Archives of Environmental & Occupational Health The ...

This article was downloaded by: [University of Florida] On: 14 November 2014, At: 07:02 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Archives of Environmental & Occupational Health Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vaeh20

The Relationship of Ambient Ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible Influence of Gender and Ethnicity a

a

a

a

Jo Ann Glad DrPH , LuAnn Lynn Brink PhD , Evelyn O. Talbott DrPH , Pei Chen Lee PhD , b

c

Xiaohui Xu PhD , Melissa Saul MS & Judith Rager MPH a

a

Department of Epidemiology , University of Pittsburgh , Pittsburgh , Pennsylvania , USA

b

Department of Epidemiology and Biostatistics , University of Florida, College of Public Health and Health Professions , Gainesville , Florida , USA c

Department of Biomedical Informatics , University of Pittsburgh, School of Medicine , Pittsburgh , Pennsylvania , USA Published online: 23 Apr 2012.

To cite this article: Jo Ann Glad DrPH , LuAnn Lynn Brink PhD , Evelyn O. Talbott DrPH , Pei Chen Lee PhD , Xiaohui Xu PhD , Melissa Saul MS & Judith Rager MPH (2012) The Relationship of Ambient Ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible Influence of Gender and Ethnicity, Archives of Environmental & Occupational Health, 67:2, 103-108, DOI: 10.1080/19338244.2011.598888 To link to this article: http://dx.doi.org/10.1080/19338244.2011.598888

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Archives of Environmental & Occupational Health, Vol. 67, No. 2, 2012 C 2012 Taylor & Francis Group, LLC Copyright

Downloaded by [University of Florida] at 07:02 14 November 2014

The Relationship of Ambient Ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible Influence of Gender and Ethnicity Jo Ann Glad, DrPH; LuAnn Lynn Brink, PhD; Evelyn O. Talbott, DrPH; Pei Chen Lee, PhD; Xiaohui Xu, PhD; Melissa Saul, MS; Judith Rager, MPH

ABSTRACT. An investigation of the relationship of air pollution and emergency department (ED) visits for asthma was an opportunity to assess environmental risks for asthma exacerbations in an urban population. A total of 6,979 individuals with a primary discharge diagnosis of asthma presented to 1 of 6 EDs in the Pittsburgh, Pennsylvania, area between 2002 and 2005. Using a case-crossover methodology, which controls for the effects of subject-specific covariates such as gender and race, a 2.5% increase was observed in asthma ED visits for each 10 ppb increase in the 1-hour maximum ozone level on day 2 (odds ratio [OR] = 1.025, p < .05). Particulate matter with an aerodynamic diameter ≤2.5 µm (PM2.5 ) had an effect both on the total population on day 1 after exposure (1.036, p < .05), and on African Americans on days 1, 2, and 3. PM2.5 had no significant effect on Caucasian Americans alone. The disparity in risk estimates by race may reflect differences in residential characteristics, exposure to ambient air pollution, or a differential effect of pollution by race. KEYWORDS: air pollution, epidemiology, ozone, particulate matter

T

here is an estimated 7% prevalence of adult asthma in the United States, with a higher prevalence in blacks (9.2%) compared with whites (6.9%).1 Recent studies have noted risk differences for asthma exacerbations associated with air pollution by age, race, and gender. This may be due to intrinsic biological differences (ie, reduced lung capacity in women compared with men and hormonal influences) or socioeconomically mediated effects related to access to care.2 Early air pollution health effect studies have focused on available health effect information: mortality and hospitalizations. These data represent only the most serious asthma episodes and may underestimate the association

with air pollution. With the increasing availability of electronic medical record data on emergency department (ED) visits, the researcher’s ability to capture a more common outcome related to asthma exacerbations (ED visits) is now possible. In the last decade, the effect of ozone, particulate matter with an aerodynamic diameter ≤2.5 µm (PM2.5 ), and other pollutants on asthma and respiratory-related exacerbations have been assessed in several investigations in different regions of the United States and in other countries.3–14 Delfino et al12 conducted a study in Montreal that reported that ED visits for respiratory illnesses were 21.8% (95% confidence

Jo Ann Glad, LuAnn Lynn Brink, Evelyn O. Talbott, Pei Chen Lee, and Judith Rager are with the Department of Epidemiology, University of Pittsburgh, Pittsburgh, Pennsylvania, USA. Xiaohui Xu is with the Department of Epidemiology and Biostatistics, University of Florida, College of Public Health and Health Professions, Gainesville, Florida, USA. Melissa Saul is with the Department of Biomedical Informatics, University of Pittsburgh, School of Medicine, Pittsburgh, Pennsylvania, USA. 2012, Vol. 67, No. 2

103


interval [CI], 9.7–33.8%) higher than average for a mean 8hour maximum increase of 38 ppb ozone among patients over the age of 64. This study demonstrated that elevated levels of some pollutants at levels below national standards negatively affect public health among sensitive subgroups. Weisel and colleagues13 found a similar relationship between ozone and asthma ED visits in a study involving 11 hospitals in New Jersey. Even after controlling for pollen, spore concentrations, and temperature, when ozone levels reached 60 ppb or greater, ED visits were found to be 26% greater than when ozone levels were below 60 ppb. Wilson et al15 studied the relationship among sulfur dioxide (SO2 ) and ozone concentrations, weather, and ED visits for respiratory disease in Maine and New Hampshire for a 3-year period. A 10% increase in ED visits for older people (65+) was observed for each quartile (6.3 ppb) increase in SO2 in Portland, Maine. Similarly, Villeneuve and associates10 conducted a case-crossover study in northern Alberta, Canada, to examine the effects of ambient air pollution on ED visits for asthma. The analysis was conducted for all months and by season for each different pollutant (sulfur dioxide [SO2 ], nitrogen dioxide [NO2 ], carbon monoxide [CO], PM2.5 , PM10 , and ozone [O3 ]). Similar analyses were conducted to examine these associations across age groups and genders. The authors found no significant effect of ambient air pollution on asthma ED visits during the winter months. Data analysis representing summer months resulted in significant findings for all pollutants except for SO2 . In addition, they reported that children 2 to 4 years of age and the elderly were the most sensitive to CO (odds ratio [OR] = 1.48 and 1.54 for 5-day average, respectively) and to NO2 (OR = 1.50 and 1.37 for 5-day average, respectively). Paulu and Smith16 reported an overall 11% increase in Maine asthma emergency department visits (2000–2002) associated with a 10-ppb increase in the 8-hour average ozone level (0- to 3-day average lag). The association with average ozone was highest among males aged 2 to 14 (17% increase; 95% CI, 3–32%) and among females aged 15 to 34 (20% increase; 95% CI, 10–31%). Such studies emphasize that there may be groups of susceptible individuals who are differentially affected by air pollution. According to the American Lung Association’s 2007 “State of the Air” report,17 the Pittsburgh region has the second highest concentration of particulate matter (PM) in the nation. Ozone levels are also high in the area, with exceedances of the 1-hour maximum and 8-hour average standards. The aim of this study was to study the relationship of asthma exacerbations within a large urban health care system and ambient air pollution concentrations of ozone and PM2.5 in the greater Pittsburgh area using a case-crossover study design. The differential effect of ambient air pollution on asthma by gender and race was also examined.

104

METHODS Study population Health data for this study were provided by a large metropolitan medical center within the greater Pittsburgh area. The data analyzed for this study represents about 60% of the coverage for this area. The study population consisted of individuals served by this large health care system within Allegheny County (AC), which has a population of 1.3 million people. The Medical Archival Retrieval System (MARS) was used to provide de-identified electronic data on emergency room visits for the medical system network. This study used deidentified data from the MARS database via an electronic medical record abstract where the demographic information is stored. Variables for analysis included date of visit, age, gender, zip code, pseudo-identifier, discharge diagnoses, and race of patient. The 6,979 patients seen in any of 6 EDs from January 1, 2002, to December 31, 2005, with primary asthma (International Classification of Diseases [ICD]9 493.x) discharge diagnosis were included in this study. To meet Health Insurance Portability and Accountability Act (HIPAA) guidelines and ensure patient confidentiality, all data were de-identified using an honest broker system. This study met the criteria for exemption of informed consent by the University of Pittsburgh Institutional Review Board. Daily air pollution and meteorological data Ambient air pollutant levels of ozone and PM2.5 from January 1, 2002, through December 31, 2005, were obtained from the Allegheny County Health Department (ACHD) Air Quality Program. There are 3 monitoring stations for ozone in AC, one at the periphery of the county and one in the central portion representing the most urban area, including the city of Pittsburgh (central Pittsburgh). We used the monitor located in the urban region of AC because the 6 hospitals representing ED visits were located within close proximity to this central site (10 miles). The AC PM2.5 monitors are intermittently operated. Only 2 PM2.5 air monitors are continuously operated and provide daily measurements. Considering the study period, the monitoring station that represented the urban/central point of the county provided more complete air monitoring data and was used in this study. These samples were taken using US Environmental Protection Agency (EPA) sampling guidelines set forth under the Clean Air Act. The daily 1-hour maximum ozone concentration and daily mean PM2.5 concentration were used in our analysis. The meteorological data, including daily mean temperature and humidity, were derived from the US National Climatic Data Center database for the monitoring site located at the Pittsburgh International Airport, Allegheny County.

Archives of Environmental & Occupational Health


Statistical analysis Analyses were conducted using the case-crossover analysis tool (C-CAT) developed by Apex Epidemiology Research in collaboration with the New York State Department of Health for use with SAS (SAS, Cary, NC).18 It provides an easy-to-use interface to SAS code that implements timestratified case-crossover analysis to estimate the association between acute health effects and transient changes in environmental exposures. Because each subject serves as his/her own control, the case-crossover approach controls for the effects of stable subject-specific covariates such as gender and race. If the case and control periods in each risk set are relatively close in time, this method also controls for potential time-varying confounders such as seasonal effects or personal habits by design.19 In this study, the case period referred to the day when asthma ED visits occurred. Controls were defined as all days within a 28-day interval with the same day-of-the-week as the case period. Therefore, a match setting consisted of 1 case period and 3 control periods. The referent day could be on the same day of the week 3 consecutive weeks before and none after; 2 weeks before and 1 week after; or 1 week before and 2 after; or 3 weeks after, all occurring within the 28 days. In order to handle recurring events and to satisfy the assumption of independence for each visit, a “washout” period was applied. The washout was set to remove any repeat visits for a person within the 28-day time period, removing incidents that may represent a continuation of the same health event.20 For example, if an asthmatic sought emergency treatment for an asthma attack on Monday, February 1, and returned at any point between February 2 and March 1, the second visit occurring later in the month would not be counted (“washed out”). C-CAT performed conditional logistic regression to calculate the risk of ozone, PM2.5 , and ozone and PM 2.5 combined. Temperature on the day of the event was evaluated alone, then as a covariate in each conditional logistic regression as we explored the relationship between air pollution and asthma ED visits in our diverse urban population. Because of the potentially lagged effect of air pollution on asthma ED visits, lagged-day exposures up to 5 days and a 6-day average (day 0 to day 5) were used in the analyses. The 6-day averages were calculated as the mean of air pollution levels on the lag 0, 1, 2, 3, 4, and 5 days before the event (or the control day), which allowed us to evaluate the potential cumulative effect of 6 days of air pollution exposure on asthma ED visits. The variables of air pollution concentration for the lag 0, 1, 2, 3, 4, and 5 days and the 6-day average were fitted into separate models and the independent health effects of air pollution on asthma ED visits were estimated for these exposure periods. Results are presented as odds ratios and associated 95% confidence intervals for every 10 ppb increase in concentration for ozone and 10 µg/m3 increase in concentration for PM2.5 . In order to examine additional effect modifiers, we stratified the population by gender and race and further evaluated 2012, Vol. 67, No. 2

Table 1.—-Individuals With a Primary Discharge Diagnosis of Asthma at Emergency Department Visit (6 University of Pittsburgh Medical Center Hospitals) by Age, Sex, and Race Any Primary asthma diagnosis for all individuals (N = 6,979)

Age (years) 0–17 18–29 30–44 45–64 65+ Sex Male Female Race African American Caucasian American Other

n

%

978 1,509 1,800 1,731 959

14.0 21.6 25.8 24.8 13.8

2,380 4,599

34.1 65.9

2,637 4,205 137

37.8 60.3 1.9

the associations between exposure to air pollution and asthma ED visits in these subpopulations. We also assessed whether these associations were similar at different age groups. RESULTS Of the 6,979 people who came to 1 of the 6 hospital emergency departments (EDs) evaluated in this analysis, 65.9% were female and 34.1% were male. The population racial breakdown was 60.3% Caucasian American and 37.8% African American (Table 1). African Americans were significantly (p < .001) younger than Caucasian Americans (mean age 35.2 and 42.4 years, respectively.) A total of 17.7% of Caucasian Americans and 7.6% of African Americans were over the age of 65. Moreover, a total of 18.9% of the African American population was under 14 years of age compared with 10.9% for the Caucasian American population. The number of visits per person during these 4 years ranged from 1 to 19, with an average number of 1.5 (± 1.3). African Americans on average had 1.6 visits during this period compared with 1.3 for Caucasian Americans, denoting a possible health care utilization differential. Ozone was highly correlated with temperature (r = .72) and moderately correlated with PM2.5 (r = .57) (Table 2). Humidity was not associated with asthma ED visits in the conditional logistic regression for any lag day. There is much greater temporal variability observed for ozone than PM2.5 . A significant risk estimate was observed for ozone in a day-2 lag model (OR = 1.025; 95% CI, 1.006–1.044) for the entire population (Table 3). This notable increase in ED visits was associated with the 10 ppb difference in ozone level 2 days prior to the ED visit. This increase in ED visits due to ozone was also noted in a model that included ozone 105

Table 2.—-Correlations Between 1-Hour Maximum Ozone, PM2.5 , and Weather Variables Range PM2.5 (µg/m3) Temperature (◦ F) Humidity (%) Ozone (ppb)


∗ Correlation

PM2.5 Temperature Humidity

0.20–55.17

1

6.50–83.17

.529∗

1

0.00–93.25

.084∗

.034

1

0.00–130

.573∗

.722∗

−.042

Ozone

1

is significant at p ≤ .01 (2-tailed).

and PM2.5 . Ozone in a day-2 lag model with PM2.5 (OR = 1.021; 95% CI, 1.001–1.042) for the entire population was noted (data not shown). A 10 µg/m3 increment in PM2.5 1 day prior to the ED visit (1-day lag) was associated with a 3.6% increased risk of ED visits (OR = 1.036; 95% CI, 1.001–1.073). When we stratified the analysis by race, we observed a small, but nonsignificant increase in risk of ED visits for ozone exposure in days-0–4 lag models for African Americans and day-1, day-2, and day-5 lag models in Caucasian Americans. We observed stronger effects of PM2.5 on ED visits in the African American population compared with the effects in Caucasian American group. For the African American strata, PM2.5 predicted visits to the ED on day-1 lag (OR = 1.055; 95% CI, 1.001–1.112); day-2 lag (OR =

1.067; 95% CI, 1.015–1.122); day-3 lag (OR = 1.053; 95% CI, 1.002–1.106); and the 6-day average (OR = 1.088; 95% CI, 1.001–1.184), whereas the predictions of PM2.5 on ED visits in Caucasian Americans were not significant. There were no significant increases in risks in Caucasian Americans or African Americans in the 2-pollutant model of ozone and PM2.5 . As previously noted, the sample of African Americans was significantly younger than Caucasian Americans. In order to adjust for age, we conducted additional analyses of the subgroup of individuals aged 18 to 64, stratified by race (Table 4). Although the sample size was further reduced by 967 to 3,447 visits for the African American population who had an ED visit, there was a significant effect of PM2.5 on day-3 lag (OR = 1.061; 95% CI, 1.003–1.123) and day-4 lag (OR = 1.058; 95% CI, 1.002–1.118) of exposure. There was no effect observed among the Caucasian American population for ozone or PM2.5 in the 18 to 64 years age group. Although not shown, we considered younger (64 years, n = 959) individuals of both races in a separate analysis. Young people exposed to PM2.5 on day 0 had an odds ratio of 1.012 (0.916–1.118) and for older individuals, it was 1.021 (0.906–1.151). Results are similar to the “all ages” and 18- to 64-year-old groups. We also stratified analyses by gender (data not shown) and found that temperature on day 0 had an independent effect on women (OR = 1.068; 95% CI, 1.034–1.103) but not men, the total population, or by racial group. However, we did not find any significant increase risk of ED visits for ozone and PM2.5 in either men or women analyzed separately.

Table 3.—-Odds Ratios and Confidence Limits of Single-Lag Modelsa for Ozone, PM2.5 , and Asthma ED Visits by Race: All Age Groupsb Total population (N = 10,183) Lag days 1-Hour max ozone 1. Day 0 2. Day 1 3. Day 2 4. Day 3 5. Day 4 6. Day 5 7. 6-Day average PM2.5 1. Day 0 2. Day 1 3. Day 2 4. Day 3 5. Day 4 6. Day 5 7. 6-Day average aAll bAll

106

Caucasian American (N = 5,583)

African American (N = 4,414)

OR

95% CI

OR

95% CI

OR

95% CI

1.001 1.016 1.025 1.008 0.998 0.996 1.022

0.982–1.021 0.998–1.036 1.006–1.044 0.990–1.026 0.981–1.017 0.978–1.013 0.990–1.055

0.982 1.012 1.015 0.989 0.999 1.006 1.001

0.957–1.008 0.985–1.039 0.989–1.041 0.965–1.014 0.975–1.025 0.981–1.031 0.958–1.046

1.021 1.011 1.026 1.025 1.001 0.982 1.034

0.992–1.051 0.981–1.043 0.998–1.055 0.997–1.053 0.974–1.029 0.956–1.009 0.985–1.086

1.005 1.036 1.032 1.017 1.006 0.994 1.040

0.970–1.040 1.001–1.073 0.999–1.068 0.984–1.051 0.974–1.038 0.963–1.027 0.984–1.100

0.999 1.026 0.997 0.972 0.972 0.977 0.975

0.970–1.040 0.977–1.077 0.952–1.044 0.929–1.017 0.930–1.017 0.934–1.022 0.904–1.053

0.983 1.055 1.067 1.053 1.038 1.002 1.088

0.933–1.036 1.001–1.112 1.015–1.122 1.002–1.106 0.989–1.090 0.954–1.053 1.001–1.184

models are adjusted for day 0 temperature. results are for 10-unit increase in ozone and PM2.5 .


Table 4.—-Odds Ratios and Confidence Limits of Single-Lag Modelsa for Ozone, PM2.5 , and Asthma ED Visits by Race: Age Between 18 and 64 Entire population (N = 7,511)


Lag days 1-Hour max ozone 1. Day 0 2. Day 1 3. Day 2 4. Day 3 5. Day 4 6. Day 5 7. 6-Day average PM2.5 1. Day 0 2. Day 1 3. Day 2 4. Day 3 5. Day 4 6. Day 5 7. 6-Day average aAll bAll

Caucasian American (N = 4,064)

African American (N = 3,447)

OR

95% CI

OR

95% CI

OR

95% CI

0.993 0.999 1.021 1.009 1.004 1.004 1.016

0.972–1.015 0.978–1.021 0.999–1.042 0.988–1.030 0.983–1.025 0.984–1.025 0.980–1.053

0.973 1.004 1.004 0.992 1.004 1.014 0.996

0.944–1.003 0.972–1.037 0.974–1.034 0.963–1.020 0.976–1.034 0.986–1.044 0.946–1.048

1.009 1.001 1.024 1.026 1.012 0.991 1.036

0.976–1.043 0.967–1.036 0.992–1.057 0.995–1.058 0.981–1.044 0.961–1.022 0.980–1.095

0.993 1.023 1.039 1.031 1.019 1.007 1.053

0.955–1.034 0.982–1.064 1.000–1.079 0.994–1.071 0.982–1.057 0.971–1.045 0.988–1.122

0.999 1.022 0.994 0.985 0.980 0.989 0.992

0.947–1.055 0.965–1.082 0.941–1.050 0.934–1.039 0.931–1.033 0.939–1.042 0.908–1.084

0.971 1.021 1.055 1.061 1.058 1.033 1.095

0.915–1.031 0.962–1.084 0.996–1.117 1.003–1.123 1.002–1.118 0.976–1.093 0.995–1.205

models are adjusted for temperature. results are for 10-unit increase in ozone and PM2.5 .

COMMENT Air pollution investigations are fraught with the issue of colinearity among environmental factors (temperature, humidity, and air pollutants). A recent National Academy of Science report21 reviewed the health effects of acute ozone exposure and other factors that affect the interpretation of ozone health effects. Of the possible confounders, ambient PM2.5 and weather have raised the most concerns about confounding. Becuase ambient ozone and PM2.5 differ substantially by season and location, the warm and cold season associations observed in some ozone health effects studies are most likely not due to confounding by ambient particles.21 Weather variables (temperature and/or dew point) are consistently considered as both independent predictors of asthma health exacerbations (ED visits) as well as a controlling variable. In this investigation, we chose to use lag day 0 temperature (day of ED visit) for all of our models, as there is the potential of temperature on that day to be a trigger for the exacerbation. Results of our investigation showed that a 10 ppb increase in the 1-hour daily maximum ozone level was significantly related to a 2.5% increase in asthma ED visits 2 days later. This was further validated with a 2.1% increase in asthma ED visits due to ozone on day 2 in a 2-pollutant model of ozone and PM2.5 . For a 1-day lag of PM2.5 , a 3.6% increase in asthma visits for a 10 µg/m3 increase was also noted in the total population (significant at p < .05). This 10-unit increase, however, is proportionally much greater for PM2.5 than for ozone. The ozone levels range was 0 to 130 ppb 2012, Vol. 67, No. 2

with a mean of 40.6 ppb. The PM2.5 range, on the other hand, was much smaller with a mean of 13.3 µg/m3, and a range of 0.2 to 55 µg/m3. These results underscore the stronger physiological effect of ozone on asthma exacerbations. We also noted a differential response in ED visits for African Americans compared with Caucasian Americans. This was the first study of its kind in the Pittsburgh region to consider the effect of PM2.5 and ozone on emergency department visits for asthma exacerbations. The population of African Americans in the city of Pittsburgh is about 27%, and in Allegheny County (including the city of Pittsburgh) is just over 12%, according to 2007 Census Bureau estimates (US Census Bureau Factfinder).22 The reasons for the racial differences in increases in ED visits between African Americans and Caucasian Americans may include urban versus suburban residence or the availability of air conditioning in the summer. Utilization and access to family physicians and specialists as a first-line defense for the treatment of asthma exacerbations has been shown to reduce the need to visit EDs. In addition, these differences may be due to intrinsic biological differences between African Americans and Caucasian Americans. Air conditioning has previously been shown to lessen the effect of air pollution.22 Among Caucasian Americans, 52.3% of all homes in the Pittsburgh Metropolitan Statistical Area (MSA) have central air conditioning,23 whereas the proportion among African Americans was reported to be 23.5%. In Pittsburgh, 32.4% of homes have central air, whereas 65% of those in the rest of the county (suburbs) are climate controlled. Therefore, the inner city 107


population may reflect a group more exposed to ambient air pollution. One of the strengths of this investigation was the quality and specificity of the outcome measure. Only individuals who had an ED discharge diagnosis of asthma were used, thereby removing those who may have come to the ED for another malady and had asthma listed as a secondary or tertiary diagnosis. In addition, our use of a time-stratified case-crossover study design controls for the effects of seasonality, secular trends, and time-invariant factors by study design itself. A general weakness of this and other air pollution and health effects studies is the lack of multiple monitors that can measure variability across regions. In addition, our sample ED visits may not be geographically representative of the population of Allegheny County, Pennsylvania, as the hospitals represented are concentrated in and around the city of Pittsburgh. In summary, using a 2-day lag, each 10 ppb increase in ozone was associated with a 2.5% increase in asthma ED visits. Stratifying the analysis by race demonstrated that African American asthmatics may be more affected by changes in air quality in the Pittsburgh region, possibly reflecting an exposure differential or a difference in health utilization behavior. Future investigations using this case-crossover methodology will be of benefit in order to consider the consistency of the findings across regions, in different populations and in different climates and geographic areas. Moreover, the ability to conduct meta-analyses will result in a more defined set of overall risk measures. The Public Health significance of this project lies in the ability to track potentially serious conditions in a particular environment and link these health effects to putative environmental exposures.

5.

6.

7.

8. 9. 10.

11. 12. 13. 14. 15. 16. 17.

********** This research was funded by CDC grant number 1U19EH000103-02. For comments and further information, address correspondence to LuAnn Lynn Brink, PhD, Department of Epidemiology, University of Pittsburgh, 130 DeSoto Street, Pittsburgh, PA 15261, USA. E-mail: [email protected]

**********

18. 19. 20.

References 1. Centers for Disease Control and Prevention. National Surveillance for Asthma—United States, 1980–2004. MMWR Surv Summ. 2007;56(SS8):1–14. 2. Deugen S, Zmirou-Navier, D. Social inequalities resulting from health risks related to ambient air quality—a European review. Eur J Pub Health. 2010;20:27–35. 3. Boutin-Forzano S, Adel N, Gratecos L, Jullian H, et al. Visits to the emergency room for asthma attacks and short-term variations in air pollution. A case-crossover study. Respiration. 2004;71:134–137. 4. Burnett RT, Brook JR, Raizenne ME, Krewski D. Association between ambient carbon monoxide levels and hospitalizations for con-

108

21. 22. 23.

gestive heart failure in the elderly in 10 Canadian cities. Epidemiology. 1997;8:162–167. Steib DM JS, Burnett RT. Meta-analysis of time-series studies of air pollution and mortality: effects of gases and particles and the influence of cause of death, age, and season. J Air Waste Manage Assoc. 2002;52:470–484. Lin M , Chen Y, Burnett RT, Villeneuve PJ, Krewski D. The influence of ambient coarse particulate matter on asthma hospitalization in children: case-crossover and time-series analyses. Environ Health Perspect. 2002;110:575–581. Lin M, Chen Y, Burnett RT, Villeneuve PJ, Krewski D. Effect of shortterm exposure to gaseous pollution on asthma hospitalization in children: a bi-directional case-crossover analysis. J Epidemiol Community Health. 2003;57:50–55. Burnett RT, Smith-Doiron M, Stieb D, et al. Association between ozone and hospitalization for acute respiratory diseases in children less than 2 years of age. Am J Epidemiol. 2001;153:444–452. Yang Q, Chen Y, Shi Y, Burnett RT, McGrail K, Krewski D. Association between ozone and respiratory admissions among children and the elderly in Vancouver, Canada. Inhal Toxicol. 2003;15:101–112. Villeneuve PJ, Chen L, Rowe BH, Coates F. Outdoor air pollution and emergency department visits for asthma among children and adults: a case-crossover study in northern Alberta, Canada. Environ Health. 2007;6:40–55. Garty BZ, Kosman E, Ganor E, Berger V, et al. Emergency room visits of asthmatic children, relation to air pollution, weather, and airborne allergens. Ann Allergy Asthma Immunol. 1998;81:563–570. Delfino RJ, Murphy-Moulton AM, Becklake MR. Emergency room visits for respiratory illnesses among the elderly in Montreal: association with low level ozone exposure. Environ Res. 1996;76:67–77. Weisel CP, Cody RP, Lioy PJ. Relationship between summertime ambient ozone levels and emergency department visits for asthma in central New Jersey. Environ Health Perspect. 1995;103(Suppl 2):97–102. Glad JAM. The Use of Electronic Medical Records Based on a Physician Diagnosis of Asthma for County-Wide Asthma Surveillance [dissertation]. Pittsburgh: University of Pittsburgh, 2008. Wilson AM, Wake CP, Kelly T, Salloway JC. Air pollution, weather, and respiratory emergency room visits in two northern New England cities: an ecological time-series study. Environ Res. 2005;97:312–321. Paulu C, Smith A. Tracking associations between ambient ozone and asthma-related emergency department visits using case crossover analysis. Journal of Public Health Manage Pract. 2008;14:581–591. American Lung Association. State of the Air: 2007 Report. Available at: http://www.lungusa.org/assets/documents/publications/state-of-theair/state-of-the-air-report-2007.pdf. Accessed June 17, 2011. Bateson JH. Case-Crossover Analysis Tool (C-CAT) Beta Version 1.1. User’s Manual. Albany: New York State Department of Health, Apex Epidemiology Research; 2006. Fung K, Krewski D, Chen Y, Burnett R, Cakmak S. Comparison of time series and case-crossover analyses of air pollution and hospital admission data. Int J Epidemiol. 2003;32:1064–1070. Haley VB, Bateson TF, Boothe VE, et al. For Environmental Public Health Air Surveillance Evaluation Team. Case Crossover Analysis of Acute Health Effects and Ambient Air Quality For Environmental Public Health Tracking. Technical Background Report. Atlanta, GA: Environmental Public Health Tracking Branch, Centers for Disease Control and Prevention; 2007. Bailar J, ed. Estimating Mortality Risk Reduction and Economic Benefits From Controlling Ozone Air Pollution. Washington, DC: National Research Council of the National Academies of Science; 2008. Bell ML, Ebisu K, Peng RD, Dominici F. Adverse health effects of particulate air pollution: modification by air conditioning. Epidemiology. 2009;20:682–686. US Department of Housing and Urban Development and US Census Bureau. Current Housing Reports, Series H170/04-13, American Housing Survey for the Pittsburgh Metropolitan Area; 2004.
