Do Methodological Flaws Invalidate a Randomized Trial of Lifestyle Modification Programs in Obese Patients?

IN RESPONSE:

Dr. Berger raises 2 criticisms of our study. First, he claims that permuted block randomization with a fixed block size of 5 was inappropriate because it could lead to prediction of future allocations. Although such prediction is a valid theoretical concern, it implies an unlikely situation in which study personnel manipulate the randomization scheme at every step. Block size was neither included in the protocol nor made known to study personnel during the conduct of our study. Allocations were implemented with a voice response system that was transparent to study personnel. Thus, prediction of allocations was mitigated, and the block size ensured balance among intervention groups within each center, an important consideration for an interventional study. As evidence of the inadequacy of our randomization scheme, Dr. Berger notes that race is poorly distributed, but in doing so, he has chosen a single unbalanced characteristic out of several possible characteristics. We formally compared baseline characteristics among intervention groups and found no significant differences. Specifically with respect to race, most participants were white, and the proportion of white persons did not significantly differ among intervention groups. The fact that the relatively small number of black persons was unevenly distributed was probably nothing more than an accident of randomization and has little, if any, effect on the results. Dr. Berger goes on to suggest that the maximal procedure, an alternative method of randomization which enforces balance on known covariates, should have been chosen over permuted blocks; however, no definitive opinion exists in the wider statistical community on the question of whether covariates should be taken into account in trial design (1).

The second criticism refers to the withdrawal rate and the lack of a pure intention-to-treat analysis. Withdrawal is a common occurrence in weight-loss studies, and the rates in this trial, which ranged from 26% to 34% across intervention groups, were moderate compared with other studies in the literature (2). Furthermore, because the most common categories of withdrawal were loss-to-follow-up and consent withdrawal, we are uncertain whether efforts to obtain 6-month weight measurements from patients who withdrew would have been successful. Dr. Berger states that we neglected to include patients who withdrew in the analysis, but this is not the case. In the primary longitudinal analysis, patients who withdrew were included to the extent to which their data were available. In this regard, the analysis satisfies the intention-to-treat principle and preserves the randomization for the purpose of statistical inference. We also conducted sensitivity analyses, including a worst-case scenario that assumes that every patient who withdrew returned to their baseline weight by 6 months. Although the magnitude of weight loss decreased in this scenario, the relative pattern among the treatment groups was maintained. Dr. Berger also questions the assumptions of the primary analysis. We do not feel that an assumption of normality is unrealistic for body weight data in a trial of this size. Furthermore, the mixed-model repeated measures approach has been recommended by experts (3, 4) as a means of dealing with missing data in longitudinal studies.