NO JOIN PREDICATE

am encountering the NO JOIN PREDICATE warning in a query plan for a particualr SQL statement. In poking around, I've come across psotings that mention the CU 4 for SQL Server 2005. The update applies to a KB article that addresses complex queries with an outer join.

But in my case, it's not an outer join causeing the problem - they are all left joins. Here is the piece of SQL that is running extremely slowly:




Code Snippet
UPDATE @TESTTABLE
SET [@TESTTABLE].SignInCancelledNoEMR =
(SELECT q.cnt FROM
(SELECT COUNT(*) AS cnt,cl.Clinic_id AS ClinicIDq,cl.CLINIC_REG_ID AS MarketIDq
FROM dbo.tblactivity AS t
LEFT JOIN [ERIIProduction].dbo.computer c on t.machineid = c.comp_windowsname
LEFT JOIN [ERIIProduction].dbo.clinic cl on cl.clinic_id = c.comp_clinic_id
LEFT JOIN [ERIIProduction].dbo.Region r ON cl.CLINIC_REG_ID = r.REG_ID
WHERE t.status = '1'
AND cast(checkindate as datetime) >= @startDate
AND cast(checkindate as datetime) < DATEADD(d, 1, @endDate)
AND cl.clinic_reg_id IN (SELECT Item FROM dbo.Split(@regionIDList,','))
AND NOT EXISTS (SELECT 1 FROM [ERIIProduction].dbo.VisitPatient vp WHERE (t.activityid = vp.VSPAT_GalvanonActivityID)) --AND (t.memberid = vp.VSPAT_GalvanonMemberID))
GROUP BY cl.Clinic_id,cl.CLINIC_REG_ID) AS q
WHERE q.ClinicIDq = ClinicID AND q.MarketIDq = MarketID)






Before I apply the CU and enable the trace flags, I'd like to know if there is something else I can do within the query itself to make it run faster. I've tried the OPTION(FORCE ORDER ) hint, but that doesn't have any effect on the queery run time.
And as a whole, the query cost relative to the entire batch jumped from 75% to 96%.

Granted the execution time of the entire batch dropped from 9 minutes to 1 minute, this query should not even take that long.

Thanks for any input...!