Comparative Evaluation of Central Venous Oxygen Saturation, Carbon Dioxide Venous Arterial Gradient, and Lactate Levels as Markers of Tissue Perfusion After Cardiac Surgery: A Prospective Exploratory Observational Study

Technical Summary: Comparative Evaluation of Perfusion Markers After Cardiac Surgery

Problem Statement
Following cardiac surgery, patients frequently experience profound metabolic disturbances and tissue hypoperfusion, particularly within the first 24 hours of Intensive Care Unit (ICU) admission. Prompt recognition of circulatory inadequacy is critical for guiding therapeutic interventions such as vasoactive support or fluid resuscitation. While various markers are routinely used to assess tissue perfusion—including serum lactate, central venous oxygen saturation (ScvO₂), and the venous–arterial carbon dioxide gradient (CO₂ gap)—there is ongoing debate regarding their relative sensitivity and specificity in predicting early postoperative recovery. Specifically, it remains unclear which of these biomarkers best discriminates patients who can be safely discharged from the surgical ICU within 48 hours versus those requiring prolonged care.

Methodology
This study is a prospective, exploratory, observational cohort study conducted at the Heart Institute of the Hospital das Clínicas, FMUSP, Brazil. The study population consisted of 100 consecutive patients who underwent cardiac surgery requiring cardiopulmonary bypass (CPB) across two collection periods (February–September 2022 and February–June 2023). No formal sample size calculation was performed; the study is explicitly designated as hypothesis-generating.

Data were collected at two time points: ICU admission and 24 hours post-admission. The primary variables evaluated were:

• ScvO₂: Central venous oxygen saturation.
• Lactate: Serum lactate levels.
• CO₂ Gap: The venous–arterial carbon dioxide gradient (calculated from venous and arterial pCO₂).
• Clinical Variables: Demographics, comorbidities (e.g., hypertension, diabetes, heart failure), and pharmacological support (vasoactive/inotropic drugs).

The primary outcome was ICU discharge within 48 hours of admission. Patients were stratified into "discharged" and "not discharged" groups. Statistical analysis included descriptive statistics, paired t-tests for temporal changes, Pearson's correlation, and logistic regression (univariate and multivariate) to assess associations with the primary outcome. The discriminatory performance of the markers was evaluated using Receiver Operating Characteristic (ROC) curves, with comparisons between Area Under the Curve (AUC) values performed using DeLong's test.

Key Results

• Temporal Trends: Over the first 24 hours, the population mean ScvO₂ significantly decreased (from 76.39% to 73.92%), and lactate levels significantly declined (from 22.86 mg/dL to 18.59 mg/dL). The CO₂ gap remained relatively stable with no significant change.
• Association with Discharge: In the exploratory multivariate analysis, ScvO₂ measured at 24 hours was the only perfusion marker showing a statistically significant association with early ICU discharge (Odds Ratio = 1.096; 95% CI: 1.020–1.180; p = 0.012).
• Discriminatory Performance:
  • ScvO₂ at 24 hours demonstrated an AUC of 0.661 (95% CI: 0.551–0.772; p = 0.007) for predicting discharge.
  • Formal DeLong's testing confirmed that ScvO₂ at 24 hours had significantly superior discriminatory performance compared to lactate at 24 hours (AUC 0.661 vs. 0.428; Z = 2.889; p = 0.004).
  • The comparison between ScvO₂ and the CO₂ gap showed a non-significant trend (AUC 0.661 vs. 0.520; p = 0.053).
  • ScvO₂ at admission, lactate, and the CO₂ gap (at both time points) did not demonstrate significant associations with the outcome in the multivariate model.
• Clinical Correlates: A history of systemic arterial hypertension was significantly associated with a lower likelihood of discharge (p = 0.032).

Significance and Claims
The authors conclude that in this specific exploratory cohort, ScvO₂ measured 24 hours after cardiac surgery is the most robust biomarker among those evaluated for predicting early ICU discharge, outperforming both serum lactate and the CO₂ gradient. The study posits that while the population-level ScvO₂ declines post-CPB due to systemic inflammatory response and rewarming, the absolute value maintained at 24 hours reflects the preservation of the oxygen delivery-to-consumption ratio and hemodynamic recovery.

However, the paper maintains a highly modest tone regarding its clinical implications. The authors explicitly state that:

1. The findings are hypothesis-generating and require validation in larger, adequately powered prospective studies.
2. The modest AUC (0.661) indicates limited discriminatory power, meaning the marker should not be used for individual clinical decision-making at this stage.
3. Limitations such as the single-center design, lack of sample size calculation, use of ScvO₂ (rather than true mixed venous SvO₂), and the use of a "soft" endpoint (ICU discharge rather than hard outcomes like mortality) preclude definitive clinical recommendations.

Ultimately, the study suggests that while ScvO₂ at 24 hours shows promise as a discriminator for recovery trajectories, its utility must be confirmed through rigorous, multicenter trials incorporating comprehensive covariate adjustments and hard clinical endpoints.