It does not have a standard error so there is no way to compute a combined estimate. It would seem that such a statistic would not be very useful anyway since it is meant to test the validity of the MH approximation for the tables themselves. A combined estimate, even if it could be computed, would not be a way to test the combined estimate anyway.
It might make sense in your case to simply check it for each of the imputed data sets.
... View more