Post on 09-Feb-2018
transcript
Association between ambient air Association between ambient air
pullution and the hospital emergency pullution and the hospital emergency
room visits for respiratory diseases: room visits for respiratory diseases:
a casea case--crossover studycrossover study
Chang Su, Yuming Guo, Xiaochuan PanChang Su, Yuming Guo, Xiaochuan Pan
Department of Occupational and Environmental Department of Occupational and Environmental
Health, School of Public Health, Peking Health, School of Public Health, Peking
University, Beijing100191, ChinaUniversity, Beijing100191, China
BackgroundBackground�� PopulationPopulation--based studies of air pollution have shown an based studies of air pollution have shown an
association with respiratory morbidity and mortality.association with respiratory morbidity and mortality.
�� PMPM1010, SO, SO22 and NOand NO22 have been associated with increased have been associated with increased
hospital visits and admissions:hospital visits and admissions:
�� Pneumonia, COPD exacerbation, asthmaPneumonia, COPD exacerbation, asthma
�� To date, the majority of research on To date, the majority of research on air pollutionair pollution and and hospital hospital
emergency room visitsemergency room visits has been conducted in developed has been conducted in developed
countries, and a limited number of studies have evaluated in countries, and a limited number of studies have evaluated in
the developing countries.the developing countries.
ObjectiveObjective
�� To explore the the association between ambient To explore the the association between ambient
air pollution(PMair pollution(PM1010, SO, SO22 and NOand NO22 ) and the ) and the
hospital emergency room visits for the hospital emergency room visits for the
respiratory diseases in Beijing, China. respiratory diseases in Beijing, China.
Methods: data collectionMethods: data collection
AA Hospital Emergency Hospital Emergency
Room Visits Room Visits
BB
CC
Air Pollutants Air Pollutants
Weather Condition Weather Condition
Data of daily average Data of daily average
temperature and relative temperature and relative
humidity were collected from humidity were collected from
the Institute of Atmospheric the Institute of Atmospheric
Physics where is about 2 km Physics where is about 2 km
away from the Peking University away from the Peking University
Third Hospital.Third Hospital.
We collected the data of We collected the data of
daily emergency room daily emergency room
visits of visits of respiratory respiratory
disease fromdisease from June 1, June 1,
2004 to December 31, 2004 to December 31,
20052005 from Peking from Peking
University Third University Third Hospital. Hospital.
Based on ICDBased on ICD--10 10
dignostic codes(J00dignostic codes(J00--J99).J99).
DataData of PMof PM1010,, SOSO22 and NOand NO22 were were
collected from Beijing Municipal collected from Beijing Municipal
Environmental Monitoring Center.Environmental Monitoring Center.
Study areaStudy area
Methods: data analysisMethods: data analysis
�� The main intention of this study is to compare each subjectThe main intention of this study is to compare each subject’’s exposure s exposure
experienced just before hospital visit with that at other times experienced just before hospital visit with that at other times when the subject when the subject
was not hospitalized. was not hospitalized.
�� The conditional logistic regression model was used in our study The conditional logistic regression model was used in our study to analyze to analyze
different exposures between the case time and the control time.different exposures between the case time and the control time.
�� Control:Control:
�� unidirectional control sampling approch: one week beforeunidirectional control sampling approch: one week before
�� bibi--directional control sampling approch: one week before and afterdirectional control sampling approch: one week before and after
Case-crossover design
Advantages of caseAdvantages of case--crossover designcrossover design
Confounding Confounding
factorsfactors
ageagesexsex
smokingsmoking
harm habitsharm habits
nutrition conditionsnutrition conditions
seasonal confoundingseasonal confounding
Steps of caseSteps of case--crossover designcrossover design
�� ToTo assess the effect of air assess the effect of air
pollution up to lags of pollution up to lags of fivefive days. days.
�� Model construction:Model construction:
SingleSingle--pollutant modelpollutant model
MultipleMultiple--pollutantspollutants modelmodel
�� PM10PM10
�� SO2SO2
�� NO2NO2
�� TheThe relevant data of air relevant data of air
temperature and humidity temperature and humidity
werewere put into the model to put into the model to
controlcontrol their effect. their effect.
distributed lag modeldistributed lag model
define the hazard perioddefine the hazard period
Single/multipleSingle/multiple pollutants modelpollutants model
ResultsResults
AA
BB
CC
DD
Air pollution and weather Air pollution and weather
conditioncondition
Hospital emergency room visitsHospital emergency room visits
SpearmanSpearman correlation analysis correlation analysis
Association of daily hospital Association of daily hospital
emergency room visits for emergency room visits for respiratoryrespiratory
diseases and air pollutants diseases and air pollutants
Summary of air pollutants and weather Summary of air pollutants and weather
condition during study periodcondition during study period
13.4113.41±±±±±±±±11.0011.00
49.1549.15±±±±±±±±20.1020.10
101.27101.27±±±±±±±±1.02 1.02
14.9014.90±±±±±±±±9.89 9.89
32.1032.10
92.0092.00
103.68103.68
7575
--8.508.50
10.0010.00
98.8898.88
00
23.2023.20
67.0067.00
102.06 102.06
1919
15.4015.40
49.0049.00
101.26 101.26
1212
3.503.50
31.7531.75
100.52 100.52
88
Weather conditionWeather condition
Temperature (Temperature (℃℃℃℃℃℃℃℃))
Relative humidity Relative humidity (%)(%)
Atmospheric pressure(pa)Atmospheric pressure(pa)
Emergency room visitsEmergency room visits
52.6252.62±±±±±±±±51.6951.69
68.6668.66±±±±±±±±25.9325.93
145.34145.34±±±±±±±±90.30 90.30
280.00280.00
214.40214.40
573.00573.00
5.715.71
17.6017.60
10.0010.00
66.0066.00
80.0080.00
189.14189.14
32.7132.71
64.0064.00
128.57128.57
18.0018.00
52.8052.80
76.0076.00
Air pollution Air pollution
SOSO22 ((μμμμμμμμg/mg/m33))
NONO22 ((μμμμμμμμg/mg/m33))
PMPM1010((μμμμμμμμg/mg/m33))
757550502525Mean Mean ±±±±±±±± SDSDMaximumMaximumMinimumMinimum
Frequency distributionFrequency distribution
Total number of ER visits for respiratory disease: 10894
Spearman correlation analysis for air pollutants Spearman correlation analysis for air pollutants
and weather data between the study periodand weather data between the study period
1.0001.000
RelativeRelative
humidity (%)humidity (%)
--0.383*0.383*1.0001.000
AtmosphericAtmospheric
pressure(pa)pressure(pa)
0.395*0.395*--0.855*0.855*1.0001.000
Temperature Temperature
((℃℃℃℃℃℃℃℃))
0.293*0.293*--0.194*0.194*0.133*0.133*1.0001.000PM10PM10
0.307*0.307*--0.0180.018--0.0190.0190.741*0.741*1.0001.000NO2NO2
--0.286*0.286*0.577*0.577*--0.717*0.717*0.349*0.349*0.480*0.480*1.0001.000SO2SO2
Relative Relative
humidity (%)humidity (%)
AtmosphericAtmospheric
pressure(pa)pressure(pa)
Temperature Temperature
((℃℃℃℃℃℃℃℃))PM10PM10NO2NO2SO2SO2
* * (P<0.05)
Association of daily ER visits for respiratory diseases Association of daily ER visits for respiratory diseases
and air pollutants at different Lagsand air pollutants at different Lags
Single pollutants modelSingle pollutants model-- association of daily ER visits for association of daily ER visits for
respiratory diseases and air pollutantsrespiratory diseases and air pollutants
1.011(1.0001.011(1.000--1.023) *1.023) *1.033(1.0191.033(1.019--1.047) *1.047) *NONO2 2 (( lag 2)lag 2)
1.012(1.0051.012(1.005--1.019) *1.019) *1.017(1.0101.017(1.010--1.025) *1.025) *SOSO2 2 (( lag 0)lag 0)
1.004(1.0011.004(1.001--1.008) *1.008) *1.011(1.0081.011(1.008--1.015) *1.015) *PMPM10 10 (( lag 0)lag 0)
bibi--directional control directional control
approchapproch
Unidirectional control Unidirectional control
approchapproch
OR (95%CI)OR (95%CI)
PollutantsPollutants
* * (P<0.05)
Multiple pollutants modelMultiple pollutants model-- association of daily ER visits for association of daily ER visits for
respiratory diseases and air pollutantsrespiratory diseases and air pollutants
1.0121.012(1.001(1.001--1.021.0222) *) *1.001(0.9831.001(0.983--1.009)1.009)+ + PMPM1010 and SOand SO
22
1.0121.012(1.002(1.002--1.022) *1.022) *1.005(0.9921.005(0.992--1.017)1.017)+ SO+ SO22
1.010(1.0001.010(1.000--1.021) *1.021) *0.997(0.9840.997(0.984--1.010)1.010)+PM+PM2.52.5
Lag Lag 22Lag Lag 22NONO22
1.011(1.0031.011(1.003--1.018)*1.018)*1.010(1.0011.010(1.001--1.018) *1.018) *+ + PMPM1010 and NOand NO
22
1.012(1.0051.012(1.005--1.019)*1.019)*1.017(1.0091.017(1.009--1.025) *1.025) *+ NO+ NO22
1.010(1.0021.010(1.002--1.017)1.017)1.009(1.0011.009(1.001--1.018) *1.018) *+ + PMPM1010
Lag 0Lag 0Lag 0Lag 0SOSO22
1.002(0.9881.002(0.988--1.005)1.005)1.010(1.0051.010(1.005--1.014) *1.014) *+SO+SO2 2 and NOand NO
22
1.004(1.0001.004(1.000--1.009) *1.009) *1.012(1.0081.012(1.008--1.016) *1.016) *+ NO+ NO22
1.002(0.9981.002(0.998--1.006)1.006)1.009(1.0051.009(1.005--1.014) *1.014) *+ SO+ SO22
Lag 0Lag 0Lag 0Lag 0PMPM1010
bibi--directional control directional control
approchapproch
Unidirectional control Unidirectional control
approchapproch
OR (95%CI)OR (95%CI)
PollutantsPollutants
* * (P<0.05)
DisussionDisussion
�� A transient increase in ambient PM10, SO2 and NO2 was associatedA transient increase in ambient PM10, SO2 and NO2 was associated
with an increased risk of hospital emergency room visits for totwith an increased risk of hospital emergency room visits for total al
respiratory diseases.respiratory diseases.
�� After controlling the influence of two pollutants, the risk effeAfter controlling the influence of two pollutants, the risk effect of the ct of the
other one air pollutant on the hospital emergency room visits foother one air pollutant on the hospital emergency room visits for total r total
respiratory diseases decreased. It indicates that the effect of respiratory diseases decreased. It indicates that the effect of the three the three
air pollutants has interaction on the related health effects.air pollutants has interaction on the related health effects.
�� The results of uThe results of unidirectional control sampling approch are differ from nidirectional control sampling approch are differ from
bibi--directional control approch. There would be bias about the timedirectional control approch. There would be bias about the time
trends of exposure in the trends of exposure in the uunidirectional control approch, the results of nidirectional control approch, the results of
bibi--directional control approch are more reliable.directional control approch are more reliable.
DiscussionDiscussion
Limitations and further studyLimitations and further study
�� Limitations: Limitations:
We used ambient monitoring data to represent the individual We used ambient monitoring data to represent the individual exposures to air exposures to air
pollutants. Our assessment of weather conditions was collected epollutants. Our assessment of weather conditions was collected entirely from ntirely from
only one monitoring only one monitoring station.station. It is possible that this type of error may introduce It is possible that this type of error may introduce
bias to our results. bias to our results.
�� Further study: Further study:
Classify the visits into different sex group, age group; anaClassify the visits into different sex group, age group; analyze the influence of lyze the influence of
season.season.
Thank You !Thank You !