Author's response to reviews Title: The effectiveness of retreatment with peginterferon alfa and ribavirin in patients with chronic viral hepatitis C genotype 2 and 3: a prospective cohort study in Brazil Authors:

Methods: From September 2003 to March 2009 a cohort of 216 patients who had previously failed therapy with a regimen of standard interferon and ribavirin, were followed in a specialized service implemented in the Brazilian Unified Health System, Rio Grande do Sul. All patients were retreated with PEG-IFN 2a or 2b per week, associated with RBV, through the oral route, with doses determined according to weight (1,000 mg if weight ≤ 75 Kg and 1,250 mg if ≥ 75 Kg) per day for 48 weeks. The HCV-RNA was tested by Polymerase Chain Reaction (PCR). Virological response (VR) within 48 weeks and SVR in the 72 weeks was considered for evaluation of treatment efficacy. Analyses were performed in patients who received at least one dose of PEG-IFN. 3. In several places within the Results section and Table 6, the abbreviation for sustained virologic response (SVR) is incorrectly written “RSV” or “RVS”. Sustained virological response (SVR) is now correctly written. 4. Several typing errors were noted, particularly in the Discussion. This made some sections of the manuscript difficult to understand. This manuscript will require additional proofreading for grammatical errors, spelling, and word choice. Grammatical errors corrected. Discretionary Revisions: 1. In the final paragraph of the Results section, the authors should consider revising the statement “control the confusion”. The authors seem to be referring to the utilization of multivariate analysis in order to control for potential bias in their assessment of predictors of SVR. This statement is noted again in the Discussion section. Methods: In multivariate analysis, the assessment of factors independently associated with SVR rate was applied to the Poisson regression analysis Results: In multivariate analysis, the assessment of factors independently associated with SVR rate was applied to the Poisson regression analysis, using as criteria for variable entry into the model p-value less than 0.20 bivariate analyses (Table 7). 2. Table 6 provides an analysis of factors associated with virologic response. The authors should consider commenting on whether any of these factors had a significant association with SVR when stratified by prior treatment response (nonresponders vs. relapsers). According to our analysis, SVR rates stratified by response to prior treatment (relapsers versus non-responders) were compared in the different groups using the chi-squared Person test, t-student and Mann-Whitney non-showing no statistically significant relationship between selected variables (age, gender, BMI, median aminotransferase levels, histological diagnosis and baseline serum HCV-RNA), probably due to the small sample size when stratifying the population between non-responders and relapsers. Level of interest: An article whose findings are important to those with closely related research interests Quality of written English: Needs some language corrections before being published Statistical review: No, the manuscript does not need to be seen by a statistician. Declaration of competing interests: I declare that I have no competing interests.