Data source The Italian VATS Group Database is a validated, risk-adjusted, prospective, outcomes-based program with 50 participating hospitals in Italy. The VATS Group Database collects data on different variables on a standardized data form: patient demographics informations, surgical interventions, medical history, cancer staging and outcome. Patients' outcomes are reviewed and records are updated the first time at 30 days after surgery, then at 180 days. Next update is recorded at 6 months from surgery and every 6 months for the first 2 years of follow-up, and annually thereafter. The VATS Group Database implements rigorous quality assurance and safety procedures to maintain a high level of accuracy and security of data.
The study has been approved by hospital ethics committee. The current analysis was reviewed and approved for scientific merit and feasibility by the VATS Group Scientific Committee and presented at the annual VATS Group meeting.
Patient population and study design Study population consisted of patients who received intended VATS lobectomy as the primary procedure for NSCLC at VATS Group participating centres and included in the VATS Group database between January 2014 and January 2020. We excluded all patients with incomplete clinical, intraoperative and postoperative data. Data entered from centers sporadically (less than 10 cases per year) were also excluded.
In order to analyse the factors related to lenght of stay, the study design provides a comparison between two groups of patients: the first group, identified as "LOS ≤ 4" including all patient disharged within 4 days after surgery, while the second group, called "LOS > 4" including all cases discharged after the fourth postoperative day.
Covariates We compared selected clinical variables to analyse the presence of a possible association with lenght of stay: data were divided into preoperative, intraoperative and postoperative.
Furthermore, three groups were made according to the number of cases entered by the different centers and considered as a possible influential variable: group A (less than 50 cases), group B (number of cases between 51 and 100) and group C (above 100 patients).
Statistical Analysis Patient baseline demographic and comorbidity variables were analyzed using frequencies and proportions for categorical variables and means, standard deviations (SDs), medians and interquartile ranges (IQR) for continuous variables.
The comparisons between groups (LOS ≤ 4 vs LOS > 4) categories were made by the following tests: Student t-test for continuous variables with approximately normal distribution, non-parametric Wilcoxon Mann Whitney test for other numerical variables and for categorical variables Pearson's chi-square test or Fisher's exact test in the case of low-volume categories were adopted.
To investigate the factors independently associated with the hospital stay outcome we used simple logistic regression models with the variables of interest, and a simplified multiple model with a stepwise backward procedure to obtain a more parsimonious model that contained only the significant factors. Covariates of interest were patients' demographics, co- morbidities, hospital lobectomy volume , operative features, pain relief techniques, type of surcial approach. A significance level of 0.2 was required to enter the model and a level of 0.05 was required to remain in the model. Adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs) were presented. All P-values are two sided with statistical significance evaluated at the 0.05 α level. All tests were performed with Stata 13.0 statistical software (Stata Corp, College Station, TX).
