Methodology in retrospective chart review, how much detail I should provide?

[libraryismyhome]

I am currently working to complete my protocol now, and I was wondering how much I should write in methodology when I am doing retro chart review.

I mean all I am doing is chart reviewing, so should I write (1) I am doing retro chart review (2) the data are obtained from the device A, which measures the whatever deficit in patients with whatever disease. (3) These are the variables and parameters that I am going to use and compare. (4) physical differences such as height differences are going to be normalized

Do I really need to describe how was data collected, what statistical method are going to be used, and etc? But I don't know what stat methods I am going to use yet. 😀

thank you in advance!!

---

Members don't see this ad.

 

[HumanBean]

The answer to most of your questions depends on who you are writing the "protocol" for. If you're just writing a protocol to hand to your PI, make sure you have established exactly what you will be investigating and how you will compare the data within the charts. Identify exclusion criteria and the parameters for the study. This will tell you how many separate groups you will have to compare, and then you will know which statistical tests you will need to analyze the data. If you have several groups / variables, I would recommend using a biostats textbook to help identify the proper statistical methodology.

When describing how the data was collected, usually this should include a description of the patient population, the time period that the data was collected, etc., Things like how height was measured can usually be excluded, unless a unique way of measuring height was used.

If you're writing this for the methods section of a manuscript, I would recommend that you look at publications from a journal that you would likely publish in.

Good luck.

 

[MNIkid87]

What is the protocol for? A paper? Just for your PI? REB approval?

If it for a paper you can probably just write "We reviewed the medical records of XXX consecutive patients presenting to DEPARTMENT for CONDITION. Data was collected on variableofinteres1, variableofinterest2, .... Then exclusion/ Inclusion criteria if any.Then write your statistical approach briefly (starting with the normalization of variables)