AbstractBackground: Australian Indigenous children continue to suffer high rates of pneumonia despite the introduction of conjugate pneumococcal vaccines. Evidence is emerging of declining hospitalisations for pneumonia among Indigenous children in Western Australia, though a disparity between Indigenous and non-Indigenous children remains . A large retrospective study found evidence of a lack of impact of 7-valent pneumococcal conjugate vaccine on radiological pneumonia among Northern Territory Indigenous children . Existing influenza surveillance systems in the Northern Territory show consistently high influenza incidence rates among Indigenous children. Australia has been gripped in a debate regarding universal annual seasonal influenza vaccination. Further informing this debate requires the potential identication of specic population groups that may be targetted for vaccination, the question of vaccine cost-effectiveness, and the risk of vaccine associated adverse events. This study asks:
• What is the population based incidence rate ratio of hospitalised radiological pneumonia between Indigenous and non-Indigenous children in the Top End of the Northern Territory?
• What are the risk factors for radiological pneumonia among hospitalised Indigenous and non-Indigenous children?• What is the population based incidence rate ratio of hospitalised influenza infection between Indigenous and non-Indigenous children in the Top End of the Northern Territory?
• What is the contribution of influenza and other respiratory viruses to the burden of radiological pneumonia among hospitalised Indigenous and non-Indigenous children?
• What is the pneumococcal serotype specific nasopharyngeal carriage rate among hospitalised Indigenous and non-Indigenous children, and what are the risk factors for such carriage?
• What is the association between pneumococcal carriage, influenza co-infection and radiological pneumonia?
• What is the cost of hospitalisation for respiratory illness among Indigenous and non-Indigenous children?
Methods: Using prospective hospital based surveillance we identied a cohort of children presenting with World Health Organization (WHO) dened radiological pneumonia and children with laboratory proven influenza. We collected clinical, laboratory, radiological, demographic and cost related data. Population based incidence rates were calculated with reference to several population datasets for the Top End of the Northern Territory of Australia. Incidence rate ratios between Indigenous and non-Indigenous children were calculated. Association between exposures and outcomes of interest were calculated as the odds ratio of exposure between children with and without the outcomes of interest using a case-control design. Extensive sensitivity analyses of the choice cases and controls were conducted. Costs were analysed from a healthcare provider perspective, using patient specific ground-up costing methods and cost predictive models were developed.
Results: Indigenous infants continue to have higher rates of pneumonia than non-Indigenous infants in the vaccine era despite high vaccine uptake in both groups, though contrary to previous work  we were able to demonstrate protective effectiveness from pneumococcal conjugate vaccine. Risk factors for pneumonia include Indigenous status and remote dwelling, though the relationship between these two factors is complex. There is some evidence for increased risk of radiological pneumonia from childcare attendance, but only among non-Indigenous children. Indigenous children are more likely to have younger mothers, to live remotely, to be exposed to smoke in antenatal and postnatal life. They are less likely to attend childcare. Indigenous and non-Indigenous infants have similar rates of vaccine uptake, and are similarly exclusively breastfed at least to age 3 months.
Annual incidence of hospitalised influenza per 100,000 population was 366.4 among Indigenous and 40.4 among non-Indigenous infants respectively, giving IRR 9.1 (95%CI: 3.0 to 36.5, p<0.01). Indigenous children have more comorbidity, more exposure to environmental tobacco smoke and are more likely to be colonised with pneumococcus, none of these factors was itself associated with influenza infection. The presence of any comorbidity more than doubled the risk of influenza but that association did not reach statistical signicance. We were unable to show an association between influenza and radiological pneumonia or severe respiratory disease, though this study was not powered for theseendpoints.
Indigenous status is a strong risk factor for pneumococcal colonisation. Remoteness does not contribute to the risk of pneumococcal colonisation over and above the risk contributed through Indigenous status. Urban Indigenous children are just as likely to be colonised with pneumococcus as remote dwelling Indigenous children. Different factors may act as risks for pneumococcal carriage for Indigenous and non-Indigenous children. Breastfeeding was not associated with pneumococcal colonisation overall but conferred different magnitude of association across ethnic groups. childcare attendance was associated with pneumococcal colonisation but only among non-Indigenous children.
Indigenous children have more costly hospital admissions, this finding is consistent even after adjusting for remote status. Indigenous children cost more not because they are Indigenous and not because they live far away but because they are sicker. Children with influenza cost more than children with other common respiratory viruses such as RSV.
Conclusions: Indigenous children in the Top End of the Northern Territory have higher rates of radiological pneumonia, higher rates of nasopharyngeal pneumococcal colonisation and higher rates of hospitalised influenza infection than non-Indigenous children. In addition they have higher prevalence of comorbid conditions that indicate influenza vaccination according to current Australian guidelines. Indigenous children also have greater exposure to other risk factors for respiratory illness such as environmental tobacco smoke, than non-Indigenous children. The hospitalisation of Indigenous children is more costly, likely because of the severity of clinical presentation. Indigenous children in the Top End should be considered for targeted annual seasonal influenza vaccination.
|Date of Award||2013|
|Supervisor||Jonathan Carapetis (Supervisor) & Ross Andrews (Supervisor)|