This procedure will list down all the indexes (Clustered as well Nonclustered Indexes) whose first key column having poor cardinality.

This procedure will estimate the cardinality of all the columns of the supplied table list. If the table list is not supplied (or supplied with NULL, the default value) then it will estimate the cardinality for all the tables in the database.

This procedure will list down all the nonclustered indexes, which includes key columns as that of the clustered index. It will also list the key columns of both – nonclustered index, and the corresponding clustered index, along with the key ordinal of the clustered index key columns in the respective nonclustered index. It is now up to you to decide which one makes sense, and which doesn’t. Take the decision accordingly and delete the ones that do not make sense. This procedure may be helpful in performance tuning assignments.