Skip to main content
. Author manuscript; available in PMC: 2013 Nov 1.
Published in final edited form as: Epidemiology. 2012 Nov;23(6):829–838. doi: 10.1097/EDE.0b013e31826c2dda

Table 1.

Components of the main inference model (described by equations (3) and (4))

Component Role
Mc (t) Weekly mortality underlying a specific cause, or a collection of causes
H 31 (t) Weekly incidence proxy for influenza A/H3N2 during the 1997–1998 through 2002–2003 seasons (before the appearance of the Fujian A/H3N2 strain)
H 32 (t) Weekly incidence proxy for influenza A/H3N2 during the 2003–2004 through 2006–2007 seasons (the Fujian A/H3N2 strain)
H 1(t) Weekly incidence proxy for influenza A/H1N1
B (t) Weekly incidence proxy for influenza B
S Forward shift operator (between one and two weeks) relating influenza incidence proxies and mortality
f Fraction of influenza incidence proxy shifted forward by 1 week in S (with a fraction(1 − f) shifted forward by 2 weeks)
βH31
Regression coefficient for S (H31)
βH32
Regression coefficient for S (H32)
βH1 Regression coefficient for S (H1) (omitted in the main inference model)
βB Regression coefficient for S (B)
Base1 (t) Weekly periodic annual baseline for mortality (with a period of 52 weeks) before week 23 of 2001 (zero thereafter)
Base2 (t) Weekly periodic annual baseline for mortality (with a period of 52 weeks) starting week 23 of 2001 (zero before that)
Trend A (low degree) polynomial in the calendar year
Noise Difference between the expected and observed mortality