The LAS VEGAS study on epidemiology, ventilator management and outcome in patients receiving intra-operative ventilation

Cohort stratified for types of surgery and airway device

Editor,

Postoperative pulmonary complications (PPCs) occur often, are associated with prolonged hospital stay and mortality after surgery,^1,2 and are preventable through the use of lung–protective intra-operative ventilation.^3–5 Recently, in the European Journal of Anaesthesiology, we reported the primary results of a prospective 1-week cross-sectional study in 29 countries on five continents in 2013, named the Local ASsessment of VEntilatory management during General Anaesthesia for Surgery (LAS VEGAS) study.⁶ In the primary report, we detailed epidemiology, ventilator management and outcome in patients stratified by risk for PPCs based on the Assess Respiratory Risk in Surgical Patients in Catalonia (ARISCAT) risk score.¹ With a clear need for additional randomised clinical trials of intra-operative management, including trials that test ventilation strategies,⁷ comes a need for a better understanding of the characteristics of cohorts stratified by types of surgery, and by the type of airway device used during intra-operative ventilation.

Here, we report the epidemiology, ventilator management and outcome in various patient cohorts. Details of the original study and the current analysis are provided in Supplemental Digital Content 1. As the intention of this communication is not to compare groups but rather to provide rough data, statistical analyses were not performed.

The LAS VEGAS cohort consists of 1769 and 1860 patients undergoing open or laparoscopic abdominal surgery, 1358 and 3072 patients undergoing upper or lower gastrointestinal surgery, 1578 patients undergoing head and neck surgery, 543 patients undergoing plastic surgery, 1042 patients undergoing orthopaedic surgery and 248 patients undergoing vascular surgery. In all, 7788 versus 1548 patients received intra-operative ventilation through either an orotracheal or nasotracheal tube, or a supraglottic device. Patient demographics, intra-operative characteristics and clinical outcome are presented in Supplemental Digital Content Tables 1 to 6, demonstrating some typical characteristics of each group. For instance, patients undergoing open abdominal surgery had a high risk of PPCs and also high American Society of Anesthesiologists’ (ASA) physical status scores; not only a large proportion of patients undergoing open abdominal surgery but also those undergoing orthopaedic or vascular surgery developed intra-operative hypotension or needed vasopressors; and PPCs occurred often in specific groups, especially patients undergoing laparotomy, who also have a long hospital stay and high mortality rate. Supraglottic devices were used mainly during elective surgical procedures and rarely in patients undergoing abdominal surgery, and patients receiving ventilation through a supraglottic device seldom received neuromuscular blocking agents and had a low risk of PPCs.

Ventilator management is detailed in Figs. 1 and 2, and in Supplemental Digital Content Tables 2 and 5. Mode and median absolute tidal volume (V_T), V_T expressed in ml kg⁻¹ of predicted body weight (PBW), positive end-expiratory pressure (PEEP) and respiratory rate were strikingly similar amongst groups according to type of surgery. Laparoscopic surgery patients often received ventilation with high peak and driving pressures. Patients with a supraglottic device frequently received ventilation with a low V_T, low PEEP, peak and driving pressures, and high oxygen fractions in inspired air.

Fig. 1: Ventilation parameters in patient groups by type of surgery. Cumulative frequency distribution (a) of tidal volume relative to predicted body weight (PBW); (b) positive end-expiratory pressure (PEEP); (c) respiratory rate; (d) peak pressure; (e) driving pressure; and (f) oxygen fraction of inspired air (F iO₂). No statistical test performed, that is groups were not compared. Dotted vertical lines represent cut-offs for each ventilation parameter as in the previous report on the LAS VEGAS study.⁶ Dotted horizontal lines represent fractions of patients at each cut-off point.

Fig. 2: Ventilation parameters in patients receiving ventilation through an endotracheal tube versus a supraglottic device. Cumulative frequency distribution (a) of tidal volume relative to predicted body weight (PBW); (b) positive end-expiratory pressure (PEEP); (c) respiratory rate; (d) peak pressure; (e) driving pressure; and (f) oxygen fraction of inspired air (F iO₂). No statistical test performed, that is groups were not compared. Dotted vertical lines represent cut-offs of each ventilation parameter as in the previous report on the LAS VEGAS study.⁶ Dotted horizontal lines represent fractions of patients at each cut-off point in each group.

Thus, although patient characteristics varied amongst the various cohorts, as did the occurrence of PPCs, there was little variation in intra-operative ventilator management, especially with regard to V_T and PEEP. Although most of these findings are not surprising, the data presented here could be helpful in planning future clinical trials. First, by providing the incidence rates of PPCs for various cohorts, researchers can improve the power calculation by using more realistic effect sizes. Second, ventilation data provided here will guide decisions on how ‘control’ groups could best be ventilated in studies of intra-operative ventilation.

We avoided statistical analyses because it was not our purpose to show differences but rather present the data as they are. In addition, we did not evaluate causal relationships, as many of the presumed differences, for example in incidences of PPCs between cohorts, can be explained by patient characteristics and associated risks of PPCs rather than the way in which the lungs were ventilated.

Of note, we do not report epidemiology, intra-operative ventilator management and outcomes for patients undergoing one-lung ventilation or neurosurgery, as these will be presented elsewhere. The LAS VEGAS investigators remain very willing to share data for more specific cohorts at any request.

Acknowledgements relating to this article

Assistance with the letter: in addition to the LAS VEGAS Study Investigators (Supplementary Digital Content: https://links.lww.com/EJA/A119), the following individuals helped to prepare the manuscript: Meta van der Woude, Sabrine Hemmes, Ary Serpa Neto and Marcus Schultz. Ary Serpa Neto performed the data analysis. We would like to thank all national and local coordinators of the LAS VEGAS study, as well as all data collectors in the participating centres, that is the anaesthesiologists, research nurses, nurse anaesthetists and other healthcare workers, and all patients who participated in this study.

Financial support and sponsorship: the European Society of Anaesthesiology endorsed and partly sponsored the original LAS VEGAS study by a research grant through their Clinical Trial Network; the Amsterdam University Medical Centers, location ‘AMC’.

Conflicts of interest: none.