The influence of aggregate data and individual data on adherence to repeat screening ('occasional participants' vs 'compliant participants'), was analysed, firstly using univariate logistic regression, then two-level (multilevel) logistic regression models with individuals (level 1) residing within the areas (level 2).
The modelling begins with individual level baseline data from the general lifestyle survey 2009 on mean weekly and peak day consumption (a proxy for individuals' scale of binge drinking) of alcohol for 11 385 people in England.
In addition, generalized linear and additive mixed models with individual-level random effects were fitted to examine the extent of within-person clustering (16, 17).
We therefore used two-level logistic models with individual-level covariates (age, gender, and history of previous hepatitis) and household-level risk covariates (education of the household head, wealth index, and source of water for cooking and bathing).
To examine diabetes prevalence across urbanisation levels, we used multilevel logistic regression models, with individual-level age and household income as fixed effects and with the community and province of residence as random effects.
We used a random-intercept logistic model with individuals (level 1) nested in areas (level 2).
A hierarchical modelling strategy was adopted [ 18, 22], with individual level factors being assessed first in the multivariable model.
b Model 1: Multivariate model with individual-level covariates only.
The final model with individual-level, hospital-level, and state-level factors contributed to 86.1% of the variability in GDM between states, with the remaining variability between states statistically significant at the 5% level (P = 0.044).
A hierarchical Bayesian model with individual-level growth parameters was fitted to the serial AAA size data using WinBUGS 1.4.3 software (see Supplementary Data online) called from R using the R2WinBUGS package.
