Adenoma Prevalence With the Use of a Computer-Aided Detection Device (CADe) by Patient Demographics

Introduction: In recent years, computerized-aided devices (CADe) have been successfully implemented in screening colonoscopies to detect colorectal cancer (CRC), the fourth most common cancer in the United States. While screening colonoscopies provide the best method of early detection, nearly one-fourth of colorectal neoplasms may be missed during this procedure, resulting in post-colonoscopy interval CRC. Most currently available CADe studies do not stratify by sex and age, two main demographic risk factors for CRC. This study aimed to identify the colorectal adenoma prevalence associated with age and sex with the use of a CADe device. Methods: We performed a post-hoc analysis of the SKOUT registration study to evaluate the efficacy of a CADe. Patients were randomized to standard or CADe colonoscopy. The primary endpoint was adenomas per colonoscopy (APC). Secondary outcome measurements included adenoma detection rate (ADR) and adenomas per positive colonoscopy (APPC). In this post-hoc analysis, we compared outcome measures by age (< 65 vs. ≥65 years) and sex (male vs. female). Results: Of 1,418 participants (standard n=710, CADe n=708), 67.1% were < 65, and 52.8% were male. Using CADe, a significant increase in mean APC was observed in comparison to the control group for participants < 65 years (0.724 vs. 0.883, P=0.029), participants ≥65 years (1.055 vs. 1.367, P=0.032), and men (1.022 vs. 1.308, P=0.009). For ADR, a significant increment was observed for men (0.478 vs. 0.547, P=0.025), and a borderline significance was observed for participants < 65 years (0.396 vs. 0.450, P=0.054). For APPC, a significant increase was observed for participants ≥65 years (2.018 vs. 2.568, P=0.009) and women (1.583 vs. 1.894, P=0.032) (Table 1). Conclusion: Our study found that colonoscopy quality metrics increased across demographic subgroups by age and sex with the use of CADe, with the largest increase in APC being seen in subjects ≥65 years and in males. Further studies using CADe comparing adenoma detection and post-polypectomy cancer incidence should address its effect on interval CRC in these groups. With real world application of CADe, our data on potential differences by sex and age can help to create risk-adapted screening strategies and inform benchmarks for colonoscopy quality metrics. Table 1. - Adenoma Prevalence Measured by Adenomas Per Colonoscopy (APC) According to Age and Sex Between Subgroups of Different Study Arms Control group CADe group mean (SD) Difference (Treatment-Control) (95% CI) P-value of difference RR (Treatment vs. Control) (95% CI) P-value of relative risk APC Age < 65 years 0.724 0.883 0.159 (-0.010, 0.321) 0.029 1.22 (1.06, 1.41) 0.006 ≥ 65 years 1.055 1.367 0.312 (0.006, 0.663) 0.032 1.30 (1.10, 1.53) 0.002 Sex Male 1.022 1.308 0.286 (0.034, 0.534) 0.009 1.28 (1.12, 1.46) < 0.001 Female 0.615 0.758 0.143 (-0.043, 0.331) 0.064 1.23 (1.03, 1.48) 0.026 ADR Age < 65 years 0.396 0.450 0.054 (-0.007, 0.117) 0.054 1.14 (0.94, 1.38) 0.20 ≥ 65 years 0.523 0.532 0.009 (-0.083, 0.104) 0.42 1.02 (0.79, 1.31) 0.89 Sex Male 0.478 0.547 0.069 (-0.004, 0.139) 0.069 1.14 (0.94, 1.40) 0.19 Female 0.388 0.400 0.012 (-0.060, 0.085) 0.39 1.03 (0.81, 1.31) 0.81 APPC Age < 65 years 1.830 1.962 0.132 (-0.152, 0.434) 0.18 1.07 (0.93, 1.24) 0.34 ≥ 65 years 2.018 2.568 0.551 (0.053, 1.022) 0.009 1.27 (1.08, 1.51) 0.005 Sex Male 2.136 2.389 0.253 (-0.090, 0.611) 0.082 1.12 (0.98, 1.28) 0.10 Female 1.583 1.894 0.311 (-0.008, 0.651) 0.032 1.20 (1.00, 1.44) 0.06 Abbreviations: CADe, computer-aided detection; CI, confidence interval; RR, relative risk; SD, standard deviation.