I'm preparing a poster to present in a scientific meeting. I've done a retrospective research in our hospital database regarding patients undergoing a particular procedure (which is used for people suffering from an acute disease) and check for their hospitalization lenght.
The outcome of this procedure determines the path the patient will follow (surgery or clinical monitoring).
I'm trying to demonstrate that undergoing that procedure earlier leads to a shorter hospitalization lenght. (Our claimings are in bold)
I found the correlation coefficient between these 2 variables: "days between entry-procedure", "total lenght of hospitalization".
I get: Sample size 37
Correlation coefficient r 0,4739
Significance level P=0,0030
95% Confidence interval for r 0,1771 to 0,6917
Here's a link to the database: https://docs.google.com/spreadsheets/d/1CAMvweHw2tVb7TBjKvcojXppP478thT4OjxyuQuomfU/edit?usp=sharing
This justify my claiming that doing the procedure early leads to shorter hospitalization lenght? Filtering for those who don't qualify for surgery I get even a higher correlation coefficient.
I also included previous surgery in the life of these patients. I found that patients who had a particular type of surgery ("0" in the table) have higher risk to enter the hospital for that disease, but have also the highest chance to resolve the issue without surgery. I don't know how to verify this claiming for statistical significance.