Volume 12, Number 12—December 2006
Research
Evaluating Detection of an Inhalational Anthrax Outbreak
David L. Buckeridge*
, Douglas K. Owens†‡, Paul Switzer‡, John Frank§, and Mark A. Musen‡
, Douglas K. Owens†‡, Paul Switzer‡, John Frank§, and Mark A. Musen‡Table 3
Sensitivity, time to outbreak detection (timeliness), proportion of outbreaks detected through syndromic surveillance before clinical case finding, and mean detection benefit of syndromic surveillance compared with clinical case finding for 3 release scenarios and 2 levels of specificity*
| Amount released (kg) | Specificity 0.900 (1 false alarm every 10 d) | Specificity 0.975 (1 false alarm every 40 d) | ||||||
|---|---|---|---|---|---|---|---|---|
| Sensitivity per outbreak | Mean timeliness, d | Proportion with detection benefit | Mean detection benefit, d | Sensitivity per outbreak | Mean timeliness, d | Proportion with detection benefit | Mean detection benefit (d) | |
| 1 | 1.00 | 3.1 (0, 5) | 0.59 | 1.0 (0, 3.3) | 0.98 | 4.3 (2, 7) | 0.28 | 0.32 (0, 1.0) |
| 0.1 | 0.99 | 3.3 (0, 6) | 0.55 | 1.0 (0, 3.5) | 0.95 | 4.7 (2, 7) | 0.24 | 0.33 (0, 1.1) |
| 0.01 | 0.94 | 3.6 (0, 7) | 0.51 | 1.1 (0, 3.7) | 0.82 | 5.1 (2, 8) | 0.19 | 0.33 (0, 1.3)
*Values in parentheses are 10th and 90th percentiles of the distribution. *Values in parentheses are 10th and 90th percentiles of the distribution. |
*Values in parentheses are 10th and 90th percentiles of the distribution.
