1	Fit the global imputation model using the SI method to find ${\widehat{\mathit{\alpha }}}_{si}$ and $\mathrm{Cov}\left({\widehat{\mathit{\alpha }}}_{si}\right)$;
2	Sample α1, …, αM independently from $\mathcal{N}\left({\widehat{\mathit{\alpha }}}_{si},\mathrm{Cov}\left({\widehat{\mathit{\alpha }}}_{si}\right)\right)$;
3	Send α1, …, αM to all sites;
4	for m ← 1 to M do
5	for k ← 1 to K do Impute the missing data at site k based on αm;
6	Fit the analysis model and obtain ${\widehat{\mathit{\theta }}}_m$ and $\mathrm{Cov}\left({\widehat{\mathit{\theta }}}_m\right)$;
7	end
8	Combine the results by Rubin’s rule to obtain $\widehat{\mathit{\theta }}$ and $\mathrm{Cov}\left(\widehat{\mathit{\theta }}\right)$;