Lumen-apposing metal stents (LAMS) versus plastic stents for EUS-guided drainage of walled-off necrosis (WON) (LVPWON): study protocol for a multicenter randomized controlled trial

The sample size calculation is based on a literature review and conference discussion regarding the clinical success rate of LAMS or DPPS drainage for WON. The LVPWON trial is a superiority trial in which the sample size calculation was based on the assumption that the incidence of the primary endpoint with LAMS and DPPS is 90% and 75%, respectively. Assuming a one-side alpha (type I error) of 0.025 and a power of 85% (beta, a type II error, was taken as 0.15), we used the following formula to calculate sample size and found that 256 patients (128 per condition) would be necessary, including a possible dropout rate of 10%.

\( n={\frac{\mu_{1-\alpha}\sqrt{2\overline{p}\left(1-\overline{p}\right)}+{\mu}_{1-\beta}\sqrt{p_T\left(1-{p}_T\right)+{p}_C\left(1-{p}_C\right)}}{{\left({p}_T-{p}_C\right)}^2}}^2 \)

In this formula, p_T represents expected success rate of the LAMS group, p_C indicates expected success rate of the DPPS group, \( \overline{p} \) is average success rate of the two groups, and μ represents the quantile of standard normal distribution.

Patient characteristics will be recorded in a CRF. An electronic data collection system (electronic data capture [EDC]) will be used to complete the trial data collection. The EDC system has been rigorously tested to fully meet the requirements of the “Quality Management Practice for Medical Device Clinical Trials” and “Technical Guidelines for Clinical Trial Data Management Work.” Before the system is officially launched, training tests will be conducted on relevant users to ensure that the system meets the trial requirements. After the official launch, related personnel will receive the account number and password. The account is bound to the user’s role and permissions: the user must properly maintain the account information, not inform others about the account information, and exercise appropriate rights for others. Clinical auditors will monitor the clinical trial center’s work at least once a month. This trial has an interim analysis once half of the required number of the patients have been enrolled, during which the data will be analyzed. If a serious AE is identified, the study will be terminated immediately.

Categorical data will be described using frequency and composition ratios. Continuous data will be described using mean, standard deviation, maximum, minimum, and median, as well as 25th and 75th quantiles. Baseline demographic analysis based on the descriptive analysis, the likelihood ratio χ2 test will be used for the comparison between the categorical data groups. Fisher’s exact probability method will be used when theoretical frequency of cells exceeding 25% is < 5. Normally distributed continuous datasets will be compared using a group t-test. For non-normally distributed continuous data, Wilcoxon rank sum test will be used for comparison between groups. This is a multi-center study. The test conditions of each center are not identical (such as operators, etc.) and the outcomes may be different. For the primary outcome, comparison between groups will be performed using the CMH (Cochran–Mantel–Haensel) χ2 analysis to adjust for the central effect. The differences between the success rates of the groups and their 95% confidence intervals (CI) will also be estimated. AEs will be described by the number and incidence of AEs; the likelihood ratios between groups will be compared using the χ2 test or Fisher’s exact probability test. Major investigators will have access to the final trial dataset and the statistician involved will be blinded to the treatment assignments. Statistical analysis will be performed with two-sided tests at a level of significance of 0.05 (except where specified). Data analysis will be conducted using SAS^®9.4 statistical software.