Is there any way to get distinct counts in a pivot table that uses a a SQL Server table (not a cube and not an Excel sheet) as it's source?