Loading...
Please wait, while we are loading the content...
Similar Documents
Evaluating the Performance of a Spatial Scan Statistic Using Simulated Outbreak Characteristics
| Content Provider | Semantic Scholar |
|---|---|
| Author | Grannis, Shaun J. Egg, James Cassa, M. Eng Christopher A. Olson, Karen Mandl, Ken Overhage, J. Marc |
| Copyright Year | 2007 |
| Abstract | OBJECTIVE To characterize the performance of a spatial scan statistic, we used SaTScanTM to measure the sensitivity and positive predictive value for detecting simulated outbreaks having varying size, case density, and syndrome type. BACKGROUND Research evaluating the use of spatial data for surveillance purposes is ongoing and evolving. As spatial methods evolve, it is important to characterize their effectiveness in real-world settings. Assessing the performance of surveillance systems has been difficult because there has been a paucity of data from real bioterrorism events. Recent efforts to assess surveillance system performance have focused on injecting synthetic outbreak data (signal) into actual background visit data. These studies focused on either temporal data, a single syndrome category, or a single bioterrorism agent. We are unaware of prior studies evaluating the performance of spatial outbreak detection for multiple syndrome categories in an operational surveillance system. METHODS We extracted respiratory, gastrointestinal, and constitutional syndromes from two year’s of actual emergency department surveillance data from 16 Indianapolis hospitals in the Indiana Network for Patient Care. Patient’s home addresses were converted to latitude and longitude using ESRI ArcGIS 8.3. For each week of visit data we used a cluster creation tool to insert one of 360 unique simulated outbreaks that varied by the number of cases (10, 25, 40), cluster radius (0.25, 0.5, 1, 3 km), and distance from the center of Indianapolis (8, 16, and 24 km). We used SaTScan version 6.0 to detect the single simulated outbreak in each of 35,280 1-week datasets for each syndrome. We calculated sensitivity and positive predictive value (PPV) for various combinations of syndrome category, cluster size, and cluster density. We performed sensitivity analyses using p-values of 0.005, 0.01, 0.1, 0.2, and 0.5. We defined a true positive SaTScan cluster as having at least 50% simulated cases with a scan statistic p-value below the prespecified cut-off. False positive clusters were defined as having less than 50% synthetic points and a p-value below the pre-specified cut-off. RESULTS Average sensitivity and PPV for all clusters was 0.97 ± 0.08 and 0.92 ± 0.10, respectively. Table 1 shows averages for SaTScan sensitivity and PPV for all synthetic clusters stratified by syndrome. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | https://staff.washington.edu/lober/www.isdsjournal.org/htdocs/articles/934.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
