Bayesian spatial modelling of childhood cancer incidence in Switzerland using exact point data: a nationwide study during 1985–2015

Table 2 Median posterior of the variance hyperparameter of the Gaussian field (\(\sigma^{2}\)) for the unadjusted and adjusted model, median posterior of variation explained (\(V\left( {Z\left( s \right)} \right)\)) and median posterior of grid specific relative risk based on residence at diagnosis

	LGCPs
	All cancers	Leukaemia	Lymphoma	CNS tumours
\(\sigma^{2}\) unadjusted^a (median, 95% CI)	0.01 (0, 0.02)	0.00 (0, 0.03)	0.01 (0, 0.04)	0.02 (0.01, 0.06)
\(\sigma^{2}\) adjusted^b (median, 95% CI)	0.01 (0, 0.03)	0.00 (0, 0.01)	0.00 (0, 0.03)	0.02 (0, 0.06)
Variation explained^c (median; 95% CI)	0.72 (0.43, 0.89)	0.81 (0.58, 0.94)	0.82 (0.60, 0.94)	0.64 (0.31, 0.84)
RR unadjusted^a (median; range^d)	0.99 (0.83, 1.13)	1.00 (0.96, 1.09)	0.99 (0.9, 1.13)	1.01 (0.82, 1.23)
RR adjusted^b (median; range^d)	1.02 (0.86, 1.08)	1.00 (0.97, 1.04)	1.00 (0.96, 1.07)	1.00 (0.87, 1.25)

CI credibility intervals, RR grid specific relative risk compared to Switzerland as a whole, LGCP log-Gaussian Cox process, CNS Central and Nervous System
^aThe unadjusted model refers to the models without any covariates
^bAdjusted for NO₂, background radiation, years of general cancer registration, linguistic region and degree of urbanicity
^cVariation explained by the covariates from the fully adjusted model, defined as \(R^{2} = \frac{{V\left( {\varvec{X}\left( s \right)\varvec{\beta}} \right)}}{{V\left( {\varvec{X}\left( s \right)\varvec{\beta}} \right) + V\left( {Z\left( s \right)} \right)}}\) where \(V\left( \cdot \right)\) denotes the variance over the \(K\) spatial units, \(\varvec{\beta}\) is the vector of intercept and covariates, \(\varvec{X}\) the design matrix and \(Z\left( s \right)\) the Gaussian field. The variation here refers to the fully adjusted model
^dRange is defined as [min, max]

ISSN: 1476-072X