Extracted Data: Add n(%)#388

Is it possible to extract both the n and % for a variable?

For example, for baseline characteristics, I would like to extract the number of male participants, n= 365 and that they are 60% of the sample.

Were you thinking of auto-calculating it (e.g. when you enter 60%, we compute .6 * 365 = 219), or literally store & show you “60%” in the future?

The latter. W’re hoping to write down n (%) for our variables of interest. Similar to this:

Thanks Dmitriy– A quick question, do the Fixed Effects %% on Quantitative Synthesis work for this? Or can you direct us toward what would be needed in addition to this / altered about this presentation?

Hi @Kevin.kallmes, are these % columns computer generated or were they user inputted?

Good question– these are automatically calculated by the software, and I’ve toggled from Random Effects to Fixed Effects auto-calculation.

I could see a potential problem with auto calculation:

Let’s say there is a variable that has subgroups, like whether a patient with ALS has limb (n=45), bulbar (n=45), or unknown (n=10) site of onset of disease. Unknown onset is not of interest so we would not abstract it, leaving limb and bulbar (n=90).

In a scenario like this, would the Quant Synthesis module calculate the % for limb onset to be 50%? (45/90)

If so, that’s a problem and we’d be better of being able to code the 45% with the variable.

Is this a reasonable concern or are there ways we can get around it?

There is already a manual method for addressing this with dichotomous and continuous variables– you actually have the ability to adjust the “N” (total population) manually in the column. This can also be used to reflect loss-to-follow-up!

For categorical variables, I recommend simply not collecting the “Unknown” category in that case’ the “N” for categorical variables is automatically calculated by adding up all groups. So, as long as you pick the data you care about, the “N” will reflect the population of interest!