J Thorac Cardiovasc Surg 2008;135:469-471
© 2008 The American Association for Thoracic Surgery
Reply to the Editor
Alberto Ruol, MD, FACS,
Giuseppe Portale, MD
Department of Gastroenterological and Surgical Sciences, Clinica Chirurgica III, University of Padova School of Medicine, Padova, Italy
We appreciate Dr Yung's concerns about the quality of scientific articles published in peer-reviewed journals such as the Journal of Thoracic and Cardiovascular Surgery, which are read by thousands of specialists in both thoracic and general surgery. It has been a privilege for us to have our data published in the Journal and we would like to reassure Dr Yung that the review process of the Journal has not failed to detect possible significant limits in our manuscript, as he suggested.
We will supply the following details.
- 1. The survival curves shown in the figures were generated with the Kaplan–Meier method, but the "reproduction" process failed to show the typical staircase appearance. We are submitting 2 new figures with the censored numbers (
Figures 1 and 2). The P values are the same because the survival estimate had already been correctly calculated with the Kaplan–Meier method.
- 2. The median survival for all patients was 29.8 for patients younger than 70 years old and 20.8 for those older. Once again, nothing was wrong with the survival analysis we had provided the readers; it was just a typing mistake, as anyone truly familiar with statistics surely has already realized.
- 3. We obviously included in the multivariate analysis variables significant on univariate analysis, plus other variables that, although not significant on univariate analysis, the reader might have been interested to see in the multivariate analysis (eg, use of neoadjuvant therapy, histologic type). We did not include variables of no clinical interest (eg, region where the patients came from) but only variables of clinical interest. We also are providing a new multivariate analysis that includes the variables suggested by Dr Yung. Clearly, pStage, as anyone could expect, remains the most significant factor predicting long-term survival (Tables 1 and 2).
- 4.
2 Is not meaningless and there are hundreds of papers with 2 rather than hazard ratio. In any case, we are supplying the data Dr Yung was interested in (Table 3). Obviously, and once again, this was clearly not an issue of statistical "competence" or "ignorance," but rather preference.
- 5. We want to reassure Dr Yung (not the American Society of Anesthesiologists [ASA], which is already aware of this) that the ASA score does work for esophageal cancer. The fact that mortality rates in younger and elderly patients are similar does not necessarily imply that there cannot be differences in the ASA class distributions among the groups. When we analyzed patients from both groups and compared mortality and morbidity of patients in ASA classes I–II versus those in ASA classes III–IV, this is what we learned (Table 4).
The ASA classes do make a difference in the risk of mortality.
We hope Dr Yung is now pleased with the completeness of the data provided. There is definitely no need to "recover" this article. We are sure the reviewers of the Journal who took care of our—not that "limping," really—manuscript were at least as precise and meticulous in their work as Dr Yung would have been. Hypercriticism for the sake of being critical does not help any progress in surgical research.
