1. Lessons for Europe from past pandemics and the North American experience so far Evolution of the pandemic of A(H1N1)v influenza European Centre for Disease Prevention and Control Based on a talk given on 11 May 2009 in Stockholm to ECDC’s Advisory Forum

2. 2 About this presentation This is an open-access ECDC Educational PowerPoint presentation arranged in modules for use by professional explaining about the new A(H1N1)v virus to other professionals and policy makers. The slides should always be viewed with their accompanying notes, and ‘cutting and pasting’ is not recommended. A number of the slides will change with time. The slides are updated at intervals and the user should periodically check for updates available on the ECDC website: http://ecdc.europa.eu/ Comments on the slides and the notes are very much welcomed to be sent to influenza@ecdc.europa.eu.influenza@ecdc.europa.eu Please state "Pandemic PowerPoints" in the subject line when writing to us. ECDC thanks the National Institute of Infectious Diseases, Japan, for the original work on Slide 3, and the Centers for Disease Control and Prevention, USA, for the original idea in Slide 27.

3. 3 Pandemics of influenza H7 H5 H9 * 1980 1997 Recorded new avian influenzas 19962002 1999 2003 195519651975198519952005 H1N1 H2N2 1889 Russian influenza H2N2 1957 Asian influenza H2N2 H3N2 1968 Hong Kong influenza H3N2 H3N8 1900 Old Hong Kong influenza H3N8 1918 Spanish influenza H1N1 19151925195519651975198519952005 18951905 2010 2015 2009 Novel influenza H1N1v Recorded human pandemic influenza (early sub-types inferred) Reproduced and adapted (2009) with permission of Dr Masato Tashiro, Director, Center for Influenza Virus Research, National Institute of Infectious Diseases (NIID), Japan. Animated slide: Press space bar H1N1 H1N1v

4. 4 The situation could be a lot worse for Europe! (Situation circa summer 2009)  A pandemic strain emerging in the Americas  Immediate virus sharing so rapid diagnostic and vaccines  Based on A(H1N1)v currently not that pathogenic  Some seeming residual immunity in a major large risk group  No known pathogenicity markers  Initially susceptible to oseltamivir  Good data and information coming out of North America  Arriving in Europe in the summer  Milder presentation initially A pandemic emerging in SE Asia Delayed virus sharing Based on a more pathogenic strain, e.g. A(H5N1) No residual immunity Heightened pathogenicity Inbuilt antiviral resistance Minimal data until transmission reached Europe Arriving in the late autumn or winter Severe presentation immediately Contrast with what might have happened — and might still happen!

5. 5 But no room for complacency (Situation and information: late May 2009)  Pandemics take some time to get going (1918 and 1968).  Some pandemic viruses have ‘turned nasty’ (1918 and 1968).  Is the ‘mildness’ and the lack of older patients because older people are resistant or because the virus is not transmitting much among them?  There will be victims and deaths — as in the US — in risk groups (young children, pregnant women and especially people with other underlying illnesses).  As the virus spreads south, will it exchange genes with seasonal viruses that are resistant: A(H1N1)-H247Y, more pathogenic A(H3N2), or even highly pathogenic A(H5N1)?  An inappropriate and excessive response to the pandemic could be worse than the pandemic itself.

6. 6 Idealised curve for planning Single wave profile showing proportion of new clinical cases, consultations, hospitalisations or deaths by week. Based on London, 2nd wave 1918. 0% 5% 10% 15% 20% 25% 123456789101112131415 Week Proportion of total cases, consultations, hospitalisations or de aths Source: Department of Health, UK InitiationAccelerationPeakDeclining Animated slide: Please wait

7. 7 One possible European scenario — summer 2009 In reality, the initiation phase can be prolonged, especially in the summer months. What cannot be determined is when acceleration takes place. 0% 5% 10% 15% 20% 25% AprMayJunJulAugSepOctNovDecJanFebMar Month Proportion of total cases, consultations, hospitalisations or deaths InitiationAccelerationPeakDeclining Animated slide: Please wait Apr

8. 8 How pandemics differ — and why they can be difficult

9. 9 For any future pandemic virus – what can and cannot be assumed? What probably can be assumed: Known knowns  Modes of transmission (droplet, direct and indirect contact)  Broad incubation period and serial interval  At what stage a person is infectious  Broad clinical presentation and case definition (what influenza looks like)  The general effectiveness of personal hygiene measures (frequent hand washing, using tissues properly, staying at home when you get ill)  That in temperate zones transmission will be lower in the spring and summer than in the autumn and winter What cannot be assumed: Known unknowns  Antigenic type and phenotype  Susceptibility/resistance to antivirals  Age-groups and clinical groups most affected  Age-groups with most transmission  Clinical attack rates  Pathogenicity (case-fatality rates)  ‘Severity’ of the pandemic  Precise parameters needed for modelling and forecasting (serial interval, R o )  Precise clinical case definition  The duration, shape, number and tempo of the waves of infection  Will new virus dominate over seasonal type A influenza?  Complicating conditions (super-infections)  The effectiveness of interventions and counter-measures including pharmaceuticals  The safety of pharmaceutical interventions

10. 10 Some of the 'known unknowns' in the 20th century pandemics  Three pandemics (1918, 1957, 1968)  Each quite different in shape and waves  Some differences in effective reproductive number  Different groups affected  Different levels of severity including case fatality ratio  Imply different approaches to mitigation

11. 11 0% 10% 20% 30% 40% 50% 60% 020406080 Age (midpoint of age class) % with clinical disease 1918 New York State 1918 Manchester 1918 Leicester 1918 Warrington & Wigan 1957 SE London 1957 S Wales 1957 Kansas City 1968 Kansas City With thanks to Peter Grove, Department of Health, London, UK Age-specific clinical attack rate in previous pandemics Animated slide: Press space bar

12. 12 Different age-specific excess deaths in pandemics 0 2000 4000 6000 8000 10000 12000 14000 16000 0-45-910-1415-1920-2425-3435-4445-5455-6465-7475+ Age group Excess deaths 0 500 1000 1500 2000 2500 3000 3500 4000 <11-22-55-1010-1515-2020-2525-3535-4545-5555-6565-7575+ Age group Excess deaths Excess deaths, second wave, 1918 epidemic Excess deaths second wave 1969 pandemic, England and Wales Source: Department of Health, UK

13. 13 1918/1919 pandemic: A(H1N1) influenza deaths, England and Wales 1918/19: ‘Influenza deaths’, England and Wales. The pandemic affected young adults, the very young and older age groups. R o = 2-3 (US) Mills, Robins, Lipsitch (Nature 2004) R o = 1.5-2 (UK) Gani et al (EID 2005) R o = 1.5-1.8 (UK) Hall et al (Epidemiol. Infect. 2006) R o = 1.5-3.7 (Geneva) Chowell et al (Vaccine 2006) Courtesy of the Health Protection Agency, UK Transmissibility: estimated Basic Reproductive Number (R o )

14. 14 Estimated additional deaths in Europe if a 1918/19 pandemic occurred now – a published worst case scenario Austria13,000Latvia13,800Netherlands23,100 Belgium14,900Lithuania18,800Poland155,200 Bulgaria47,100Germany116,400Portugal25,100 Czech Rep34,100Greece27,400Romania149,900 Cyprus1,900Hungary37,700Slovenia5,000 Denmark7,300Ireland6,700Slovakia20,600 Estonia6,100Italy95,200Spain87,100 Finland8,100Luxembourg500Sweden13,300 France89,600Malta1,100UK93,000 Iceland420Norway5,800 EU total: 1.1 million Murray CJL, Lopez AD, Chin B, Feehan D, Hill KH. Estimation of potential global pandemic influenza mortality on the basis of vital registry data from the 1918–20 pandemic: a quantitative analysis. Lancet. 2006;368: 2211-2218.

15. 15 1957/1958 pandemic: A(H2N2) — especially transmitted among children R o = 1.8 (UK) Vynnycky, Edmunds (Epidemiol. Infect.2007) R o = 1.65 (UK) Gani et al (EID 2005) R o = 1.5 (UK) Hall et al (Epidemiol. Infect. 2006) R o = 1.68 Longini et al (Am J Epidem 2004) 0 200 400 600 800 1,000 6 132027 3 10172431 7 142128 5 121926 29 162330 7 142128 4 111825 18 1522 JulyAugustSeptemberOctoberNovemberDecemberJanuaryFebruary Week number and month during the winter of 1957/58 Recorded deaths in England and Wales from influenza 1957/58: ‘Influenza deaths’, England and Wales Courtesy of the Health Protection Agency, UK Transmissibility: estimated Basic Reproductive Number (R o )

16. 16 1968/1969 pandemic: A(H3N2) — transmitted and affected all age groups R o = 1.5-2.2 (World) Cooper et al (PLoS Med.2006) R o = 2.2 (UK) Gani et al (EID 2005) R o = 1.3-1.6 (UK) Hall et al (Epidemiol. Infect. 2006) 1968/69: GP consultations, England and Wales Courtesy of the Health Protection Agency, UK Initial appearance Seasonal influenza Transmissibility: estimated Basic Reproductive Number (R o )

17. 17 Differing attack rates determined by serology: serological attack rate observed in the UK 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0-910-1920-2930-3940-4950-5960-6970-79 1969 (first wave)1970 (second wave)1957 Courtesy of the Health Protection Agency, UK

18. 18 Idealised curves for local planning In reality, larger countries can experience a series of shorter but steeper local epidemics. 0% 5% 10% 15% 20% 25% 123456789101112131415 Week Proportion of total cases, consultations, hospitalisations or de aths Animated slide: Press space bar

19. 19 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1918 New York State 1918 Leicester 1918 Warrington and Wigan 1957 SE London 1968 Kansas City clinical attack rate (%) Numbers affected in seasonal influenza epidemics and pandemics (overall clinical attack rate in previous pandemics) Seasonal influenza

20. 20 Seasonal influenza compared to pandemic — proportions of types of cases Asymptomatic Clinical symptoms Deaths Requiring hospitalisation Seasonal influenza Pandemic Asymptomatic Clinical symptoms Deaths Requiring hospitalisation

21. 21 Initial experience in North America 2009

22. 22 Emerging themes in North America, early June 2009 (1)  Early epidemic: –increased influenza-like illness reports due to increased consultations; –many cases attributable to seasonal influenza until mid-May.  Infection rate for probable and confirmed cases highest in 5−24 year age group.  Hospitalisation rate highest in 0−4 year age group, followed by 5−24 year age group. –Pregnant women, some of whom have delivered prematurely, have received particular attention but data inadequate to determine if they are at greater risk from H1N1v than from seasonal influenza as already established.  Most deaths in 25−64 year age group; most with known risks for severe disease. –Obesity suggested as risk but may be indicator for pulmonary risk.  Adults, especially 60 years and old, may have some degree of preexisting cross-reactive antibody to the novel H1N1 flu virus.  Transmission persists in several regions of the US with increased or rising incidence in New York area and northeastern US.

23. 23 Emerging themes in North America, early June 2009 (2)  Containment impossible with multiple introductions and R 0 1.4 to 1.6.  Focus on counting laboratory-confirmed cases changing to seasonal surveillance methods. –Outpatient influenza-like illness, virological surveillance (including susceptibility), pneumonia and influenza mortality, pediatric mortality and geographic spread.  Serological experiments and epidemiology suggest 2008–2009 seasonal A(H1N1) vaccine does not provide protection.  Preparing for the autumn and winter when virus is expected to return: –communications: a pandemic may be 'mild' yet cause deaths; –25% of U.S. stockpile deployed to states (includes medication and equipment); –determining if and when to begin using vaccine; –school closures being analyzed to determine effectiveness; –other domestic and international investigations of public health questions.

24. 24 Measuring the severity of a pandemic

25. 25 There is an expectation that pandemics should be graded by severity But there are difficulties:  severity varies from country to country;  it can change over time;  some relevant information is not available initially;  key health information includes medical and scientific information: –epidemiological, clinical and virological characteristics.  There are also social and societal aspects: –vulnerability of populations; –capacity for response; –available health care; –communication; and –the level of advance planning.

26. 26 What is meant by 'mild' and 'severe'? Not a simple scale  Death ratio. Expectation of an infected person dying (the Case Fatality Ratio).  Number of people falling ill with respiratory illnesses at one time — 'winter pressures'. Pressure on the health services' ability to deal with these — very related to preparedness and robustness.  Critical service functioning. Peak prevalence of people off ill or caring for others.  Certain groups dying unexpectedly, e.g. children, pregnant women, young healthy adults.  Public and media perception  Conclusions. Not easy to come up with a single measure.  May be better to state what interventions/countermeasures are useful and justifiable (and what are not). http://www.who.int/csr/disease/swineflu/assess/disease_swineflu_assess_20090511/en/index.html and http://www.who.int/wer/2009/wer8422.pdf

27. 27 Arguments for and against just undertaking mitigation and not attempting delaying or containment

28. 28 Policy dilemma — mitigating vs. attempting delaying (containing) pandemics? Arguments for just mitigating and not attempting delaying or containment:  Containment specifically not recommended by WHO in Phases 5 & 6.  Was not attempted by the United States for this virus.  Delaying or containment cannot be demonstrated to have worked — would have seemed to have worked in 1918 and 1968 without doing anything.  Very labour-intensive — major opportunity costs.  Will miss detecting sporadic transmissions.  Overwhelming numbers as other countries ‘light up’.  When you change tactic, major communication challenge with stopping prophylaxis.

29. 29 Policy dilemma — mitigating vs. attempting delaying (containing) pandemics? Arguments for case-finding, contact tracing and prophylaxis:  Countries are then seen to be doing something.  Recommended in one specific circumstance by WHO (the rapid containment strategy).  There are some places it would work in Europe (isolated communities).  It is what public health people do for other infections.  Public may expect it.

30. 30 With interventions Aims of community reduction of influenza transmission — mitigation  Delay and flatten epidemic peak  Reduce peak burden on healthcare system and threat  Somewhat reduce total number of cases  Buy a little time Daily cases Days since first case No intervention Animated slide: Press space bar Based on an original graph developed by the US CDC, Atlanta