In the dark : an experimental study in distinguishing an authentic herbal substance from a sham substance

Publication Type:
Issue Date:
Full metadata record
Background: The randomised controlled trial (RCT) is currently the “gold standard” for assessing the efficacy of a therapeutic intervention. One important aspect of an RCT is the blinding of various parties involved in the trial. Blinding is similar to randomization, in that its goal is to distribute participant expectancies evenly across the two or more groups. The assessment of blinding has been fraught with complications and associated with methodological difficulties. A review of the literature found very poor and limited reporting of the success of blinding in not only pharmaceutical studies, but also many Chinese medicine RCTs. In addition the special qualities of Chinese herbal medicine such as taste, odour, color and texture make it very difficult to blind trial participants. Because of these reasons it was decided to undertake a series of experiments to determine whether a credible sham substance could be developed for use in a clinical trial for Chinese herbal medicine. Methods: Three separate studies were undertaken in sequential order between 2009 and 2011, with the second and third study attempting to improve on each previous study. In all three studies the primary aim was to evaluate whether a greater proportion of participants could correctly identify an authentic herbal substance when comparing to a sham substance using visual, odour and taste assessments. The first study was a pilot involving eleven participants. The aim of this study was to pilot the data collection process and determine whether the initial materials were suitable for the study. The second study involved 81 participants. This study improved on the initial pilot study in several ways. There was an increase in participants and an opaque eye pad was used during the tasting phase to effectively blind the participant. In addition a closer colour matching of the three sham substances to the herbal substance was achieved. The third and final study involved only two substances, a refined sham substance and the herbal substance. Participants were randomly given only one of the two substances and asked whether they believed the substance to be a’ herbal’ or ‘placebo’ substance or ‘didn’t know’ using a questionnaire. In this study there was no comparison with more than one substance as occurred in study one and two. This study improved on the first and second study by improving the taste and colour of the sham substance in order to be comparable to the herbal substance. Results: For study one a one proportion test found that for both odour (p=0.484) and visual appearance (p=0.077) the number of participants correctly selecting capsule B as the herbal substance was not significantly different from what may have been selected by chance. This was not the case for taste where significantly more participants correctly selected capsule B as the herbal substance (p=0.004) compared to the other three substances. For study two, in all three assessments (visual, odour and taste) a statistically greater percentage of participants selected the herbal substance than the comparator sham substances (p<0.0001 in all three cases). For the final study a similar proportion of subjects in both groups believed they were observing and smelling an authentic herbal substance (p=0.435 for observing and p=0.443 for odour) suggesting effective blinding. However, similar to the two earlier studies a statistically larger proportion of subjects receiving the authentic herbal substances believing they were tasting a herbal substance (p=0.002) compared to the sham herbal substance group. Conclusion: This is the first time that a series of studies have been undertaken that have attempted to identify important cues and features that may threaten participant blinding in a herbal medicine trial. It was shown that with refinement it was possible to blind for visual and odour assessments but the sense of taste remained in all three studies a difficult human sense to blind. This obviously presents a problem for any herbal clinical trial where there is the opportunity for participants to break open a capsule and taste the contents. The studies have further demonstrated the need for pre-trial blinding for Chinese herbal studies and that a concerted effort should be made to develop a similar looking, smelling and tasting sham substance compared to the verum intervention arm. If possible inert materials, flavouring and colouring agents should be used. If this is not possible potentially active materials that have been shown to be irrelevant for the particular condition can be an option.
Please use this identifier to cite or link to this item: