Bias and quality assessments of the

Bias and quality assessments of the included studies

The risk of bias in each included RCT was evaluated using the Cochrane Collaboration's ‘Risk of bias’ tool21. The quality assessment of the comparative cohort studies was performed using the Newcastle-Ottawa scale (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp), which included three main categories: (1) selection of cohort, (2) comparability of cohort, and (3) determination of outcomes.
Statistical analysis

Our meta-analysis was performed using the RevMan software (version 5.20, The Nordic Cochrane Centre, Copenhagen, Denmark) and Stata software (version 11.0, Stata corporation, College Station, TX, USA). For relapse-free survival, a hazard ratio (HR) and its 95% confidence interval (CI) were applied for analysis. For dichotomous outcomes (adverse events and response rate), risk ratios (RRs) were calculated, and these RRs were then pooled. Continuous variables were analyzed using mean differences (MDs) and 95% CIs. The heterogeneity of the included studies was analyzed using the Cochrane Q test and the I2 statistic, and P < 0.1 or I2 > 50% represented significant heterogeneity. If there was significant heterogeneity, we used a random effects model for the data analysis. Otherwise, we used a fixed effect model.
Results

Introduction•
Methods•
Results•
Discussion•
References•
Acknowledgements•
Author information

Our literature search identified 600 articles, of which 243 were from PubMed, 349 from Web of Science, and 8 from the Cochrane Library. Using Endnote software, 190 repeated studies were removed. After screening the titles and abstracts, 400 studies were excluded, and the remaining 10 articles underwent full-text screening. Five studies were excluded for the following reasons: two studies compared treatment effects before and after using rituximab in the same series of patients19, 22; another study was a comparison of two groups that both used rituximab, with or without mycophenolate mofetil23; the other two studies compared different doses of rituximab infusion24, 25. Finally, five studies were included for our analysis, three of which were RCTs18, 26, 27; one study was a retrospective comparative control study28, and one was a prospective comparative control study29. The study selection process is shown in Figure 1. The basic characteristics of the included studies are listed in Table 1.