A cluster randomized controlled trial to evaluate the efficacy of esophageal and gastric cancer screening in mortality reduction in a non-high-incidence area: methodology and initial results.

Background: A cluster randomized controlled trial of endoscopy-based screening for esophageal cancer (EC) and gastric cancer (GC) was conducted to evaluate the efficacy and feasibility of this strategy in a non-high-incidence rural area of China. The trial design and baseline findings are presented here. Methods: A total of 33 eligible villages in Luoshan County in Henan Province were assigned randomly to the intervention or control group in a 1:1 ratio by a computer-generated randomization list. Local residents aged 40 to 69 years were enrolled from the villages. Participants in the intervention group were risk-stratified with a questionnaire, and high-risk individuals were subsequently screened by endoscopy. The primary outcomes were EC and GC mortality. The secondary outcomes comprised the detection rate, stage distribution, and the treatment rate. In this study, baseline characteristics were assessed by a questionnaire. Multivariate logistic regression analysis was performed to explore factors associated with endoscopy compliance. Results: Trial recruitment was completed in 2017, and ultimately, there were 12,475 and 11,442 participants allocated to the intervention (17 clusters) and the control group (16 clusters), respectively. We included 23,653 participants in the analysis, with 12,402 in the intervention group and 11,251 in the control group. A total of 6,286 (50.7%) participants in the intervention group were estimated as high-risk individuals, and 2,719 (43.3%) underwent endoscopy. Multivariate logistic regression analysis demonstrated that some factors including age, gender, education, personality and mental health, and upper gastrointestinal diseases or symptoms might affect endoscopy compliance. The detection rates for positive cases of EC and GC were 0.22% and 0.55%, respectively. The rates for esophageal and gastric precancerous lesions were 0.70% and 2.35%, respectively. The early detection rates for EC and GC were 50.0% and 33.3%, respectively. Additionally, the overall treatment rate for positive cases was 90.0%. Conclusions: The diagnostic yield of endoscopy-based screening for EC and GC was relatively low in a non-high-incidence rural area. The study may offer clues for the improvement of endoscopy compliance and the optimization of screening strategies for upper gastrointestinal cancer in non-high-incidence areas. Trial Registration: Chinese Clinical Trial Registry ChiCTR-EOR-16008577.