July 2011 as a Influenza Prevention Measure David Z. Roth,1,2 and Bonnie Henry,1,2 1BC Centre for Disease Control; 2School of Populaton and Public Health, University of British Columbia

Social distancing measures minimize influenza transmission by reducing contact between susceptible and infectious individuals, and include school closures, travel restrictions, and restrictions on mass Key Points gatherings. Here we review the recent • School Closures: Mathematical simulations provide evidence literature to evaluate the effectiveness of that school closures can reduce influenza transmission in an ideal this approach as a pandemic prevention setting. What is less clear is if these effects are still evident when measure. key assumptions are only partially met. Empirical evidence from the last pandemic suggests that school closures reduce commu- nity transmission; however, few studies systematically compare community transmission in areas with and without this interven- Social Distancing tion. Furthermore, school closures are likely not effective during Social distancing (SD) measures reduce severe (high R ) and not economically or socially ac- influenza transmission by limiting the 0 ceptable during mild pandemics (low R0). As a result, widespread contact frequency between infected proactive school closures are likely not an effective prevention and susceptible individuals (1). SD measure during an influenza pandemic. measures include school or workplace • Travel Restrictions and Border Control: Stringent travel restric- closures, the restriction or prohibition tions and border control may briefly delay imminent pandemics; of mass gatherings, travel restrictions, however, these approaches are neither economically nor socially measures to reduce community contacts, feasible except in unique settings (e.g. small island). and/or border control. These non- • Mass Gathering Restrictions: There is no recent evidence pharmaceutical approaches, also known outlining the effectiveness of mass gathering prohibition. While as public health measures, are currently such approaches should logically reduce influenza transmission, part of the World Health Organization they are not socially acceptable in most situations, especially for (WHO), U.S. Centers for Disease religious gatherings. Resources should instead be dedicated to Control and Prevention (CDC) and case identification and patient treatment and . other pandemic response plans (2). Widespread use of these measures likely with guaranteed availability during the plans, there is limited evidence on the helped reduce mortality rates during the early stages of a novel influenza pandemic effectiveness of such interventions (9); 1918-1919 influenza epidemic (3, 4, 5) before vaccine development (7, 8). the majority of evidence that does exist and to quell the recent 2009 pH1N1 focuses on seasonal influenza, modeling Despite the importance given to outbreak in Mexico (6). Furthermore, of past pandemics (1918, 1958, 1967), SD measures are the only intervention SD measures in influenza pandemic

knowledge that’s contagious! Des saviors qui se transmettent! acute respiratory infection (ARI) rates in communities with and without an 8-day school closure reveals compelling Key unanswered questions preliminary results. The odds of having an ARI during the period of closure were • Are proactive school closures more effective than reactive closures? 50% lower than before the closure in • How long must schools remain closed, and when is it is safe to re- the intervention community (OR=0.50, open schools? p<0.01), while the odds increased by 64% in the non-intervention community • Is there a threshold number of school closures required to reduce over the same time period (OR=1.64, community transmission? p<0.01) (24). • How important are hygiene practices in the school setting? Non-influenza-driven school closures also provide information on the effectiveness of SD measures. Pandemic H1N1 or more recently, the H5N1 avian children under 13 years of age reduced the infection rates in children in France 1 influenza. The 2009 pH1N1 pandemic reproductive number (R0) from 1.7 to decreased by 20%-29% during holiday represents an opportunity to re-evaluate 1.5, which decreased further to 1.1 after periods (no observable impact on adults), the effectiveness of SD approaches used all schools were closed for the summer and holidays are estimated to cause a as part of the pandemic response plans. (20). This drop in R0 likely resulted from 16%-18% reduction in seasonal influenza an estimated 70% reduction in intra- in all individuals (25). Extrapolating age-group transmission caused by school to prolonged school closures during a School Closures closures. However, no comparison data pandemic suggests a potential 13%- Background/Rationale from locations without school closures 17% decrease in cumulative cases in Prevention measures are most effective is presented and secular trends may be the community, with peak attack rates when they are directed at those at greatest responsible for the observed patterns. decreasing by 39%-45% (47%-52% risk of infection (8). Children and young Furthermore, school closures in previous in children). However, contact patterns adults experienced more severe illness Hong Kong winters did not appear during holidays differ from those occurring during the pH1N1 pandemic (10), and to affect community transmission of during pandemics and such extrapolations up to 20% of transmission occurred seasonal influenza (21). should be evaluated with care (25). in the school setting (11). School start Prefect-wide school closures implemented Hospital visit data from Israel show that dates were correlated with the timing in Osaka, Japan in response to the ratio of influenza-like diagnoses to of pH1N1 outbreaks (12), and school- approximately 100 ill students in May non-respiratory diagnoses decreased most centred outbreaks were reported in France 2009 decreased the number of newly rapidly two weeks after an elementary (13), England (14, 15), New York (16, reported cases from 30 on May 17 to 0 by school strike in Macabbi, during which 17), Canada (18) and China (19). It is May 25 (22). Only 13 schools reported a time 80% of children aged 6-12 years estimated that each infected school child single new case during this period. School remained home (26). This decrease was had spread the virus to 2.4 other children closures are also thought to have helped significant for both school-age children (95% CI: 1.8-3.2). School closures contain Japan’s first pH1N1 outbreak and adults without children. Furthermore, are therefore one potential method of (23). Modeling of these data shows influenza rates rebounded after the strike reducing influenza transmission early that school closures were effective when ended, indicating that school closures in an epidemic before the creation of a combined with post-exposure prophylaxis reduce community transmission (26, 27). vaccine and/or distribution of antivirals and home isolation, but only slowed It should be noted that children comprise (7, 8). transmission when used in isolation, 34% of Maccabi’s population and study doing little to reduce the total number of findings may not hold in locations with infected individuals (23). proportionately fewer children (26). Evidence: Observational Studies A published abstract of work from the In contrast, no difference in pre/post School closures in Hong Kong for Dallas-Fort Worth area comparing absentee rates, a proxy measure for

1The number of secondary cases caused by a single index case in a susceptible population.

2 National Collaborating Centre for Infectious Diseases influenza activity, was observed between its effectiveness, contrasting with previous Strict movement restrictions prohibiting schools with and without a winter break work (29, 30). students, professors and staff from during the peak of an influenza outbreak leaving their institutions (fengxiao) is Simulations of the Australian community in King County, Washington (28). These an alternative measure used in China of Albany (population 30,000) show that findings were still valid in a sub-analysis (34). Simulation findings indicate that school closures of 2 weeks could decrease of elementary schools only. However, these movement restrictions delay the influenza attack rates by 19%, while the school closure reported here likely epidemic peak if implemented early, 4-week closures combined with antiviral occurred too late in the influenza season but are less effective than local measures treatment reduce influenza attack rates to alter the viral transmission patterns focused on and hygiene. from 32.5% to 9% (32). However, (28). Furthermore, fengxiao may lead to more school closures are most effective for mild severe outbreaks in a university setting if epidemics (R =1.5), with effectiveness 0 not applied properly (34). Evidence: Mathematical Simulations increasing as school closure duration lengthens (33). Attack rates can be The importance of early implementation Simulations created prior to the pH1N1 reduced by up to 15% for a variety of to school closure effectiveness, coupled with the economic costs of the pandemic indicate that school closures R0s by combining school closures and can reduce attack rates, but that the antivirals. These models agree with intervention, mean that practical triggers effectiveness of such closures decreases previous findings (31) in suggesting are needed to guide school administrators with later implementation and with that closing individual schools is more and public health personnel. increasing R0 (29); recent studies support practical than system-wide closures, Retrospective analysis of school closure these findings. Simulations of a pH1N1 with the former being less dependent on data from Japan indicate that a threshold outbreak in a small Ontario city show the timing of their initiation and more of 5% single-day influenza-related that 7-day rolling school/daycare closures responsive to where and when cases are absenteeism, double-days >4%, or triple- reduce attack rates from 21.7% to 4.5% in occurring. days >3% are optimal for alerting school the absence of vaccination, and that such administrators to consider school closure Additional simulations of Albany support closures are more effective if implemented (35). However, use of such triggers may the effectiveness of the concurrent use of early in the course of the epidemic (30). result in schools being closed too late in multiple SD measures. The application of Here the combination of early school the pandemic to affect viral spread (27). school closures, isolation of symptomatic closures and vaccinations successfully Careful consideration must also be given individuals in their households, workplace halted transmission, but school closures to the timing of reopening schools, as a non-attendance, and reduction of contact are not required if the population has secondary peak in influenza cases may in the wider community reduces attack pre-existing immunity or if vaccines are occur if schools are re-opened before herd rates from 33% to 10% if introduced in distributed early in the pandemic (30). immunity has been reached or before the first 6 weeks of a simulated epidemic immunization of students and the general Agent-based simulations2 using data from with an R of 1.5 (7). Such prevention 0 population has occurred (27). Pennsylvania show that school closures approaches must be applied sooner at The variation in the estimated effect of longer than 8 weeks may delay the an R0 of 2.5, with 2, 3 and 4-week starts epidemic peak by up to 1 week, providing resulting in final attack rates of 7%, 21%, school closures on attack rates is likely due to differences in model assumptions additional time to distribute antivirals and 45% respectively. At an R0>3.5, and vaccinations (31). However, school all prevention measures are ineffective. regarding the timing of intervention closures of less than 8 weeks have limited Others show that a combination of implementation, the degree of contact effectiveness and closures of under 2 adult and child SD, antiviral treatment or mixing outside of the school setting, weeks may increase outbreak severity. and prophylaxis can reduce the total and the duration of the school closure (7, Interestingly, this model predicts no number of cases in the population from 36). However, in general, model findings meaningful difference between closing 35% to 10% for a low severity epidemic indicate the following: schools individually when sick children (R0<1.6 and a case fatality rate of <0.5%), • Short duration school closures (<2 present themselves, and system-wide compared to a decrease from 35% to weeks) are unlikely to be effective closures (31). Furthermore, the timing of 22% if only SD and school closures are in reducing community transmis- the intervention had minimal impact on implemented (1). sion;

2A simulation approach used to model dynamic systems in which “agents” are programmed to follow specific rules. These simulations have minimal assumptions and allow the system to evolve over time.

Centre de collaboration nationale des maladies infectieuses 3 Table 1 Summary of empirical work on social distancing interventions and their impacts on influenza transmission.*

Measure Author Intervention Design Location Finding

School Closures Wu, J. 2010 (21) 2 week school Age-structured Hong Kong Drop in R0 from 1.7-1.5 with 2 week closure for susceptible-infec- closure. R0 decreased to 1.1 after full children < 13, tious-recovered summer closure. increased to entire transmission summer. model Ryosuke et al., Closure of 270 Evaluation of Osaka Prefect, 13 schools reported only 1 case during 2009 (22) high schools/ simple case Japan school closure, compared to >100 526 junior high counts cases in a single school prior to school schools in Osaka closure. No further outbreaks. Prefect from May 18- May 24. An- tivirals prescribed to students with infection. Copeland et al., 8 day school Comparison of Dallas-Fort Worth Odds of having ARI during closure 2010 (24) closure acute respira- were 50% lower in intervention com- tory infection in munity than before closure, while schools with and increasing by 64% in non-intervention without closure community Cauchemez et al., School holidays Combined analy- France pH1N1 rates decreased by 20-29% 2008 (25) sis of surveillance during holidays (no decrease in data and the tim- adults). ing of holidays. Heymann et al., School Strike, Regression on Israel Ratio ILI to non-ILI decrease most 2009 (26) 80% children ratio of influenza rapidly 2 weeks after strike. Influenza aged 6-12 kept like illness to rates rebounded after strike. home. non-respiratory illness for school children, children household members, and all children aged >12. Rodriguez et al., Winter Break dur- Comparison of King Country, No difference in absenteeism rates. 2009 (28) ing peak influenza pre-post absentee- Washington State Results held for sub-analysis of season ism rates elementary schools only. Air Travel Re- Hsu and Shih, Travel restrictions Dynamic trans- Global Delay influenza pandemic by 1-3 strictions 2010 (54) to top 50 airports. mission models weeks. Restrictions of 99% delay Restrictions of up pandemic 1-2 months. to 99% of flights. Entry Screening Cowling et al., Entry Screen- Examination of Multiple countries Entry screening may delay local trans- 2009 (57) ing of individual entry screening mission 7-12 days. travelers policy in relation to first reported case Border Control Nishiura et al., Quarantine of in- Modeling of Global Quarantine 99% effective in prevent- 2009 (62) coming travellers epidemiological ing infection. for >8.6 days. characteristics of influenza

* Simulation studies are not found in the table because the complexity of such simulations does not allow for easy summarization.

4 • The effectiveness of school Economic and Social Consequences (9). Prolonged school closures also affect closures in reducing transmis- of School Closures educational continuity, which most sion decreases the later they are The decision to close schools depends negatively impacts children who are implemented; not only on the effectiveness of the struggling academically (9). People in • School closures in isolation are intervention, but must be balanced by the lower wage occupations, many of whom ineffective at R0>2.5; social and economic costs. School closures may be single parents, are typically unable • The concurrent use of multiple are estimated to result in absenteeism of to work from home or to miss work (48). SD measures, or pharmaceutical 16% of the entire UK workforce (42) and Underprivileged children are therefore and non-pharmaceutical mea- from 6%-19% (43) to 21% (27) of health forced to care for themselves if a parent sures, has a greater impact than care personnel. Surveys in New South is unable to remain home. Self-care the use of any in isolation [see Wales indicate that up to 37% of public situations are associated with high-risk (37) for a review]; health personnel may miss work due to behaviors such as drug use and alcohol school closures (44). Such absenteeism is consumption (9) and may have long-term • Individual school closures are especially problematic during pandemics societal implications. more practical than system-wide when hospitals may be running at closures; and capacity (27). • School closures do not halt trans- Travel, Border Control, and Entry mission, but likely provide addi- Absenteeism makes school closures a Screening more costly option than pharmaceutical tional time for the distribution of Movement of infected individuals within measures despite the latter’s higher initial antivirals and vaccination. and between countries, especially by air costs (1). An influenza pandemic could travel, facilitates influenza transmission decrease the gross domestic product by: 1) placing passengers in close School Closures and (GDP) of select countries by 0.5%-2%, contact for extended periods of time, Contact Patterns but these costs could double or triple and 2) increasing the connectedness School closures are effective only if student with the addition of school closures (45); of populations. Recent molecular contact is reduced, and many models associated costs would double again if evidence identified a single viral strain assume limited mixing of children outside closures were extended from 4 weeks at in six passengers traveling from the U.S. of the school setting (27). However, the peak to the entire wave (13 weeks) to Europe, which indicates in-flight survey-based estimates of contact patterns (46). A single week of school closures in transmission had likely occurred (49). during school closures suggest there is the U.S. could cost between 10 and 47 One-quarter of 116 case patients in continued student interaction (38, 39, billion dollars (0.1%-0.3% of GDP) (1), Singapore with travel-associated infection 40, 41). In Australia, 74% of students while a 26-week school closure could traveled after the onset of illness, and participated in activities outside of the cost 6% of GDP, 14-21 times the cost of 15% became ill while traveling (50). home during school closures, with an using only targeted antiviral prophylaxis Risk of in-flight transmission appears average of 3.7 out-of-home activities per or vaccines (47). The estimated total dependent on proximity (51, 49), which student/week, including sporting events, cost of a combined intervention using explains why transmission is higher outdoor recreation, shopping and parties adult and child SD, school closure, and in Economy than in First Class (52). (38). In Pennsylvania, students remained antiviral treatment for a community of However, no clustering was observed in at home for 77% of their days during 10,000 individuals is estimated to be as 9 of 123 positive passengers travelling to a one-week elementary school closure, high as 12.4 million dollars, with 74% of China (53). but 69% of students visited at least one that cost resulting from absenteeism (1). other location (39). In Boston, students In such a setting the addition of school Worldwide transmission of influenza interacted with the community and with closures to pharmaceutical interventions is greatly accelerated once the virus other students during a week-long school becomes cost-effective only for severe spreads to the top 50 global airports, and break, with contacts being more frequent strategies that apply control measures to outbreaks (R0>2.0, case fatality 1%) (1). in older students (40). In 39 U.S. states, these airports may help contain the virus Closing schools also has social 56% of parents responded that their (54). International air travel restrictions consequences that disproportionately children participated in at least 1 activity could delay influenza pandemics by affect the disadvantaged. School closures involving persons outside the home 1-3 weeks, while restricting 99% of air interrupted school lunch programs that during a three-day school closure (41). travel could provide an additional 1-2 fed 29 million U.S. children in 2004 months for vaccine administration (55).

Centre de collaboration nationale des maladies infectieuses 5 However, such restrictions are not from previously identified high-volume laboratory testing, epidemiological economically feasible and are predicted to cities of origin may enhance global analysis), reducing the spread of infection delay viral spread but not impact overall awareness of infectious disease threats (infection control), managing and morbidity (24, 56). (65). treating ill persons, and disseminating Identification and quarantine of sick relevant public health messages (69, 70). Consideration must also be individuals at points of entry may Mass Gatherings delay the introduction of influenza to given to contingency planning, surge Mass gatherings are events that attract select countries. Screening consists of capacity, staffing, space for quarantine/ sufficient numbers of participants temperature checks, health declaration isolation, and equipment for detection to “strain the planning and response surveys, observing arriving passengers for and prevention (71). In many mass resources of the community, city or nation symptoms, and/or thermal scans (57). gatherings, the density of individuals hosting the event” (66). Examples include Novel screening methods that quickly makes SD approaches impossible; hence sporting events, concerts, World Youth identify ill travellers by measuring and facemasks, cough etiquette and hand Days, and the Hajj. Mass gatherings analyzing an individual’s heart rate, washing should be actively promoted and breathing rate, and facial temperature readily available (72). have also been recently reported (58). A Pandemic H1N1 was detected during the review of first-recorded pH1N1 dates in Asian Youth Games in Singapore (71), relation to entry screening policies show Mass gatherings in a music festival and an international that entry screening may delay local developing countries, or sporting event in Serbia (73), and a transmission by 7-12 days (57). However, involving populations music festival in Belgium (74). Frequent simulations from Japan suggested that from poorer nations, pose temperature checks were used to identify border detection and quarantine had a particular problem due cases in Singapore, while the Serbian only limited effectiveness, with the initial to the limited resources and Belgian strategy focused on self- case being detected after >100 cases had available for surveillance identification facilitated through the use already entered the country (59). and treatment. of posters containing information on In general, border control has limited signs and symptoms, communication effectiveness in large countries with with the general public, sensitizing porous borders. Given the current scale medical personnel, and deployment of 24/7 mobile epidemiological teams in of air transport, effective border control bring together participants and influenza the Serbian case. Individual isolation of would require unrealistic detection rates strains from a variety of locations; six identified cases was used in Singapore in order to delay or limit transmission viral strains were detected at World Youth (71) but not in Serbia and Belgium, (60, 61). However, border control may Days in 2008 in Australia alone (67). which instead recommended self- be effective in small island settings with Prohibition of public gatherings has been isolation (73, 74). All identified cases a limited number of travellers where linked to reduced mortality and delays in were provided with antiviral treatment in quarantine of incoming travellers for reaching peak mortality during the 1918 the above situations. >8.6 days could have 99% effectiveness influenza pandemic (4, 68). However, in preventing the release of infectious national governments generally do not Mass gatherings in developing countries, individuals into the community (62). recommend prohibiting mass gatherings or involving populations from poorer However, few island nations could rely on during mild influenza epidemics because nations, pose a particular problem such measures in isolation (63). of potential social disruption. due to the limited resources available for surveillance and treatment. The The importance of air travel in spreading The principles used to reduce seasonal Hajj, the annual Muslim pilgrimage influenza does, however, mean that influenza transmission generally hold to Mecca, represents one of the largest analysis of air traffic data can help predict for mass gatherings, yet additional mass gatherings worldwide. In 2008, potential infectious disease hotspots (64). cooperation and coordination between 2.5 million pilgrims from 140 countries Real-time web-based disease surveillance public health agencies is required converged in Mecca, 11.3% of whom systems combining information on to optimize the implementation of came from low-income countries with worldwide patterns of commercial air prevention efforts (69). Planning limited ability to provide influenza traffic with RSS feeds that monitor news considerations should focus on detection vaccines (75). The incidence of vaccine- stories for references to infectious diseases and monitoring (screening, surveillance,

6 National Collaborating Centre for Infectious Diseases preventable influenza-like illness in 2,070 References M Assossou, and Marie Paule Kieny. Pakistani pilgrims at the 1999 Hajj 1. D.J. Perlroth, R.J. Glass, V.J. Davey, The 2009 a (H1N1) influenza vi- D. Cannon, A.M. Garber, and D.K. rus pandemic: A review. Vaccine, was 22/100 (76), while the influenza Owens. Health Outcomes and Costs 28(31):4895–4902, July 2010. attack rates in 115 participants of the of Community Mitigation Strategies PMID: 20553769. 2003 Hajj was 38% upon returning to for an Influenza Pandemic in the 11. Yang Yang, Jonathan D. Sugimoto, London (77). Memish et al. (70) provide United States. Clinical Infectious M. Elizabeth Halloran, Nicole Diseases, 50(2):165–174, 2010. E. Basta, Dennis L. Chao, Laura a review of recommendations aimed at 2. Centre for Disease Control and Matrajt, Gail Potter, Eben Kenah, improving public health preparedness Prevention. Community strategy and Ira M. Longini. The transmis- during the Hajj, one of the primary for pandemic influenza mitigation. sibility and control of pandemic http://pandemicflu.gov/professional/ influenza a (H1N1) virus. Science, recommendations being that those community/commitigation.html, 326(5953):729–733, October 2009. pilgrims with risk factors should stay 2010. 12. Dennis L Chao, M Elizabeth home. However, survey results indicate 3. Martin C. J. Bootsma and Neil M. Halloran, and Ira M Longini. School that European pilgrims are unlikely to Ferguson. The effect of public health opening dates predict pandemic measures on the 1918 influenza pan- influenza A(H1N1) outbreaks in follow these recommendations (78). The demic in U.S. cities. Proceedings of the United States. The Journal of scale of such an event and the potential the National Academy of Sciences, Infectious Diseases, 202(6):877–880, for high attack rates, coupled with the 104(18):7588 –7593, May 2007. September 2010. PMID: 20704486. 13. P. Carrillo-Santisteve, S. Renard- limited feasibility of SD measures in 4. H. Markel, H.B. Lipman, J.A. Navarro, A. Sloan, J.R. Michalsen, Dubois, G. Cheron, M. Csaszar- such settings, point towards the need for A.M. Stern, and M.S. Cetron. Goutchkoff, M. Lecuit, O. international vaccination support (72). Non-pharmaceutical interventions Lortholary, and P. Bello. 2009 pan- implemented by US cities during demic influenza A (H1N1) outbreak the 1918-1919 influenza pandemic. in a complex of schools in Paris, JAMA, 298(6):644–654, 2007. France, June 2009. Euro surveillance: Research Priorities 5. N. Tomes. “Destroyer and Teacher”: bulletin europ«een sur les maladies The current lack of strong empirical Managing the Masses During the transmissibles= European communi- cable disease bulletin, 15(25), 2010. evidence results from both the rarity of 1918–1919 Influenza Pandemic. Public Health Reports, 125(Suppl 14. I. Kar-Purkayastha, C. Ingram, H. influenza pandemics and the difficulty 3):48–62, 2010. Maguire, and A. Roche. The im- of implementing robust epidemiological 6. C. Franco-Paredes, P. Carrasco, and portance of school and social activi- study designs at the community level. J.I.S. Preciado. The first influenza ties in the transmission of influenza A (H1N1) v: England, April-June While simulation studies are helpful, pandemic in the new millennium: lessons learned hitherto for current 2009. Euro surveillance: bulletin research priorities should focus on well- control efforts and overall pandemic europeen sur les maladies transmis- designed epidemiological studies (with preparedness. Journal of Immune sibles= European communicable dis- ease bulletin, 14(33), 2009. inclusion of control groups ideally) in a Based Therapies and Vaccines, 7(1):2, 2009. 15. L Calatayud, S Kurkela, P E Neave, A community setting. Practical information 7. Joel K Kelso, George J Milne, and Brock, S Perkins, M Zuckerman, M on how to effectively implement SD Heath Kelly. Simulation suggests that Sudhanva, A Bermingham, J Ellis, R rapid activation of social distancing Pebody, M Catchpole, R Heathcock, measures, specifically school closures, and H Maguire. Pandemic (H1N1) is required to guide public health, can arrest epidemic development due to a novel strain of influenza. 2009 virus outbreak in a school in governmental agencies, and school BMC Public Health, 9:117–117, London, April-May 2009: an obser- personnel during a pandemic and other 2009. PMID: 19400970 PMCID: vational study. Epidemiology and 2680828. Infection, 138(2):183–191, February infectious disease outbreak. 2010. PMID: 19925691. 8. Jacco Wallinga, Michiel van Boven, and Marc Lipsitch. Optimizing in- 16. Justin Lessler, Nicholas G Reich, fectious disease interventions during Derek A T Cummings, H P Nair, H T an emerging epidemic. Proceedings Jordan, and N Thompson. Outbreak of the National Academy of Sciences of 2009 pandemic influenza A of the United States of America, (H1N1) at a New York City 107(2):923–928, January 2010. school. The New England Journal PMID: 20080777. of Medicine, 361(27):2628–2636, December 2009. PMID: 20042754. 9. Benjamin E Berkman. Mitigating pandemic influenza: the ethics of 17. Anne Marie France, Michael Jackson, implementing a school closure Stephanie Schrag, Michael Lynch, policy. Journal of Public Health Christopher Zimmerman, Matthew Management and Practice: JPHMP, Biggerstaff, and James Hadler. 14(4):372–378, August 2008. Household transmission of 2009 PMID: 18552649. influenza A (H1N1) virus after a school-based outbreak in New York 10. Marc P Girard, John S Tam, Olga City, April-May 2009. The Journal

Centre de collaboration nationale des maladies infectieuses 7 of Infectious Diseases, 201(7):984– of school closure on influenza trans- and Zhien Ma. Community-Based 992, April 2010. PMID: 20187740. mission from Sentinel data. , Measures for Mitigating the 2009 18. Jennifer Cutler, Emily Schleihauf, 452(7188):750–754, 2008. H1N1 Pandemic in China. PLoS Todd F. Hatchette, Bev Billard, 26. A D Heymann, I Hoch, L Valinsky, ONE, 5(6):e10911, June 2010. Gaynor Watson-Creed, Ross E Kokia, and D M Steinberg. School 35. A. Sasaki, A. Gatewood, H.S. Al Davidson, Yan Li, Nathalie Bastien, closure may be effective in reducing Ozonoff, N. Tanabe, N. Seki, R. Shelly Sarwal, and the Nova transmission of respiratory viruses Saito, and J.S. Brownstein. Scotia Human Swine Influenza in the community. Epidemiology Evidence-based Tool for Triggering Investigation Team. Investigation of and Infection, 137(10):1369–1376, School Closures during Influenza the first cases of human-to-human October 2009. PMID: 19351434. Outbreaks, Japan. Emerging infection with the new swine-origin 27. Simon Cauchemez, Neil M Ferguson, Infectious Diseases, 15(11):1841– influenza A (H1N1) virus in Canada. Claude Wachtel, Anders Tegnell, 1843, 2009. CMAJ, 181(3-4):159–163, August Guillaume Saour, Ben Duncan, and 36. George J. Milne, Joel K. Kelso, 2009. Angus Nicoll. Closure of schools Heath A. Kelly, Simon T. Huband, 19. Yang Huai, Jinyan Lin, Jay K. Varma, during an influenza pandemic. and Jodie McVernon. A small com- Zhibin Peng, Jianfeng He, Chen The Lancet Infectious Diseases, munity model for the transmission Cheng, Haojie Zhong, Yuansheng 9(8):473–481, August 2009. PMID: of infectious diseases: Comparison Chen, Yingdong Zheng, Yuan Luo, 19628172. of school closure as an intervention Wenjia Liang, Xiaoling Wu, Zhenyu 28. Carla V Rodriguez, Krista Rietberg, in Individual-Based models of an Huang, Jeffrey McFarland, Zijian Atar Baer, Tao Kwan-Gett, and Jeffrey influenza pandemic. PLoS ONE, Feng, Timothy M. Uyeki, and Duchin. Association between school 3(12):e4005, December 2008. Hongjie Yu. Original Article: A pri- closure and subsequent absenteeism 37. Vernon J Lee, David C Lye, mary school outbreak of pandemic during a seasonal influenza epidemic. and Annelies Wilder-Smith. 2009 influenza A (H1N1) in China. Epidemiology (Cambridge, Mass.), Combination strategies for pandemic Influenza and Other Respiratory 20(6):787–792, November 2009. influenza response - a systematic re- Viruses, 4(5):259–266, 2010. PMID: 19770773. view of mathematical modeling stud- 20. Joseph T. Wu. School Closure and 29. Kathryn Glass and Belinda Barnes. ies. BMC Medicine, 7:76, 2009. Mitigation of Pandemic (H1N1) How much would closing schools PMID: 20003249. 2009, Hong Kong. Emerging reduce transmission during an 38. Paul V Effler, Dale Carcione, Infectious Diseases, 16(3):538–541, influenza pandemic? Epidemiology, Carolien Giele, Gary K Dowse, 2010. 18(5):623–628, 2007. Leigh Goggin, and Donna B Mak. 21. Benjamin J. Cowling. Effects of 30. M. Z. Gojovic, B. Sander, D. Fisman, Household responses to pandemic School Closures, 2008 Winter M. D. Krahn, and C. T. Bauch. (H1N1) 2009-related school closures, Influenza Season, Hong Kong. Modelling mitigation strategies for Perth, Western Australia. Emerging Emerging Infectious Diseases, pandemic (H1N1) 2009. Canadian Infectious Diseases, 16(2):205–211, 14(10):1660–1662, 2008. Medical Association Journal, February 2010. PMID: 20113548. 22. Ryosuke Kawaguchi, Masaya 181(10):673–680, 2009. 39. T.L. Gift, R.S. Palekar, S.V. Sodha, Miyazono, Tetsuro Noda, Yoshihiro 31. Bruce Y Lee, Shawn T Brown, Philip C.K. Kent, R.P. Fagan, W.R. Archer, Takayama, Yasunori Sasai, and Cooley, Maggie A Potter, William P.J. Edelson, T. Marchbanks, A. Hiroyasu Iso. Influenza (H1N1) D Wheaton, Ronald E Voorhees, Bhattarai, D. Swerdlow, et al. 2009 outbreak and school closure, Samuel Stebbins, John J Grefenstette, Household effects of school closure Osaka Prefecture, Japan. Emerging Shanta M Zimmer, Richard K during pandemic (H1N1) 2009, Infectious Diseases, 15(10):1685, Zimmerman, Tina-Marie Assi, Pennsylvania, USA. Emerging October 2009. PMID: 19861075. Rachel R Bailey, Diane K Wagener, Infectious Diseases, 16(8):1315– 23. H Yasuda and K Suzuki. Measures and Donald S Burke. Simulating 1317, 2010. against transmission of pandemic school closure strategies to miti- 40. J.C. Miller, L. Danon, J.J. O’Hagan, H1N1 influenza in Japan in 2009: gate an influenza epidemic. Journal E. Goldstein, M. Lajous, and M. simulation model. Euro Surveillance: of Public Health Management and Lipsitch. Student Behavior during a Bulletin Europeen Sur Les Maladies Practice: JPHMP, 16(3):252–261, School Closure Caused by Pandemic Transmissibles = European June 2010. PMID: 20035236. Influenza A/H1N1. PLoS ONE, Communicable Disease Bulletin, 32. Nilimesh Halder, Joel K Kelso, and 5(5), 2010. 14(44), 2009. PMID: 19941778. George J Milne. Analysis of the effec- 41. GK Steelfisher and MM Bekheit and 24. D Copeland, A Kurian, W Chung, tiveness of interventions used during N LIddon and E Kahn and R Shieber R Basurto-Davila, D Fishbein, the 2009 A/H1N1 influenza pan- and K Lubell. Parental Attitudes H Lipman, I McCullum, P. demic. BMC Public Health, 10:168– and Experiences During School Szymanowski, J Zipprich, and F 168, 2010. PMID: 20346187 Dismissals Related to 2009 Influenza Averhoff. Effectiveness of a school PMCID: 2853510. A (H1N1) — United States, 2009. closure intervention for 2009 pan- 33. Nilimesh Halder, Joel Kelso, and MMWR Morb Mortal Wkly Rep, demic influenza a (h1n1) in two George Milne. Developing guide- 59(35):1131–1134, September communities. In International lines for school closure interventions 2010. Conference on Emerging Infectious to be used during a future influenza 42. M.Z. Sadique, E.J. Adams, and Diseases, page 186, 2010. pandemic. BMC Infectious Diseases, W.J. Edmunds. Estimating the costs 25. S. Cauchemez, A.J. Valleron, P.Y. 10(1):221, 2010. of school closure for mitigating an Boelle, A. Flahault, and N.M. 34. Sanyi Tang, Yanni Xiao, Youping influenza pandemic. BMC Public Ferguson. Estimating the impact Yang, Yicang Zhou, Jianhong Wu, Health, 8(1):135, 2008.

8 National Collaborating Centre for Infectious Diseases 43. H. Lempel, J.M. Epstein, and R.A. 51. M. G. Baker, C. N. Thornley, C. 60. N.M. Ferguson, D.A.T. Cummings, Hammond. Economic Cost and Mills, S. Roberts, S. Perera, J. Peters, C. Fraser, J.C. Cajka, P.C. Cooley, Health Care Workforce Effects of A. Kelso, I. Barr, and N. Wilson. and D.S. Burke. Strategies for School Closures in the US. PLOS Transmission of pandemic A/H1N1 mitigating an influenza pandemic. Cuffents, 1(RRN1051), 2009. 2009 influenza on passenger aircraft: Nature, 442(7101):448–452, 2006. 44. Craig B Dalton, David N Durrheim, retrospective cohort study. BMJ, 61. JG Wood, N. Zamani, CR and Michael A Conroy. Likely im- 340:c2424, 2010. MacIntyre, and NG Beckert. Effects pact of school and childcare closures 52. Bradley Wagner, Brian Coburn, and of internal border control on spread on public health workforce during Sally Blower. Calculating the poten- of pandemic influenza. Emerging in- an influenza pandemic: a survey. tial for within-flight transmission of fectious diseases, 13(7):1038–1045, Communicable Diseases Intelligence, influenza a (H1N1). BMC Medicine, 2007. 32(2):261–262, June 2008. PMID: 7(1): 81, 2009. 62. H. Nishiura, N. Wilson, and M.G. 18767427. 53. M Liu. Outbreak of pandemic Baker. Quarantine for pan- 45. Marcus Richard Keogh-Brown, H1N1 influena in an international demic influenza control at the bor- Richard D Smith, John W Edmunds, aircraft, Fujian, China, May 2009. ders of small island nations. BMC and Philippe Beutels. The macroeco- In 2010 International Conference on Infectious Diseases, 9(1):27, 2009. nomic impact of pandemic influenza: Emerging Infectious Diseases, page 63. Martin Eichner, Markus Schwehm, estimates from models of the United 132, 2010. Nick Wilson, and Michael Baker. Kingdom, France, Belgium and The 54. Chaug-Ing Hsu and Hsien-Hung Small islands and pandemic Netherlands. The European Journal Shih. Transmission and control of influenza: Potential benefits and lim- of Health Economics: HEPAC: an emerging influenza pandemic itations of travel volume reduction Health Economics in Prevention in a small-world airline network. as a border control measure. BMC and Care, December 2009. PMID: Accident Analysis & Prevention, Infectious Diseases, 9(1):160, 2009. 19997956. 42(1):93–100, January 2010. 64. Kamran Khan, Julien Arino, Wei Hu, 46. Marcus R. Keogh-Brown, Simon 55. Marta Luisa Ciofi degli Atti, Stefano Paulo Raposo, Jennifer Sears, Felipe Wren-Lewis, W. John Edmunds, Merler, Caterina Rizzo, Marco Ajelli, Calderon, Christine Heidebrecht, Philippe Beutels, and Richard D. Marco Massari, Piero Manfredi, Michael Macdonald, Jessica Liauw, Smith. The possible macroeconomic Cesare Furlanello, Gianpaolo Scalia Angie Chan, and Michael Gardam. impact on the UK of an influenza Tomba, and Mimmo Iannelli. Spread of a novel influenza a (H1N1) pandemic. Health Economics, Mitigation measures for pandemic virus via global airline transporta- 19(11):1345–1360, 2010. influenza in italy: An individual based tion. The New England Journal Of 47. Beate Sander, Azhar Nizam, Louis model considering different scenari- Medicine, 361(2):212–214, July P. Garrison, Maarten J. Postma, os. PLoS ONE, 3(3), 2008. PMID: 2009. M. Elizabeth Halloran, and Ira M. 18335060 PMCID: 2258437. 65. K. Khan, C. C. Freifeld, J. Wang, Longini. Economic Evaluation of 56. Pascal Crepey and Marc Barthelemy. S. R. Mekaru, D. Kossowsky, A. L. Influenza Pandemic Mitigation Detecting Robust Patterns in the Sonricker, W. Hu, J. Sears, A. Chan, Strategies in the United States Spread of Epidemics: A Case Study and J. S. Brownstein. Preparing for Using a Stochastic Microsimulation of Influenza in the United States infectious disease threats at mass Transmission Model. Value in and France. American Journal of gatherings: the case of the Vancouver Health, 12(2):226–233, 2009. Epidemiology, 166(11):1244 –1251, 2010 Olympic Winter Games. 48. Philip Blumenshine, Arthur December 2007. Canadian Medical Association Reingold, Susan Egerter, Robin 57. Benjamin Cowling, Lincoln Lau, Journal, 182(6):579–583, 2010. Mockenhaupt, Paula Braveman, and Peng Wu, Helen Wong, Vicky Fang, 66. World Health Organization. James Marks. Pandemic Influenza Steven Riley, and Hiroshi Nishiura. Communicable disease alert and re- Planning in the United States from Entry screening to delay local trans- sponse for mass gatherings. http:// a Health Disparities Perspective. mission of 2009 pandemic influenza www.who.int/csr/mass gathering/en/ Emerging Infectious Diseases, A (H1N1). BMC Infectious Diseases, index.html, June 2008. 14(5):709–715, May 2008. PMID: 10(1):82, 2010. 67. Christopher C. Blyth. Influenza 18439350 PMCID: 2600245. 58. Takemi Matsui, Yukiya Hakozaki, Outbreaks during World Youth Day 49. P.L. Ooi, F.Y.L. Lai, C.L. Low, Satoshi Suzuki, Takahiro Usui, 2008 Mass Gathering. Emerging R. Lin, C. Wong, M. Hibberd, Takehito Kato, Kousuke Hasegawa, Infectious Diseases, 16(5):809–815, and P.A. Tambyah. Clinical and Youhei Sugiyama, Masami Sugamata, 2010. Molecular Evidence for Transmission and Shigeto Abe. A novel screening 68. Richard J. Hatchett, Carter E. of Novel Influenza A (H1N1/2009) method for influenza patients using a Mecher, and Marc Lipsitch. Public on a Commercial Airplane. Archives newly developed non-contact screen- health interventions and epi- of internal medicine, 170(10):913– ing system. Journal of Infection, demic intensity during the 1918 915, 2010. 60(4):271–277, April 2010. influenza pandemic. Proceedings of 50. P. Mukherjee, P.L. Lim, A. Chow, T. 59. H. Sato, H. Nakada, R. Yamaguchi, the National Academy of Sciences, Barkham, E. Seow, M.K. Win, A. S. Imoto, S. Miyano, and M. Kami. 104(18):7582 –7587, May 2007. Chua, Y.S. Leo, and M.I.C. Chen. When should we intervene to control 69. Interim planning considerations Epidemiology of travel-associated the 2009 influenza A (H1N1) pan- for mass gatherins in the context of pandemic (H1N1) 2009 infection in demic? Euro surveillance: bulletin pandemic (h1n1) 2009 influenza. 116 patients, Singapore. Emerging europeen sur les maladies transmis- Technical report, World Health Infectious Diseases, 16(1):21–26, sibles= European communicable dis- Organization, November 2009. 2010. ease bulletin, 15(1), 2010.

Centre de collaboration nationale des maladies infectieuses 9 70. ZA Memish, SJN McNabb, F. H Van Oyen, C Gerard, E Robesyn, 79. Public Health Agency of Canada. Mahoney, F. Alrabiah, N. Marano, J Bots, M T Faidherbe, and F 2010-07-29. Public Health Guidance QA Ahmed, J. Mahjour, RA Hajjeh, Wuillaume. Community transmis- for Post Secondary and Boarding P. Formenty, FH Harmanci, et al. sion of influenza A (H1N1)v vi- Schools Regarding the Prevention Establishment of public health rus at a rock festival in Belgium, and Management of Influenza-Like- security in Saudi Arabia for the 2-5 July 2009. Euro Surveillance: Illness (ILI), including the Pandemic 2009 hajj in response to pandemic Bulletin Europ«een Sur Les (H1N1) 2009 Influenza Virus. influenza A H1N1. The Lancet, Maladies Transmissibles = European http://www.phac-aspc.gc.ca/alert- 374(9703):1786–1791, 2009. Communicable Disease Bulletin, alerte/h1n1/hp-ps/psili-eng.php . 71. Hoon Chin Lim, Jeffery Cutter, 14(31), August 2009. PMID: Date Acc Weng Kee Lim, Adrian Ee, Yoong 19660245. Cheong Wong, and Boon Keng Tay. 75. K. Khan, Z.A. Memish, A. Chabbra, The influenza A (H1N1-2009) expe- J. Liauw, W. Hu, D.A. Janes, J. Sears, rience at the inaugural Asian Youth J. Arino, M. Macdonald, F. Calderon, Games Singapore 2009: mass gather- et al. Global public health implica- ing during a developing pandemic. tions of a mass gathering in Mecca, British Journal of Sports Medicine, Saudi Arabia during the midst of an 44(7):528 –532, June 2010. influenza pandemic. Journal of travel 72. Shahul H. Ebrahim, Ziad A. Memish, medicine, 17(2):75–81, 2010. Timothy M. Uyeki, Tawfik A. M. 76. H. Qureshi, B.D. Gessner, D. Khoja, Nina Marano, and Scott J. N. Leboulleux, H. Hasan, S.E. Alam, McNabb. Pandemic H1N1 and the and L.H. Moulton. The incidence 2009 Hajj. Science, 326(5955):938– of vaccine preventable influenza-like 940, November 2009. illness and medication use among 73. G Loncarevic, L Payne, P Kon, V Pakistani pilgrims to the Haj in Saudi Petrovic, D Dimitrijevic, T Knezevic, Arabia. Vaccine, 18(26):2956–2962, S Medi´c,c, N Milic, J Nedelijkovi´K 2000. Seke, and D Coulombier. Public 77. Haitham El Bashir, Elizabeth health preparedness for two mass Haworth, Maria Zambon, Shuja gathering events in the context of Shafi, Jane Zuckerman, and Robert pandemic influenza (H1N1) 2009– Booy. Influenza among U.K. pilgrims Serbia, July 2009. Euro Surveillance: to Hajj, 2003. Emerging Infectious Bulletin Europeen Sur Les Diseases, 10(10): 1882–1883, MaladieTransmissibles = European October 2004. PMID: 15515248. Communicable Disease Bulletin, 78. P. Gautret, P. Parola, and P. Brouqui. 14(31), August 2009. PMID: Risk factors for H1N1 influenza 19660246. complications in 2009 Hajj pilgrims. 74. I Gutierrez, A Litzroth, S Hammadi, Lancet, 375(9710):199–200, 2010.

413-445 Ellice Aveune, Winnipeg, MB R3B 3P5 204.943.0051 [email protected] www.nccid.ca

Production of this document has been made possible through a financial contribution from the Public Health Agency of Canada. The views expressed herein do not necessarily represent the views of the Public Health Agency of Canada.

10 National Collaborating Centre for Infectious Diseases