Prediction of Postoperative Pulmonary Complications in Thoracic Surgery

The issue of postoperative pulmonary complications following major lung resection (such as lobectomy or segmentectomy) is a central topic in anesthesia and thoracic surgery. Postoperative morbidity and mortality after this type of surgery have drastically decreased in recent years with advances in anesthesia and resuscitation, as well as minimally invasive surgery, but remain high compared to other types of surgery, particularly due to postoperative pneumonia. The etiology of postoperative pneumonia is multifactorial (atelectasis, postoperative ventilation, inadequate analgesia), but the patient's immune system plays a predominant role in each individual case. Therefore, identifying inflammatory biomarkers predictive of postoperative pulmonary complications in a given patient could optimize their management and reduce the risk of postoperative pulmonary cancer (PPC). The objective of this study is to identify preoperative inflammatory biomarkers predictive of PPC after major lung resection. It will use machine learning methods specific to these data to define an immune signature of PPC. This immune signature will be validated using standard analytical techniques to facilitate the clinical translation of a diagnostic test.

• Evaluation of the incidence of pulmonary complications [Срок оценки: 30 days]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and the length of hospital stay [Срок оценки: 3 months]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and the number of reintubations recorded [Срок оценки: 30 days]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and the Number of unplanned hospitalizations in intensive care recorded [Срок оценки: 30 days]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and the Preoperative anxiety score assessed [Срок оценки: 48 hours]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and the Preoperative anxiety score assessed [Срок оценки: 48 hours]
• Evaluation of the correlation between the prognostic score defined using a machine learning method and The cost of care [Срок оценки: 3 months]
• Evaluation of the prognostic performance of the score calculated by the machine learning method on Post-operative Pulmonary Complications (PPC) assessed by the Melbourne composite score (Melbourne Group Scale (MGS) >=4) [Срок оценки: 7 days]
• Evaluation of the prognostic performance of the score calculated by the machine learning method on the severity of postpartum bleeding (PPB) [Срок оценки: 30 days]
• Evaluation of the prognostic performance of the score calculated by the machine learning method on Postoperative mortality assessed at 30 days [Срок оценки: 30 days]
• Evaluation of the prognostic performance of the score calculated by the machine learning method on Postoperative mortality assessed at 90 days [Срок оценки: 90 days]
• Evaluation of the prognostic performance of the score calculated by the machine learning method on Pre- and postoperative pain [Срок оценки: 90 days]