|
Activity Number:
|
127
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #307934 |
|
Title:
|
Assessing Spatial Performance of Surveillance Systems
|
|
Author(s):
|
Lance Waller*+ and Yuemei Wang and Carol Lin
|
|
Companies:
|
Emory University and Emory University and Bristol-Myers Squibb Company
|
|
Address:
|
1518 Clifton Road NE, Atlanta, GA, 30322,
|
|
Keywords:
|
spatial statistics ; surveillance ; system performance ; sensitivity ; specificity
|
|
Abstract:
|
Many disease outbreak surveillance systems are designed based on concepts from statistical process control and assess performance based on temporal summaries such as the expected time until false or true alarms. While these important temporal aspects of performance play key roles in assessing performance, spatial aspects of system performance also merit attention. More specifically, most environmental surveillance systems often monitor health outcomes or exposures across a defined geographic space, but this space is not homogeneous with respect to local sample size and local uncertainty. We explore aspects of system performance addressing spatial summaries of statistical power (specificity) and sensitivity (1 - false alarm rates), and illustrate the importance of evaluating both when and where events of interest are likely to occur in determining system performance.
|
