Epidemiology of respiratory syncytial virus associated acute lower respiratory infection in young children
Introduction Acute lower respiratory infection (ALRI) remains as a leading cause of childhood morbidity and mortality. With the continued universal vaccination campaign against bacterial pathogens, an increase in relative proportion of respiratory viruses contributing to ALRI is anticipated. Respiratory syncytial virus (RSV) has been recognised as the most common pathogen identified in young children presenting with ALRI as well as an important cause of hospital admission. This thesis aims to estimate the aetiological roles and attributable fractions of common respiratory viruses among ALRI cases and investigate the risk factors for RSV associated ALRI in young children. It also aims to estimate the global and regional incidence of RSV associated ALRI in both community and hospital based settings, and the possible boundaries for RSV associated ALRI mortality in children younger than five years old. Methods Systematic reviews were carried out separately for the following three research questions: aetiological roles of RSV and other common viruses in ALRI cases, risk factors for RSV associated ALRI and global/regional burden of RSV associated ALRI, formulating an overall picture of epidemiology of RSV associated ALRI in young children. They all focused on children younger than five years old. The identified studies were selected according to pre-defined inclusion and exclusion criteria. The whole process was conducted following the PRISMA guidelines for systematic review and meta-analysis. Unpublished data from RSV Global Estimates Network (RSV GEN) were collected from 45 leading researchers on paediatric pneumonia (primarily in developing countries). They either reanalysed data from their already published work with the pre-defined standardised case definitions or shared hitherto unpublished data from ongoing studies. Data from both systematic reviews and RSV GEN working group were included into further meta-analysis. Random effects model was consistently applied in all meta-analyses. Results There were 23 studies identified through literature search satisfying the eligibility criteria, investigated the viral aetiology of ALRI in young children. Strong evidence was observed for RSV in support of its causal contribution in children presenting with ALRI and the association was significant measured in odds ratio: 9.79 (4.98-19.27). Thus, the corresponding attributable fraction among the exposed was estimated as 90% (80%-95%), which means around 90% of RSV associated ALRI cases were in fact attributed to RSV in a causal path. In total, 27 studies (including 4 unpublished studies) were included and contributed to the analysis. Across these studies, 18 risk factors were described and 8 of them were observed to have significant associations with RSV infection: prematurity - gestational age <37 weeks, low birth weight (<2.5 kg), being male, having siblings, maternal smoking, history of atopy, no breastfeeding and crowding - >7 persons in household. Overall, 304 studies met the selection criteria and were included to estimate the global and regional burden of RSV associated ALRI in young children. These included 73 published articles identified through Chinese language databases and 76 unpublished studies provided by RSV GEN working group, mainly from developing countries. It is estimated that in 2015, there were 33.0 (95% CI 20.6-53.2) million episodes of RSV associated ALRI occurring in children younger than 5 years old across the world. 30.5 (95% CI 19.5-47.9) million of them were in developing countries. 3.0 (95% CI 2.2-4.0) million cases were severe enough and warranted hospitalisation. Around 60,000 children died in the hospital settings with 99% of these deaths occurring in developing countries. The overall mortality from RSV associated ALRI was estimated about 131,000. Conclusion This thesis not only enhanced the epidemiological understanding of RSV in young children, but also provided important information for public health decision makers. It incorporated both data through systematic reviews of published articles in the past 20 years and more than 70 unpublished data sets shared by RSV GEN working group. The population based incidence, hospitalisation, mortality and risk factor data are essential to assess the various severity of illness in a specific age group and region, and inform local public health professionals regarding appropriate and prompt cases management, prevention and vaccine allocation strategies. National sentinel systems of RSV surveillance gathering structured and reasonably representative data are needed. Within the surveillance system, a universal definition regarding disease severity in various settings should be developed, and diagnostic methods with higher sensitivity and specificity should be applied.