Arterial-venous PCO2 gradient in early postoperative hours following myocardial revascularization.

OBJECTIVE: To investigate the utility of the arterial-venous PCO2 gradient (P(a-v)CO2) as a marker of the increased risk of postoperative complications in the early postoperative hours following myocardial revascularization. EXPERIMENTAL DESIGN: Prospective study. SETTING: The Postoperative Intensive Care Unit (ICU) of a University Hospital. PATIENTS: Thirty patients (28 males and 2 females; aged 39-70) that consecutively underwent myocardial revascularization. INTERVENTIONS: None. MEASURES: Thirty minutes following arrival at the ICU the hemodynamic parameters were recorder; the arterial and mixed venous hemogasanalyses were obtained; the mixed venous blood hemoglobin saturation (SvO2) and the O2 consumption (VO2) were calculated; and plasma lactate was determined. The arterial and mixed venous hemogasanalyses were determined again 90 minutes after the admission to the ICU. RESULTS: P(a-v)CO2 at 30 minutes was 8.1+/-2.3 mmHg and was only slightly lower at 90 minutes (7.5+/-2.3 mmHg) so that any significant influence of patient transport to the ICU could be ruled out. P(a-v)CO2 did not significantly relate with cardiac index, mixed venous blood O2 saturation, and blood lactate. Twenty-one patients (70%) showed P(a-v)CO2 values higher than 7 mmHg at 30 minutes: in comparison with the others they were characterized by higher arterial blood PCO2 (PACO2) (37+/-5 vs 32+/-3 mmHg; p<0.05) in spite of similar ventilatory variables, by higher mixed venous blood PCO2 (PVCO2) (47+/-6 vs 37+/-3 mmHg; p<0.01), and by lower cardiac index values (2.0+/-0.3 vs 2.3+/-0.6 1/min/m2; p<0.05). The patients that presented abnormally high P(a-v)CO2 values showed a higher rate of postoperative complications, including inadequate cardiac performance, cardiac arrhythmias, prolonged mechanical ventilation, increased plasma creatinine, and jaundice (11 patients out of 21 vs 1 patient out of 9; p<0.05). Finally P(a-v)CO2 was related with arterial-mixed venous O2 content difference (regarded as an index of O2 consumption), hematocrit, blood temperature and PACO2 by multiple linear regression (R=0.74; p<0.01). The coefficients of all factors but hematocrit were significant; hence, apart from the cardiac index, P(a-v)CO2 was influenced by the metabolic rate, the body temperature (possibly because of CO2 release during rewarming), and the impaired CO2 elimination through the lungs. CONCLUSIONS: P(a-v)CO2 represents a useful even if aspecific parameter to monitor patients during the early postoperative period after myocardial revascularization.