Re: Database design

If I read your message correctly, an approach you MIGHT be able to look at would be to identify genericly needed information then add a number of generic definition attributes. As an example, you'll need a CTID attribute (clinical trial id) and maybe some kind of Trial Date attribute. All records will require these two pieces of data regardless of the CT. Then there would be specific data that only relates to a specific CT. These would just be identifed, generically, as Value1, Value2, Value3, ..... In a "control file" you would put what the contents of attribute Value1 would represent. Say CTID 001 would have Value1 data being a specific drug's units of measure, Value2 would be the quantity of the units of measure and Value3 would be a comment of that result. You could use the "control file" information for reports, entry screen labels. For doing ad-hoc reports, I'd probably look at creating some kind of statement builder that would produce the proper select statements, or whatever your environment needs. Shouldn't be too tough. The only problem I can think of is setting the maximum number of generic attributes but that depends on your environment.

Food for thought. Hope it helps.
Woody

On Wed, 12 Mar 2003 22:05:32 +0100, "proxy.mzh.de" <evgens_at_aport.ru> wrote:

>Hi,
>basing on the info you provided, the advise one can give you depends on an
>answer to the question: should the information for the trial be structured
>in any way or it might be a free text.
>In the last case the BLOB (CLOB) column would help you. There you can hold
>any unstructured informationÀ
>The requirement to have information structured, makes your life much more
>complicated. There are different approaches you can use, each depends on the
>business task you have (how the information should be structured). If you
>are in this case, pls provide more information to give you some valueable
>advise
>
>regards
> Evgen
>
>
>"john" <nospam_at_nospam.com> wrote in message
>news:wO0ba.11862$Lq.945096_at_stones...
>> Hi,
>>
>> I am creating a database for a web application. The idea for the
 application
>> is for a company administrator to be able to log into the site and
 allocate
>> staff to a project which will be a clinical trial at a particular
 practice.
>> Staff will be able to log in and update the status of the project they are
>> working on and clients will be able to log in and book a member of staff.
>> Administrators will be able to add / edit / delete - staff / clients /
>> clinical trials.
>>
>> Here is my database diagram:
>>
>> http://www.mania.plus.com
>>
>> My problem is each different clinical trail will need to have different
 data
>> collected for it. How could I handle this in the database model? The
>> administrator would like to be able to edit the data collected for each
>> trial via the site...
>>
>> Any other comments on my database model would be appreciated as this is my
>> weakest area of web development by far!
>>
>>
>>
>
Received on Sun Mar 16 2003 - 09:36:18 CET