Participants were randomised using blocked randomisation to ensure similar group sizes throughout the study; every nine participants within each employer or health plan contract were randomly assigned with an allocation ratio for the three groups of 3 3 3.
Patients were randomised, using variable block randomisation stratified for centre, by an independent call-centre.
Secondary outcomes analysis will use the treatment groups as randomised using the CC population.
Two hundred and ten patients were randomised using minimisation, with a random component of 80%.
Likewise, when a stratified analysis was used after patients were randomised using stratified permuted blocks, error rates were nominal.
Patients were randomised, using a computer-generated code, to receive either usual care (UC) or holistic therapy (HT).
When patients were randomised using minimisation, both a covariate adjusted analysis and a stratified analysis gave valid type I error rates, regardless of the size of the interaction.
Patients were randomised, using a Registration And Medication Ordering System (RAMOS; GSK, Brentford, UK), 1 1 to receive either UMEC/VI or FP/SAL.
Immediately after this first visit, patients were randomised using the baseline data (information on the recruiting hospital and any previous experience of the patient with UV therapy were used to determine the randomisation strata).
We analysed all available cases as randomised using χ tests to compare dichotomous outcomes, t tests to compare continuous outcomes, and Kruskal Wallis tests to compare ordered categorical outcomes that is, categorised scores on the hospital anxiety and depression scale.
