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

Table 2 Survival functions applied to simulate each transition and each selected configuration

	Constant $HR {(t)}^{*}$	Increasing $HR (t)$	Decreasing $HR (t)$	Increasing then
				decreasing $HR (t)$
Transition $1 \to 2$	Exponential ⁽¹⁾	Exponential ⁽¹⁾	Exponential ⁽¹⁾	Exponential ⁽¹⁾
	$(λ)$	$(λ)$	$(λ)$	$(λ)$
	$(0.0050)$	$(0.0075)$	$(0.0025)$	$(0.0020)$
Transition $1 \to 3$	Weibull ⁽²⁾	Weibull ⁽²⁾	Loglogistic ⁽³⁾	Weibull ⁽²⁾
	$(λ_{0}, γ)$	$(λ_{0}, γ)$	$(μ, σ)$	$(λ_{0}, γ)$
	$(0.0039, 1.1881)$	$(0.0022, 1.1881)$	$(5.7146, 0.2390)$	$(0.0018, 1.1881)$
Transition $2 \to 3$	Weibull ⁽²⁾	Weibull ⁽²⁾	Loglogistic ⁽³⁾	Loglogistic ⁽³⁾
	$(λ_{0}, γ)$	$(λ_{0}, γ)$	$(μ, σ)$	$(μ, σ)$
	$(0.0039, 1.1881)$	$(0.0028, 1.5439)$	$(5.6778, 0.2463)$	$(5.9858, 0.4971)$

⁽¹⁾ $\bar{S} (t) = exp (- λt);^{(2)} \bar{S} (t) = exp (- {(λ_{0} t)}^{γ});^{(3)} \bar{S} (t) = {[1 + exp (\frac{ln (t) - μ}{σ})]}^{- 1}$ .
^*We simulated the same functions and parameters for the second Constant H R(t) except for Transition $2 \to 3$ where λ ₀=0.0069.

ISSN: 1471-2288