Effectiveness of interventions to increase colorectal cancer screening: a systematic review, meta-analysis and meta-regression

  • A. Tsipa
  • D. O'Connor
  • M. Conner
  • L. Hall
  • B. Sykes-Muskett
  • D. Branley
  • F. Day


Background: Colorectal cancer (CRC) is the third most commonly diagnosed cancer worldwide. Screening is linked to significant reductions in CRC incidence and mortality, however, uptake rates remain alarmingly low. Methods: We conducted a systematic review, meta-analysis and meta-regression to examine the effectiveness of interventions aimed at increasing CRC screening uptake. We conducted a sensitivity analysis of studies that recruited vulnerable populations (minority, underserved, poor). Meta-regression examined whether intervention type, duration, gender, age, number of behaviour change techniques (BCTs) and other components of the intervention or methodology modified the overall intervention effect. Findings: A total of 102 RCTs involving 1.94 million participants were included. Meta-analysis showed significant benefit of all interventions combined (g= 0.26; OR 1.60; 95% CI 1.49-1.71, p<0.001); sensitivity analysis showed that interventions were significantly more beneficial when restricted to vulnerable population studies (k= 36, g=0.43, OR 2.18; 95% CI 1.78-2.67, P<0.001). Meta-regression showed that age, duration, number of BCTs and type of intervention did not modify intervention effects. Interventions that incorporated the BCTs ‘prompts/cues’ and techniques from the categories of ‘Social Support’, ‘Goals and Planning’ and ‘Shaping Knowledge’ were more effective than other interventions. Quality of the majority of the studies was unclear or poor (k = 86, 84.3%). Discussion: Techniques reducing structural barriers to screening were effective in increasing uptake, especially among vulnerable populations. The meta-analysis and meta-regression led to the identification of specific intervention components and BCTs potentially effective in increasing CRC screening uptake for different populations. A lack of psychosocial and theory-driven interventions was observed.
Oral presentations