I've tried for a month now, hence the question. Any ideas? I can or everything in keywords for each issue and reduce loops but not eliminate them. What am I missing?
Thanks

This process works for one record of the keyword table, but we have a hundred.

I am currently looping through each record in keyword table and inserting results into a report table, from there I can query for the report. But, I guess, cursor concurrency seems to be an issue and response time can range from 2 minutes to 2 hours on approximately the same dataset. During non production times with no other processes on these tables.
...

I can query using full text, and am getting the results I need, but with a cursor, and previous iterations of the loop are keeping locks, so one time it will be fast next it will never return.

DATA
text table
key department date text
1 1 3/1/2013 customer tripped and fell
2 1 3/2/2013 satan visited
3 1 3/3/2013 susan fell

Issue table
...

I have a table with 40 million rows containing a column defined as varchar(max) with a full text index on this column, other columns are department, date, and key. We have a list of issues and their associated keywords that we need to determine how many times and in which records these issues/keywords appear within a given timeframe for a specific department. An issue can have multiple sets of keywords, since the issue can be described in several...