• Matthew J Martin, Elizabeth FitzSullivan, Ali Salim, Carlos V R Brown, Demetrios Demetriades, and William Long.
• Division of Trauma and Surgical Critical Care, Keck School of Medicine, University of Southern California and the Los Angeles County and University of Southern California Medical Center, Los Angeles, CA, USA. docmartin2@yahoo.com
• Am. J. Surg. 2006 May 1;191(5):625-30.

PurposeBoth lactate and base deficit (BD) are used as predictors of injury severity and mortality. We examined the significance of these measures when used in combination, and particularly when they provide conflicting data.MethodsWe reviewed all intensive care unit patients with simultaneously obtained lactate and BD measurements. The ability to predict mortality and hospital stay was compared alone, in combination, and when there was disagreement between the measures. Receiver operating characteristic curves were generated to compare predictive abilities.ResultsThere were 1,298 patients with 12,197 sets of paired laboratory data; 1,026 trauma patients and 272 surgical patients. Lactic acidosis was present in 41% and a significant BD level (> 2) was found in 52%. Nonsurvivors had higher admission lactate (6.2 vs. 3.3) and base deficit (6.1 vs. 3.2) levels than survivors (both P < .01), with a modest correlation (r = .52) between the measures. The admission lactate and BD levels had similar predictive ability for mortality, with areas under the receiver operating characteristic curve of .7 and .66, respectively (both P < .01). However, the predictive ability of the BD level decreased significantly during the intensive care unit stay (area, .5) compared with lactate level (area, .68). Lactate and BD levels disagreed in 44% of all laboratory sets. In patients with a normal lactate level (< 2.2), the BD level had no predictive ability for mortality (area, .48; P = .26). However, in patients with a normal BD level (< 2.0), the lactate level retained its predictive ability for mortality (area, .67; P < .01). Lengths of stay were longer among patients with an increased lactate level, even when the BD level was normal. There was no improvement in predictive ability using a combination of the 2 measures.ConclusionsBoth lactate and BD levels may be used to identify lactic acidosis and predict mortality at admission. Increased lactate levels predict mortality and a prolonged course regardless of the associated BD level, whereas an increased BD level has no predictive value if the lactate level is normal.

27 keywords...

hide…