= Marginal incidence trend with clustered observations
{title2=$\log\mathbb E[Y_{it}]=\log P_{it}+\beta_0+\beta_1t$}
A population-averaged log-rate model with exposure $P_{it}$ gives an annual <rate ratio> $e^{\beta_1}$. A <generalized estimating equation> can account for repeated observations within clusters through a working correlation and <sandwich covariance matrix>. Independent Poisson standard errors do not account for clustering. An exposure-weighted population rate and an equally weighted average of cluster rates are different estimands.
Back to article page
