Skip to main content

Table 1 λ i (t, Z i ) estimations with the HP and LWA models are associated to the HR(t) to be estimated according to the adjustment for covariates

From: Matching methods to create paired survival data based on an exposure occurring over time: a simulation study with application to breast cancer

Exposure covariate

Models

H R( t)

 

H P

LWA

 
 

λ i ( t,Z i )

λ i ( t,Z i )

 

Time-dependent

and no time-dependent effect

λ 0 i ( j ) (t)exp γ E i t

λ 0 (t)exp γ E i t

HR ¯ (t)

  With adjustment for Z

 

λ 0(t)exp(γ E i (t)+β Z i )

HR ¯ a (t)

  and with interaction

 

λ 0(t)exp(γ E i (t)+β Z i +α (t)Z i E i (t))

H R i (t)

Time-dependent

and time-dependent effect

λ 0 i ( j ) (t)exp γ t E i t

λ 0 (t)exp γ t E i t

HR ¯ (t)

  With adjustment for Z

 

λ 0(t)exp(γ(t)E i (t)+β Z i )

HR ¯ a (t)

  and with interaction

 

λ 0(t)exp(γ(t)E i (t)+β Z i +α (t)Z i E i (t))

H R i (t)

  1. λ i (t,Z i ) is the instantaneous risk function of event for the pair P j , which is composed of the exposed subject j and the non-exposed subject i matching on Z i .