We thank the editor for giving us the opportunity to respond to the letter by Lubarsky et al. , and appreciate their critical appraisal of our article.1
Lubarsky et al. conclude that our study did not bring hard evidence that goal-directed colloid fluid therapy is the best method of managing major abdominal surgery. We did not mean to indicate that our study would bring such hard evidence. Rather, as indicated in our introduction, the purpose of our study was to “study if goal-directed fluid therapy with colloids increases perianastomotic tissue oxygen tension and perfusion in comparison to a goal-directed crystalloid and a restricted crystalloid fluid therapy.”1Our conclusion states: “Goal-directed colloid fluid therapy significantly increased microcirculatory blood flow and tissue oxygen tension in healthy and injured colon compared to crystalloids.”1We thus feel that Lubarsky et al. considerably overinterpreted our data. Our study’s aim was to investigate physiologic mechanisms that may explain some of the benefits of the already demonstrated superiority of goal-directed colloid therapy in a multitude of well-conducted clinical studies2–4and in a recent metaanalysis.5
Lubarsky et al. were concerned that our animals were hypovolemic. During preparation and before randomization, all animals received 3 ml · kg–1· h–1of Ringer’s lactate, reflecting a typical restrictive fluid therapy used in clinical studies.6Lubarsky et al. also note that fluid therapy with 250 ml of colloids is not equivalent to 250 ml of crystalloids. We agree that a 250 ml bolus of crystalloids every 30 min may appear conservative if we were treating severely hypovolemic or septic subjects. However, at this stage of the experiments, after completing surgery and instrumentation, the animals were hemodynamically stable. They had minimal blood and fluid loss (the abdominal wound was closed to limit fluid evaporation from the wound) and good diuresis. Our aim was to mimic clinical conditions and treatments, and we therefore administered 250 ml of crystalloids when mixed venous oxygen saturation was below 60%, which is comparable to approximately 600–700 ml in adult patients. At our institution a typical intravenous fluid bolus in clinical anesthesia is 500 ml of Ringer’s lactate or 500 ml of colloids. If we had chosen a protocol as suggested by Lubarsky et al. , the crystalloid goal-directed therapy (GDT) group would have received a 750-ml bolus, which would be comparable to a ≈1,700-ml bolus in humans. Such large fluid therapy could have led to fluid overload in these animals and would have been considered clinically not applicable. We have shown in an earlier study in a similar pig model as used in the present study7that even larger amounts of crystalloids administered (20 ml · kg–1· h–1) than in the present crystalloid GDT group did not affect tissue oxygen tension in the colon.
Concerning blood volume in the two GDT groups, we believe that it was similar in the two groups as judged by the hemoglobin values. The hemoglobin values were comparable in the two GDT groups before and after the experiment and significantly lower than in the fluid restricted group at the end of the study, suggesting similar-grade hemodilution. In addition, data on pulse pressure variation and stroke volume (measured with PICCO; Pulsion Medical Systems GmbH, Munich, Germany) that were not presented in this paper support this opinion, since pulse pressure variation and stroke volume were virtually identical in the two GDT groups at the end of the study. In the fluid-restricted group, pulse pressure variation remained high and stroke volume low throughout the experiments.
Concerning the choice of mixed venous oxygen saturation as a main goal during fluid therapy, we agree there are other, more clinically practical methods available for human studies, and we certainly do not suggest that clinicians should insert pulmonary artery catheters in patients undergoing routine colon surgery. However, for the purpose of this study we considered this method reliable, as the parameter has been shown to be independently associated with clinical outcome.8In our pilot studies measuring mixed venous oxygen saturation resulted in minimal variability and reproducible results. We concede that the target of 60% for mixed venous oxygen saturation seems rather low in patients, but it is ambitious in pigs, as they have a distinctly lower hemoglobin concentration, a species-specific higher hemoglobin oxygen affinity, and an increased body temperature as compared with humans.9
Finally, we disagree with the statement by Lubarsky et al. that “no threshold tissue oxygen tension with anastomotic breakdown is established.” We know at least of two well-designed studies10,11that deal with this very question and that have been referenced in our publication. In these studies, gut tissue oxygen tension is directly correlated to anastomotic breakdown, and a critical value of 20–25 mmHg was established. This critical value was also used in our study to standardize anastomotic conditions.
We are convinced that neither the editorial by Kehlet and Bundgaard-Nielsen12nor our original article are a disservice done to the anesthesia community, and conclude paraphrasing the words of the great scientist John Tukey “An approximate answer to the right problem is worth a good deal more than an exact answer to an approximate problem .”
*Medical University of Vienna, Vienna, Austria. firstname.lastname@example.org