Evaluating the Predictive Value of Lactate in Patients With Cancer Having Septic Shock.
J Intensive Care Med. 2018 Jan 01;:885066618788821
Authors: Nazer LH, Rimawi D, Hawari FI
PURPOSE: Limited studies evaluated the predictive value of serum lactate (LA) in critically ill patients with cancer. The main objective of this study was to evaluate the predictive validity of LA single measurements as well as LA clearance in predicting mortality in patients with cancer having septic shock. The study also aimed to determine the LA measurement over the first 24 hours with the highest predictability for hospital mortality.
MATERIALS AND METHODS: A retrospective cohort study of adult patients with cancer having septic shock and LA measurements during the first 24 hours. Three receiver-operating characteristic (ROC) curves were constructed to evaluate the predictive validity for hospital mortality of LA at baseline, at 6 hours and at 24 hours after identifying septic shock. The ROC with the largest area under the curve was analyzed to determine LA level with the highest predictability for hospital mortality. In addition, the ability of LA normalization (LA <2 mmol/L at 6 hours and at 24 hours) and the degree of LA elimination (>10% and >20% at 24 hours) to predict hospital mortality were evaluated by determining the predictive values for each clearance end point.
RESULTS: The study included 401 patients. LA >2.5 mmol/L at 24 hours showed the largest area under the ROC curve to predict hospital mortality (ROC area: 0.648; 95% confidence interval: 0.585-0.711) with a sensitivity of 58.4% and specificity of 62.8%. The LA normalization, LA clearance >10%, and LA clearance >20% were also predictors of hospital mortality, with the highest sensitivity for LA normalization at 6 hours (74%) and LA normalization at 24 hours (73.4%).
CONCLUSION: In patients with cancer having septic shock, LA >2.5 mmol/L at 24 hours of septic shock had the highest predictability for hospital mortality. The LA normalization and clearance were also predictors of hospital mortality. However, all LA end points were not strong predictors.
PMID: 30037273 [PubMed - as supplied by publisher]