A six-month crossover chemoprevention clinical trial of tea in smokers and non-smokers: Methodological issues in a feasibility study

Abstract

Background: Chemoprevention crossover trials of tea can be more efficient than parallel designs but the attrition and compliance rates with such trials are unknown.Methods: Attrition (dropouts) and compliance with treatment were assessed in a 25-week randomized, placebo controlled, crossover, feasibility clinical trial of four tea treatments to investigate the effect of tea on oral cancer biomarkers. Each treatment lasted 4 weeks with 2 weeks of washout in between. Participants were 32 smokers and 33 non-smokers without any evidence of premalignant oral lesions. The interventions consisted of packets of green tea, black tea, caffeinated water, or placebo. Participants were assigned to each treatment for four weeks, and were instructed to drink five packets per day while on the treatment. Dropout from the trial and compliance (consumption of ≥ 85% of the prescribed treatment packets) are the main outcome measures reported.Results: There was a high rate of dropout (51%) from the study, and the rates were significantly higher among smokers (64%) than non-smokers (36%). Among participants who completed the study the rate of compliance was 72%. The highest rates of dropouts occurred between the first and second treatment visits in both smokers (38% dropout) and non-smokers (18% dropout). Throughout the study smokers were more likely to dropout than non-smokers. Black tea treatment was associated with the highest rates of dropout among smokers (37%), but was associated with the lowest rate of dropout among non-smokers (4%).Conclusions: In a study conducted to test the feasibility of a four-treatment crossover tea trial, a high rate of dropout among smokers and non-smokers was observed. Multi-arm crossover tea trials might pose a higher burden on participants and research is needed to improve adherence and treatment compliance in such trials.Trial registration number: ISRCTN70410203. © 2012 Dash et al.; licensee BioMed Central Ltd.