Response to: ‘Postoperative pulmonary complications, pulmonary and systemic inflammatory responses after lung resection surgery with prolonged one-lung ventilation. Randomised controlled trial comparing intravenous and inhalational anaesthesia’

Abstract

Editor—We read with great interest the findings presented in the paper by de la Gala and colleagues comparing sevoflurane vs propofol in patients undergoing lung resection surgery.1de la Gala F. Pinerio P. Reyes A. et al.Post-operative pulmonary complications, pulmonary and systemic inflammatory response after lung resection surgery with prolonged one-lung ventilation. Randomised controlled trial comparing intravenous and inhalational anaesthesia.Br J Anaesth. 2017; 119: 655-663Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar The authors have conducted an apparently methodologically robust randomised controlled trial (RCT) comparing the role of sevoflurane and propofol on the pulmonary and systemic inflammatory responses after prolonged one-lung ventilation. Interestingly, they present as their main finding the observation that sevoflurane markedly reduces the incidence of postoperative pulmonary complications (PPCs) in comparison with propofol (14% vs 28%) and as a consequence results in a significant reduction in 1 year mortality. We are concerned, however, by some apparent inconsistencies in the reporting of their study. Firstly, the author's trial (NCT 02168751),2National Library of Medicine (US) Identifier NCT 02168751, Inflammatory response secondary using intravenous anesthesia versus inhalation anesthesia with halogenated agents.2014 Jun 20https://clinicaltrials.gov/ct2/show/NCT02168751?term=02168751&rank=1Google Scholar registered on https://clinicaltrials.gov (available for public review), states clearly the primary outcome as ‘a change in inflammation markers in plasma and bronchoalveolar lavage’. Surprisingly, this is not the primary outcome as defined in the published article. In fact, inflammatory cytokine concentrations were clearly defined in the published manuscript as secondary outcomes. We therefore ask the authors at what time point in the conduct of the study did they change their primary outcome and what was the rationale behind the change? A recent article by Jones and colleagues, comparing registered and reported outcomes in RCTs in the ‘top six’ anaesthesia journals, worryingly demonstrates that 92% of trials registered in 2015 had at least one primary or secondary outcome discrepancy, with more than half of these favouring statistical significance.3Jones P. Chow J. Arango M. et al.Comparison of registered and reported outcomes in randomised controlled trials published in anesthesiology journals.Anesth Analg. 2017; 125: 1292-1300Crossref PubMed Scopus (39) Google Scholar We are concerned that the article by de la Gala and colleagues continues this trend. Secondly, according to the published article, the sample size was calculated based on a predicted reduction in the incidence of PPCs from 40% to 20% in the sevoflurane group, rather the initially documented ‘inflammatory’ primary outcome. Studies comparing sevoflurane vs propofol have demonstrated a variable effect on postoperative complications and mortality with some studies showing no evidence of a statistically significant change in a variety of anaesthetic subspecialities.4Beck-Schimmer B. Bonvini J. Braun J. et al.Which anesthesia regimen is best to reduce morbidity and mortality in lung surgery? A multicenter randomized controlled trial.Anesthesiology. 2016; 125: 313-321Crossref PubMed Scopus (48) Google Scholar, 5Nieuwenhuijs-Moeke1 G.J. Nieuwenhuijs V.B. Seelen M.A.J. et al.Propofol-based anaesthesia versus sevoflurane-based anaesthesia for living donor kidney transplantation: results of the VAPOR-1 randomized controlled trial.Br J Anaesth. 2017; 118: 720-732Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar, 6Likhvantsev V.V. Landoni G. Levikov D.I. Grebenchikov O.A. Skripkin Y.V. Cherpakov R.A. Sevoflurane versus total intravenous anesthesia for isolated coronary artery bypass surgery with cardiopulmonary bypass: a randomized trial.J Cardiothorac Vasc Anesth. 2016; 30: 1221-1227Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar, 7De Conno E. Steurer M.P. Wittlinger M. et al.Anesthetic-induced improvement of the inflammatory response to one-lung ventilation.Anesthesiology. 2009; 110: 1316-1326Crossref PubMed Scopus (210) Google Scholar We suggest to the authors therefore that the predicted effect size used in the power calculation (reflecting an absolute risk reduction of 50%) is larger than one would expect given the results of other studies performed and therefore ask the authors the rationale for estimating an effect size of this magnitude? Thirdly, the study concludes with the statement that ‘the end result’ of using sevoflurane over propofol for maintenance of anaesthesia in lung resection surgery ‘is reduced mortality’. We argue that despite an increase in PPCs in the propofol group, there was no associated increase in critical care stay, hospital stay, or 30 day mortality and therefore suggest that the assumption that increased mortality at 1 year as a direct consequence of PPCs is a stretch too far. When interpreting the results of clinical trials the clinician is forced to weigh up whether the study's findings are adequate to warrant a change in clinical practice. We suggest given the inconsistencies demonstrated in this study and the inherent fragility of the result (fragility index of 2, calculated using ClinCalc.com) the answer is ‘not yet’. None declared.