Urgent, retrospective study help

SDN Members don't see this ad. (About Ads)

Hi,

First of all thank you for hosting these forums for helping lost students like myself.

I am a MS3 and working on a small research project. Here is my question:

1- I scoured through the data charts of all patients who had disease xyz and surgical treatment abc, and I found 59 such patients lets say. But it was a mix bag, some had follow-ups less than 6 months, some were repeats, so I had to exclude them, so about 12 patients were excluded. That is FOR NOW (not sure about future results as PI might want to study these excluded patients too in the same paper?), for the sake of a poster, I am interested in patients with >6 month follow up and no previous surgeries.

2- So what data should I include for descriptive analysis. This might be a stupid question but I am lost. Should I include in descriptive data analysis (age, gender, disease etiology etc) the 59 patients, OR 59 - 12 = 47 patients. It makes sense to not include data from the excluded patients. But is not this "playing around" with data and we cannot then make inferences about prevalence from descriptive data?

3- Moreover, if in future, for the sake of publication, we want to study these excluded patients, we would have to include them in our descriptive analysis, and that might change the descriptive results, compared to the poster I would have presented previously? Is it acceptable?

I would appreciate ANY responses.

you're not likely going to get any help here; these q's are really something you should discuss w/ your PI.

Originally Posted by patbingsoo

you're not likely going to get any help here; these q's are really something you should discuss w/ your PI.

Thank you for your answer. I did not want to discuss it with PI as I would come across as unintelligent, naive, stupid, not-deserving-of-being-in-med-school, and everything else under the Sun.

Anyways, I found a tactful way to ask the PI. I hope such a question wont effect my reference letter which I need from her.

Thanks.

Its much better to ask for clarification from PI, than it is to risk doing it wrong.

I don't think it'll make you seem stupid, naive etc., but rather wanting to do it right.