PakMediNet Discussion Forum : Biostatistics : What type of study design can I use?
What type of study design can I use?
I dont know if I can use a case control study in the following settings or not?
the situation is like this:
There are about 1000 records of the blood donors who their blood tested positive for HIV at a blood donation center from 2002 to 2008.
all of them came in that period to the consultation department and we collected data regarding sociodemographic, and risk factors of HIV infection( unsafe sex, IV drugs,...). so we have all these information in their records at our consultation department.
Now, can I use that records as the data for the Case Group in a case-control study without recruiting them again and asking them questions? I can recruit a control group similar to that case group except that they are HIV negative.
if I do such a study, the name of this study will be Retrospective Case-Control Study?
or I can not use the secondary data from the patients records at consultation clinic as the data for Case group?
If I can not do that case-control study, and I just describe the prevalence and distribution of risk factors among those patients from their recorded data without control group, then this study will be a Cros Sectional study using the secondary data? or will be a case series study?
Posted by: dokijoonPosts: 4 :: 28-03-2009 :: | Reply to this Message
Re: What type of study design can I use?
In a case control study, you have the disease group and a control group, and then you look back in the past for a particular risk factor. Keeping in view about your study, it can a case control study if you have a suitable control for it. Ideally, both cases and controls should be identical as possible to prevent confounding factors except that the cases are HIV positive. You need to follow both cases and controls back in the past (retrospective) to see their risk factors. Case control study will establish cause-effect relationship perfectly.
If you cannot find control group, then a cross-sectional study is a reasonable design. It is not powerful than case control but is very easy. Cross-sectional study will not eastablish cause-effect relationship but can tell about the association.
I hope this would help you.
Posted by: docosamaPosts: 333 :: 31-03-2009 :: | Reply to this Message