Newly Published
Correspondence  |   July 2020
Impact of Closed-loop Anesthesia on Cognitive Function: Reply
Author Notes
  • Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium and Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, Université Paris-Saclay, Hôpital De Bicêtre, Assistance Publique Hôpitaux de Paris (AP-HP), Le Kremlin-Bicêtre, France (A.J.). joosten-alexandre@hotmail.com
  • Accepted for publication June 19, 2020.
    Accepted for publication June 19, 2020.×
Article Information
Correspondence
Correspondence   |   July 2020
Impact of Closed-loop Anesthesia on Cognitive Function: Reply
Anesthesiology Newly Published on July 14, 2020. doi:https://doi.org/10.1097/ALN.0000000000003469
Anesthesiology Newly Published on July 14, 2020. doi:https://doi.org/10.1097/ALN.0000000000003469
We would like to thank Dr. Sneyd and Prof. Evered for their interesting comments regarding our recent article which assessed the feasibility and potential impact of automated closed-loop anesthesia management on short- and mid-term cognitive function after noncardiac surgery.1,2 
We agree with these authors that an alternative interpretation of our results could be that automated anesthetic management using the combination of three controllers had no impact on delayed neurocognitive recovery. In correspondence with the Journal’s Statistical Editor, the most valid treatment estimate for change in cognition score (the 30-item Montreal Cognitive Assessment) from baseline would be an analysis of covariance, using preoperative cognition score as a covariate and group assignment as a fixed effect. In line with Journal policy, this analysis was labeled as a post hoc sensitivity analysis, as it was requested after examining the data. When analyzed in this manner, there was no statistical difference between the two groups at 1 week (point of estimate 0.7 with 95% CI, −0.2 to 1.6; P = 0.14) and 3 months (point of estimate 1.1 with 95% CI, 0.0 to 2.2; P = 0.056) postsurgery follow-up. The difference between this analysis and the originally planned analysis is that the groups were somewhat imbalanced at baseline, and this imbalance could account for approximately 30% of the treatment effect when analyzed as a change score. As requested by the letter authors, we have added herein the median cognition scores in both groups at their baseline, short-term (1 week) follow-up, and mid-term (3 months) follow-up (table 1), and we have added figure 1 that describes changes from baseline cognition score for the two groups. It is worth noting that the point estimate from the analysis of covariance approach is similar in magnitude to the original approach, but with increased imprecision (i.e., wider CIs).