Correspondence  |   August 2018
In Reply
Author Notes
  • Erasme University Hospital, Université Libre de Bruxelles, Brussels, Belgium (A.J.). Alexandre.Joosten@erasme.ulb.ac.be
  • (Accepted for publication May 8, 2018.)
    (Accepted for publication May 8, 2018.)×
Article Information
Correspondence
Correspondence   |   August 2018
In Reply
Anesthesiology 8 2018, Vol.129, 386-387. doi:10.1097/ALN.0000000000002292
Anesthesiology 8 2018, Vol.129, 386-387. doi:10.1097/ALN.0000000000002292
We would like to thank Drs. Slagt and van Eijk for their interesting comments regarding our recent publication.1  These authors suggested that the complexity of surgery, resulting in a longer procedure (1 h longer in the crystalloid group), and not the type of fluid was responsible for the higher incidence of postoperative complications in the crystalloid group, due to a higher inflammatory response. We have no data to support more complex surgeries in the crystalloid group beyond the surgical time because surgical procedures and incidences of high-risk surgery were comparable in the two groups. Additionally, blood loss was also not different between the two groups, further supporting similar surgical complexity among the two groups. If the inflammatory response related to the surgical procedure was responsible for the higher fluid balance in the crystalloid group, we might also have expected a significantly higher fluid balance on postoperative day 1 in this group, which was not demonstrated. In order to take into account the difference in surgical duration, we originally presented our results in ml · kg–1 · h–1 and observed a significantly higher fluid administration in the crystalloid group. It is true that we did not directly measure any parameter, which may have indicated a more severe inflammatory response in the crystalloid group than in the colloid group. As a result, we could not completely rule out the hypothesis of Drs. Slagt and van Eijk. To further investigate their hypothesis, we did go back to the data of the 102 patients who underwent a gastrointestinal anastomosis to compare the surgical duration between those who either did or did not have an anastomotic leakage postoperatively. We observed that surgical duration was not different between these two groups (anastomotic leakage, 240 min [204 to 387] vs. no anastomotic leakage, 268 min [185 to 336]; P = 0.850). Interestingly, fluid balance was significantly higher (6.0 ml · kg–1 · h–1 [5.1 to 8.4] vs. 3.1 ml · kg–1 · h–1 [1.7 to 5.0]; P = 0.021) among patients developing an anastomotic leakage. These data confirm that surgeries with postoperative complications had higher intraoperative fluid requirements, unrelated to length of surgery. It should also be noted that there have been experimental studies demonstrating that goal-directed colloid therapy significantly increases microcirculatory blood flow and tissue oxygen tension in perianastomotic colon tissue compared to a goal-directed crystalloid fluid therapy.2  Finally, the surgical duration could still be a result of the groups and not a confounding factor. As a result, although we cannot formally exclude that a prolonged surgical duration might have contributed to our results, we remain confident that our results are mainly related to the type of fluid used to optimize hemodynamic management as part of a closed-loop–assisted goal-directed fluid therapy.