Du kan slutte dig til den samme tabel eller bruge underforespørgsel til at kontrollere for dubletter. Dette burde være lettere for SQL-serveren, da det ikke ville tælle alle dubletter. Noget som dette:
SELECT
t1.column_name
FROM
table_name AS t1
WHERE
NOT EXISTS (
SELECT
*
FROM
table_name AS t2
WHERE
t2.column_name = t1.column_name
AND t2.id != t1.id
)
ELLER
SELECT
t1.column_name
FROM
table_name AS t1
LEFT JOIN table_name t2 ON (
t2.column_name = t1.column_name
t2.id != t1.id
)
WHERE
t2.column.name IS NULL