Randomised controlled trial and economic evaluation of a targeted cancer awareness intervention for adults living in deprived areas of the UK

Cancer outcomes are poorer in socioeconomically deprived UK communities, partly due to delayed help-seeking resulting from low symptom awareness. Existing mass media campaigns often fail to reach these groups. This necessitates targeted interventions delivered within community networks, leveraging trust and personal connections. The ABACus Health Check intervention, co-produced with local stakeholders, aimed to improve cancer symptom awareness and prompt timely help-seeking. This trial evaluated its effectiveness in deprived areas of Yorkshire and Wales, building upon previous feasibility evidence demonstrating successful engagement and potential for scalability.

Existing literature highlights socioeconomic inequalities in cancer survival and stage at diagnosis, with delayed presentation being a contributing factor. Studies show that low awareness of cancer symptoms is linked to longer help-seeking delays. Behavioral factors such as perceived barriers, fear, and fatalism influence presentation. While mass media campaigns may improve outcomes, their reach in deprived communities is limited. Community-based behavioral interventions facilitated by trusted peer supporters are suggested as a promising approach to engage underserved populations and normalise earlier help-seeking. The ABACus Health Check was developed based on these insights, integrating behavior change techniques and social support.

This was an unblinded, individually randomized controlled trial involving adults aged 40+ years recruited from community and healthcare settings in deprived areas of South Yorkshire and South-East Wales. Participants were randomized to either the ABACus Health Check intervention (personalized behavioral advice from a trained lay advisor) or usual care. Data were collected at baseline, two weeks, and six months post-randomization. The primary outcome was total cancer symptom recognition at two weeks, assessed using an adapted Awareness and Beliefs about Cancer (ABC) measure. Secondary outcomes included anticipated symptom presentation, barriers to presentation, beliefs about cancer, state anxiety, intervention costs, and healthcare resource use. Intention-to-treat analysis was used, with mixed-effects models accounting for clustering by lay advisor. A health economic evaluation compared costs and outcomes between groups.

234 participants were randomized (117 per arm). The difference in total symptom recognition at two weeks (adjusted mean difference (AMD) 0.6, 95% CI: -0.03, 1.17, p = 0.06) was not statistically significant. However, at six months, the intervention group showed significantly increased symptom recognition (AMD 0.8, 95% CI: 0.18, 1.37, p=0.01) and earlier intended presentation (AMD -2.0, 95% CI: -3.02, -0.91, p < 0.001). Recognition of "lesser-known" symptoms was higher in the intervention arm at both two weeks and six months. The implementation cost per participant was £91.34, with no significant between-group differences in healthcare resource use post-intervention. Baseline symptom awareness was high, suggesting a ceiling effect, especially for "well-known" symptoms. There were no significant effects on perceived barriers, beliefs about cancer, or state anxiety.

The lack of short-term effects on total symptom recognition, despite longer-term benefits, might be due to measurement reactivity, a common challenge in cancer awareness intervention trials. The study's success in engaging a hard-to-reach population and achieving longer-term effects highlights the potential of personalized, facilitated interventions. The low cost and lack of additional resource use make it a viable option. The improved recognition of "lesser-known" symptoms is crucial, as these are often delayed in presentation.

The ABACus Health Check is a viable, low-cost intervention to increase cancer awareness and prompt earlier symptom presentation in deprived communities. Further research is needed to confirm longer-term effects on diagnosis and outcomes, as well as to explore scalability and adaptation for diverse populations.

Self-report measures might underestimate actual behavior. The sample was predominantly white British, limiting generalizability. The adapted ABC measure had low internal consistency. The study did not account for potential exposure to other cancer awareness campaigns in the control group.