Should baseline be a covariate or dependent variable in analyses of change from baseline in clinical trials?

Abstract

In randomized clinical trials, a pre-treatment measurement is often taken at baseline, and post-treatment effects are measured at several time points post-baseline, say t=1, ..., T. At the end of the trial, it is of interest to assess the treatment effect based on the mean change from baseline at the last time point T. We consider statistical methods for (i) a point estimate and 95 per cent confidence interval for the mean change from baseline at time T for each treatment group, and (ii) a p-value and 95 per cent confidence interval for the between-group difference in the mean change from baseline. The manner in which the baseline responses are used in the analysis influences both the accuracy and the efficiency of items (i) and (ii). In this paper, we will consider the ANCOVA approach with change from baseline as a dependent variable and compare that with a constrained longitudinal data analysis (cLDA) model proposed by Liang and Zeger (Sankhya: Indian J. Stat. (Ser B) 2000; 62:134-148), in which the baseline is modeled as a dependent variable in conjunction with the constraint of a common baseline mean across the treatment groups. Some drawbacks of the ANCOVA model and potential advantages of the cLDA approach are discussed and illustrated using numerical simulations.

Clinical, Functional, and Radiographic Benefits of Longterm Adalimumab Plus Methotrexate: Final 10-year Data in Longstanding Rheumatoid Arthritis

“…Radiographic progression was assessed in patients with mTSS assessments available at baseline and Year 10 (or at early termination, if occurring after Year 9). Differences in mean ∆mTSS between treatment groups were assessed using a constrained longitudinal data analysis 19 , which adjusts for the observed baseline difference in estimating treatment effects. Observed changes in mTSS were plotted as cumulative probabilities both for the entire 10-year period (from baseline) and for the 9-year open-label period (from Year 1).…”

A randomized controlled trial of attention modification for social anxiety disorder

“…A random-effects model, which is a commonly used to analyze clustered data (Hedeker & Gibbons, 2006;Liu, Lu, Mogg, Mallick, & Mehrotra, 2009) was used to test whether treatment condition or treatment location were associated with change in the selfreport measures over time (i.e., the SIPS, the CES-D, the STAI-T). We used a random-effects model for the analysis of change over time because this analysis does not require complete data at all measurement occasions.…”

Statistical methods to estimate treatment effects from multichannel electroencephalography (EEG) data in clinical trials

“…the values of the response variable measured before start of treatment, are the most likely candidates for inclusion into the model". Different approaches to incorporate baseline data into the model have been proposed (Liu et al, 2009;Crager, 1987). For example, the ANCOVA method (Crager, 1987) includes the baseline as a covariate, while the constrained longitudinal data analysis (cLDA) method (Liu et al, 2009) models the baseline as a dependent variable.…”