Cause of metabolic acidosis in prolonged surgery

Objective: The intraoperative development of metabolic acidosis is frequent ly attributed to hypovolemia, tissue hypoperfusion, and lactic acidosis. In this study, dilutional acidosis was evaluated as a possible mechanism for the routine development of intraoperative acidosis in noncardiac, nonvascul ar surgery patients. Design: Prospective, observational study. Setting: University-affiliated Veteran's Affairs Medical Center and a staff model, health maintenance organization hospital. Patients: Twelve patients undergoing prolonged surgical procedures expected to last greater than or equal to 4 hrs were enrolled in the study. Interventions: Perioperative management was based on the judgment of the at tending anesthesiologist and surgeon without knowledge of the study's inten t. Measurements and Main Results: Arterial blood gas parameters, serum electro lytes, and urine electrolytes were measured pre- and postoperatively. Pulmo nary artery catheters were placed for hemodynamic measurement and oxygen de livery calculations. Plasma volume was measured both pre- and postoperative ly, using the Evans blue dye dilution technique. Although significant changes in lactate level (1.1 +/- 0.6-1.8 +/- 1.0) occ urred, the change was not large enough to explain the degree of change in b ase excess (0.8 +/- 2.3 to -2.1 +/- 2.9). Chloride levels significantly inc reased (106 +/- 3-110 +/-: 5) with a correlation (r(2) = .92; p < .0001) be tween the degree of change in chloride and the degree of change in base exc ess. Plasma volume did not change. Furthermore, a correlation between the v olume of normal saline administered and the change in base excess was found (r(2) = .86; p < .0001), although no correlation was found with Ringer's l actate solution. An even stranger correlation was noted when the total chlo ride amount administered was compared with the change in base excess (r(2) = .93; p < .0001). Conclusions: In this patient population, a common source of increasing base deficit is related to chloride administration. The largest source of chlor ide is usually normal saline. Classically, dilutional acidosis would explai n the predominance of this acidotic change; however, no increase in plasma volume occurred. The absence of plasma volume change would suggest that the mechanism postulated to result in dilutional acidosis is incomplete. The c ommon treatment of administering more fluid for intraoperative acidosis may be inappropriate, may have caused the acidosis, and may further exacerbate the acidosis. Chloride levels should he assessed whenever a metabolic acid osis is seen perioperatively.