Regional and Patient-Level Determinants of Endoscopic Utilization in Rural Healthcare: A Multi-Level Analysis

Muye Ma¹Pengfei Li¹Zhengyang Lu¹Nan Zhang²Suzhen Wang¹Youhua Lu²Jinming Yu^2*

• ¹School of Public Health, Shandong Second Medical University, Weifang, Shandong, China
• ²Shandong Cancer Hospital and Institute, Shandong First Medical University & Shandong Academy of Medical Science, Jinan, Shandong, China

Objective: Despite numerous studies on endoscopic services in urban settings, tailored assessments in rural healthcare remain limited, creating a gap in our understanding of resource-constrained environments. To address this gap, this study innovatively applied the Andersen Behavioral Model to systematically quantify endoscopic examination uptake and identify both patient-level (e.g., occupation, health history) and region-level (e.g., infrastructural challenges, socioeconomic indicators) factors influencing service utilization in rural China.Methods: We employed a multi-level logistic regression model with random intercepts to account for intraregional correlation and fixed effects for individual predictors. A multi-stage stratified random sampling approach was employed across 6 prefectures, yielding a sample of 1118 patients. We initially used descriptive statistics to summarize basic sample characteristics. Univariate analysis was then conducted to identify potential factors associated with endoscopic examination utilization. To further quantify these associations, we applied single-level and multi-level logistic regression model to account for potential regional effects and provide more robust analysis.Results: Of the 1,118 surveyed patients, 62.3% underwent endoscopic examinations, and among these, 77.9% received services at county-level institutions. In single-level binary logistic regression, region, occupation, household size, history of gastritis/esophagitis, and lesion location emerged as significant predictors (P < 0.05). In the multi-level logistic regression model, region remained a key factor, with the western region exhibiting 0.661 times lower odds (95% CI: 0.392–1.115) and the central region revealing 1.398 times higher odds (95% CI: 1.006–1.943) of service utilization compared to the eastern region. Additionally, unemployed status was associated with a 20% increased likelihood, and smaller household size correlated with a 87% increase in screening uptake.Conclusions: Our findings underscore the importance of addressing regional disparities through targeted resource allocation and localized health education programs to improve endoscopic service uptake among rural populations. These insights can inform policy interventions aimed at early cancer detection and optimized healthcare delivery in resource-limited settings.