Postoperative Patient-Reported Outcomes After Uniportal Video-Assisted Thoracoscopic Surgery And Prediction For Post-Discharge Recovery Using The Perioperative Symptom Assessment For Lung Surgery Scale

Background:With the continuous advancement of surgical techniques,video-assisted thoracoscopic surgery(VATS)is widely carried out around the world.At present,almost all lung operations can be performed without thoracotomy.VATS has less chest wall damage,avoids severe neuromuscular trauma compared with traditional thoracotomy,can significantly improve recovery speed,and postoperative quality of life of patients,and reduce postoperative complications.Patient-reported outcomes(PROs)can be used for proactive symptom management,assess the effectiveness of different surgical treatment options,and facilitate shared decision-making between physicians and patients.Therefore,the characteristics of postoperative PROs changes in patients with lung surgery and their related influencing factors have also received more and more attention.The aim of this study was to use a novel perioperative-specific scale for lung surgery patients to assess postoperative symptoms and functional status and to identify factors that may influence these symptoms,ultimately constructing predictive models to assess postoperative recovery in individual patients risk.Methods:We conducted a single-center longitudinal observational cohort study using a new scale,the Perioperative Symptom Assessment for Lung Surgery(PSA-Lung)to assess symptom recovery and functional status in patients treated with uniportal VATS.We divided the patients into two groups according to the PRO score at the fourth week after operation,and compared the clinical characteristics under each PRO index.We compared the PRO scores of different surgical resection and lymph node dissection methods in the fourth postoperative week.In addition,we conducted qualitative interviews on cough at 4 weeks postoperatively.Finally,we ranked the importance of each PRO score of patients on POD 1-4 by establishing a random forest machine learning prediction model,and screened out the top 10 variables for multivariate logistic regression analysis.Based on this,we constructed a nomogram to predict the risk of postoperative complications in patients.Results:A total of 202 patients were included in the study.On the POD 1,"pain"symptoms were the most severe PRO item.with a median of 5 points and an average of 5.39 points.Most of the patients recovered well in the fourth week after surgery,and the severity decreased greatly compared with the first day after surgery,with a median of 1 point and an average of 1.73 points.The "cough" symptom was not obvious on the POD1 and POD2,the medians were 2 points and 2 points,and the average points were 2.31 points and 2.23 points,but the scores began to increase on POD2,and It reached its peak in the second week after surgery,with a median of 3 points and an average of 3.44 points,ranking first among all PRO items.At postoperative week 4,the two highest-scoring PRO items were "shortness of breath"(median 2,mean 2.43)and "cough"(median 2,mean 2.34).Preoperative FEV1 and preoperative FEV%values were statistically significant under all PRO items.Preoperative MVV and preoperative MVV%had significant statistical significance under the items of "pain","cough","shortness of breath","fatigue","distress","difficulty walking" and "difficulty in daily activities".Different surgical methods and different lymph node dissection methods had different PRO scores in the fourth postoperative week.The nomogram constructed in the study included 5 postoperative PRO score indicators:"pain" POD1,"difficulty in daily activities" POD1,"pain" POD2,The "cough" scorePOD3 and the "cough" POD4 were used to predict the probability of poor postoperative recovery due to complications after discharge from lung surgery,with an AUC value of 0.83(95%CI,0.75-0.91).Conclusions:Using the PSA-Lung scale,we characterized perioperative symptomatic recovery after Uniportal lung surgery.It was found that "shortness of breath" and"cough" were the most severe symptoms in the fourth week after uniportal VATS,while"pain" was lower scored at the postoperative week 4,indicating that patients recovered quickly from painful symptoms after uniportal VATS.Although "cough" was not the most important symptom in the early postoperative period,its score increased after discharge and peaked in the second postoperative week.The preoperative pulmonary ventilation function indexes FEV1,MVV,FEV1%and MVV%were important factors affecting postoperative symptoms and functional recovery of patients.The lobectomy group had more symptom burden than segmental resection and wedge resection.The systematic lymph node dissection group had more symptom burden than the sampled and no dissection groups.Finally,a random forest machine learning prediction model was built to rank the importance of each PRO score of patients on POD1 to POD4,we developed a nomogram model to predict the risk of post-discharge complications.