The Impact of Serum Glucose on the Predictive Value of Serum Lactate for Hospital Mortality in Critically Ill Surgical Patients.

backgroundLactate has been widely used as a risk indicator of outcomes in critically ill patients due to its ready measurement and good predictive ability. However, the interconnections between lactate metabolism and glucose metabolism have not been sufficiently explored, yet. In this study, we aimed to investigate whether glucose levels could influence the predictive ability of lactate and design a more comprehensive strategy to assess the in-hospital mortality of critically ill patients.

methodsWe analyzed the clinical data of 293 critically ill patients. The primary outcome was in-hospital mortality. The logistic regression analysis and the area under the receiver operating characteristic curve (AUROC) were applied to evaluate the predictive ability of lactate in association with glucose.

resultsThe lactate level showed significant association with in-hospital mortality, and its predictive ability was also comparable to other prognostic scores such as the SOFA score and APACHE II score. We further divided 293 patients into three groups based on glucose levels: low-glucose group (<7 mmol/L), medium-glucose group (7-9 mmol/L), and high-glucose group (>9 mmol/L). The lactate level was associated with in-hospital mortality in the low- and high- glucose groups, but not in the medium-glucose group, whereas the SOFA score and APACHE II score were associated with in-hospital mortality in all three glucose groups. The AUROC of lactate in the medium-glucose group was also the lowest among the three glucose groups, indicating a decrease in its predictive ability.

conclusionsOur findings demonstrated that the predictive ability of lactate to assess in-hospital mortality could be influenced by glucose levels. In the medium glucose level (i.e., 7-9 mmol/L), lactate was inadequate to predict in-hospital mortality and the SOFA score; the APACHE II score should be utilized as a complementation in order to obtain a more accurate prediction.