Skip to main content

Table 2 Summary of signals generated by EARS C1, C2 and C3 cusum methods.

From: Outbreak detection algorithms for seasonal disease data: a case study using ross river virus disease

Method & parameters No. signals of specified duration Details of signals lasting ≥4 wks
  1 wk 2 wks 3 wks ≥4 wks  
C1 4 wk BL k = 3 5 6 1 5 1992 (wk 5-11), 1996 (3-12), 2003 (10-20), 2004 (4-14), 2007 (38-45)
C1 4 wk BL k = 4 5 4 0 5 1992 (wk 5-9), 1996 (5-9), 2003 (10-18), 2004 (4-7), 2007 (38-43)
C2 4 wk BL k = 4 7 5 2 5 1992 (wk 5-13), 1996 (5-14), 2003 (11-25), 2004 (4-17), 2006 (2-7)
C2 4 wk BL k = 5 7 1 1 4 1992 (wk 5-12), 1996 (5-13), 2003 (11-23), 2004 (4-15)
C2 4 wk BL k = 6 5 2 1 4 1992 (wk 5-11), 1996 (5-12), 2003 (11-21), 2004 (4-8)
C3 4 wk BL k = 5 8 1 2 4 1992 (wk 5-12), 1996 (5-12), 2003 (11-15), 2004 (4-15)
C1 8 wk BL k = 3 6 3 1 3 1992 (wk 5-13), 1996 (5-14), 2003 (11-19)
C1 8 wk BL k = 4 6 0 0 3 1992 (wk 5-10), 1996 (5-11), 2003 (11-15)
C2 8 wk BL k = 4 0 4 0 6 1992 (wk 5-15), 1996 (5-18), 2001 (13-16), 2003 (11-21), 2004 (11-14), 2006 (2-8)
C2 8 wk BL k = 5 1 4 0 3 1992 (wk 5-13), 1996 (5-16), 2003 (11-18)
C3 8 wk BL k = 4 0 4 0 6 1992 (wk 5-15), 1996 (5-18), 2001 (13-16), 2003 (11-21), 2004 (11-14), 2006 (2-8)
C3 8 wk BL k = 5 1 4 0 3 1992 (wk 5-13), 1996 (5-16), 2003 (11-18)
  1. Parameter values applied were h = 2 with 4 or 8 weeks of baseline (BL) data and 3 ≤ k ≤ 6.