Residency Project Question

[hopeful_dr]

I have a question about the year-long residency project.

Does the project necessarily have to have "before/after" data?

Example: You have a hypothesis that "XYZ" is being done at the hospital, which you (as the resident) may or may not necessarily agree with. You think a change needs to be made and whether it be through education or implementation of a guideline, you find a way to change the way things are currently being done to "ABC". After intervening, you analyze the results of that intervention.

After looking at posters and projects of residents from previous years, to me the projects seem to have to do with more retrospective trend analyses, comparing patient outcomes in certain disease states for traditionally used medications versus newer medications, ways of improving medication adherence, and that sort of thing. There are some quality improvement research projects, but those were performed some time after a protocol was implemented.

To me, "before/after" sounds fairly unreasonable to do within the time frame that we have to collect and analyze the data pre-intervention and have data available post-intervention. I'm not sure if even "intervening" is necessary to include as part of the research project. However, the preceptors at the institution I am currently at are adamant about this so-called "before/after" approach to the research project, which I don't necessarily agree with.

Am I wrong?

---

Members don't see this ad.

 

[bacillus1]

I helped out with the project of a resident at my school last year. It was not "before/after". It was a retrospective analysis of inpatient prescribing of a certain disease state, and showing areas where medication delivery in this patient population could be improved. No intervention was made.

I have a question about the year-long residency project.

Does the project necessarily have to have "before/after" data?

Example: You have a hypothesis that "XYZ" is being done at the hospital, which you (as the resident) may or may not necessarily agree with. You think a change needs to be made and whether it be through education or implementation of a guideline, you find a way to change the way things are currently being done to "ABC". After intervening, you analyze the results of that intervention.

After looking at posters and projects of residents from previous years, to me the projects seem to have to do with more retrospective trend analyses, comparing patient outcomes in certain disease states for traditionally used medications versus newer medications, ways of improving medication adherence, and that sort of thing. There are some quality improvement research projects, but those were performed some time after a protocol was implemented.

To me, "before/after" sounds fairly unreasonable to do within the time frame that we have to collect and analyze the data pre-intervention and have data available post-intervention. I'm not sure if even "intervening" is necessary to include as part of the research project. However, the preceptors at the institution I am currently at are adamant about this so-called "before/after" approach to the research project, which I don't necessarily agree with.

Am I wrong?

 

[KARM12]

Depends on the program...but I think the majority of projects are retrospecitve in nature. It can be difficult to do a prospective study and get a large enough sample size for it to be significant. It takes longer for a prospective study to be approved by IRB, implemented, etc. One year usually isn't enough.

 

[xiphoid2010]

Totally agree with Karma. Steer clear a prospective study unless something has already been partially set up (ie. last year's resident ran out of time, lol) or you are staying on to completing PGY-2 at the same place.

Coming from a research background, I warned my fellow residents against taking on a prospective study in a 1 year residency. However couple of the residents at the university medical center next door didn't get the memo, and it was a disaster. 1 was able to enroll 1 patient, 1 got 3 patients... it was just plain sad.

Especially if you have not done any real research previously, take a conservative approach. Focus on retrospective design, do a preliminary check to insure you got a decent (100-200) patient sample size, data type should be numeric/categorical making it easy for statistical analysis.