歯 PDF

Transcription

1

2 Short- term Effect of Air Pollution on Respiratory Disease in Seoul : A Case- Crossover Design

3

4

5

6 T able 1. Distribution of daily emergency respiratory disease visits, 24hr mean s of air pollutant s, temper ature and r elative humidity 18 T able 2. Pearson Correlation Coefficients for 24hr means of air pollutants 19 T able 3. Pear son Corr elation Coefficient s for 24hr maxium of air pollutants 20 T able 4. Pearson Correlation Coefficients for 8hr means of air pollutants 20 T able 5. Relative risk at one pollutant model of O3, SO2, NO2, CO, T SP, PM 10 with various lag times 26 T able 6. Relative risk at tw o pollutant s m odel of O3, SO2, NO2, CO, T SP, PM10 with variou s lag times using 24hr mean values 27 T able 7. Relative risk at tw o pollutant s m odel of O3, SO2, NO2, CO, T SP, PM10 with variou s lag times using 24hr maximum values 28 T able 8. Relative risk at tw o pollutant s m odel of O3, SO2, NO2, CO, T SP, PM 10 with various lag times using 8hr mean values 29 T able 9. Relative risk at one pollutants model of retrospective selection with variou s lag times using 24hr mean values 30 T able 10. Relative risk at one pollutant s model of pr ospectiv e selection with variou s lag times using 24hr mean values 31 T able 11. Relative risk at one pollutants model of retrospective selection with variou s lag times using 24hr max values 32 T able 12. Relative risk at one pollutant s model of pr ospectiv e selection with various lag times u sing 24hr m ax values 33 T able 13. Relative risk at one pollutants model of retrospective selection with variou s lag times using 8hr mean values 343 T able 14. Relative risk at one pollutant s model of pr ospectiv e selection with variou s lag times using 8hr mean values 35

7 Figur e 1. Contr ol selection method 11 Figur e 2. Daily count s of r espir atory emer gency visit s 17 Figur e 3. Daily mean O3 concentr ation 21 Figur e 4. Daily mean SO2 concentration 22 Figur e 5. Daily mean CO concentr ation 22 Figur e 6. Daily mean NO2 concentration 23 Figur e 7. Daily mean T SP concentr ation 23 Figur e 8. Daily mean PM10 concentr ation 24

8 , , 14, , 2, 3 1.,,,. 6 30ppb (95%CI = ). 24 O3 SO2, O3 NO2, O3 CO, O3 T SP, O3 PM i -

9 1.87(95%CI= )... :,,,, - ii -

10 Meuse vally, Donora, London, Los Angeles. (,,, ),, (Logan, 1953;, 1989;, 1997).,,,,,, (, ).,,,,,,,,,,,,, - 1 -

11 (, 1997)... (Lippmann, 1989).,.,..,.,.,.,

12 . 2. (1) (smog ),.,.,. (NOx ). (NOx ) (HC), HC NOx.(, 1991) ppb,, 13,000ppb, 1,500-2,000ppb 2, 2,500ppb,, (Dix, 1981).,. Bates (1983) 1, - 3 -

13 . Ponka (1991), / (Schw artz, 1993). White (1994) ppm 37% (Schw artz, 1994). Romieu (1995),. Stieb (1996) ,519 30ppb (Burmett, 1997). 12 (O3, NO2, SO2, PM 10 ) 10 / (W ong, 1999)

14 (1999) General Additive Poisson Model 100ppb (2000),, LA,, (Kimney, 1991) ,, 10 percentile 90 percentile 3.5% (95% 1.7%- 5.3%), 3.5% (95% 1.04%- 6.1%), 5.4% (95% 0.4%- 10.7%) (Ander son, 1996). (1999),,, (, 1998). (1999) General Estimating Equation

15 (2), (SOX ) 1., EPA (, T SP ) 10 (PM 10 ). PM 10, 1995 PM 10. Pennsylvania Philadelphia T SP 100 / 7% (95% CI = ), 100 / 5%. T SP (Schw artz, 1992). T SP 100 / 6% (95% CI = ) (Schwartz, 1994) 65 PM 10 (Saldiva, 1995)., - 6 -

16 (T ouloumi, 1994) (95% CI = ), (Dockery, 1993).. Ohio. T SP (24 ) 100 / 3% (Samet, 1981). 5 ( ), (Sunyer, 1993). Ponka (1994) ,, SO2 NO2. (1998) PM 10, (1997),,. (1994),,. (3) - 7 -

17 (NO) (NO2 ) (NOx ). NO2 (hydrocarbon ) 2..,,,.. (4).,.,,,. carboxy hemoglobin (COHb)

18 (Cas e - c ro s s ove r de s ig n) 1 ) 1991 Maclure the Onset study" - - (case- crossover design ) (Maclure, 1991).,,,,, (Mittleman, 1995) 2 ) - (control period). (1) - 9 -

19 (Unidirectional Pair - matched and multiple int erv al approach) 2, ( ) ( ). - 1: , 4,, 1, 1:N -. (2) (Bidirectional Case- crossover design ) Navidi(1998) RR ( ),

20 Fig 1. Contr ol selection m ethod

21 ,., (medical utilization data of unscheduled visit by respiratory diseases)

22 . 1. ( ) 20 ( ),,,, 5, (telementry system, T MS ) (real time) ,,,,,,,,,,,,,,,,,,,.,,,, 9 10 (PM 10 )

23 , (2) m. 3. (3).,,,, /,,,,,. (medical utilization data of unscheduled visit by respiratory diseases). 1996,,.,,

24 ,.. ICD- 10 code. J20; J21; J40; J41; J42; J43; J44; J45; J46; J96; J99; 1996, , ,952., 2, 18,

25 ,. (Lee, 1999). Navidi(1998) (bidirectional control period selection ), 7, 14, , , 24, (one pollutant model) 6 2 (tw o pollutants model).., 24, 24, 8 1, 2, 3.,,, 30ppb,, 30 /

26 (T able 1) (Figure. 2.) F igure 2. Daily count s of respiratory em ergen cy visit s,,,,, 15.6ppb, 13.9ppb, 32.9ppb, 12.6ppb, 86.4 /, /, 12.7, 64.5%.,,

27 (T able 1). T able 1. Distribution of em erg ency respirat ory disease v isit s, 24hr m ean s of air V ariable pollut ant s, t em perature an d relativ e hum idity M ean St andard deviation Min 1). P ercentile 25th 50th 75th M ax 2 ). Em erg ency visit s O3 (ppb ) S O2 (ppb ) NO2 (ppb ) CO (ppb ) T SP ( / ) PM 10 ( / ) T emperature ( ) Relativ e hum idity (% ) ) Min. : m inim um 2) M ax. : m axim um T able 2, 3, 4 (O3, SO2, CO, NO2, T SP, PM 10 ) , ,,

28 - 0.5, , T able 2. Pearson Correlation Coefficients for 24hr means of air pollutants V ariable O3 S O2 NO2 CO T SP PM 10 T emperature Hum idity O * * * S O NO CO T SP PM T em perature Hum idity 1 P - v alue P - v alue

29 T able 3. Pearson Correlation Coefficients for 24hr maxium of air pollutants V ariable O3 S O2 NO2 CO T SP PM 10 T em perature Hum idity O S O NO CO T SP PM T emperature Hum idity 1 P - v alu e P - v alu e T able 4. Pearson Correlation Coefficients for 8hr means of air pollutants V ariable O3 S O2 NO2 CO T SP PM 10 T emperature Hum idity O S O NO CO T SP PM T em perature Hum idity 1 P - v alue P - v alue

30 (Figure. 3) F igure 3. Daily m ean O3 con centration (Figure. 4), (Figure. 5). (Figure. 6), (Figure. 7) (Figure.8)

31 F igure 4. Daily m ean S O2 concentration F igure 5. Daily m ean CO concentration

32 F igure 6. Daily m ean NO2 con centration F igure 7. Daily m ean T SP concentration

33 F igure 8. Daily m ean PM 10 con centration 2. 6,. 24, 8, one pollutant model two pollutants model., 24 1 one pollutant model , (SO2, NO2, CO, T SP, PM 10)

34 . T w o pollutant s model 24 O3 SO2, O3 NO2, O3 CO, O3 T SP O3 PM 10, (95%CI= ). 24 O3 SO2, O3 NO2, O3 CO, O3 T SP O3 PM , 2 3. SO2 CO CO ( ). NO2 CO 3 CO 1.060(95%CI= ). T able 3. CO SO2 NO2 0.49, O3 SO2, O3 NO2, O3 CO, O3 T SP O3 PM 10 1 O

35 T able 5. Relativ e risk at on e pollut ant m odel of O3, S O2, NO2, CO, T SP, PM 10 w ith v ariou s lag tim es P ollut ant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) S O2 1) 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) 1) N O2 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) CO 1 ) 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) T SP 2 ) 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) 2 ) PM 10 24hr m ean ( ) ( ) ( ) 24hr m ax 5 ) ( ) ( ) ( ) 8hr m ean ( ) ( ) ( ) 1) O3, S O2, NO2, CO : RR for the in crease of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : R elativ e Risk 4) CI : Confiden ce Interv al 5) m ax : M axim um

36 T able 6. Relativ e risk at tw o pollut ant s m odel of O3, S O2, NO2, CO, T SP, PM 10 w ith v ariou s lag tim es u sin g 24hr m ean v alu es P ollutant RR 3 ) (95% CI 4 ) ) (24hr m ean ) lag 1 day lag 2 day lag 3 day O ( ) ( ) ( ) S O ( ) ( ) ( ) O ( ) ( ) ( ) N O ( ) ( ) ( ) O ( ) ( ) ( ) CO ( ) ( ) ( ) O ( ) ( ) ( ) T SP ( ) ( ) ( ) O ( ) ( ) ( ) P M ( ) ( ) ( ) S O ( ) ( ) ( ) N O ( ) ( ) ( ) S O ( ) ( ) ( ) CO ( ) ( ) ( ) S O ( ) ( ) ( ) T SP ( ) ( ) ( ) S O ( ) ( ) ( ) P M ( ) ( ) ( ) N O ( ) ( ) ( ) CO ( ) ( ) ( ) N O ( ) ( ) ( ) T SP ( ) ( ) ( ) N O ( ) ( ) ( ) P M ( ) ( ) ( ) CO ( ) ( ) ( ) T SP ( ) ( ) ( ) CO ( ) ( ) ( ) P M ( ) ( ) ( ) 1) O3, S O2, NO2, CO : RR for the increase of 30 ppb 2) T SP, PM 10 : RR for the increa se of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Interv al

37 T able 7. Relativ e risk at t w o pollut ant s m odel of O3, S O2, N O2, CO, T SP, PM 10 w ith v ariou s lag tim es u sing 24hr m ax im um v alu es P ollut ant RR 3 ) (95% CI 4 ) ) (24hr m ax 5 ) ) lag 1 day lag 2 day lag 3 day O ( ) ( ) ( ) S O ( ) ( ) ( ) O ( ) ( ) ( ) NO ( ) ( ) ( ) O ( ) ( ) ( ) CO ( ) ( ) ( ) O ( ) ( ) ( ) T SP ( ) ( ) ( ) O ( ) ( ) ( ) PM ( ) ( ) ( ) S O ( ) ( ) ( ) NO ( ) ( ) ( ) S O ( ) ( ) ( ) CO ( ) ( ) ( ) S O ( ) ( ) ( ) T SP ( ) ( ) ( ) S O ( ) ( ) ( ) PM ( ) ( ) ( ) NO ( ) ( ) ( ) CO ( ) ( ) ( ) NO ( ) ( ) ( ) T SP ( ) ( ) ( ) NO ( ) ( ) ( ) PM ( ) ( ) ( ) CO ( ) ( ) ( ) T SP ( ) ( ) ( ) CO ( ) ( ) ( ) PM ( ) ( ) ( ) l) O3, S O2, NO2, CO : RR for the increase of 30 ppb 2) T SP, PM 10 : RR for th e in crease of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Int erv al 5) m ax : m ax im um

38 T able 8. Relativ e risk at tw o pollut ant s m odel of O3, S O2, NO2, CO, T SP, PM 10 w ith v ariou s lag tim es u sing 8hr m ean v alues P ollutant RR 3 ) (95% CI 4 ) ) (8hr m ean ) lag 1 day lag 2 day lag 3 day O ( ) ( ) ( ) S O ( ) ( ) ( ) O ( ) ( ) ( ) NO ( ) ( ) ( ) O ( ) ( ) ( ) CO ( ) ( ) ( ) O ( ) ( ) ( ) T SP ( ) ( ) ( ) O ( ) ( ) ( ) PM ( ) ( ) ( ) S O ( ) ( ) ( ) NO ( ) ( ) ( ) S O ( ) ( ) ( ) CO ( ) ( ) ( ) S O ( ) ( ) ( ) T SP ( ) ( ) ( ) S O ( ) ( ) ( ) PM ( ) ( ) ( ) NO ( ) ( ) ( ) CO ( ) ( ) ( ) NO ( ) ( ) ( ) T SP ( ) ( ) ( ) NO ( ) ( ) ( ) PM ( ) ( ) ( ) CO ( ) ( ) ( ) T SP ( ) ( ) ( ) CO ( ) ( ) ( ) PM ( ) ( ) ( ) 1) O3, S O2, N O2, CO : RR for the incr ea se of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Int erv al

39 T able 9. Relativ e risk at on e pollut ant s m odel of retrospectiv e selection w ith v ariou s lag tim es u sin g 24hr m ean v alu es P ollutant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) S O2 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1 ) NO2 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) CO 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) T SP 2 ) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 2 ) PM 10 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1) O3, S O2, N O2, CO : RR for the incr ea se of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Int erv al

40 T able 10. Relativ e risk at one pollut ant s m odel of prospectiv e selection w ith v ariou s lag tim es u sin g 24hr m ean v alu es P ollutant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) S O2 1) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 1 ) NO2 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) CO 1) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) T SP 2 ) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 2 ) PM 10 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 1) O3, S O2, N O2, CO : RR for the incr ea se of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Int erv al

41 T able 11. Relativ e risk at one pollutant s m odel of retrospectiv e selection w ith v ariou s lag tim es u sing 24hr m ax im um v alu es P ollutant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) S O2 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1 ) NO2 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) CO 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) T SP 2 ) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 2 ) PM 10 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1) O3, S O2, N O2, CO : RR for the incr ea se of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Int erv al

42 T able 12. Relativ e risk at one pollut ant s m odel of prospectiv e selection w ith v ariou s lag tim es u sing 24hr m ax im um v alu es P ollutant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) S O2 1) 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) 1 ) N O2 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) CO 1 ) 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) T SP 2 ) 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) 2 ) P M 10 7day s after ( ) ( ) ( ) 14day s after ( ) ( ) ( ) 21day s after ( ) ( ) ( ) 1) O3, S O2, NO2, CO : RR for the increase of 30 ppb 2) T SP, PM 10 : RR for the increa se of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Interv al

43 T able 13. Relativ e risk at one pollutant s m odel of retrospectiv e selection w ith v ariou s lag tim es u sing 8hr m ean v alues P ollutant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) S O2 1) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1 ) N O2 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) CO 1 ) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) T SP 2 ) 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 2 ) P M 10 7day s before ( ) ( ) ( ) 14day s before ( ) ( ) ( ) 21day s before ( ) ( ) ( ) 1) O3, S O2, NO2, CO : RR for the increase of 30 ppb 2) T SP, PM 10 : RR for the increa se of 30 / 3) RR : Relativ e Risk 4) CI : Confidence Interv al

44 T able 14. Relativ e risk at one pollut ant s m odel of prospectiv e selection w ith v ariou s lag tim es u sing 8hr m ean v alues P ollut ant RR 3 ) (95% CI 4 ) ) lag 1 day lag 2 day lag 3 day O3 1) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) S O2 1) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 1) N O2 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) CO 1 ) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) T SP 2 ) 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 2 ) PM 10 7day s after ( ) ( ) ( ) 14day s aft er ( ) ( ) ( ) 21day s aft er ( ) ( ) ( ) 1) O3, S O2, NO2, CO : RR for the in crease of 30 ppb 2) T SP, PM 10 : RR for the increase of 30 / 3) RR : R elativ e Risk 4) CI : Confiden ce Interv al

45 Maclure ( ),,,, (Navidi, 1998). - ( ) ( ).. (selection bias), (information bias ), (confounding bias ) 3., (selection bias)

46 . -.., (information bias ) (misclassification ). (nondifferential misclassification ).. (differential misclassification ) (recall bias )... -,

47 ., (confounding bias) Navidi(1998) PM 10. Neas (1999) Lee (1999) -,

48 . (2000),,,.. (, 1998) 1 100ppb 2.18, 100 / , 65% 100ppb 2.16 (p- value 0.001), 100 / 1.16 (p- value 0.01). 1 30ppb 1.91 (24, 1 ). 2 24,...,

49 ( ) ( ) ( 1, 2, 3 ). (T able.5) (T able.9,11,13) (T able 10,12,14). 1, 2, 3 (Fig 1. A ). 1, 2, 3., 1, 2, 3 (Fig 1. B). 1, 2, 3 (Fig 1. C) (95%CI= ), (95%CI= ), (95%CI= ), (95%CI= ), (95%CI= ), (95%CI= )

50 (, 1999) 24, 24, 8 3.,, (, 1998;, 2000) 10.1% 85.0%.,. (National Ambient Air Quality Standard),

51 (, 1991)..,, (Kleeberger, 1995)., (Van Rooij, 1994)., (, 1995;, 1996; 1997).,.,,, (Mumtaz, 1995)

52 (Liu, 1997),..,,,,.,,. (, ). (, 1997),

53 .,,

54 , -. SO2., (95%CI= ) 30ppb O3 SO2, O3 NO2, O3 CO, O3 T SP, O3 PM (95%CI= ).,

55 ,, ; 13(5) : , ; 32(2) : ,, ; 11(3) : , ; 23(1) : ; 7(1) : ( ) ; 16(1) : 6-19,, ; 33(1) :

56 ,, ; 31(3) : , 1992,,. ( ). 1998; 31(1) : 82-90, Dockery DW,. ( ). 1999; 32(2) : ,, ; 10(3) : ,, ; 12(2) : , 1997; , ; 7(3) :

57 ; 32(12) : , , 1999., ; 17(1) : 64-75, ; 12(4) : , 1994; , 1997.,

58 Ander son HR, Leon AP, Bland JM, et al. Air pollution and daily mortality in London : British Medical Journal 1996; 312 : Bates DV, Sizto R. T he Ont ario air pollution study : identification of the cau sative agent. Envir on Health Per spect 1989; 79 : Burnett R, Br ook JR, Yung WT, et al. A ssociation betw een ozone and hospitalization for respiratory disease in 16 Canadian cities. Envir on Res 1997; 72 : Dix HM. Enviromental air pollution. John Wiley, New York 1981; 286 Dockery DW, Pope CA, Xu X, et al. An association betw een air pollution and mortality in six US cities. New England Journal Medicine 1993; 329 : Euler LG, Hodgkin J. Chronic ob structive pulnonary disease symptom effect s of long - term cumulativ e exposure to ambient lev els of total suspended particulates and sulfur dioxide in California Sev enth - day adv entist Resident s. Arch Environl Health 1987; 41(4) : Greenland S. Confounding and exposure tr ends in case- crossov er and case- time- contr ol design s. Epidemiology 1996; 7 :

59 Kinney PL, Ozkaynak H. A s sociation of daily mortality and air pollution in Los Angeles County. Environ Res 1991; 54 : Kleeberger SR. Genetic su sceptibility to ozone exposur e. T oxicol Lett. 1995; 82 : Lee JT, Schw artz J. Reanaly sis of air pollution on daily m ortality in Seoul, Korea: A case- crossov er design. Environ Health Per spect 1999; 107(8) : Lippman M. Health effect of ozone: A critical review. J Air Pollut Cont A ssoc 1989; 39 : Liu LJS, Delfino R, Koutrakis P. Ozone exposure assessment in a Sputhem California community. Environ Health Per spect 1997; 105(1) : Logan WPD. Mortality in London fog incident. Lancet 1953; 1 : Maclure M. T he case- crossov er design a m ethod for studying tr an sient effet s on the risk of acute ev ent s. Am J Epidemiol 1991; 133 : Mittleman MA, Maclur e M, Robin s JM. Control sampling strategies for case- cr os sover studies : an assessm ent of r elative efficiency. Am J Epidemiol 1995; 142(1) :

60 Mittlem m an MA, M aculure M, Nachnani M, et al. Edu cational att ainm ent, an ger, and the risk of trigg erin g m y ocardial infarction on set. Arch Int ern M ed 1997; 157 : Navidi W. Bidrectional case- crossov er design for exposur e w ith tim e tr ends. Biometrics 1998; 54 : Neas LM, Schw artz J, Dockery D. A case- crossov er analy sis of air pollution and mortality in Philadelphia. Envir on Health Per spect 1999; 107(8) : Ponka A. A sthma and low lev el air pollution in Helsinki. Ar ch Environ Health 1991; 46 : Ponka A, Virtanen M. Chronic br onchitis, emphy sem a, and low - level air pollution in Helsinki, Envir on Res 1994; 65 : Pope CA, Schw artz J, Ran som MR. Daily mort ality and PM10 pollution in Ut ah Valley. Arch Envir on Health 1992; 47(3) : Romieu I, Meneses F, Sienna - Monge JJL, et al. Effects of urban air pollutants on emergency visits for childhood asthma in Mexico City. Am J Epidemiol 1995; 1416 : Saldiva PHN, Pope CA, Schw artz J, et al. Air pollution and mortality in

61 elderly people a time- series study in Sao Paulo, Br azil. Ar ch Envir on Health 1995; 50(2) : Samet JM, Speizer FE, Schw artz J, et al. T he r elation ship betw een air pollution and emer gency room visit s in an industrial community. J Air Pollut Cont A ssoc 1981; 31 : Schw artz J, Dockery DW. Increased mortality in Philadelphia associated with daily air pollution concentration. Am Rev Respir Dis 1992; 145 : Schw artz J, Slater D, Lar son T V, et al. Particulate air pollution and hospital emergency room visit s for asthma in Seattle. Am Rev Respir Dis 1993; 147 : Schw artz J. Air pollution and hospital admis sion s for the elderly in Detr oit, Michigan. Am J Respir Crit Car e Med 1994; 150 : Schw artz J. Air pollution and daily mort ality : A review and m eta analy sis. Envir on Res 1994; 64 : Stieb M, Burnett RT, Bev eridge RC, et al. A ssociation betw een ozone and asthm a emer gency departm ent visit s in Saint John, New Brun swick, Canada. Envir on Health Per spect 1996; 104(12) :

62 Suissa S. T he case- time- cotr ol design : Further as sumption and condition s. Epidemiology 1999; 9 : Suny er J, Anto J, Murillo C, et al. Effect s of urban air pollution on emer gency room admis sion for chr onic obstructiv e pulmonary disease. Am J Epidemiol 1991; 134 : Suny er J, Saez M, Murilo C, et al. Air pollution and emer gency room admission s for chr onic ob structive pulm onary disease: A 5- year study. Am J Epidemiol 1993; 137(7) : T ouloumi G, Pocock SJ, Katsouyanni, et al. Short - term of air pollution on daily mort ality in Athen s: a time- series analy sis. Int J Epidemiol 1994; 23(5) : Van Rooij JG, Veeger MM. Bodelier- Bade MM, et al. Smoking and dietary int ake of polycyclic ar om atic hy dr ocarbon as sour ces of interindividual variability in the baseline ex cr etion of 1- hydroxypyr ene in urine. Int Arch Occup Environ Health 1994; 66(1) : W hit e M C, Et zel RA, W ilcox W D, et al. Ex acerbation of childhood a sthm a and ozon e pollution in Atlant a. Environ Res 1994; 65 : Wong T W, Lau T S, Yu T S, et al. Air pollution and hospital admissions for

63 respir atory and car diovascular desease in Hong Kong. Occup Environ Med 1999; 56(10) :

64 = Abs tra c t = S h ort - term E ffe ct of A ir P ollution on Re s pirat ory D is e as e in S e ou l : A Cas e - Cro s s ov er D e s ig n Young - Ju Lee Department of Envir onment al Science Graduate School of Health Science and Management Yonsei University (D ire ct e d by P rof e s s or Jon g - T ae Le e ) It is w ell know n that short - term increase in air pollution such as those that occurr ed in developed countries during this century, are associated with an increased number of death, hospital admission and hospit al em ergency room visit s especially among those with pr eexisting illnes ses like cardiov ascular disease and chr onic respir atory disease. T his study w as designed and conducted to evaluat e the hypothesis that increased emer gency r oom visit s for medical utilization data of un scheduled visit by r espir atory diseases, and if so, to quantify the strength of association betw een those. Daily count s of emer gency r oom visit s for r espirat ory disease, air quality, and w eather data w ere collected. W e complied daily r ecords of hospital emergency room visit s in Seoul from Nov ember to Oct ober Daily respiratory disease em ergency r oom visit s w ere r egres sed on temperature, humidity, day of w eek indication s seasonal variation indication s and air pollution. Each pollutant w as fir st ex amined individually and then tw o pollutant s m odel w ere fitted. Lately the case- crossov er design that w as newly developed by Malculare has been being introduced in the fields of air pollution epidemiology. T his study,

65 u sing case- crossov er design sear ch a r elation ship betw een air pollut ant s and emergency r oom visit s for r espir atory disease. We analy zed the dat a by conditional logistic regres sion u sed when m atched case- control study is analy sed. Control periods w er e chosen by bidir ectional pair ed mat ching technic 7, 14, 21day s before and after case periods. Only ozone w as as sociated with increased risk of emer gency room visit s for respir atory disease. T he relativ e risk of respir atory emer gency r oom for 30ppb increase in ozone (24hr m ean, lagged 1day ) w as 1.91(95% confidence interv al; ) and w e observed that. In conclu sion, there w as statistically significant association betw een ambient levels of ozone and daily count s of emer gency room visit s for respiratory disease. So, w e have to ex ert our effort s to minimize the air pollution effectiv ely to pr otect public health fr om air pollution. Key w ords : air pollution, ozone, conditional logistic regres sion, case- cr os sover design, emer gency room visit