According to Stitzer, each condition, including the usual care comparison, will be tested several times on a rotating basis during the two-year project, and up to 1,000 residential clients will be able to participate in the project.
Only Objective 1 will be assessed in terms of non-inferiority; all other comparisons will be tests for superiority.
In addition, under the policy, photo-comparison software will be tested in hopes of keeping people from getting multiple licenses under different names.
Comparisons between groups will be tested by use of the χ test.
Comparisons between the service models will be tested for statistical significance.
Data not normally distributed will be analysed using medians and IQR; comparisons between the two models will be tested for statistical significance using the Mann-Whitney test.
The 2.5% significance value is chosen to adjust for the multiple comparisons, because the primary endpoint will be tested on the patient and partner separately.
While these are different specific comparisons, the same general hypothesis will be tested: whether information meetings led by pharmacists are effective in changing physicians' behaviour (comparing to no visit on the same topic).
Differences between intervention and comparison participants on health care utilization will be tested by both chi-square tests and logistic regression analysis.
