Introduction
Colorectal cancer (CRC) screening reduces cancer risk, yet real-world adherence to screening remains variable. We aimed to leverage a large healthcare organization longitudinal cohort to evaluate factors that influence the timeliness of adherence to screening guidelines.
Aims & Methods
We included average-risk Maccabi Healthcare Services members aged 50–75 years who were alive on 1 January 2025. Demographics, comorbidities, hemoglobin values, and dates of fecal immunochemical test(FIT), CT-colonography (CTC) and colonoscopy were extracted.
Adherence was defined as annual FIT, colonoscopy every 10 years, or CTC every 5 years.
Patients covered for >50 % versus ≤50 % of follow-up time by screening guidelines were compared with Wilcoxon rank-sum (continuous) and χ²/Fisher’s exact (categorical) tests.
In order to study factors associated with adherence times, longitudinal adherence was modelled with a gap-time Andersen–Gill Cox regression. The dependent variable was time (in periods) to the recurrent event active adherence. Adjusted hazard ratios (HRs) and 95 % confidence intervals (CI) were estimated with robust variance, clustering on patient ID to account for within-subject correlation.
Results
Our cohort comprised 590,550 adults at screening age, contributing 6,581,044 person-years of follow-up. In a baseline comparison between individuals whose proportion of time covered was less than 50%(n=282,347) vs more than 50%(n=308,203), the latter group was older (60.00 [55.00, 68.00] years vs 62.00 [57.00, 67.00] years, p-value<0.001) with a longer median follow up (9.0 [4.0, 17.0] years vs 11.0 [6.0, 16.0] years, p-value<0.001). Higher prevalence of hypertension (17.5 % vs 12.0 %, p-value<0.001), diabetes mellitus (8.2 % vs 5.8 %, p-value<0.001), pre-diabetes (22.4 % vs 17.5 %, p-value<0.001), and cardiac disease (4.8 % vs 3.4 %, p-value<0.001) were seen among the more compliant group. Consistent with greater adherence, endoscopic referral was more common at baseline in the more compliant group (6.2 % vs 1.0 %, p-value<0.001). Interestingly. Although the differences in distributions of sector, district and socioeconomic rank categories were significant, the absolute percentage differences across strata were small and unremarkable (<3 percentage points for any category).
A gap-time survival model demonstrated that each additional year of age was associated with longer intervals and a lower aHR for active adherence (aHR 0.94, 95 % CI 0.94–0.94). Men undergone screening tests marginally sooner than women (aHR 1.05, 95% CI 1.04–1.05). Adherence was lower in the orthodox sector (aHR 0.84, 95% CI 0.84–0.85). Metabolic or cardiovascular conditions translated into earlier adherence: pre-diabetes (aHR 1.63, 95% CI 1.62-1.63), diabetes (aHR 1.54, 95% CI 1.54-1.55), hypertension (aHR 1.22, 95% CI 1.21-1.23), cardiac disease (aHR 1.14, 95% CI 1.14-1.15), and obesity (aHR 1.10, 95% CI 1.10-1.11). Severe conditions and subjects included in severe disease registry were associated with lower adherence (aHR 0.33, 95% CI 0.30-0.38; and aHR 0.63, 95% CI 0.59-0.67, respectively). Recent higher hemoglobin levels delayed active adherence (aHR 0.97 per each increase of 1 g/dL, 95% CI 0.97-0.98).
Conclusion
Our results suggest that long term compliance for CRC screening is related to age as well as for chronic medical conditions, particularly metabolic and cardiovascular. Screening efforts should therefore be aimed at healthy subjects or those with minor contact with medical providers.
