Effects on the Australian economy of a moderate and severe H1N1 epidemic George Verikios Centre of Policy Studies, Monash University James McCaw Murdoch Childrens Research Institute and School of Population Health, The University of Melbourne Tony Harris Centre of Health Economics, Monash University
Introduction I Early 2009: emergence of H1N1 epidemic June 2009: global pandemic (WHO) Jan 2010: –H1N1 activity had peaked in most regions of the world but pandemic is ongoing –In Australia, 37,562 confirmed cases, 191 deaths, but H1N1 activity is low Confirmed cases per 100,000 population suggest swine flu in Australia was more severe than many other regions
Introduction II Hospitalisations per 100,000 population also higher than most other regions –Of these, ICU admissions higher than expected (13%) SARS 2003: large temporary economic effects on China, Hong Kong, Taiwan, etc What about 2009 swine flu in Australia? –MONASH-Health
Model: MONASH-Health I Based on the MONASH model of the Australian economy (Dixon & Rimmer 2002) –dynamic (annual) baseline (forecast) policy (forecast + swine flu) deviation (effects of swine flu) –detailed (100+ industries/commodities) –computable general equilibrium household consumption government consumption investment exports imports foreign assets/liabilities –model outputs: macro and micro variables
Model: MONASH-Health II A simplified representation of the MONASH input-output table
Model: MONASH-Health III Health treatment activities in MONASH-Health Medical services in MONASH-Health
Model: MONASH-Health IV Input-output structure of health treatment activities in MONASH-Health Sales structure of health treatment activities in MONASH-Health
MONASH-Health V quarterly capital idling (Dixon & Rimmer 2009) –rental rates are sticky endogenous unemployment –real wage rates are sticky => lagged adjustment to shocks
Model shocks possible economic effects of a major epidemic: –temporary reductions in inbound and outbound international tourism and business travel; –temporary upsurge in sick leave and widespread school closures requiring withdrawal of parents from the labour force; –large surge in demand for hospital and other medical services; –some deaths with a related permanent reduction in the labour force; –temporary cessation of large public gatherings (sporting events, etc).
Swine flu scenarios outbreak 2.A severe outbreak epidemiological aspects of scenarios constructed using SEIR model of infectious disease transmission (Kermack & McKendrick 1927)
SIR model of infectious disease Between S and I, the transition rate is β I, where β is the contact rate, which -roughly speaking - takes into the account the probability of getting the disease in a contact between a susceptible and an infectious subject. Between I and R, the transition rate is ν (simply the rate of recovery). If the duration of the infection is denoted D, then ν = 1/D, since an individual experiences one recovery in D units of time. Assume that the permanence of each single subject in the epidemic states is a random variable with exponential distribution.
SIR system The SIR system can be expressed by a set of ordinary differential equations The key to the dynamics of an epidemic is the ratio β/v – the reproduction ratio In this simulation we take observed confirmed cases and work back to the number infected. The cumulative number of cases that are admitted to hospital and ICU and that die are based on publicly available data but the exact timing is modelled given a typical β for influenza
2009 outbreak I 4.3m Australians infected and experience symptoms in 2009Q3 & 2009Q4 –3.9m seek no med attention but spend $5 on drugs –0.44mm seek medical attention but aren’t hospitalised ($61 per capita) –4,500 hospitalised ($3,564 per capita) –857 hospitalised in an ICU ($12,356 per capita) 670 of these survive, 187 die $73m increase in medical expenses over 2009Q3 & 2009Q4; 4.1% increase in demand for respiratory treatments over 2009Q3 & 2009Q4 Medical expenditure returns to baseline in 2010Q1 Demand-contracting effect
2009 outbreak II Workers miss –3.2m workdays over 2009Q3 & Q4 due to illness –148,000 workdays caring for children who are sick or at home due to school closures, or are absent from work due to own-illness Equivalent to 0.48% fall in labour productivity Labour productivity returns to normal in 2010Q1 Cost-increasing effect
2009 outbreak III Of the 187 deaths, 126 are workers Permanent reduction in workforce of 0.001% over 2009Q3 & Q4 Supply-reducing effect
2009 outbreak IV During 2009Q3, inbound/outbound tourism fall by 7.9% (Dwyer et al. 2006) During 2009Q4, inbound/outbound tourism fall by 1.2% (Dwyer et al. 2006) Tourism recovers smoothly to basecase over 2010Q1 and 2010Q2 Cancelled outbound tourism expenditures are saved Demand-contracting effect
2009 outbreak: results I (percentage deviations from baseline)
2009 outbreak: results II (percentage deviations from baseline)
2009 outbreak: results III Effects of individual shocks on aggregate employment (percentage deviations from baseline)
Severe outbreak I $1.6b increase in medical expenses over 2009Q3 & 2009Q4; 31% increase in demand for respiratory treatments over 2009Q3 & 2009Q4 Medical expenditure returns to baseline in 2010Q1
Severe outbreak II Permanent reduction in workforce of 0.32% over 2009Q3 & Q4 medical expenditure returns to baseline in 2010Q1 1.7% fall in labour productivity Labour productivity returns to normal in 2010Q1
Severe outbreak III During 2009Q3, inbound/outbound tourism fall by 65% (Pine & McKercher 2004; Wilder-Smith 2006) Tourism recovers smoothly to basecase over next 4 quarters Cancelled outbound tourism expenditures are saved
Severe outbreak: results I (percentage deviations from baseline)
Severe outbreak: results II (percentage deviations from baseline)
Severe outbreak: results III Effects of individual shocks on aggregate employment (percentage deviations from baseline)
Conclusion H1N1 epidemics are very short term: need quarterly model Significant macroeconomic effects: short and sharp –2009 outbreak: return to baseline in 4 quarters –severe outbreak: return to baseline in 5 quarters limitations of severe scenario –capacity constraints for hospitals –change in risk-modifying behaviour –policy responses: vaccination; prohylactic anti-viral medications and their costs