Download This Paper
Preprints with The Lancet is a collaboration between The Lancet Group of journals and SSRN to facilitate the open sharing of preprints for early engagement, community comment, and collaboration. Preprints available here are not Lancet publications or necessarily under review with a Lancet journal. These preprints are early-stage research papers that have not been peer-reviewed. The usual SSRN checks and a Lancet-specific check for appropriateness and transparency have been applied. The findings should not be used for clinical or public health decision-making or presented without highlighting these facts. For more information, please see the FAQs.
An Endoscopic Ultrasound-Based Application System for Predicting Endoscopic Resection-Related Outcomes and Diagnosing Subepithelial Lesions: A Multicenter Prospective Study
23 Pages Posted: 3 Feb 2023
More...Abstract
Background: Subepithelial lesions (SELs) are associated with various endoscopic resection (ER) outcomes and diagnostic challenges. We aimed to establish a tool for predicting ER-related outcomes and diagnosing SELs and to investigate the predictive value of endoscopic ultrasound (EUS).
Methods: Phase 1 (system development) was performed in a retrospective cohort (n=837) that underwent EUS before ER for SELs at eight hospitals. Ten EUS parameters were collected as potential predictors. Prediction models for post-ER ulcer management, procedure time, hospital stay, medication costs, and diagnosis were developed using logistic regression. Models with satisfactory internal validation performance were included in a mobile application system—SEL endoscopic resection predictor (SELERP). In Phase 2, the models were externally validated in a prospective cohort of 200 patients. Risk prediction models were assessed using C-statistics and calibration. Diagnostic models were evaluated for sensitivity and specificity.
Findings: SELERP was developed using EUS characteristics, which included 10 models for five key outcomes: post-ER ulcer management, short procedure time, long hospital stay, high medication costs, and diagnosis of SELs. EUS features significantly contributed to all models. In Phase 1, 10 models were derived and validated (C-statistics, 0·67–0·99; calibration-in-the-larges, -0·14–0·10; calibration slopes, 0·92–1·08). In Phase 2, the derived risk prediction models showed convincing discrimination (C-statistics, 0·64–0·73) and calibration (calibration-in-the-large, -0·02–0·05; calibration slopes, 1·01–1·09) in the prospective cohort. The sensitivities and specificities of the five diagnostic models were 68·3%–95·7% and 64·1%–83·3% respectively.
Interpretation: We developed and prospectively validated an EUS-based application system for the prediction of ER procedural outcomes and differential diagnosis of SELs, which could aid clinical decision-making and facilitate patient-physician consultation.
Trial Registration: http://www.chictr.org.cn (ChiCTR2000040118).
Funding: National Key R&D Program of China (2021YFE0202000); Shandong Provincial Key Research and Development Program (Major Scientific and Technological Innovation Project) (NO.2021CXGC010506); China Scholarship Council.
Declaration of Interests: Yonghang Lai and Yiyan Zhang are employed by Qingdao Medicon Digital
Engineering Co., Ltd. The other authors declare no conflict of interest
Ethics Approval: This study was approved by the Ethics Committee on Scientific Research of Qilu
Hospital of Shandong University (KYLL-2020-448), followed by approval of the institutional review
board of each participating institution.
Keywords: subepithelial lesion, submucosal tumor, endoscopic ultrasound, endoscopic resection, prediction, diagnosis
Suggested Citation: Suggested Citation