Randomization codes for individual patients (A or B) were generated using an online random number generator available on the website randomization.com (http://www.randomization.com, seed 20243) using permuted-block randomization (variable-sized blocks of 4, 6 and 8).
Subjects were randomly allocated to two group with BMI, age and time since last menses as randomization variables.
Randomization variables included gender, weightSDS at birth, heightSDS at age 1 year, GHmaxAITT (during an arginine insulin tolerance test), GHmax24h profile (during 24h spontaneous GH sampling), age and heightSDS at start, the child's heightSDS difference to its MPHSDS (diff MPHSDS) at start, and predicted 1st year Δ heightSDS[ 23].
The mean (± SD) of the total costs in the study groups was determined and the univariate and multivariate linear regression was used to determine the impact of the interventions, the pre and post randomization variables in predicting the total mean costs.
The full analysis set excluded patients who failed to satisfy major eligibility criteria based on the pre-randomization variable, patients without any post-randomization data, and those who withdrew consent.
For example, the allocation ratio ρ in adaptive randomization is a random variable with unknown distribution, and the binomial distribution of n1 assumed in model (7) is not valid even when the null hypothesis is true.
In any randomized clinical trial the primary device for equalizing populations between treatment groups is random assignment of treatments, and this includes subgroups defined by pre-randomization variables.
303 "A secondary analysis of the primary endpoint will adjust for those pre-randomization variables which might reasonably be expected to be predictive of favorable outcomes.
Multivariable regression analysis (OLS, with log transformation) was also used to predict total costs across the cost categories using pre-randomization variables, the alternate treatment strategies and other covariates that relate to resource consumption.
A recent systematic review examining 35 studies of methodological quality and trends in various medical disciplines [24] found that progress over time in these disciplines in key variables such as allocation concealment, blinding and randomization was variable.
