Test of proportions; study design and sample size

ted

New Member
#1
I want to look to see if there is an association between two conditions. Basically, patients come into dermatology and complain of an itch between their shoulder blades and are diagnosed with this thing called "NP". I believe the itch may be due to a muscle/skeletal imbalance hence something that should be treated not by dermatology but elsewhere (osteopathy).

So the two conditions are (1) patients that complain of an itch between their shoulder blades and diagnosed with NP and (2) display somatic dysfunction which is basically characterized by four tests (tissue texture, asymmetry, restriction, and tenderness...a yes/no diagnostic) and if they have at least one they have somatic dysfunction.

I am anticipating a sort of test of proportions because there are going to be people who have somatic dysfunction for a variety of reasons. So, I need to collect data on the non-NP patients to get an estimate of that population's proportion of somatic dysfunction. Then collect data on all the NP patients I diagnose and estimate that population's proportion of somatic dysfunction. I can only collect data on patients from my practice. I will try to get a preliminary estimate of sample sizes by estimating the proportions of these two groups just by a convenience sample of patients I see in the next few weeks.

Suggestions on study design, sample size calculations, or alternative methods to answer my question would be most welcomed.