Sunday, August 24, 2008
South/West Halls
In 2003 South Carolina was one of the first states to adopt the NEDSS Base System (NBS). Both sophisticated and simple tools are used in parallel to analyze data from the NBS and to make these available to disease control epidemiologists. The Disease Control Information System (DCIS), presented in its initial iteration at a previous PHIN meeting, has been assembled from small-footprint, open-source or public domain software such as EpiInfo for DOS, Epidata, SQLite, and DOS batch files. These work together "behind the scenes" while a freeware menuing system, Epiglue, provides a GUI for end-users. The system can run from a USB pocket drive. DCIS provides a user friendly "front end" for data which are integrated daily following extraction from the NBS, from legacy NETSS data, as well as from TIMS and several other surveillance data sets. Initial iterations of DCIS focused on traditional data presentation such as case line-lists, statistical summaries, and menu-driven export of data subsets of interest to Excel, Access, SAS, EpiInfo or Epidata according to user preferences. In this presentation we will focus on second-generation DCIS tools including the R public public domain package as used here for daily automated creation of surveillance data graphic summaries. The flexibility of R allows these to go beyond standard graphics (e.g. basic bar or pie charts) and additionally offer more sophisticated data visualization and display. Examples include "process control charts" and "rate-analysis regression charts". DCIS also takes advantage of the oft underutilized capabilities MS-Office macros. For example, R graphics are daily auto-incorporated into a menu-driven PowerPoint file which is then simply copied to end-users PCs whenever they launch DCIS. Continued refinements of the DCIS system demonstrate the ongoing utility of simple small-footprint free software tools even in the sophisticated IT era of NEDSS and PHIN.