Hi there,

I'm hoping you can help. I'm setting up a small scale study looking at an intervention to reduce drug prescription error rates. I will have two randomised groups - one prescribing with a control method and the other group with the intervention. A previous study has recorded a drug error rate of 8.4%. We hope to detect a reduction in error rates of at least 25%.

My main question at this time is regarding power calculation/sample size. If I want a power of 80 how do i work out what my required sample size is. I have tried keying in some numbers into online calculators but I don't seem to be getting reliable results. I hope I have given you enough info. Any help appreciated.