Table 3 Posterior estimates of Bayesian model parameters and covariable effects on malaria.
Parameters | Posterior estimate | |
|---|---|---|
Null model | Full model | |
Bayesian model validation | ||
Posterior mean of the deviance (\(\stackrel{-}{\mathrm{D}}\)) | 465,681 | 444,267 |
Deviance information criterion (DIC) | 466,109 | 449,843 |
Effective number of parameters (PD) | 4289 | 5567 |
Spatial fractional variance (%) | 99 | 74 |
Precision of random effects of the model (median, 95% CrI) | ||
\({\uptau }_{\mathrm{\vartheta }}\) (parameter for zero-inflation) | 2.39 (2.35–2.43) | 1.67 (1.66–1.69) |
\({\uptau }_{\mathrm{u}}\) (spatial structural random effects) | 2.95 (2.39–3.45) | 1.31 (0.97–1.80) |
\({\uptau }_{\mathrm{v}}\) (spatial unstructured random effects) | 4.30 (3.50–5.27) | 7.22 (4.92–8.86) |
\({\uptau }_{\upgamma }\) (temporal structural random effects) | 3.53 (3.28–3.72) | 2.04 (1.17–2.69) |
\({\uptau }_{\upphi }\) (temporal unstructured random effects) | 9.43 (7.27–10.87) | 9.67 (7.94–11.29) |
\({\uptau }_{\mathrm{S}}\) (seasonal random effects) | 3.87 (3.73–4.03) | 9.43 (7.14–11.10) |
\({\uptau }_{\updelta }\) (space–time random effects) | 5.82 (5.51–6.07) | 5.89 (5.58–6.14) |
\({\uptau }_{\mathrm{T}}\)(annual temporal random effects) | 9.09 (6.40–10.77) | 9.59 (7.02–11.18) |
Fixed effect on antilog scale, relative risk (median, 95% CrI) | ||
Duration (years) of seasonal chemoprevention of malaria | 0.93 (0.92–0.94) | |
Availability of malaria rapid diagnosis test | 1.13 (1.09–1.18) | |
Proportion of households in the lowest wealth quintile | 1.05 (1.00–1.10) | |
Main effects of free-of-charge of health care | ||
Period before free healthcare | 1 | |
Period after free-of-charge of health care | 2.27 (2.03–2.54) | |
Main effects of health facility attendance rate | 5.81 (5.67–5.96) | |
Interaction effect between the presence of free healthcare and facility attendance rate | 0.46 (0.45–0.48) | |
