Pandemic influenza in Papua New Guinea: a modelling study comparison with pandemic spread in a developed country

Milne, George J; Baskaran, Pravin; Halder, Nilimesh; Karl, Stephan; Kelso, Joel

It has been shown under laboratory conditions that the highly pathogenic H5N1 avian influenza strain may re-assort or mutate to become readily transmissible between humans. [2] In response to the threat of influenza pandemics, most countries have developed pandemic preparedness plans. [3, 4, 5, 6, 7, 8] . However the level of completeness of these plans with regards to World Health Organization (WHO) guidelines [9] varies dramatically between developed high income countries and developing, low-income countries. [7, 8] Low-income countries suffer the highest rates of morbidity and mortality from influenza pandemics. Data from the 1918/19 pandemic suggests much higher mortality rates in India compared to Europe and the USA. While in northern Europe and the USA mortality rates were consistently below 1%, over 4.3% average mortality was reported from India, with some Indian provinces exceeding mortality rates of 7%. [10, 11] Even in the mild H1N1 pandemic of 2009, 59% of the estimated 284,500 deaths occurred in Africa and South-East Asia. [2] These differences are most likely due to the significant association of influenza mortality with comorbidities usually absent in the developed countries (such as chronic malnutrition, malaria, tuberculosis, pneumonia and HIV). [12] Furthermore it has been shown that the increased mortality rate due to an influenza pandemic is strongly associated with per capita income. [11] It is apparent that the burden of the next pandemic will also be focused overwhelmingly on the developing world, a point made by the WHO and others. [5, 11, 13, 14, 15] A substantial number of pandemic influenza models have been constructed in the past decade. A key goal of this effort is to determine the effectiveness of interventions, in terms of their ability to reduce infection and consequential mortality rates. Although it has been recognized that the major burden of mortality from influenza pandemics is suffered by low-income countries [5, 11, 16] the primary focus of the vast majority of these models is on examining intervention effectiveness in developed (e.g., the USA and European countries [17, 18, 19, 20] ) and middle income countries (e.g., Thailand [21, 22] ). Given the markedly different demographic and community structure, as well as resource and infrastructural constraints in comparison with developed countries, these models have limited capability to predict the impact of an influenza pandemic on low-income countries [23] . In addition, many low-income countries are located in tropical or subtropical regions where the timing and impact of influenza epidemics are more poorly understood than in temperate regions, where they are known to have a well-defined winter season. [15] To address this 'modelling gap' an individual-based disease transmission model of pandemic influenza has been developed for Papua New Guinea (PNG), based on the methods used previously to develop a pandemic influenza model of Albany, Western The model was used to describe the spread of a novel influenza strain after introduction into the community. The resulting illness attack rate was determined together with data indicating where person-to-person transmission occurred. An unmitigated outbreak and two outbreaks with PNG-feasible intervention strategies activated were simulated and the results compared to those obtained using the Albany, Australia model.