Should I Take Aspirin? (SITA): randomised controlled trial of a decision aid for cancer chemoprevention

Abstract

Background Australian guidelines recommend that people aged 50–70 years consider taking low-dose aspirin to reduce their risk of colorectal cancer (CRC). Aim To determine the effect of a consultation with a researcher before an appointment in general practice using a decision aid presenting the benefits and harms of taking low-dose aspirin compared with a general CRC prevention brochure on patients’ informed decision making and low-dose aspirin use. Design and setting Individually randomised controlled trial in six general practices in Victoria, Australia, from October 2020 to March 2021. Method Participants were recruited from a consecutive sample of patients aged 50–70 years attending a GP. The intervention was a consultation using a decision aid to discuss taking aspirin to reduce CRC risk while control consultations discussed reducing CRC risk generally. Self-reported co-primary outcomes were the proportion of individuals making informed choices about taking aspirin at 1 month and on low-dose aspirin uptake at 6 months, respectively. The intervention effect was estimated using a generalised linear model and reported with Bonferroni-adjusted 95% confidence intervals (CIs) and P -values. Results A total of 261 participants (86% of eligible patients) were randomised into trial arms ( n = 129 intervention; n = 132 control). Of these participants, 17.7% ( n = 20/113) in the intervention group and 7.6% ( n = 9/118) in the control group reported making an informed choice about taking aspirin at 1 month, an estimated 9.1% (95% CI = 0.29 to 18.5) between-arm difference in proportions (odds ratio [OR] 2.47, 97.5% CI = 0.94 to 6.52, P = 0.074). The proportions of individuals who reported taking aspirin at 6 months were 10.2% ( n = 12/118) of the intervention group versus 13.8% ( n = 16/116) of the control group, an estimated between-arm difference of −4.0% (95% CI = −13.5 to 5.5; OR 0.68 [97.5% CI = 0.27 to 1.70, P = 0.692]). Conclusion The decision aid improved informed decision making but this did not translate into long-term regular use of aspirin to reduce CRC risk. In future research, decision aids should be delivered alongside various implementation strategies.