Comparing multilevel and Bayesian spatial random effects survival models to assess geographical inequalities in colorectal cancer survival: a case study

Table 4 Estimated area-level random effects from multilevel discrete-time survival models

Model	Description¹	ml _u_j²(95% CrI)	p ³	MOR (95% CrI)⁴
1	Null (no covariates) only individual & area-level random effects	0.025 (0.014, 0.039)	<0.001	1.16 (1.13, 1.21)
2	Individual-level covariates	0.011 (0.006, 0.018)	0.04	1.10 (1.08, 1.14)
3	Individual-level covariates & area-remoteness	0.007 (0.003, 0.014)	0.10	1.08 (1.05, 1.12)
4	Individual-level covariates & area-disadvantage	0.006 (0.001, 0.014)	0.08	1.08 (1.03, 1.12)
5	Full model: all individual- & area-level covariates	0.005 (0.001, 0.012)	0.12	1.07 (1.03, 1.11)

CrI: Credible Interval.
1. Models 2–5 adjusted for all individual-level covariates; Model 4 also adjusted for area disadvantage; Model 5 also adjusted for area remoteness and area disadvantage.
2. The residual area-level variance from the MCMC simulations for multilevel analysis.
3. From Wald χ2 test.
4. Median odds ratio-Refer to text and Appendix 1 for details.

ISSN: 1476-072X