Flooding and Asthma in Kentucky ASTHO Environmental Public Health Tracking Fellowship Phase II Report Submitted By: Colleen Kaelin, MSPH, RS Epidemiologist II Kentucky Department for Public Health 275 East Main Street Frankfort, KY 40621 Submitted To: Association of State and Territorial Health Officials Environmental Public Health Tracking: State-to-State Peer Fellowship Program 2231 Crystal Drive, Suite 450 Arlington, VA 22202 September 2013
I. Project Description After consulting with the Office of Administrative and Technical Services, it was decided that the resources that would be required to construct a demonstration tracking portal for the state of Kentucky in the time frame specified by the Phase II RFP were not available. The Kentucky Environmental Public Health Tracking Work Group decided that a study examining the effect of flooding in Kentucky on asthma hospitalization would be a beneficial data linkage project. Kentucky is often affected by flooding, usually in the spring, and the prevalence of asthma in Kentucky is higher than the national average. In 2008, the overall prevalence of asthma in Kentucky was 9.5%, compared to 8.7% for the United States as a whole. The rate of asthma hospitalizations for Kentucky was also higher than the national rate for the same year. According to the Environmental Public Health Tracking Network, the percent of children in Kentucky ever diagnosed with asthma increased from 13.7% in 2004 to 14.7% in 2010. Exposure to mold and wet conditions are known risk factors for asthma and other respiratory conditions. An International Study of Asthma and Allergies in Childhood (ISAAC) of 46,051 children in 20 countries established the link between residential dampness and respiratory symptoms. In addition to the data linkage project, a second goal of the Phase II project was to establish a Technical Advisory Group (TAG) for Kentucky s Environmental Public Health Tracking Program. The goals of the Kentucky TAG are to develop a plan to present data on environmental hazards, exposures, and chronic diseases based on the Nationally Consistent Data and Measures in a format that is accessible to the public and to develop a communications plan to promote environmental health tracking in our state. The Communications Plan will be modeled after the 2012 Tracking Network Communication Standards and Recommendations along with the Risk Communication Plans of the Florida and Maryland Tracking Programs. II. Activities The first step in conducting the data linkage project was to select a graduate student to serve as the project s Principal Investigator. After interviewing several candidates, we choose Magan Meade, an MPH candidate from Eastern Kentucky University. She immediately began accessing the necessary data elements from the Office of Health Policy and the National Weather Service. The outcome and data analysis of the project will be addressed in the next section. The next activity was the formation of the Technical Advisory Group. The initial TAG meeting took place on August 14 th, 2013. Magan Meade attended in person along with Colleen Kaelin, the environmental epidemiologist, Jimmie Patterson, the representative of the Office of Administrative and Technical Services and Deepa Valvi, the Department for Public Health s Asthma epidemiologist. Attending by conference call were Dr. John Braggio of the Maryland Tracking Program and Sara Walsh, of the Foundation for a Healthy Kentucky. The Foundation is a non-profit organization founded in 2001 with the mission to address the unmet health care needs of Kentuckians. This organization sponsors the Kentucky Health Facts website, which
presents data on some of the same content areas as the Environmental Public Health Tracking Network, such as smoking and asthma prevalence. On August 27 th, 2013, Ms. Colleen Kaelin attended the ASTHO Fellows Networking session at the Chamblee Campus of the Centers for Disease Control in Atlanta, Georgia. The process of submitting Kentucky s data for the Community Water content area to the National Environmental Public Health Tracking Network was discussed. The other fellows discussed their data linkage and capacity building projects as well. After the presentations, the fellows discussed the possibility of collecting and submitting hospitalization data to the national tracking program. It was decided that the fellows would commit to this goal as soon as possible. The Kentucky Department for Public Health conducted a virtual meeting with the Florida tracking staff on September 13 th, 2013. The topics discussed included content development and user tools for the Florida tracking portal, Florida s risk communication plan and developing a communication plan for Kentucky, submitting hospitalization data to the national tracking program, and the ways that Florida has reached out and used the tracking program to promote public health. A recording of the presentation and the accompanying slides are available on the Kentucky Tracking Program Share Point website and at http://www.youtube.com/watch?v=e4wraz9w-hm&feature=youtu.be. On October 30 th, the Building Epidemiologic Capacity in Kentucky (aka BECKY) workgroup will hold its annual meeting in Frankfort. Colleen Kaelin will give an update of the national tracking program and a progress report on environmental public health tracking in Kentucky to the workgroup. This should help generate support for public health tracking in Kentucky and form new contacts for the tracking and technical advisory workgroups. Also, the next annual conference of the Kentucky Public Health Association is scheduled for April 14-17, 2014. The tracking workgroup will distribute promotional materials about the environmental public health tracking network and the various communication toolkits at the conference. If possible, Colleen Kaelin will repeat her presentation from the BECKY meeting at this conference. III. Outcomes/Data Analysis The purpose of the Phase II data linkage project was to examine the association between flooding and hospitalization for asthma and related diagnoses in Kentucky, particularly in the Western region of the state. Magan Meade, the principal investigator, selected the variables for the study and designed the data analysis. Flooding data was gathered by the National Weather Service (NWS). A flood is defined by NWS as any high flow, overflow, or inundation by water which causes or threatens damage. Flash flooding is defined as a rapid and extreme flow of high water into a normally dry area, or a rapid water level rise in a stream or creek above a predetermined flood level, beginning within six hours of a causative event on a widespread or localized basis. The storm event database provided the number of days of flood
and flash flood events and flood and flash flood warnings per county from 2006 to 2012. Data on hospitalization for asthma and related diagnoses was provided by the Office of Health Policy in the Kentucky Department for Public Health. The diagnostic codes collected were ICD9 493-493.99, 490, 491, 4910, 4911, 4912, 49120, 49121, 49122, 4918, 4919, 492, 4920, 4928 and 496. Case counts were provided for 2000-2012, but the analysis of the data was limited to 2006-2012 since flooding data for 2000-2006 was not available. Data was organized by year, geographic region, county, month, and season. Case counts of less than 5 were suppressed. To estimate the rate of hospitalizations per 100,000 persons where the data was suppressed, the total annual number of cases per county was divided by the number of months with a case count of less than five and the result was extrapolated into the empty cells. Smoking and socioeconomic status, as determined by median income per county, were included in the study as potentially confounding variables. Data on the percent of smokers per county was provided by the Behavioral Risk Factor Surveillance System, a telephone survey of health status sponsored by the Centers for Disease Control and administered by state health departments. Data on median household income per county was collected using the U.S. Census tool, State and County Quick Facts, at http://quickfacts.census.gov/qfd/states/21000.html. To simply the analysis of the data and more accurately determine the impact of flooding on asthma hospitalization; the counties were organized into the six geographic regions of the state. According to the University of Kentucky, the geographic regions are the Eastern Coal Field, the Bluegrass Region, the Knobs Arc, the Pennyrile region, the Western Coal Field, and the Jackson Purchase region. The map below displays the state of Kentucky with county boundaries and geographic regions: We hypothesized that flooding has a significant impact on the rate of hospitalization for asthma and related diagnosis apart from the effect of smoking and socioeconomic status. It is a known fact that moisture from flooding causes an increase in mold and particulates in the air, both inside and outside the home, which causes lung irritations and allergic reactions related to
asthma. We also theorize that climate change has caused an increase in the number of floods and flash floods in Kentucky, and that the number of floods and flash floods varies significantly between the geographic regions of the state. As expected, there was significant variation in the number of flood and flash flood events from season to season and year to year. The highest number of flood events occurred in the spring months of March, April and May, while the highest number of flash flood events occurred in the summer months of May, June and July. The number of flood events also varied significantly from region to region. The Jackson Purchase and Western Coal Field regions had the highest percent of flood events, while the Jackson Purchase and Eastern Coal Field regions had the most flash floods. Using the percent of total flood events took into account the large variation in the number of counties in each geographic region. The percent of total flood events by geographic region and year is displayed in the following graph:
There were also wide variations in the average rate of hospitalization for asthma and related diagnosis by season, year and geographic region. The average annual rate of asthma hospitalization per 100,000 persons by season is displayed in the graph below. The geographic variation in asthma hospitalization for the time period of 2006-2012 is displayed in the graph below:
The highest average rates of asthma hospitalization were in the Eastern Coal Fields and Jackson Purchase regions. The counties with the highest rate of asthma hospitalization were Monroe County in the Pennyrile region (122.27 cases per 100,000), Bell County in the Eastern Coal Field region (73.19 cases per 100,000) and Fulton County in the Jackson Purchase region (61.33 per 100,000). The mean rate of asthma hospitalization by year and geographic region is displayed in the graph below:
The impact of flood events on asthma was analyzed using a series of linear regression models with the SPSS computer program. Since flooding is a fairly rare outcome and numbers were small, the data was also analyzed using Poisson regression. Calculating the impact of flood events on hospitalization for asthma and related diagnosis gave a correlation coefficient of 0.3 (3%). Therefore 3 percent of the variability of asthma hospitalizations between geographic regions is accounted for by floods. Similarly, the impact of flash flood incidents on asthma hospitalization gave a correlation coefficient of 0.2 (2 percent). However, combining the data on flood and flash flood incidents gave a correlation of 0.001, which shows no association. Analysis by Poisson correlation shows a decrease in hospitalizations for asthma and related diagnoses during the months of a flash flood event. We can speculate that the large amount of rain during such brief and severe storm events reduces the amount of air particulates that exacerbate asthma and other respiratory conditions. Analysis of the data determined that higher socioeconomic status has a negative association with the incidence of asthma hospitalization. As the socioeconomic status increased, the rate of hospitalization of asthma and related conditions decreased. The correlation coefficient for
this variable was 0.26, so 26 percent of the variability in asthma hospitalization between the geographic regions could be accounted for by differences in socioeconomic status. The correlation coefficient for smoking and the rate of hospitalization was 0.19; therefore almost 20 percent of the variability in hospitalization rates could be accounted for by differences in the percent of smokers by geographic region. A linear regression model with all the variables was used to examine the interaction of the factors included in the study on the outcome, i.e. hospitalization for asthma and related diagnoses. The model using median income, percent of smokers, and number of flood and flash flood events accounted for only 7 percent of the variability among hospitalizations, but the association was statistically significant at the <0.5 level. Therefore, the model with these variables indicates that flooding, combined with smoking prevalence and median income are significantly related to the variation in rates of hospitalization for asthma and related diagnosis, but a large portion of the variation is left unaccounted for. There are several limitations in this study. First, the percentage of smokers per county may be underreported since the Behavioral Risk Factor Surveillance Study is a selfreported health survey. Also, median income alone may not fully represent the true socioeconomic status of each county. Other factors which are known to affect the incidence of asthma hospitalization, such as age and gender, were not included in this study. The relationship between climate and public health is a topic that will become more important in the coming years, and it will continue to be an essential component of the environmental public health tracking program. IV. TAG Recommendations and Action Plan The first topic discussed at the August 14 th, Technical Advisory Group Meeting was the various options for presenting Kentucky s tracking data and the advantages and disadvantages of each. The first option is to complete the content areas on the national Environmental Public Health Tracking Network as contained in the Nationally Consistent Data and Measures document. The incomplete content areas for Kentucky on the national tracking portal are Birth Defects and Hospitalization for Asthma, Carbon Monoxide Poisoning and Myocardial Infarction. The advantages of this option are that since community water data has already been submitted by Kentucky and is currently displayed on the national portal, contacts within the national tracking program and the necessary infrastructure are already in place. The national tracking portal also has established rules for data suppression and confidentiality, so we will not need to develop rules of our own. There will be no additional cost for personnel, computer software or office supplies with this option. The disadvantages of this option are that data sharing agreements will have to be developed with the data shareholders of the remaining non-reported content areas, and that data is submitted to the national program on their schedule. Advanced query options for Kentucky are sometimes not available or give very little usable information due to the small population size and numbers of health events. Sub-county level data is not available on the national portal for this reason. Finally, there is not a great deal of awareness in Kentucky among the general public or public health decision makers regarding the environmental public health tracking program or the national tracking portal.
The next option to present tracking data for Kentucky is participation in the mid-atlantic regional tracking network sponsored by the Maryland tracking program. The Mid-Atlantic regional portal utilizes the same CDC approved software and infrastructure as the Maryland Environmental Public Health Tracking Portal. The regional tracking portal offers several advanced on the fly mapping and analysis options at the county and sub-county levels which can be controlled or turned off by the user. The regional tracking program has the capacity to build a Kentucky specific tracking portal at our request, if we provide the funding. Using this option, we would be able to submit and upload data at our convenience. The disadvantages of the Mid-Atlantic portal are that there would be some cost for processing and/or storing the Kentucky data, either on data servers located in Maryland or on site in Kentucky. The advanced data query options would probably not be widely used by the general public, and the display of sub-county level data would be limited due to the small population and case numbers. Finally, the a detailed memorandum of understanding between Kentucky and Maryland would have to be drafted and approved by the administration of both states at several levels in order for data to be submitted to this network. The final option for presenting Kentucky tracking data to the public is to add data to the Kentucky Health Facts website sponsored by the Foundation for a Healthy Kentucky, a nonprofit organization founded in 2001 as a result of a settlement agreement between the Commonwealth of Kentucky and Anthem, Inc. following Anthem s merger with Kentucky Blue Cross/Blue Shield. The kentuckyhealthfacts.org website is query able and presents data by location and by topic on demographics, social and behavioral indicators, health outcomes, access to care, maternal and child health, and senior health. Several content areas on this website are similar to or the same as the national Environmental Public Health Tracking Network. Data is presented in table, graph or map form, as in the national tracking portal, and can be presented at county or Area Development District level. This website is already well known and used frequently by the public, with a few thousand visitors each month. The Foundation is experienced and active in public outreach, and a Memorandum of Understanding between the Foundation and the Kentucky Cabinet for Health and Family Services is already in place. There would be no expense to add tracking data to this portal other than the time and energy of the data shareholders and the environmental epidemiologist. The disadvantages to adding data to the Kentucky Health Facts website is that the focus of the site is presenting data, rather than analyzing it, so that the more advanced options for searching and mapping data that are available on the national and Mid-Atlantic portals would not be available. The data shareholders would have to determine their own standards for data suppression and confidentiality, and sub-county level data would not be available. The TAG has not yet made any final determinations on which option would be best for presenting Kentucky environmental public health tracking data. Given the current funding status, experience of the Kentucky tracking staff and IT infrastructure, completing the content areas on the national portal while submitting additional data on health outcomes to the Kentucky Health Facts website is the most viable option. It is likely that all three options will be utilized at some point in the future.
The next subject addressed by the Technical Advisory Group was the development of a Risk Communication/Public Outreach plan for the Kentucky Environmental Public Health Tracking program. Florida and Maryland have both shared their plans with the Technical Advisory Group and they will be used to develop an outline/template for Kentucky s plan. The proposed topics of the Kentucky plan will include, but not be limited to: -Definition of Environmental Public Health Tracking and the History of the Environmental Public Health Tracking Network -How do we use the environmental public health tracking portal? -Success Stories: How environmental public health tracking is used to protect and improve public health -General risk communication principles and strategies -Intended target audiences and messages The Communication Toolkits and other outreach materials developed by the national tracking program will be used and adapted by the Kentucky EPHT workgroup to inform the public and health policy decision makers in Kentucky about the benefits of environmental public health tracking at the national, state and local level. Since the initial meeting, other interested parties have been invited to participate in the Technical Advisory Group, including the Department for Public Health Office of Communications and the Schools of Public Health at Eastern Kentucky University and the University of Kentucky. The next meeting of the TAG is scheduled for October 16 th, 2013. V. Contacts and Collaborations Several organizations have participated in the environmental public health tracking activities of the Kentucky Department of Public Health, both in the first phase and in the ongoing capacitybuilding activities of the current phase. Within the Kentucky Department for Public Health, the Office of Health Policy has provided hospitalization data for both of the data linkage projects sponsored by the ASTHO fellowship program. Representatives of the Office of Maternal and Child Health, where the Kentucky Birth Surveillance Registry is located, have also participated in the tracking workgroup and are aware that data from their program is an element of the EPHT Network as is the Office of Vital Statistics. Other state government agencies that have been crucial in our tracking efforts have been the Office of Administrative and Technology Services and the Kentucky Department for Environmental Protection. Involving the Office of Communications in the Cabinet for Health and Family Services in our state tracking activities will be a primary focus of the next phase of capacity building. There are several groups in the Department for Public Health that have been involved in and made aware of environmental public health tracking, including the Building Epidemiologic Capacity in Kentucky (BECKY) group, the Data Users Workgroup, the epidemiologists in the regional health departments and the Kentucky Asthma Partnership. These groups meet regularly and contain representatives of several tracking data shareholders.
External agencies that have been supportive of environmental public health tracking in Kentucky have included the EPHT programs in Florida and Maryland and the schools of public health at several Kentucky universities. The Kentucky Cancer Registry and the Regional Poison Control Center have also participated in the state s tracking work group. The second phase of capacity building will involve communication and collaboration with the Foundation for a Healthy Kentucky, the Kentucky Public Health Association and the Kentucky Medical Association. VI. Accomplishments and Challenges The primary accomplishments of the first phase of the Fellowship program were the formation of a Kentucky Environmental Public Health Tracking Workgroup, with an EPHT informational web page and Share Point website, the publication of a data linkage project examining the effect of exposure to a set of environmental hazards on a group of chronic disease conditions, and the submission of community water data to the national Environmental Public Health Tracking Network. The challenges of the first phase were primarily to develop awareness of environmental public health tracking within the Kentucky Department for Public Health by making contact with the data shareholders and informing them of how the EPHT portal could make their data accessible to the public and to decision makers, thereby improving public health. There were many technical challenges involved in the data linkage project and submitting data to the national network, including making contact with the national tracking staff and gaining permission to access the Secure Access Management System and the national EPHT Share Point website. There were many new experiences for the environmental epidemiologist in analyzing multiple variables for the data linkage project, as well as learning how to format, process and submit data to the national tracking portal. Guidance from the established tracking programs in Florida and the author of the drinking water SAS program in Colorado were critical in accomplishing the goals. The challenges of the second phase have been similar to the first. The planning and analysis of the second data linkage project has been a complex process involving collaboration between several agencies and many steps of data collection and statistical analysis. The selection of a graduate student was a lengthy process, most especially because there was no response to the initial announcement of an available internship position. This delayed the entire second phase of the fellowship project substantially. The subject of the data linkage study was not entirely within the scope of the Nationally Consistent Data and Measures, therefore the available data was limited and no prior studies of this kind could be referred to for a method of analysis. Awareness of environmental public health tracking and the EPHT web portal is still limited both among the general public as well as key decision makers who have the ability to promote cooperation from data shareholders. The remaining content areas contain information on individual medical diagnosis, so it will be necessary to address the data shareholders concerns regarding confidentiality and security of the information. Aside from the ASTHO Fellowship Program, there is no funding source for environmental public health tracking, so all tracking activities to date have utilized existing infrastructure and resources within the Department for Public Health and publically available through the national tracking program. This has proven
to be an effective strategy so far, but a funding source specifically dedicated for this purpose will eventually be necessary for Kentucky to have a fully functioning environmental public health tracking program. VII. Lessons Learned The importance of communication and collaboration both within state Department for Public Health agencies and other external partners in moving our tracking program forward cannot be overstated. Regular contact with data shareholders such as the Department for Environmental Protection and external partners such as the Florida environmental tracking program is the critical element not only in acquiring the raw elements of the nationally consistent data and measures but also for gaining expertise and resources in communication and data processing. Exploring and using existing resources in IT, public outreach, and risk communication, like the Kentucky Health Facts website and the risk communication plans shared by the tracking grantees, will be essential in building environmental public health tracking in Kentucky. Another lesson is to keep the number of variables and data analysis as simple as possible when linking environmental exposures and health impacts. Extremely complicated analysis is often necessary when examining multiple variables in health and exposure, and the results almost never point to a clear case of cause and effect. Finally, we have learned that building capacity in environmental public health tracking without outreach to the general public and public health decision makers is of little value. Environmental public health tracking is only useful when the information is used by members of the public to improve their everyday health decisions and by elected officials to make informed health policy decisions. VIII. Recommendations for Improvement of the Grant In order to build an environmental public health tracking portal, most state health departments will have to acquire new IT infrastructure and personnel. It would be beneficial if a checklist of necessary tracking program components and their approximate price along with position descriptions could be made available so that localities who were considering building a tracking portal could compare their existing resources with the list and estimate the cost and time they would need to initiate a tracking program. Giving fellowship members access to the national EPHT Share Point website would allow them to review the data dictionaries, how to guides, and data submission webinars so that they could learn how data is sent to the national network and decide what content areas of data they can collect, send and display. Adding a section for ASTHO fellows to the Share Point site would allow the fellowship alumni to share their contacts, resources, questions and experiences with other fellowship members, their tracking mentors, and the national EPHT staff. Regular conference calls or virtual meetings of the fellowship alumni groups would also be helpful in accomplishing this goal and promoting environmental public health tracking in non-grantee state and local health departments.
Communication Plan Outline I. Purpose or Goal Statement The goal of the Kentucky Environmental Public Health Risk Communication Plan will be to inform the general public of the availability and purpose of the Environmental Public Health Tracking Network and to encourage key public health decision makers to use the network in making better and more informed health policy decisions. Audience Activities Timeline Notes Evaluation EPHT October awareness 16 th 2013 questionnaire Epidemiologists in KY DPH and Regional HD s On the Building Epidemiologic Capacity in Kentucky contact list The Building Epidemiologic Capacity in Kentucky Workgroup The Kentucky Public Health Association EPHT Update presentation EPHT Update presentation and/or fact sheet distribution October 30 th, 2013 April, 2014 This activity will gage awareness of the Environmental Public Health Tracking Network among epidemiologists in Kentucky and determine what tracking areas/ activities they consider relevant to their jobs. Colleen Kaelin will give a presentation on new developments in the national tracking program and Kentucky s progress Colleen Kaelin will either repeat the BECKY presentation and/or The response rate to the questionnaire and the responses. The number of attendees at the BECKY conference and the conference feedback The number of fact sheets distributed and/or feedback from the The results will be discussed at the BECKY conference on 10-30-13
Environmental Health Group Symposium at EKU Student/faculty environmental health group Meeting with Colleen Kaelin March 2014 distribute communication materials from the national program at one of the informational booths Colleen Kaelin will meet with the Environmental Public Health Group at EKU to discuss EPHT and its uses conference. Number of attendees.
References Asthma Fact Sheet-Kentucky vs. United States, 2008. Kentucky Asthma Program. http://chfs.ky.gov/nr/rdonlyres/51b15d9d-5214-403c-879f- A62412151DAA/0/USvsKentuckyFactSheet.pdf Dampness and moulds in relation to respiratory and allergic symptoms in children: results from Phase Two of the International Stud of Asthma and Allergies in Childhood (ISSAC Phase Two). Weinmayr, G., et. al. Children s Environmental Health Network Article of the Month, September 2013 Issue. http://www.cehn.org/files/september_aom_2013.pdf Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service. (2007). National weather service instruction 10-1605 (NWSPD 10-16). Retrieved from website: www.nws.noaa.gov/directives/ University of Kentucky, (2013). Kentucky s geographic regions. Retrieved from National Digital Newspaper Program website: http://www.uky.edu/libraries/ndnp/regions.html.