We read with interest the comments and suggestions made by Dr MC Kendall and Dr LJ Castro-Alves with regard to our article titled "Predictors for 30-day readmission after pulmonary resection for lung cancer." We would like to thank them for giving us the opportunity of rebuttal.We are afraid we disagree with your statement that acute pain is a frequent cause of emergency readmission after lung resection surgery. In our study, acute pain was not a cause of readmission in any case, as it was not in most studies assessing causes of readmission after lung resection surgery. 1-3 We only found some reference to acute pain as a cause of readmission, as a rare cause in the study of Bhagat et al 4 and in another study conducted by Stitzenberg et al 5 where the authors documented chest pain as a cause of readmission in a retrospective cohort of registered patients registered between 2001 and 2007.Controlling postoperative pain is crucial in the management of patients undergoing lung resection surgery. As in most thoracic surgery university hospitals, our analgesic protocol includes epidural anesthesia and the analgesics acetaminophen, dexketoprofen, and morphine as rescue medication. At discharge, pain is controlled with oral analgesics for at least 2 weeks. In our opinion, postoperative pain should not be a cause of readmission after lung resection surgery, as it can be successfully managed in an ambulatory care setting.We agree that patients' health literacy may be a factor of influence in readmission; however, the educational level was not associated with readmission in a univariate analysis performed in a recent study. 6 Although this factor could have some influence on the readmission in our study, this factor was not included, as our patients received detailed information on the whole process, regardless of their health literacy.In relation to the third comment, it is worth noting that only seven variables were included in our multivariate analysis to select five factors of interest for the final predictive model. We did not include length stay, mortality at 90 days or reintubation, and ventilatory support in our predictive model. The reason was that these factors are associated with other variables, such as pneumonia or atelectasis, which would have reduced the possibility of collinearity in our model. In agreement with other authors, we think that all predictive models with a high discriminative power can result in overestimation. Overestimation may reduce the discriminative power of the model in cohorts of patients with other characteristics. However, in the Discussion section, we specify that, as it occurs with any risk prediction model, the validity of the model must be confirmed by external validation. 7
CONFLICTS OF INTERESTThe authors declare that there are no conflicts of interest.
ORCID
FlorencioQuero-Valenzuela REFERENCES 1. Freeman RK, Dilts JR, Ascioti AJ, Dake M, Mahidhara RS. A comparison of length of stay, readmission rate, and facility reimbursement after lobectomy of the lung. Ann Thorac Surg. 2013;...