Non-invasive estimation of respiratory depression profiles during robot-assisted laparoscopic surgery using a model-based approach

Introduction: Robot assisted laparoscopic surgeries are becoming the standard procedure for radical prostatectomies (RALRP). General anesthesia, Trendelenburg positioning and capnoperitoneum during RALRP affect patient’ gas exchange, leading to possible complications in the postoperative phase, such as hypoxemia. The aim of this paper is to examine the changes in pulmonary gas exchange through the perioperative period for RALRP using a mathematical model approach.Methods: Measurements were performed with the Automatic Lung Parameter Estimator (ALPE) system, which include a mathematical model of pulmonary gas exchange capable of quantifying shunt and ventilation to perfusion (V̇A/Q̇) mismatch. In total, 20patients (ASA physical status I-III) with a mean age of 63.8 ± 6.6 years scheduled for elective RALRP at Aalborg University Hospital, where included in this study. Local procedures for anesthesia, ventilator settings and operation were followed throughout the study. Intraoperative measurements were performed before (T1) and during 30◦ Trendelenburg position and capnoperitoneum (T2-T3), as well as after exsufflation when the patients were returned to the supine position (T4).Results: Patients with ASA-score >1 had significantly higher shunt during and after surgery (T2-T4) compared to T1 (P<0.001).In the ASA=1 group there was no statistically difference between the levels. Moreover, the level of shunt at the end-point of surgery (T4) was significantly higher in the ASA>1 group compared to ASA=1 (P=0.02).At T1 there was no statistically differences in shunt between the groups. The level of V̇A/Q̇ mismatch did not increase significantly in the two groups, although when analyzed together in one group, there was a significant increase from T1 to T3 and T4(P=0.002).There was no differences between the level of V̇A/Q̇ mismatch between the groups at any timepoint.Discussion: In this a mathematical model approach was used to describe the perioperative development of shunt and V̇A/Q̇ mismatch for RALRP patients. The results showed an increase in both shunt andV̇A/Q̇-mismatch throughout the intraoperative period, with different patterns of development of shunt with the ASA score. This concurswith previous findings of oxygenation during anesthesia. This study provides an indication for the use of intraoperativeinterventions, such as increased PEEP and/or lung recruitment for patients with intraoperative V̇A/Q̇-mismatch and shunt, guided by a model-based quantification of the problems.