Change-to-Open-Label Design. Proposal and discussion of a new design for clinical parallel-group double-masked trials

The "Change-to-Open-Label Design" (COLA-design) is proposed to overcome some of the ethical and organizational problems of the usual double-masked design in certain situations while preserving its scientific rigor. Patients are randomized on a doubled-masked basis into treatment groups. During the trial the patient or the treating physician may ask for a change from the masked treatment to any open-label treatment of their choice (experimental or not), if the masked treatment is thought to be unsatisfactory. It seems to be easier to obtain a patient's informed consent to participate in placebo controlled double-masked trials especially for long term studies. The main outcome variable in COLA-design is the time until a patient demands for such a change. "Survival analysis" is a powerful statistical method to evaluate this outcome variable. Even patients lost from observation can be included in the evaluation: they still deliver censored outcome values. Evidently, in trials with a COLA-design the patient's personal impression of the therapy is the most important factor in its assessment. Therapeutical success is mingled with undesirable effects into the outcome measure; therefore this design is especially applicable to investigate treatments supposed to increase quality of life.