Joint modeling of success and treatment discontinuation in in vitrofertilization programs: a retrospective cohort study

Table 3 Estimations of the links between the success and treatment discontinuation processes for the Parisian and the Medium-sized city center

	Posterior mean (posterior standard deviation)^a	95% credible interval	Test of λ^b
λ _Paris	0.01 (0.11)	[-0.2,0.2]	L = 2.70 ^c - NS
λ _{Medium-sized city}	-0.21 (0.12)	[-0.5,0.0]	L = -5.26 ^d - S

^a the models were adjusted for female age and center.
^b using partial Bayes factors (see Methods section).
^c The competitive models were M₀: [λ _Paris = 0 and λ _{Medium-sized city} =0] vs M₁: [λ _Paris ≠ 0 and λ _{Medium-sized city} =0 ]. There is a positive evidence against λ_Paris ≠ 0 meaning that λ_Paris is non-significantly (NS) different from 0 (H₀ is not rejected).
^d The competitive models were M₀: [λ _Paris = 0 and λ _{Medium-sized city} = 0] vs M₁: [λ _Paris = 0 and λ _{Medium-sized city} ≠ 0]. There is a positive evidence against λ_{Medium-sized city} =0 meaning that λ_{Medium-sized city} is significantly (S) different from 0 (H₀ is rejected).

ISSN: 1471-2393