Central venous-to-arterial carbon dioxide partial pressure difference in early resuscitation from septic shock

BACKGROUND: Central venous-to-arterial carbon dioxide partial pressure difference (ΔPCO2) can be used as a marker for the efficacy of venous blood in removing the total CO2 produced by the tissues. To date, this role of ΔPCO2 has been assessed only in patients after resuscitation from septic shock with already normalised central venous oxygen saturation (ScvO2 ≥70%). There are no reports on the behaviour of ΔPCO2 and its relationship to cardiac index (CI) and clinical outcome before normal ScvO2 has been achieved. OBJECTIVES: To investigate the behaviour of ΔPCO2 and its relationship to CI, blood lactate concentration and 28-day mortality during resuscitation in the very early phase of septic shock. To examine whether patients who normalise both ΔPCO2 and ScvO2 during the first 6 h of resuscitation will have a greater percentage decrease in blood lactate concentration than those who only achieve normal ScvO2. DESIGN: Prospective observational study. SETTING: Intensive Care Unit (ICU) in a university hospital. PATIENTS: Eighty patients with septic shock were consecutively recruited. INTERVENTIONS: Patients were resuscitated in accordance with the recommendations of the Surviving Sepsis Campaign. MAIN OUTCOME MEASURES: Blood lactate concentrations, and haemodynamic and oxygen-derived variables were obtained at ICU admission (T0) and 6 h after admission (T6). Lactate decrease was defined as the percentage decrease in lactate concentration from T0 to T6. All cause 28-day mortality was also recorded. RESULTS: Data are presented as median (interquartile range). At T0, there were significant differences (P < 0.0001) between normal (ΔPCO2 ≤0.8 kPa) and high ΔPCO2 groups for CI (3.9 3.3 to 4.7 vs. 2.9 2.3 to 3.1 l min m) and ScvO2 (73 65 to 80 vs. 61 53 to 63%). The correlation between changes in CI and ΔPCO2 was r = -0.62, P < 0.0001. Patients who reached a normal ΔPCO2 at T6 had larger decreases in blood lactate concentration and Sequential Organ Failure Assessment scores on day 1. The lactate decrease was greatest in the subgroup achieving both normal ScvO2 and ΔPCO2 at T6. Lactate decrease, unlike ΔPCO2 and ScvO2, was an independent predictor of 28-day mortality. CONCLUSION: Monitoring ΔPCO2 may be a useful tool to assess the adequacy of tissue perfusion during resuscitation. The normalisation of both ΔPCO2 and ScvO2 is associated with a greater decrease in blood lactate concentration than ScvO2 alone. The lactate decrease is an independent predictor of 28-day mortality. Further research is needed to confirm this hypothesis.