Skip directly to search Skip directly to A to Z list Skip directly to page options Skip directly to site content

Volume 11, Number 2—February 2005

Dispatch

Comparing Aberration Detection Methods with Simulated Data

Lori Hutwagner*Comments to Author , Timothy Browne*, G. Matthew Seeman*, and Aaron T. Fleischauer*
Author affiliations: *Centers for Disease Control and Prevention, Atlanta, Georgia, USA

Main Article

Table A2

Sensitivity (sen) and time to detection (ttd) by method and outbreak type

Type of outbreak Seasonally adjusted CUSUM*
Historical Limits†
C1–MILD‡
C2–MEDIUM§
C3–ULTRA¶
sen (%) ttd** sen (%) ttd** sen (%) ttd** sen (%) ttd** sen (%) ttd**
Spike†† 65.55 0.0 10.0 0.0 60.2 0.0 59.9 0.0 63.3 0.0
Spike‡‡ 48.3 0.0 7.6 0.0 43.3 0.0 43.1 0.0 47.2 0.0
Inverted log normal§§ 92.2 1.3 52.6 2.4 67.2 1.2 82.6 1.4 92.8 1.3
Inverted log normal¶¶ 75.9 1.4 30.4 2.3 54.4 1.3 66.0 1.5 80.8 1.5
Inverted log normal## 98.0 2.4 77.2 4.7 66.4 2.5 83.3 2.6 95.1 2.4
Inverted log normal*** 88.8 2.8 50.7 4.8 59.4 3.0 71.8 3.1 88.7 3.0
Log normal††† 92.6 0.6 52.5 1.7 70.6 0.4 82.6 0.6 92.5 0.6
Log normal‡‡‡ 77.4 0.8 31.4 1.7 54.8 0.6 65.1 0.7 80.1 0.8
Log normal§§§ 97.2 1.0 75.8 2.6 67.9 0.8 85.4 1.0 94.3 1.0
Log normal¶¶¶ 88.2 1.6 49.9 2.9 56.9 1.2 71.8 1.4 87.2 1.4

*The seasonally adjusted CUSUM method sums the positive differences of the current value from the mean for a period similar to the current value over 5 years.
†The historical limits method compares the current sum of 4 periods to the mean of the sum of 15 sets of 4 periods surrounding the current point of interest over 5 years.
‡The C1–MILD method is based on CUSUM, but the calculations reduce to the current value being greater than the mean plus 3 standard deviations (SD), with the mean and SD based on the past 7 days.
§The C2–MEDIUM method is based on CUSUM, but the calculations reduce to the current value being greater than the mean plus 3 SD, with the mean and SD based on the past 7 days shifted by 2 days.
¶The C3–ULTRA method is based on CUSUM, summing the positive difference of the current value from the mean for 3 days, with the mean and SD based on the past 7 days shifted by 2 days.
**95% confidence values were <0.03. Time to detection must be interpreted with sen because ttd does not include missed outbreaks.
††This type of outbreak represents an increase of 2 SD for 1 day.
‡‡This type of outbreak represents an increase of 3 SD for 1 day.
§§This type of outbreak represents an increase of 2 SD with a fast rise in cases with incubation <7days.
¶¶This type of outbreak represents an increase of 3 SD with a fast rise in cases with incubation <7days.
##This type of outbreak represents an increase of 2 SD with a fast rise in cases with incubation >7days.
***This type of outbreak represents an increase of 3 SD with a fast rise in cases with incubation >7days.
†††This type of outbreak represents an increase of 2 SD with a slow rise in cases with incubation <7days.
‡‡‡This type of outbreak represents an increase of 3 SD with a slow rise in cases with incubation <7days.
§§§This type of outbreak represents an increase of 2 SD with a slow rise in cases with incubation >7days.
¶¶¶This type of outbreak represents an increase of 3 SD with a slow rise in cases with incubation >7days.

Main Article

TOP