Help with my research project

[HIK]

Hi everyone,

Currently I am working on a research project in the Netherlands, while studying Pharmacy.

My study consists of 2 groups of different sizes. group 1: 50 patients (patients using the drug metformin) and group 2: 100 patients (control group, non diabetic).
I want to investigate if there is a significant difference in uptake of a radioactive pharmacon in the intestines. Next I also want to investigated is metformin discontinuation period and metformin dosage influence the uptake within the metformin group.
The characteristics BMI, age, sex and blood glucose differ between group 1 and group 2. I thought I could make a correction for these possible confounders but I am not sure.

Someone out there who can help me?

Thanks!

 

[Karabiner]

There is something missing in your description. If you conduct a study with patients, then it is required to involve a competent statistician, and to describe the plan for data analysis in advance.

With kind regards

Karabiner

 

[HIK]

previous studies did not show significant differences between characteristics. We did a pilot but we had no insight in metformin use and thus in patients with diabetes type 2. So it was not really clear what the outcome of the characteristics would be. I am obviously a student and the research project does not require patient intervention of any sort. My current supervisors are quite hard to reach so that is why I am asking for more insight.

I am thinking about matching patients in regard to age and gender. This way BMI and blood glucose level will be the remaining confounders. It is extremely hard to do this because the data is really scattered. So this will take a lot of time, which I am restricted to because of my study. So therefore my question if you can analyze 4 confounding factors in these sample sizes.

 

[GretaGarbo]

If you conduct a study with patients, then it is required to involve a competent statistician, and to describe the plan for data analysis in advance.

I just want to say that I agree.

 

[hlsmith]

Unclear, so do you have type II diabetics on and not on metformin? If the control group is not deemed diabetic, you will have many many more confounders and issues. If they are diabetic, they have to be comparable to the metformin group but not on metformin otherwise you are likely missing out on the biological mechanism behind their insulin resistance. And if they aren't diabetic, why are they in the health system and are these patients homogenious in ailments. And if controls are diabetic and similar, why aren't they on metformin, provider differences. Please clarify.

You can control for some confounders for unless you know the biological mechanism for the radioactive component, you may have sample selection issues or miss confounders.