Allocation bias happens when researchers don’t use an appropriate randomization technique, leading to marked, systematic differences between experimental groups and control groups. It can also happen further down the line, if clinical staff don’t follow the procedures set in place by the researchers.
One study (Salonen et al, 1992) highlighted a problem with hospital staff (particularly on the night shift), who bypassed the randomization procedure set in place to assign more “interesting” patients to the teaching service. Researchers and staff are also pressured by study participants, who want to receive a new treatment rather than a placebo.
Avoiding Allocation Bias
The key to avoiding allocation bias is randomization, which ensures confounders are spread across groups. When a randomized clinical trial isn’t properly randomized, the trial is neither randomized nor ethical (Dunford, 1990).
If participants are aware in advance which group they are being allocated to, this greatly increases the probability of allocation bias. Therefore, steps should be taken to ensure that patients are unaware of their group status. This could be achieved with blinding. Herbert et al. suggest a simple way to blind both the patient and the researcher from knowing a person’s allocation status: place the allocation in a sealed envelope, only to be opened when the study is over. A second way is to conduct the allocation offsite, with an independent company or facility.
Dunford, F. (1990). Random Assignment: Practical Considerations from Field Experiments. Evaluation and Program Planning.
Herbert, R. et al., (2011). Practical Evidence-Based Physiotherapy – E-Book. Elsevier Health Sciences.
Kirch, W. (2008). Encyclopedia of Public Health. Springer Science and Business Media.
Salonen et al., (1992) High store iron levels are associated with excess risk of myocardial infarction in Eastern Finnish Men. Circulation.