Search

Background Ambient particulate matter (PM) has been linked to adverse health outcomes, but evidence is still relatively limited for emergency department visits (EDVs). Most prior studies used daily mean PM as exposure assessment regardless of within-day variations, which might underestimate the potential acute health effects. In this study, we evaluated the hourly associations between PM and EDVs, so as to further understand the very short-term impacts of PM on EDVs. Methods Hourly data for all-cause EDVs, air pollutants and meteorological factors were collected from Shenzhen, China, between March 1, 2015 and February 28, 2018. A time-stratified case-crossover design with conditional logistic regression analysis was performed to assess the hourly associations between PM and EDVs, adjusting for hourly mean temperature and relative humidity. Subgroup analyses stratified by gender, age and clinic department were conducted to identify vulnerable populations. We further classified the whole year into warm (April to September) and cold (October to March of the next year) months to evaluate seasonal effect modification. Results Hourly mean all-cause EDVs peaked at 9 a.m. and 7 p.m., while PM2.5 and PM10 peaked at 8 a.m. and 8 p.m. The hourly associations between PM and EDVs appeared immediately and persisted for about 10 hours. All-cause EDVs increased 0.72% (95% confidence interval (CI): 0.33, 1.12) and 0.72% (95% CI: 0.45, 0.99) associated with per 10 μg/m3 rise in PM2.5 and PM10 at lag 0–10 h, respectively. Stronger effects of PM2.5 and PM10 on EDVs were observed among younger groups (age Conclusions Our study provided hourly perspective for the short-term associations between PM exposure and increased EDVs. The adverse impacts of PM on EDVs were only identified in cold season, and young people suffered more from both PM2.5 and PM10 exposures. These findings may have valuable implications for public policy promotion on air cleaning, as well as hospital management in optimizing allocation of resources on PM-sensitive conditions.