Wearable Sweat Sensor for High-Concentration Lactate Monitoring in Critically Ill Patients

Over the past decade, the wearable electronics industry has seen exponential growth in demand for continuous health monitoring technologies. From smartwatches to electronic skin, wearable devices have expanded their focus beyond biophysical signals to include biochemical signals such as small-molecule substances and macromolecular proteins in sweat. However, wearable sweat sensing technology still faces significant challenges, including discomfort and skin irritation caused by sweat-inducing materials, effective sweat extraction and collection capability, continuity and reliability of sweat sensing data, sensor calibration, and understanding of sweat analysis methods to discover clinically valuable new biomarkers. Therefore, how to translate wearable sweat monitoring devices into stable, continuous clinical use-matching the needs of health monitoring, disease diagnosis, and precision medicine-remains a research hotspot and challenge.

Lactate is abundant in sweat and is a hallmark product of anaerobic metabolism. Abnormal lactate levels are associated with multiple diseases. As an important marker of tissue perfusion, blood lactate level is closely associated with the prognosis of patients with severe sepsis and septic shock. Clinical guidelines explicitly recommend blood lactate monitoring in septic patients and lactate-guided fluid resuscitation when lactate levels are elevated. Dynamic monitoring of serum lactate changes and calculation of lactate clearance can more accurately reflect dynamic changes in tissue oxygen supply and consumption, and evaluate tissue perfusion and oxygen metabolism. Studies have shown that a lactate clearance rate ≥10% within 6 hours of resuscitation may predict lower mortality in septic patients. Clinical research has also shown that serum lactate concentration can be used to guide clinical treatment decisions, serving as both a therapeutic target and a prognostic indicator. However, most existing wearable sweat sensors have a detection limit of only 0-20 mmol/L, suitable only for exercise monitoring under normal physiological conditions, and cannot meet the clinical need for monitoring hyperlactatemia patients, whose lactate levels can reach up to 100 mmol/L. In addition, current wearable sweat lactate sensors suffer from poor reproducibility and are easily affected by environmental factors and complex sweat composition, posing major obstacles to clinical translation.

Based on this, the investigators have designed an intelligent wearable device for high-concentration sweat lactate monitoring, which includes a lactate detection electrode, a pH detection electrode, a temperature detection electrode, a sweat-stimulating electrode, a microfluidic sweat collection device, a hydrogel patch for inducing sweating, a flexible integrated circuit, and an intelligent display system. This wearable sweat lactate sensor achieves a detection range of up to 100 mmol/L, with a sensitivity of 89 nA/(cm²·mM) and a response time of 150 seconds, and achieves non-exercise-induced sweating through electrical stimulation. In addition, given the sensitivity of the lactate oxidase sensor to temperature and pH, corresponding temperature and pH sensors have been designed for calibration purposes.

This study is a prospective cohort study to be conducted in the surgical intensive care unit (ICU). Critically ill patients with blood lactate concentration >1.5 mmol/L will be enrolled and divided into two groups: (1) a sweat collection group, in which sweat will be collected using a commercial sweat collection device (HELA) for laboratory (ELISA) analysis; and (2) a wearable device monitoring group, in which patients will wear the investigator-designed wireless wearable sweat sensor for real-time monitoring. Clinical data, including medical history, imaging, and laboratory results, will be collected concurrently. The correlation between sweat and blood lactate concentrations, as well as the accuracy and safety of the wearable device, will be evaluated using Pearson correlation analysis.