Statistical analysis
Baseline characteristics of the studied population were described as medians (ranges) for quantitative variables and numbers (percentages) for qualitative variables. The time to OMT second events was plotted using Kaplan-Meier curves. Baseline factors associated with OMT second events were identified by univariate analysis using the log-rank test if they were qualitative or by the Cox model if quantitative (such as age at diagnosis). Univariate analysis on the type of surgery was based on the population without missing data (6 cases excluded). Factors that were associated with a p<0.10 in univariate analysis were considered in a multivariate Cox model using a stepwise procedure for entering and staying in the model. All statistical analyses were performed with SAS software V9.4 (SAS Institute, NC, Cary), and a p-value <0.05 was considered to denote statistical significance.
This study was supported by the board of ethics committee (CLERS) and conducted in accordance with the Declaration of Helsinki. Institutional review board approval was obtained.