Reconsidering lactate as a sepsis risk biomarker

Area of Science:

• Critical Care Medicine
• Biomarker Research
• Sepsis Pathophysiology

Background:

• Lactate is a key biomarker in sepsis and septic shock.
• Traditional methods for assessing biomarker added-value, such as odds ratio (OR) and area under the receiver operator characteristic curve (AUC-ROC) difference, have limitations.

Purpose of the Study:

• To evaluate the added value of lactate and its derivatives in predicting hospital mortality in sepsis.
• To compare different methods for assessing biomarker added-value in a sepsis cohort.

Main Methods:

• A sepsis cohort (n=717) was identified from an ICU database.
• Logistic regression models incorporated admission lactate, lactate clearance, and lactate-time curve (LTC).
• Added-value was assessed using OR, AUC-ROC difference, net reclassification index (NRI), and net benefit (NB).

Main Results:

• The base model had an AUC-ROC of 0.81. Lactate indices did not significantly improve AUC-ROC or NRI.
• Over-time models showed AUC-ROC of 0.86, but lactate clearance and LTC did not significantly improve prediction.
• Net benefit was not increased by any lactate index, suggesting limited added-value with standard metrics.

Conclusions:

• Lactate's added-value assessment in sepsis is model-dependent.
• Incorporating risk reclassification and net benefit measures is crucial for accurate biomarker evaluation.
• Standard metrics may underestimate the true predictive contribution of lactate in sepsis.