Date of Award
Master of Public Health (MPH)
Douglas W. Roblin, PhD
Lisa M. Casanova, PhD
Christina H. Fuller, ScD
INTRODUCTION: Syndromic surveillance is a method of rapid disease detection based on categories of syndromes, or signs, experienced before the full onset of disease. It is increasingly being used by government agencies and health departments to identify disease outbreaks in a timely manner. Environmental exposures are known to induce respiratory and gastrointestinal symptoms, tend to have a seasonality component, and adversely affect the health of millions of people.
OBJECTIVE: In this study, we assess the availability of environmental exposure data for air pollution (PM2.5, ozone, and NO2), pollen, and water contaminant exposure for use in a syndromic surveillance project. We also evaluate: 1) the general proximity of HMO populations to monitors, and 2) distribution of SES characteristics of the area populations with respect to monitor locations.
METHODS: We collected exposure data, patient population data, and Census tract SES data for two metropolitan areas where Kaiser Permanente (KP) provides medical services: Atlanta, Georgia and the northern Virginia, District of Columbia (DC), and Baltimore area. Exposure data for air pollution and pollen were collected for 2013-2014. Straight-line distance from a monitor to the nearest KP clinic, and from each Census tract centroid, to the nearest air pollution or pollen monitor was computed using the Euclidean distance formula.
- Air pollution is routinely monitored by a Federal mandate, is universally available, and easily obtained. Pollen data is collected by private entities, which in some cases hinders access. Water quality data is generally publically available, but it is collected at the source and not easily traceable to water delivery endpoints.
- In both Atlanta and DC, Maryland, and Virginia most of the clinics (78% and 94%, respectively) are located within 10 miles of an air pollution monitor; approximately 83% and 94% of the KP populations were located within 10 miles of an air pollution monitor.
- SES populations differ substantially by race, age, income, and education with respect to the nearest monitor. However, the median and interquartile range of various air pollutants does not differ much across the monitors – indicating that, on average, there is little SES gradient in type of level of air pollution exposure.
CONCLUSIONS: Overall, this study adds knowledge regarding future considerations about the coverage of environmental monitors and to what extent exposure measure estimates can be assigned to certain populations located near monitors.
Johnson, Nolan, "An Assessment of the Feasibility of Environmental Exposure Data for Syndromic Surveillance." Thesis, Georgia State University, 2015.