Few research studies have identified predictors of continued adherence with colorectal cancer screening recommendations. The aim of this study was to describe multiple patterns of participatory behaviour in three faecal occult blood test (FOBT) screening rounds and to determine social cognitive, demographic and background variables predictive of variations in adherence.
A random sample of 1,941 South Australian men and women aged 50-74 participated in a baseline behavioural survey followed by a three year screening program offering free annual FOBT (survey response rate 48%). Patterns of participation across the three screening rounds were recorded and described as one of five screening behaviours; consistent reparticipation (adherent with all screening rounds), consistent refusal (adherent with no screening rounds), drop out (adherent with earlier but not later rounds), intermittent reparticipation (adherent with alternate rounds) and delayed entry (adherent with later but not the initial round(s)). Multivariate generalised linear models determined survey variables predictive of categories of non-adherence relative to consistent reparticipation.
Adherence with three rounds was 55.8%. Self-efficacy, response efficacy and perceived barriers were marginally predictive of drop out and consistent refusal behaviours. Satisfaction with prior screening was predictive of drop out (RR= 0.58, p<.001, 95% CI 0.48-0.71), delayed entry (RR=0.77, p=.007, 95% CI 0.63-0.93) and consistent refusal behaviours (RR=0.74, p=.003, 95% CI 0.61-0.90). Unique demographic characteristics were associated with different non-adherent subgroups. No health insurance predicted drop out behaviour (RR=1.44, p=.020, 95% CI 1.06-1.96), male gender predicted delayed entry (RR= 1.71, p=<.001, 95% CI 1.28-2.29) and fewer general practitioner visits predicted intermittent reparticipation (RR=0.80, p=.015, 95% CI 0.67-0.96). Marital status, employment and age were also predictive of non-adherence.
This is the first study to apply a multi-level behavioural framework to explain longitudinal rescreening adherence. Unique demographic and behavioural characteristics were associated with different non-adherent behaviours.