Epidemic disease global information system unit

Goals / Objectives
To serve as a resource to facilitate the control of infectious diseases of domesticated animals in the Commonwealth of Pennsylvania by creating and maintaining a GIS database for the State which will contain the location of animal agriculture enterprises in the State. Relevant information pertaining to each animal type will be collected and recorded in the GIS database. Spatial models for the dynamics of the transmission of specific animal disease threats will be created. These models will utilize specific geographical information concerning location, size and type for each animal enterprise in the State.

Project Methods
An Epidemic Disease Geographical Information System (EDGIS) has been established at our facility. Information that will comprise the system database is being collected from a variety of sources and particular attention is being paid to maintaining a level of confidentiality for the information in the database. Ways in which information relevant to disease control and mitigation can be best abstracted and used are being explored.

Progress 01/01/02 to 12/31/02

Outputs
An epidemic disease GIS (EDGES) research unit has been established. The main objective of the EDGIS unit was to create and maintain a spatial database, using ArcView software, of all the livestock farms in Pennsylvania. Currently, the EDGIS unit has collected data for beef, dairy, poultry, and swine farms. We are now in the process of collecting data for sheep and deer. We have placed nearly 30 percent of the farms onto a map through address matching known as geocoding. To improve on the matching percentage an Avenue script was created which queried the best results of the two geocoding services (Street Map and GDT). As a secondary function, the script removed all duplicate records from the database. After processing the database through the script, the matching percentage increased on average about 3 percent per database. Although there was not a significant increase in the number of farms matched the number of farms that matched with greater accuracy (80-100 percent accuracy) was significant. Matching the remaining unmatched farms is our next challenge. For a variety of reasons many of the addresses in the animal data base cannot be geo-coded. For example, fifty percent of the records in the animal databases are rural routes or unknown roads. Those records that are unknown roads can be discarded because we cannot process those records. With regard to rural routes, we have been able to improve the proportion of geo-codable addresses by submitting the data set to ClientLogic,to correct address lists for areas that have undergone permanent address conversions. The improvement was small but significant and we shall be able to repeat the process as more rural route addresses undergo permanent address conversions. Given the difficulties referred to above, we have spent the last two months in a quality assurance exercise in which we attempted to measure the reliability of our current mapping by a direct on-the-ground comparison with existing farms. We chose to use the Brucellosis milk ring test address list provided to us by the PDA as our basic data set. We randomly selected four quadrats (about 4.5 miles square. Each quadrat was visited by one or two research staff for a period of two to three days each and portable GPS units used to map the road-side location of every dairy farm found. Our analysis of the results is in progress

Impacts
(N/A)

Publications

No publications reported this period