This audit's true numbers and details have been changed to abide by confidentiality standards.
An audit of the patients in a hospital of a certain specialty is carried out to assess whether you are more likely to be prescribed prophylactic drugs in line with some local guidelines if you are based in the main ward for that specialty rather than on a general/'other' ward.
10/10 of the patients on the specialty ward have been prescribed for correctly (ie following guidelines). In comparison, 3/5 of the patients on the other wards have been prescribed for correctly.
What statistical test would be suitable here, bearing in mind I would like to take into account a lower power (due to low quantity of data)? Likewise, each piece of data is binomial (a patient has either been prescribed for 'correctly' or 'not correctly').
Any advice would be appreciated.