKenzie A, possible relationship to 1988, Wellington: Waller M, Brundage J. Low the 1889–92 infl uenza Allen and Unwin. but highly variable mortal- pandemic. J Infect Dis., 34. Rice GW. Black Flu 1918: ity among nurses and 2014; 210(6):993–995. The story of New Zealand’s physicians during the infl u- 25. Wilson N, Boyd M, Nisa S, et worst public health disas- enza pandemic of 1918–19. al. Did exposure to a severe ter. 2017, Christchurch, Infl uenza Other Respir outbreak of pandemic New Zealand: Canterbury Viruses. 2011; 5:213–219. infl uenza in 1918 impact University Press. 18. Wilson N, Barnard Telfar L, on long-term survival? 35. Pool DI. The effects of the Summers J, et al. Differ- Epidemiol Infect., 2016. 1918 Pandemic of Infl uenza ential Mortality Rates by 144(15): p. 3166–3169. on the Maori Population Ethnicity in 3 Infl uenza 26. Wilson N, Rice G, Thomson of New Zealand. Bull Hist Pandemics Over a Century, G, Baker M. Re-evaluating a Med. 1973; 47:273–281. New Zealand. Emerg Infect local public health control 36. Rice GW. A disease deadlier Dis. 2011; 18(1):71–77. measure used in New than war. N Z Med J. 2013; 19. Shanks GD, Summers JA, Zealand for the pandemic 126(1378):12–14. Wilson N. Large mortal- infl uenza of 1918. N Z Med 37. Rice G. Microgeography of ity differences between J. 2005; 118(1224):U1714. Historical Infl uenza Mortal- Australian and New 27. Summers JA, Wilson N, and ity: A New Zealand example Zealand soldiers during Baker MG, Ethnic gradient from the 1918 Pandemic, in 1918–19 infl uenza pandem- in mortality amongst the 1990, One Thousand Years ic., in 61st Annual Meeting New Zealand military of New Zealand Population: of the American Society personnel in World War Heeding the Past and of Tropical Medicine and One [poster presentation], Planning for the Future., Hygiene. 2012, Ameri- in Australasian Epide- N.Z.D.S. Conference, Editor. can Journal of Tropical miological Association 1989: University of Waika- Medicine and Hygiene: Annual Scientifi c Meeting. to, Hamilton, New Zealand. Atlanta, USA. p. 72. 2010: Sydney, Australia. 38. Charania NA, Turner NM. 20. Wilson N, Summers J, 28. Champtaloup S. A study 100 years since the 1918 Baker M. Injury Epidemi- of the infl uenza epidemic infl uenza pandemic—prog- ology and New Zealand in New Zealand 1918: ress made, yet questions military forces in World epidemiology, administra- remain. A synopsis of the War One. Inj Prev., 2012. tion and bacteriology, with 4th New Zealand Infl uenza 18 (Suppl 1)(A153). a report on the pathol- Symposium, February 21. Summers JA, Shanks ogy of twenty-fi ve fatal 2018. N Z Med J. 2018; GD, Baker MG, Wilson cases. 1920, University 131(1481):50–55. N. Severe impact of of Edinburgh. 39. Rice G. Maori mortality the 1918–19 pandemic 29. Maclean FS. Challenge for in the 1918 infl uenza infl uenza in a national Health: A History of Public epidemic. N Z Population military force. N Z Med J. Health in New Zealand. Rev., 1983; 9(1):44–61. 2013; 126(1378):36–47. 1964, Wellington: R.E. 40. Bennett C. Now the War 22. Summers J, Wilson N, Baker Owen, Government Printer. is Over, We Have Some- MG, Gottfredsson M. The 30. Wright-St Clair RE. Infl uen- thing Else to Worry Us”: infl uenza pandemic of za in New Zealand and the New Zealand Children’s 1918–1919 in two remote doctors who died from it. N Responses to Crises, 1914– island nations: Iceland and Z Med J. 1983; 96:765–768. 1918. J History Childhood New Zealand. N Z Med J. 31. Rice GW. Infl uenza in Youth. 2014. 7(1):19–41. 2013; 126(1373):74–80. New Zealand before 1918: 41. Lange KJ. “Child’s Play?”: 23. Summers J, Stanley J, Baker a preliminary report. The Role of Children in the MG, Wilson N. Risk factors Am J Epidemiol. 2018. 1918 Infl uenza Epidemic for death from pandemic 32. Rice GW. ‘That Terrible in New Zealand. 2015, infl uenza in 1918–1919: Time’: Refl ections on the Dickinson College. a case–control study.

42. Rice GW. Why did Welling- the wartime use of enhancing educational ton suffer nearly double the mustard gas. N Z Med value around pandemic death rate of Christchurch J. 1948; 47(257):4–16. preparedness. N Z Med J. in the 1918 infl uenza 53. Brundage JF, Shanks GD. 2017; 130(1465):53–70. pandemic?, in In: The 2018 What really happened 61. Edwards V. Pestilence from Jim Gardner Memorial during the 1918 Infl uenza abroad? The 1918 infl u- Lecture. 2018, Canterbury Pandemic? The importance enza epidemic. N Z Med History Foundation: of Bacterial Secondary J. 1986; 99(812):809–812. Christchurch, New Zealand. Infections. J Infect Dis. 62. Bradshaw W, Learning the 43. Rice GW. Christchurch in 2007; 196(11):1717–1718. lessons of the 1918 Spanish the 1918 Infl uenza Epidem- 54. Carbery ARD. The New infl uenza epidemic. Nurs ic: a preliminary study. Zealand Medical Service in N Z., 2005; 11(10):22–23. NZJH., 1979; 13(2):109–131. the Great War, 1914–1918. 63. Roddick JA. Delirium and 44. Bryder L. “Lessons” of the 1924; Available from: http:// fever in the antipodes: 1918 infl uenza epidemic www.nzetc.org/tm/schol- Nursing and the 1918 in Auckland. NZ J History. arly/name-140805.html infl uenza epidemic in 1982; 16(2):97–121. 55. Wilson N, Summers J, Dunedin Hospital, New 45. Bryder L. The 1918 Baker MG. Impact of the Zealand. J Res Nurs. Infl uenza Epidemic in 1918–19 infl uenza pandem- 2006; 11(4):357–370. Auckland. 1980, Univer- ic on the New Zealand 64. Wood PJ, Managing bound- sity of Auckland. Military and persisting aries between professional 46. Bulling GM. Preventable lessons for pandemic and lay nursing following deaths?: the 1918 infl uenza control. Microbiol Aust., the infl uenza pandemic, pandemic in Nightcaps. 2014; 35(3):138–140. 1918–1919: insights for 1991, University of Otago, 56. Wilson N, Summers J, professional resilience Dunedin, New Zealand. Baker M, Rice G. A bizarre today? J Clin Nurs. 2017; 47. Jamieson JPS. Notes of case of drowning as a result 26(5–6):805–812. the recent epidemic of of pandemic infl uenza 65. Roddick JA. When the Flag Infl uenza in Nelson. N Z (1918 pandemic). N Z Flew at Half Mast: Nursing Med J. 1919; 18:46–51. Med J. 2009; 122(1302). and the 1918 Infl uenza 48. Rice G. Crisis in a country 57. McLane JR. Setting a Barri- Epidemic in Dunedin. 2005, town: the 1918 epidemic cade Against the East Wind: Otago Polytechnic, in Temuka. Historical Western Polynesia and the Dunedin, New Zealand. News. 1985; 51:7–13. 1918 Infl uenza Pandemic. 66. Rice GW. The making 2012, University of Otago, 49. Lovell Gregg J. How of New Zealand’s 1920 Dunedin, New Zealand. Coromandel kept the Health Act. N Z Med Infl uenza Epidemic at bay. 58. McLeod MA. Baker M, J. 1988; 22(1):3–22. N Z Med J. 1919; 18:51–53. Wilson N, et al. Protec- 67. Fitchett F, Bowie JT, tive Effect of Maritime 50. Rice G. Australia and New Champtaloup ST, Drennan Quarantine in South Zealand in the 1918–19 AM. A Report on the recent Pacifi c Jurisdictions, infl uenza pandemic. “Infl uenza Epidemic” in 1918–19 Infl uenza Occas Pap Med Hist Dunedin, New Zealand. Pandemic. Emerg Infect Aust. 1990; 4:67–74. N Z Med J. 1919; 18. Dis. 2008; 14(3):468–470. 51. Rice GW. Japan and New 68. Archives New Zealand, 59. Elliot G, Wilson T, Sefton Zealand in the 1918 Books of Active Service Moorhouse WH. H-31c infl uenza pandemic: Casualty Forms, New SAMOAN EPIDEMIC comparative perspectives Zealand Expeditionary COMMISSION (REPORT on offi cial responses and Force 1914–1918, in OF) APPENDIX TO THE crisis management, in Agency: AABK Series: JOURNALS OF THE HOUSE The Spanish Infl uenza 519. 1914–1918. OF REPRESENTATIVES, Pandemic of 1918–19., H. 69. NZ Government, The Great 1 JANUARY 1919. 1919, Phillips and D. Killingray, War, 1914–1918 : New Wellington, New Zealand. Editors. 2003, Routledge: Zealand Expeditionary Taylor and Francis Group: 60. Wilson N, Ferguson C, Rice Force Roll of Honour. London and New York. G, et al. Remembering the 1924, Government 1918 infl uenza pandem- 52. Richardson GM. The onset Printer: Wellington. ic: national survey of of pneumonic infl uenza 70. New Zealand Department memorials and scope for in 1918 in relation to of Statistics, The New

Zealand Offi cial Yearbook. of health inequality in 82. Dow DA. Safeguarding 1924, Wellington: Eyre New Zealand: occupational the Public Health: A and Spottiswoode. class and lifespan in History of the New 71. New Zealand Department the late 1800s and early Zealand Department of of Statistics, The New 1900s. Aust N Z J Public Health. 1995, Wellington: Zealand Offi cial Yearbook. Health. 2018; 42:175–179. Victoria University Press. 1920, Wellington: Eyre 77. Markel H, Lipman HB, 83. Boyd M, Baker MG, and Spottiswoode. Navarro JA, et al. Nonphar- Mansoor OD, et al. 72. Killick E, Financing the maceutical interventions Protecting an island nation Infl uenza Epidemic. implemented by US cities from extreme pandemic Journal of the Department during the 1918–1919 threats: Proof-of-concept of Public Health, Hospi- infl uenza pandemic. JAMA. around border closure as tals, and Charitable 2007; 298(6):644–654. an intervention. PLoS One. Aid. 1919. 2:6–12. 78. Bloom-Feshbach K, 2017; 12(6):e0178732. 73. Mamelund SE. A social- Simonsen L, Viboud C, 84. Boyd M, Mansoor OD, ly neutral disease? et al. Natality decline Baker MG, Wilson N. Individual social class, and miscarriages asso- Economic evaluation household wealth and ciated with the 1918 of border closure for a mortality from Spanish infl uenza pandemic: the generic severe pandemic infl uenza in two socially Scandinavian and United threat using New Zealand contrasting parishes in States experiences. J Infect Treasury methods. Aust N Kristiania 1918–19. Soc Sci Dis. 2011; 204(8):1157–64. Z J Public Health. 2018. Med. 2006; 62:923–940. 79. Chandra S, Yu YL. Fertil- 85. Archives New Zealand. 74. Bengtsson T, Dribe M, ity Decline and the 1918 Archway Search Engine. Eriksson B. Social class Infl uenza Pandemic in 2012; Available from: and excess mortality in Taiwan. Biodemography http://www.archway. Sweden during the 1918 Soc Biol. 2015; 61(3):266–72. archives.govt.nz/ infl uenza pandemic. 80. Chandra S, Sarathchandra 86. Wellcome Library. The Am J Epidemiol. 2018. D. The infl uenza pandemic library at Wellcome 75. Grantz KH, Rane MS, Salje of 1918–1919 in Sri Lanka: Collection. 2017 [cited H, et al. Disparities in its demographic cost, 2017 November]. infl uenza mortality and timing, and propagation. 87. Moher D, Liberati A, transmission related to Infl uenza Other Respir Tetzlaff J, et al. Preferred sociodemographic factors Viruses. 2014; 8(3):267–73. Reporting Items for within Chicago in the 81. Yuen CY, Tarrant M. A Systematic Reviews pandemic of 1918. Proc comprehensive review of and Meta-Analyses: The Natl Acad Sci U S A. 2016; infl uenza and infl uenza PRISMA Statement. PLoS 113(48):13839–13844. vaccination during preg- Med. 2009. 6(7):e1000097. 76. Wilson N. Clement C, Boyd nancy. J Perinat Neonatal M, et al. The long history Nurs. 2014; 28(4):261–70.