Combined gas exchange and pulse wave monitoring for detecting anaerobic metabolism in critically ill patients

Introduction

The anaerobic metabolism causes lactate to increase. This is accompanied by a reduction in bicarbonate concentration in the blood, causing CO₂ production (VCO₂) to accelerate, evidenced as an increased respiratory CO₂ output. Over the physiological range of CO₂ content the CO₂ tension is linearly related to CO₂ content, and the venoarterial CO₂ tension difference (DPCO₂) could be used as a surrogate for the difference between mixed venous and arterial CO₂ contents. Under anaerobic metabolism the respiratory quotient increases. The venoarterial CO₂ tension difference/arteriovenous O₂ content difference ratio [DPCO₂/C(a-v)O₂] therefore increases. We tested the DPCO₂/C(a-v)O₂ ratio and mixed venous oxygen saturation (SvO₂) values as potential predictors of anaerobic metabolism.

Methods

Fifty-one consecutive adult patients (head injury, subarachnoid haemorrhage, sepsis) admitted to our ICU were prospectively studied. The DPCO₂/C(a-v)O₂ ratio, SvO₂, oxygen delivery index (DO₂I), and arterial lactate (Lac) values were collected at T1 (ICU admission) and at T2 (after 24 hours). The DO₂I was calculated using the cardiac index (CI) measured by the Pressure Recording Analytical Method (PRAM). PRAM is a pulse contour system for beat-to-beat quantification of the CI. It does not need calibrating factors, and allows the stroke volume (SV) to be calculated avoiding the inaccuracies derived from instant variations of arterial impedance. Radial (80%) or femoral (20%) arteries were used for the blood pressure analysis. The presence of anaerobic metabolism (e.g. hyperlactatemia, Lac+) was defined by an increase in Lac >2 mmol/l. Linear correlations and the ROC test were applied.

Results

For a threshold value of DO₂I > 330 ml/min/m², an inverse relationship (R² = 0.69; P < 0.05) between Lac and SvO₂ at T1 and at T2 (R² = 0.88; P < 0.05) was found. For a DO₂I < 330 ml/min/m², good direct correlations between Lac and the DPCO₂/Da-vO₂ ratio calculated at T1 (R² = 0.89; P < 0.05), were found. The SvO₂ did not show any significant relationship with Lac for a DO₂I < 330 ml/min/m². ROC curves to predict Lac+ were constructed. The areas under the ROC curves (AUC) were 0.78 and 0.47 for the DPCO₂/Da-vO₂ ratio and SvO₂, respectively. The AUC for the DPCO₂/Da-vO₂ ratio was significantly greater that that for SvO₂ (P < 0.05). From the ROC curve an optimal cutoff value of 1.5 (sensitivity = 0.78, specificity = 0.75) was determined for the DPCO₂/Da-vO₂ ratio predicting the presence of Lac+.

Conclusion

Pulse wave analysis plays a key role in the continuous monitoring of critically ill patients. It provides beat-to-beat values of the SV, and consequently allows the DO₂I to be frequently calculated. The DPCO₂/Da-vO₂ ratio is simple and quick to calculate and would be a valuable approach in clinical practice. Our findings showed that DPCO₂/Da-vO₂ ratio values are directly related to Lac. For DO₂I < 330 ml/min/m² the DPCO₂/Da-vO₂ ratio seems a more reliable predictor of anaerobic metabolism than SvO₂. Combined gas exchange and pulse wave monitoring might be a very useful approach to detect anaerobic metabolism in ICU patients.