Ways to use interventions to increase participation in mail-out bowel cancer screening: a systematic review and meta-analysis

Abstract

The impact of colorectal cancer can be reduced through nationwide fecal occult blood test (FOBT) screening. Unfortunately, participation in screening programs are low with interventions only increasing participation modestly. This meta-analysis explores if intervention effectiveness can be increased by targeting specific subpopulations with specific interventions or by combining interventions. Six databases were searched for studies aiming to increase participation in mail-out FOBT screening. To investigate if interventions are more effective for certain subpopulations, the difference in (log) Risk Ratios (RRs) between alternate subpopulations (male vs. female; low vs. high Socioeconomic Status (SES); <65 vs. ≥65 years) was assessed. To investigate if interventions should be combined, uptake rates for single interventions were compared to uptake rates for combined interventions. Cochrane Collaboration tools were used to assess the risk of bias. Searches found 3,436 articles, with 32 meeting the inclusion criteria. These contained 30 trials that reported uptake rates within subpopulations and 17 trials that combined interventions. Most differences in intervention effects between subpopulations were nonsignificant. Combining interventions led to greater participation, RR = 1.06, confidence interval [1.03; 1.10]. As interventions rarely affect subpopulations differently, targeting them at specific subpopulations may be an ineffective strategy. While individual interventions show modest effects, these results indicate that future programs might overcome this by combining interventions together. Care is needed when selecting interventions to combine as adding some interventions (e.g., additional print materials) can reduce the effectiveness of a combined strategy. Future research should examine methods for effectively combining interventions in nationwide programs to maximize participation.