Evaluating practitioner-blinding in Chinese herbal medicine research: Findings from a randomised feasibility study in the United Kingdom

Abstract

Purpose: Practitioner-blinding is often carried out in randomised controlled trials (RCTs) of Chinese herbal medicines (CHMs) yet evaluation of blinding is infrequently conducted. We aimed to evaluate the feasibility of practitioner-blinding within a UK study and identify reasons for practitioner guesses. Methods: We conducted a practitioner and patient-blind feasibility study exploringCHM for polycystic ovary syndrome, randomising 40 women to standardised CHM or individualised CHM for 24 weeks. We evaluated practitioner-blinding at Week 4, 12 and End of Study (EoS). This questionnaire invited a treatment allocation guess (Standardised/Individualised) and certainty rating (Not at all sure, just guessed/Fairly sure/Entirely sure). This was used to calculate a Bang Blinding Index (BBI). The final item asked for reasons for their answer, analysed using content analysis. Results: Completion rates of blinding-questionnaire was excellent (mean=86%). Practitioner-guessing and BBI at Week 4 was standardised random/individualised unblinded (standardised -0.11,95%CI -0.35 to 0.14; individualised 0.47,95%CI 0.2 3 to 0.71), at Week 12 standardised random/individualised unblinded (standardised -0.24,95%CI -0.54 to 0.07; individualised 0.50,95%CI 0.12 to 0.88); EoS standardised opposite/individualised unblinded (standardised -0.56,95%CI -0.9 1 to -0.20; individualised 0.61,95%CI 0.30 to 0.92). ‘Presence of effects’ was the highest ranking reason for treatment guess (52% of responses) and consistently led to a guess of ‘individualised treatment’/‘not at all sure’, and ‘absence of effect’ consistently led to a guess of ‘standardised treatment’/‘not at all sure’. This can be interpreted as ‘wishful thinking’ scenario whereby the practitioner consistently believed treatment responsewas due to individualised treatment, suggesting that blinding was likely secure. Conclusion: We have demonstrated that practitionerblinding is feasible and likely secure in this feasibility study. To our knowledge, this is the first time practitioner-blinding has been rigorously evaluated in a CHM study in the UK. Qualitative data has provided further insight into practitioner reasons for treatment guess which will be used to maximise practitioner-blinding in future studies. Contact: Lily Lai, l.y.w.lai@southampton.ac.uk