Alternating Assignment was Incorrectly Labeled as Randomization

A Letter Regarding “Effects of Composite Supplement Containing Astaxanthin and Sesamin on Cognitive Functions in People with Mild Cognitive Impairment: A Randomized, Double-Blind, Placebo-Controlled Trial”

We read with interest the article describing a study by Ito et al., investigating the use of dietary supplements as a means to slow the progression of cognitive decline among older adults [1]. The authors describe what they label a randomized controlled trial (RCT), the gold standard of experimental design. However, the method of participant group allocation described in this study is not randomization. The method in this study assigned participants to either the supplement or placebo using “an alternating quasi-randomization allocation method.”

By sequentially enrolling participants using alternating assignment, the researchers and enrolling physicians in this study were able to know to which group the next participant would be assigned, and there is no allocation concealment. Allocation concealment is described by the Cochrane Collaboration as the measures “taken to secure strict implementation of that schedule of random assignments by preventing foreknowledge of the forthcoming allocations [2].” The allocation method employed by Ito et al. allows the research team to determine in which group a participant would be assigned, and thus could (unintentionally) manipulate the enrollment. For group allocation to be truly random, researchers would need to have generated an unpredictable, random sequence and use the sequence so that group allocation is not known until participants are enrolled. The use of sequential, alternating assignment is in violation of both of these principles and is recommended to be “completely avoided” [3]. Alternating assignment, or similarly using patient chart numbers, days of the week, date of birth, etc., are nonrandom methods of group allocation, and should not be used in place of randomly assigning participants [4]. Using nonrandom methods is “not preferred because factors other than chance can influence group assignment or group assignment can be predicted, and therefore subverted, by knowing the assignment method and the groups to which patients have already been assigned [5].” Alternating assignment in place of randomization permits confounding (dependency between one or more covariates and treatment assignment in the data generating process [i.e., at the population level]) by observed or unobserved covariates. Notably, the fact that covariates selected in the present study (age, body mass index, geriatric depression scale, and a dementia scale), were not found to be significantly different between the two groups does not eliminate this possibility. Randomization ensures that both measured and unmeasured covariates are balanced in the data generating process [6].

The study by Ito et al. cannot be classified as an RCT. Group allocation that is not random should not be reported as such. Alternating assignment can lead to “serious problems” in the trial and lead to spurious results [6]. There are a number of disciplines (i.e., public health, community interventions, etc.) which commonly employ nonrandomized intervention evaluation studies, and these can be conducted with rigor [7]. It is crucial for researchers conducting these nonrandomized trials to report procedures accurately. We request that an erratum be issued to correct the scientific record and change the title of this study to reflect that the trial was not randomized.

Additionally, the authors do not provide sufficient information to adequately judge the primary results for which they report statistically significant outcomes, namely psychomotor speed and processing speed. Given the reported data, we do not have sufficient information to reproduce these results. The International Council on Medical Journal Ethics publication guidelines state that it is best practice to “describe statistical methods with enough detail to enable a knowledgeable reader with access to the original data to judge its appropriateness for the study and to verify the reported results” [8]. Similarly, the APA Publication Manual states “when reporting p values, report exact p values (e.g., p = 0.031) to two or three decimal places” [9]. We therefore request that the authors report the standard errors of the changes in the dependent variables (rather than just displaying them graphically) and the exact p-values, so that their results may be verified.

Bridget A. Hannon¹, J. Michael Oakes², David B. Allison^3
1Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, USA
²Robert Wood Johnson Foundation Interdisciplinary Research Leaders Program, Division of Epidemiology, University of Minnesota, Minneapolis, MN, USA
³Department of Epidemiology and Biostatistics, School of Public Health, Indiana University-Bloomington, Bloomington, IN, USA
Correspondence to: David. B. Allison, allison@iu.edu

ACKNOWLEDGMENTS & DISCLOSURE
Supported in part by NIH grants R25DK099080 and R25HL124208. The opinions expressed are those of the authors and do not necessarily represent those of the NIH or any other organization. Dr. Allison or his institution has received grants, contracts, consulting fees, or promises for same from Ikea; The Law Offices of Ronald Marron; Nestle Research Center; Tamasik, Kotin, & Kasserman; Biofortis; Fish & Richardson, P.C.; The Dairy Management Institute; Herbalife; and WW (formerly Weight Watchers).

REFERENCES
[1] Ito N, Saito H, Seki S, Ueda F, Asada T (2018) Effects of composite supplement containing astaxanthin and sesamin on cognitive functions in people with mild cognitive impairment: A randomized, double-blind, placebo-controlled trial. J Alzheimers Dis 62, 1767-1775.
[2] Green S, Higgins J (2005) Cochrane handbook for systematic reviews of interventions.
[3] Dettori J (2010) The random allocation process: Two things you need to know. Evid Based Spine Care J 1, 7-9.
[4] Hartung DM, Touchette D (2009) Overview of clinical research design. Am J Health Syst Pharm 66, 398–408,
[5] Lang TA., Secic, M, Lang T (2006) How to report statistics in medicine: An annotated guide for authors, editors, and reviewers. ACP Press.
[6] Simon SD (2001) Is the randomized clinical trial the gold standard of research? J Androl 22, 938-943.
[7] Des Jarlais DC, Lyles C, Crepaz N, Trend Group (2004) Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: The TREND statement. Am J Public Health 94, 361-366.
[8] International Committee of Medical Journal Editors (1997) Uniform requirements for manuscripts submitted to biomedical journals. Pathology 29, 441-447.
[9] DeCleene KE, Fogo J (2012) Publication Manual of the American Psychological Association. Occup Ther Health Care 26, 90-92.