Download:
-
Nitrogen dioxide (NO2) has been widely associated with respiratory morbidity and mortality. However, its association with hospitalization expenditure and length of hospital stay (LOS) is scarcely analyzed. Thus, we analyzed this association in Shanxi Province, which has famously high fuel combustion with residential and industrial regions centralized in basins and valleys, making it difficult for air pollutants to disperse (1). Daily air pollution exposures were estimated for each case using an inverse distance approach and then averaged for each city. City-specific associations of NO2 with hospital admission, hospitalization expenditure, and LOS were assessed by general additive models. The overall effects were then pooled through a random-effects meta-analysis. The potentially avoidable values and population attributable fraction (PAF) for total respiratory disease, asthma and chronic obstructive pulmonary disease (COPD) were assessed by reducing the concentration to the standards of the World Health Organization (WHO) (25 µg/m3) and Chinese government (40 µg/m3). Significant associations between NO2 and respiratory hospitalization, hospital cost, and LOS were observed. We estimated 198.73 million CNY [95% confidence interval (CI): 119.07–278.40] in hospital costs, and 272.00 thousand days (95% CI: 179.22–364.79) of LOS could be potentially avoided under the WHO’s standards for Shanxi during 2017–2019. This study provides evidence for the association between NO2 exposure and respiratory burden, suggesting that continually diminishing ambient NO2 could effectively reduce respiratory burden.
This study collected hospital admission data for respiratory diseases from all the secondary and tertiary hospitals in Shanxi Province from February 2017 to November 2019. The information included general characteristics, residential address, date of admission and discharge, the principal diagnosis (based on the International Classification of Diseases, ICD-10), and hospital expenditures. The hospital expenditures included fees for drugs, clinical examinations, nursing care, and clinical operations. The LOS was calculated based on the admission date, and the LOS for each case was summed and regarded as the value for that day. We included the participants who had resided in Shanxi Province and excluded cases with missing information and those who had only received emergency or outpatient treatment and were then discharged.
Daily ambient air pollutant concentrations of NO2, fine particulate matter, sulfur dioxide, and ozone at all state-controlled air quality monitoring stations in Shanxi from 2017 to 2019 were also obtained. Individual level exposure was assessed using reverse distance weighting (IDW) by geocoding patient addresses and monitoring station locations and averaging an individual’s daily exposure for each city (1-2).
The daily mean temperature and relative humidity were extracted from ERA5-L and reanalysis data (3). A bilinear interpolation approach was used to estimate the daily meteorological exposure for each participant, and then average the exposure for each city.
The city-specific and overall associations between NO2 and hospitalization, hospital cost, and LOS for respiratory diseases, COPD, and asthma were assessed using a two-stage statistical approach based on the time-series model. At the first stage, a generalized additive model (GAM) with a quasi-Poisson was built to explore the effect of NO2 on daily hospital admissions for each city. Considering the normal distribution of LOS and cost, the GAM with a Gaussian link was applied to examine the association between NO2 and those two indicators. Time trend was controlled by a penalty smoothing spline function of 6 degrees of freedom (df) per year. Daily mean temperature and relative humidity were adjusted using the same spline function with 6 df and 3 df. Public holidays and days of the week (DOW) were treated as categorical variables.
Considering the lag effect, the associations on single lag days of 0 to 5 and moving average days of 01 to 03 were also examined. Further, a random-effect meta-analysis based on the associations of all 11 cities was applied to obtain the overall effects of NO2.
At the second stage, this study estimated the attributable value and PAF of respiratory hospitalization, cost, and LOS for overall respiratory disease, asthma, and COPD to elucidate the potential benefits by controlling NO2 concentration to a certain level. The recommended concentrations by the WHO (25 µg/m3) and Chinese government (40 µg/m3) were used as reference concentrations. The formula and sensitivity analysis can be found in the
Supplementary Materials .This analysis included 666,189 hospital admissions due to respiratory diseases, corresponding to a total of 5.54 million days of LOS and 4.58 billion CNY in hospital expenditures.
Figure 1 illustrates the positive effects of NO2 on daily hospitalizations, hospital expenditures, and LOS across different lag days and moving average days at the provincial level. The effect was highest in lag03, every 10 μg/m3 increment in NO2 was associated with a 2.14% (95% CI: 1.39–2.89) increase in respiratory hospitalization, 13,167.36 CNY (95% CI: 7,697.29–18,637.43) in hospital expenditures, and 17.94 days (95% CI: 11.74–24.14) increase of LOS.
Supplementary Tables S1–S3 displayed the city-specific attributable values which vary across different cities.
Figure 1.The provincial overall association of daily NO2 (10 μg/m3) with (A) hospital admission, (B) hospitalization expenditure, and (C) length of hospital stay (days) at different lag days including single lag days (lag0–5) and moving average days (lag01–03) from 2017 to 2019.
Abbreviation: CNY=Chinese Yuan; LOS=length of hospital stay.The potentially avoidable value and PAF by meeting different air quality standards are shown in Table 1. Under the WHO’s standards, about 18.10 thousand (95% CI: 11.78–24.54) hospital admissions, 198.73 million CNY (95% CI: 119.07–278.40) hospital costs, and 272.00 thousand days (95% CI: 179.22–364.79) of LOS could be avoided.
Variable Potential avoidable value (95% CI) PAF (%, 95% CI) 25 µg/m3 40 µg/m3 25 µg/m3 40 µg/m3 Hospitalizations (thousand) 18.10
(11.78–24.54)7.06
(4.60–9.55)2.72
(1.77–3.68)1.06
(0.69–1.43)Hospitalization expenditure (million CNY) 198.73
(119.07–278.40)72.81
(43.62–101.99)4.34
(2.60–6.08)1.59
(0.95–2.23)Hospital stay (thousand days) 272.00
(179.22–364.79)99.65
(65.66–133.64)4.91
(3.23–6.58)1.80
(1.18–2.41)Abbreviation: WHO=World Health Organization; PAF=population attributable fraction; CI=confidence interval; CNY=Chinese Yuan. Table 1. Potentially avoidable values and population attributable fraction of hospital admission, hospitalization expenditure and hospital stay due to respiratory disease by reducing ambient NO2 concentration to the recommended concentration by WHO (25 µg/m3) and Chinese government (40 µg/m3), 2017–2019.
The potentially avoidable cost and LOS per capita based on the two standards are demonstrated in Table 2. About 298.31 CNY (95% CI: 178.73–471.90) in hospital costs and 0.41 days (95% CI: 0.27–0.55) of LOS could be avoided for each respiratory case under the standards of WHO.
Variable Potentially avoidable values for each case (95% CI) 25 µg/m3 40 µg/m3 Hospitalization expenditure (CNY) 298.31 (178.73–417.90) 109.29 (65.48–153.10) Hospital stay (days) 0.41 (0.27–0.55) 0.15 (0.10–0.20) Abbreviation: WHO=World Health Organization; CI=confidence interval; CNY=Chinese Yuan. Table 2. The potentially avoidable hospitalization expenditure and length of hospital stay for each respiratory case by controlling daily NO2 concentrations to criteria of WHO (25 µg/m3) and Chinese government (40 µg/m3), 2017–2019.
Supplementary Table S4 delineates the overall attributable burden of COPD and asthma due to NO2. About 5.69 million CNY (95% CI: 0.72–10.65) for asthma and 50.66 million (95% CI: 25.06–776.26) for COPD could be prevented by decreasing ambient NO2 concentrations to 25 µg/m3, corresponding to PAFs of 5.40% (95% CI: 0.69%–10.11%) and 4.79% (95% CI: 2.37%–7.22%). -
This analysis firstly demonstrated that higher NO2 exposure was associated with increased respiratory hospitalization, higher hospital cost, and longer LOS. Furthermore, the results indicate that numbers of respiratory hospitalization, hospital costs, and length of hospital stay could be potentially saved by reducing the NO2 concentrations. These findings provided ample evidence for formulating air pollution control policies, suggesting that continuously reducing NO2 concentrations can decrease the burdens of respiratory diseases in Shanxi Province and other similar regions with high NO2 pollution.
Our results indicate an adverse effect of ambient NO2 on hospital burden due to respiratory diseases, which is in line with previous studies. For example, one study found an association between life expectancy of people living with respiratory diseases and air pollution (4). Furthermore, in Pennsylvania, the regions with more serious air pollution were found to have a higher hospital expenditure for respiratory diseases (5).
The potentially avoidable values and PAFs for cause-specific respiratory diseases in our analysis were also aligned with previous studies. For example, Guo et al. estimated that the asthma-induced economic cost attributed to NO2 was 17.15 million dollars per year for Shanghai, China (6). Notably, the similar PAFs suggested that ambient NO2 had an adverse effect on both chronic and allergic respiratory diseases.
Several mechanisms could help explain the observed adverse effect of NO2 on the respiratory system. In one clinical trial, NO2 caused mild airway inflammation in healthy subjects, which affected blood cells and increased susceptibility to respiratory virus infections in airway epithelial cells (7). Moreover, exposure to higher ambient NO2 was confirmed to be associated with lung function, which could trigger or aggravate COPD (8).
Limitations should be considered. First, this was an ecological study and it is difficult to make a causal inference on the association between ambient NO2 exposure and respiratory burden. Second, this study only included cases from secondary and tertiary hospitals in Shanxi Province but excluded those from primary hospitals and patients without hospital admission, which may create a representativeness issue and lead to an underestimate of the respiratory burden in Shanxi Province.
In conclusion, this study elaborated on the association between ambient NO2 exposure and respiratory hospitalization, hospital cost, and hospital stay. The findings in this analysis identified that government’s continued implementation of NO2 concentration control can potentially reduce hospital admissions, saving a considerable amount of hospital expenditure for people in regions with high NO2 pollution.
-
No conflicts of interest.
HTML
| Citation: |
|