Serum Lactate and Mortality in Emergency Department Patients with Cancer.

INTRODUCTION: Patients with malignancy represent a particular challenge for the emergency department (ED) given their higher acuity, longer ED length of stay, and higher admission rate. It is unknown if patients with malignancies and hyperlactatemia are at increased risk of mortality. If serum lactic acid could improve detection of at-risk patients with cancer, it would be useful in risk stratification. There is also little evidence that "alarm" values of serum lactate (such as >/=4 mmol/L) are appropriate for the population of patients with cancer. METHODS: This was a continuous retrospective cohort study of approximately two years (2012-2014) at a single, tertiary hospital ED; 5,440 patients had serum lactic acid measurements performed in the ED. Of the 5,440 patients in whom lactate was drawn, 1,837 were cancer patients, and 3,603 were non-cancer patients. Cumulative unadjusted mortality (determined by hospital records and an external death tracking system) was recorded at one day, three days, seven days, and 30 days. We used logistic regression to examine the risk of mortality 30 days after the ED visit after adjusting for confounders. RESULTS: In an unadjusted analysis, we found no statistically significant difference in the mortality of cancer vs. non-cancer patients at one day and three days. Significant differences in mortality were found at seven days (at lactate levels of <2 and 4+) and at 30 days (at all lactate levels) based on cancer status. After adjusting for age, gender, and acuity level, 30-day mortality rates were significantly higher at all levels of lactic acid (<2, 2-4, 4+) for patients with malignancy. CONCLUSION: When compared with non-cancer patients, cancer patients with elevated ED lactic acid levels had an increased risk of mortality at virtually all levels and time intervals we measured, although these differences only reached statistical significance in later time intervals (Day 7 and Day 30). Our results suggest that previous work in which lactate "cutoffs" are used to risk-stratify patients with respect to outcomes may be insufficiently sensitive for patients with cancer. Relatively low serum lactate levels may serve as a marker for serious illness in oncologic patients who present to the ED.