Tabulation of predictors used in the treatment propensity model, distinguishing between baseline variables and variables subsequent to baseline.

‘Change from baseline’ refers to the difference between the baseline value of any pain score and the last measurement prior to the treatment decision, or end of follow-up if no treatment. ‘Lagged changes’ are the differences between the corresponding pain score measures at the last two visits prior to treatment, or end of follow-up. The ‘Treatment’ column gives 95% confidence intervals for mean differences between individuals treated and left untreated by end of the trial. The columns ‘Roland’ and ‘Likert’ give 95% confidence intervals for the regression coefficient of each of these two scores in a univariate regression on each predictor individually.
Tabulation of predictors used in the treatment propensity model, distinguishing between baseline variables and variables subsequent to baseline.