Sample size calculation

PASS 11.0 software was used to obtain sample size. As a parallel-group randomized controlled trial, the Log-rank Test (Lakatos) was used for the sample size calculation based on the following parameters or assumptions:

The survival probability of treatment group was estimated for 77.9%. According to the latest national multicenter survey data in China, 31.6% patients with a history of DFU developed new ulcer during 1 year follow-up [5]. Thus, the survival probability of control group was assumed to be 64.8%. We assume that the treatment group could have a 30% higher survival probability than control group.

The accrual period for patients’ recruitment was estimated for 12 months. The accrual pattern would be equal. The follow-up period would be 12 months. Therefore, the total time of this trial would last 24 months.

The power was set at > 0.8, while α was 0.05. The dropout rate of 0.2 was assumed. No switching between two groups. Two groups would have same sample size.

Therefore, the total sample size calculated using PASS was at least 314 cases. In this trial, we would recruit a total of 320 patients (160 for each group).