BACKGROUND AND GOAL OF STUDY: 
The goal of our study was compare predictive model to detect high risk patients to develop postoperatory pulmonary complications (PPC) after lung resection surgery (LRS). 

MATERIALS AND METHODS: 
Substudy RCT (NCT 02168751). We retrospectively  classified patients in two group: PPC and Non-PPC. PPCs were defined as in ARISCAT study. We did three logistic regression models: Model 1: Only clinical (pre and intraoperative) variables. Model 2: Intra and early postoperative biomarkers and Model 3: Combined model 1 and 2. Comparisons between models were based on the receiver operating characteristics (ROC) and tests of integrated discrimination improvement. Significance was set at p< 0.05.

RESULTS AND DISCUSSION:
 37/174 (21.3%) patients had PPCs. Patients PPC group had poor pulmonary function before surgery, they were older and the duration of surgery was longer. Lung and systemic inflammatory response was higher in patients of PPCs group. Onset of the complication was a mean (SD) 3.32 (1.76) days after surgery. No PPCs were detected on day 0 (day the patient underwent surgery).  The area under the ROC curve of the model 1, 2, and 3 for prediction of PPC was 0.79 (95%CI, 0.71-0.87) 0.80 (95%CI, 0.72-0.88), and 0.93 (95%CI, 0.88-0.97) respectively. Model 3 compare with model 1 had an integrated discrimination improvement of 0.29 (p< 0.001) and a net reclassification index of 0.28 (p=0.007).

CONCLUSION(S): 
Combining systemic and pulmonary biomarkers increases the ability of models based only on clinical data to identify PPCs early and could prove useful when deciding which hospital resources will be used after LRS.

Usefulness of combining clinical and biochemical parameters for prediction of postoperative pulmonary complications after lung resection surgery