Dear Editor,

We were interested to read the article “Prognostic factors and survival in Ewing’s sarcoma treated by limb salvage surgery” authored by Alvarez-SanNicolas et al. published in Clinical and Translational Oncology [1]. The authors aimed to determine survival and prognostic factors in Ewing’s sarcoma (ES) patients undergoing limb salvage (LS). If some of the statistical problems are clearly stated, the quality of the article would be better.

Cox proportional hazards analysis and univariate analysis were used to identify predictions EFS factor. Some variables in multivariate analysis were directly imported into the Cox’s multivariate analysis without first being tested by univariate analysis. Regarding the number of available samples taken into consideration, variables for inclusion should be selected carefully, to ensure parsimony of the final model [2, 3].

Prerequisites for the use of Cox proportional hazards models, the proportional effect of the variables on the results remained unchanged during the follow-up period. The authors should first test the proportional hazard assumption. Extended Cox model is an option if the proportional hazard assumption is violated.