To the Editor:
The recent article in Anesthesiology compared intraoperative fluid management by crystalloids versus colloids.1 The authors analyzed postoperative sequelae 1 yr after surgery of the same group of patients whose early postoperative results were reported in a previous publication.2 Fluid management was based on hemodynamic optimization according to noninvasive cardiac output measurements, and it was provided by a closed-loop system. The study groups were defined by the type of fluids infused by means of a goal-directed fluid therapy strategy that consisted of multiple 100-ml mini-fluid challenges using either a crystalloid solution or a synthetic colloid solution (hydroxyethyl starch).
Eight of the 80 patients in the crystalloid group had anastomotic leakage and five had bleeding that required reoperation in comparison with none of the 80 patients in the colloid group.2 In addition, the operation took longer in the crystalloid group.
The power analysis that the authors performed was based on changes in Post-Operative Morbidity Survey under the hypothesis that the only difference between treatment groups will be the type of fluid management. However, there could hardly be any argument that a patient’s recovery from surgery depends on numerous factors other than fluid management.
A recent review that appeared in Anesthesiology suggests that the effect of various types of fluid management should be evaluated in addition to the complexity of the surgery.3 In the current study, the authors performed the analysis by intention-to-treat, with which we concur. However, when 16% of patients in one group had significant surgery-related events versus 0% of patients in the other group, a different analysis is needed to prevent potentially misleading conclusions of the study findings. Although the authors provided a list of the specific sources of surgical complications that were encountered (i.e., bleeding that required reoperation, anastomotic leakage, peritonitis, and reoperation), the data analysis ascribed all of the postoperative sequelae solely to fluid management. Reoperation attributable to bleeding or anastomotic leakage requires prolonged treatment and could affect kidney function as well as the patients’ overall health condition no less than—and possibly more than—the type of fluids given intraoperatively. Moreover, the issue of group differences in surgical risk was not addressed.3
The second article analyzes patient wellbeing and renal function 1 yr after the surgery, and the authors used the power analysis of the Post-Operative Morbidity Survey on postoperative day 2 from the first study.1 The intention to connect long-term recovery to intraoperative fluid management is understandable; however, the underlying medical condition and the type of surgery that the patient underwent a year earlier must also be taken into account.
The authors declare no competing interests.