In Reply:—
We appreciate the comments of Drummond 1and Story et al. 2Both letters address issues that clarify the report by Scheingraber et al. 3
First, Drummond 1appropriately calls additional attention to the important study by his colleagues at the Royal Infirmary in Edinburgh. 4Both McFarlane and Lee 4and Scheingraber et al. 3conducted randomized clinical trials comparing 0.9% saline balanced salt solutions. The two studies differ in that McFarlane and Lee 4enrolled patients undergoing “major hepatobiliary or pancreatic surgery” averaging approximately 3.5 h in duration, whereas Scheingraber et al. 3enrolled patients undergoing “lower abdominal gynecologic surgery” averaging approximately 2.25 h. Despite the somewhat shorter duration of surgery, the gynecologic patients randomized to the saline group received a slightly greater total volume 3(71 ± 14 ml/kg during the first 120 min of infusion) than the patients receiving saline in the study by McFarlane et al. 4(14.6 ± 41 ml · kg−1· h−1during an interval of 219 ± 77 min). As a consequence, the increase in plasma chloride and the decrease in plasma bicarbonate were greater in the gynecologic patients. Together, the reports suggest that hyperchloremic acidosis is a dose-dependent consequence of saline administration. Whether this acid–base abnormality is in fact harmful remains unclear, although we 5were unable to cite any compelling evidence of adverse effects. We are skeptical that differences in outcome, if any, related to the choice of saline or balanced salt solution would justify the cost of a randomized clinical trial.
However, one noteworthy characteristic of Plasmalyte 148, the balanced salt solution used by McFarlane and Lee, 4is that the sodium concentration is 140 mEq/l. Consequently, in contrast to lactated Ringer’s solution, infusion of substantial volumes does not decrease serum sodium and serum osmolality, and does not raise the same theoretical concerns about increases in brain water. 5
We agree with Story et al. 2that the Stewart approach 6,7offers interesting insights into acid–base chemistry; however, we disagree that the relative merits of the Stewart approach versus the conventional Henderson-Hasselbalch approach constitute a “central issue.” Regardless of its attractive biochemical features, the Stewart approach has not yet become popular for routine clinical use, perhaps because it is less simple to quantify at the bedside and because it prompts no important differences in treatment.
