Abstract
Background
To avoid selection bias in clinical trials, random allocation concealment is crucial to ensure that participants and or researchers remain unaware of assignments.
Purpose
We aimed to design an allocation concealment method that reduced the possibility of selection bias for a randomized, open-label, crossover trial to evaluate device function of four female condom (FC) types.
Methods
Using scratch card technology, we devised a simple method of concealment, whereby the treatment sequence was printed on a single card for each participant, and the codes for each treatment in the sequence were concealed beneath foil squares on a stiff A6-sized card. On the first and subsequent follow-up visits, the foil corresponding to that visit was scratched from the square to reveal the condom type allocation for the next condom-use period. Staff in the South African and Chinese trial sites were trained in use and care of the card, and on completion of the study completed a questionnaire on their experience of use.
Results
Research staff in both countries found the card easy to use and those who had previously used the sequentially numbered, opaque, sealed envelopes (SNOSE) system for random allocation reported the scratch card easier to use. Research staff most commonly used a coin to remove the foil square and some used their fingernails. In both South Africa and China, no errors in allocation sequence were found during study monitoring.
Limitations
Scratch card system of allocation cannot be printed in-house.
Conclusions
This novel, effective method of concealment for a crossover random allocation was well liked by study staff. The most important advantage of this method is the ability to conceal consecutive allocations of a crossover design using a single card, thus eliminating the need for multiple envelopes per participant. While we used this method in a clinical trial of FCs, it could be employed in a range of other clinical trials and other randomized studies.
Get full access to this article
View all access options for this article.