Abstract
<title>Abstract</title> <p> Purpose To compare differences in inpatient costs between endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) for the treatment of patients with colorectal neoplastic lesions (including precancerous lesions and early-stage cancer), analyze the associated influencing factors, and conduct short-term and long-term cost-effectiveness evaluations, so as to provide a basis for clinical procedure selection and optimization of healthcare resource allocation. Methods Clinical data were retrospectively collected from 144 patients treated with ESD and 95 patients treated with EMR. Univariate analyses were performed using the Mann-Whitney U test and the Kruskal-Wallis H test. Multivariate linear regression and quantile regression models were further used to explore the influencing factors across different levels of cost distribution. Results Both univariate and multivariate linear regression analyses indicated that occupation, lesion site, maximum tumor diameter, and length of hospital stay were independent factors influencing hospitalization costs for ESD procedures ( <italic>P</italic> < 0.05); pathological findings, maximum tumor diameter, number of adenomas, and length of hospital stay were independent factors influencing hospitalization costs for EMR procedures ( <italic>P</italic> < 0.05). Quantile regression results showed that in the ESD group, hospitalization costs for patients who were specialized technical personnel decreased significantly at the 50th quantile ( <italic>P</italic> = 0.044), while unemployed patients exhibited a negative effect on costs at the 25th and 90th quantiles ( <italic>P</italic> = 0.05). Compared to lesions located in the right half of the colon, costs for lesions in the left half of the colon were significantly lower at the 25th percentile, and costs for rectal lesions were significantly lower at the 50th percentile ( <italic>P</italic> < 0.05). A maximum tumor diameter ≥ 2 cm significantly increased hospitalization costs at the 50th percentile ( <italic>P</italic> = 0.007); a hospital stay of ≥ 11 days had a positive effect on costs at the 10th, 25th, 50th, and 75th percentiles ( <italic>P</italic> < 0.05). Quantile regression analysis in the EMR group showed that a pathological diagnosis of early-stage cancer significantly increased hospitalization costs at all cost percentiles ( <italic>P</italic> < 0.01); a maximum tumor diameter ≥ 2 cm significantly increased costs at the 10th to 75th percentiles ( <italic>P</italic> < 0.05); 3 or more adenomas significantly increased costs only at the 25th percentile ( <italic>P</italic> = 0.032); a length of hospital stay ≥ 11 days had a significant positive effect on hospitalization costs at the 25th, 50th, and 75th percentiles ( <italic>P</italic> < 0.05). Conclusion Hospitalization costs for ESD and EMR treatments of colorectal lesions are influenced by multiple factors, and these factors exhibit heterogeneity across different cost quantiles. Length of hospital stay is the most robust positive cost driver for both surgical procedures, with a particularly pronounced effect in the mid-cost range. This suggests that in clinical practice, developing individualized treatment plans based on tumor size, lesion location, and patient occupational background, and shortening hospital stays through Enhanced Recovery After Surgery (ERAS) strategies, are key to achieving optimal allocation of healthcare resources. </p>