Skip to main content

Table 2 Timeliness in detection of nine known tuberculosis outbreaks by local authorities and statistical method

From: Using statistical methods and genotyping to detect tuberculosis outbreaks

Outbreak

Spoligotype

MIRU

Earliest date of detection

Earliest detection method

   

Local authoritiesa

County-based LLR alert b

CUSUM alert

SaTScan alert c

 

A

777776777760601

224325143323

05/04/09

07/31/07

06/30/07

07/31/07

CUSUM

B

700036777760771

222325133223

05/05/09

06/30/08

06/30/08

06/30/08

LLR, CUSUM, and SaTScan

C

740777607760771

223315193323

06/12/09

05/31/09

04/30/09

04/30/09

CUSUM and SaTScan

D

000000000003771

223325173533

06/01/08

None

06/30/07

None

CUSUM

E

676177607760771

224326153323

07/01/08

08/31/08

08/31/08

08/31/08

Local authorities

F

770000770000000

224125153322

08/01/09

06/30/09

06/30/09

06/30/09

LLR, CUSUM, and SaTScan

G

777776770000000

225325133324

02/07/08

03/31/08

04/30/08

03/31/08

Local authorities

H

777776757760771

223325143324

09/01/09

12/31/06d

04/30/05

12/31/06d

CUSUM

I

477777777720771

227325153323

08/28/09

None

None

None

Local authorities

  1. Definition of abbreviations: CUSUM = cumulative sums; LLR = log-likelihood ratio; MIRU = Mycobacterial Interspersed Repetitive Units; TB GIMS = Tuberculosis Genotyping Information Management System.
  2. a Date that local public health authorities first noticed a problem.
  3. b Using a 2-year time window.
  4. c Using a 2-year time window and a 50 km maximum search radius.
  5. d Earliest possible date of alert is 12/31/06 (genotyping data in TB GIMS incomplete before 01/01/05).