Air pollution in Hanoi, Vietnam : evaluating effects on hospital admissions of children

Background
Worldwide, pneumonia is responsible for approximately 1.4 million paediatric deaths per year. The prevalence of hospitalization due to pneumonia has been increasing globally, contributing to a large proportion of Disability-Adjusted Life Years (DALY). About 75% of the burden of disease occurs in low-and middle-income countries. In Vietnam, pneumonia is one of the leading causes of childhood hospitalization and deaths, thus, there is a need to determine optimal interventions for preventing childhood pneumonia. To do this, more information about potential risk factors is needed.
Air pollution, particularly outdoor air pollution, is known to be among the leading causes of pneumonia. Evidence of the link between outdoor air pollution and the risk of hospitalization and/or premature death due to pneumonia can be found in numerous epidemiological studies in high income countries. In developing countries, like Vietnam, such data is still sparse. This lack of sufficient specific evidence hinders the public health sector from implementing and promoting different interventions against air pollution in these countries.
The overall aim of the present PhD research was to investigate the short-term effects of ambient air pollution on the daily number of hospital admissions of children and adolescents in Vietnam. Specifically, the work aimed to a) study pneumonia rate trends in Vietnam; b) investigate the existing evidence linking pneumonia hospitalization and outdoor air pollution, globally; and to c) estimate the association between outdoor air pollution and the daily number of hospital admissions of Vietnamese children.
Method
Data source
Daily means of pollutants, including particulate matter less than 10, 2.5 and 1 microns in diameter (PM10, PM2.5, and PM1, respectively), sulfur dioxide (SO2), nitrogen dioxide (NO2), nitrogen oxides (NOx), carbon monoxide (CO) and 8-hour moving average of ozone (O3) were derived from fixed monitoring stations in Hanoi, namely Lang, Nguyen Van Cu. All pollutants were expressed in µg/m3. Daily means for temperature (0C), relative humidity (%) and wind speed (m/s) were obtained by averaging values from four stations: Lang, Ba Vi, Son Tay and Ha Dong.
Study approach
Data from the Vietnam National Children’s Hospital, covering the period 2007–2014, were analysed to calculate the emergency visit and hospital admissions rates caused by pneumonia in Hanoi’s paediatrics unit.
The short-term association between ambient air pollution on hospital admissions of children was determined using two broad approaches: (i) a literature review and meta-analysis of the peer-review publication up to 2017, and (ii) an epidemiological study based on information from the hospital database in Hanoi, Vietnam.
First, the review study summarized existing evidence of the relationship between exposure to air pollution and hospitalization for pneumonia. Then, a time-series analysis with quasi-Poison regression was conducted to investigate the association between ambient air pollution and daily hospital admissions of Hanoi children during the period 2007–2014.
In the summary below, the associations between air pollutants and outcome are presented in percent change (excess risk-ER%).
Results
Among the respiratory diseases affecting Hanoi children, pneumonia was responsible for 24.5% of emergency visits (2012–2014), and 54.1% of hospital admissions (2007–2014). Infants made up the group most frequently hospitalized due to pneumonia. Moreover, pneumonia was also the biggest cause of both prolonged hospitalization time and deaths in the hospital among Hanoi children. Rates of infant hospitalization due to pneumonia increased from 19.2 per 1000 children in 2009 to 32.4 per 1000 children in 2014.
Seventeen studies reporting the association between hospitalization for pneumonia and ambient air pollution from Web of Science and PubMed were derived and summarized. Eight of these studies were conducted in high income countries. Meta-analysis using random effect models indicated the positive association between air pollutants and hospital admissions of children worldwide, except CO. The excess risk in percent (ER%) of every 10µg/m3 increment of PM10 and PM2.5 were 1.5% (95% confidence intervals (CI): 0.6% to 2.4%) and 1.8% (95% CI: 0.5% to 3.1%), respectively. The ER% of 1000ppb increment of CO, however, was 0.9% (95%CI: 0.0% to 1.9%).
Time series analyses showed positive associations between ambient air pollution and daily numbers of children (aged 0-17) admitted to hospital for pneumonia in Hanoi. The strongest effect was observed for NO2; an interquartile increase in seven-day average NO2 (21.9µg/m3) resulted in a 6.1% (95%CI: 2.5% to 9.8%) increase of hospital admissions for pneumonia. This was followed by PM10, whereby the ER% per 66.5 µg/m3 increase in seven-day average concentration of PM10 in Hanoi was a 5.8% (95%CI: 2.8% to 9.0%) increase in hospitalizations for pneumonia. The effects of NO2 were stable across both pollutant models. The effects of PM10, however, were attenuated upon inclusion of O3. Stronger effects among children aged 1–5 years were observed, compared with infants. The findings also indicate a seasonal influence on the association between ozone and hospitalization due pediatric pneumonia for all ages (0-17).
Conclusion
This work provides evidence of the increasing trend of emergency and hospital admission rates among Vietnamese children. Pneumonia infection accounted for a large proportion of hospital admissions, prolonged hospitalization and was a primary cause of deaths at hospital among Vietnamese children.
The findings from the meta-analysis concluded that studies from North America and Europe document associations between exposure to ambient air pollution and pneumonia. This study also suggests that time series analysis is an easy and cost effective study approach for investigating the effects of ambient air pollution, as it uses routine data (i.e. hospital records). It is, thus, an appropriate method to apply in developing countries, such as Vietnam.
The time series study in Hanoi has demonstrated the link between air pollution and hospitalization for pneumonia among Vietnamese children, in which all pollutants except SO2 and O3 were found to have significant associations with an increase in childhood hospitalizations. It is also recommended that NO2 and PM10 levels are used as predictors for hospital admission for pneumonia in Vietnam