Analysing a change in plan

I am carrying out a study into how often an intervention in the form of expert opinion alters the plan made by an individual.
At the time of intervention they are asked what they would like to do, choosing from one of 5 options.
They then give the reasons for this before an expert will either agree with or change their plan to a different option.
I am interested in analysing the frequency with which this intervention results in a different plan to the original.

Essentially I want to know how often the intervention moves someone from one category to another, beyond a simple frequency. Are odds ratios (with 95%CI) the best way to present this data? How can these be calculated?

Happy to provide more information if required! Many thanks!


Well-Known Member
Why not CI of a simple ratio? p± Z sqrt( p(1-p)/n )

Despite the fact that there is a direct connection between P and odds I feel that P is more intuitive and there is no specific reason using odds (like there is for example in a logistic regression)