Group assignment
After institutional approval by the local ethics committee and preoperative written informed consent, 57 patients were considered eligible for this randomized, controlled clinical trial from February 2004 until November 2004. Patients were allocated to the trial groups according to a computer-generated random list. One patient had to be excluded from analysis as the donated autologous blood showed multiple clots and could not be retransfused. Informed consent was withdrawn by one patient and one patient had to be excluded due to the preoperative decision for a combined surgical procedure. In total, 54 patients (28 in the 25% Hct group and 26 in the 20% Hct group) remained for statistical analysis according to the Full Analysis Set (Intention To Treat).
Inclusion criteria were age >18 and <75 years, elective coronary artery bypass graft surgery, weight >70 kg and preoperative Hct >36% (hemoglobin >12 g/dl).
Exclusion criteria were withdrawal of consent, Jehova's Witnesses, stroke in patient's history or with persistent neurological residue, unilateral occlusion of carotid artery >70% or bilateral occlusion of carotid artery >50%, combined cardiac procedure, left ventricular ejection fraction <40%, unstable angina, a left main stem stenosis >70%, ventricular arrhythmia >LOWN IVa, symptomatic chronic pulmonary disease requiring long-term medication or FEV1 <70% or FEV1/VC max <70% or partial pressure of oxygen in arterial blood (paO2) <60 mmHg, known acute or chronic hepatitis or hepatic disease with impaired synthesis of coagulation factors or bilirubin >2.0 mg/dl, known inflammatory bowel disease, known renal insufficiency or anuric renal failure or creatinine >1.5 mg/dl, ingestion of aspirin or clopidogrel until 3 days prior to surgery, and treatment with glycoprotein-receptor antagonists within 2 days before surgery. Furthermore, patients were excluded from the study if the Hct could not maintained within the targeted range during CPB, if an emergency situation (for example, cardiopulmonary resuscitation, acute right or left ventricular failure) occurred before initiation of CPB and if the autologous blood could not be retransfused.
Anesthetic and CPB technique, isovolemic hemodilution and management in the intensive care unit
The standard anesthetic practice was an opioid-based anesthetic supplemented with midazolam and isoflurane as required. In all patients, a femoral artery was cannulated with a 4-Fr.-cannula (Pulsiocath, Pulsion, Munich, Germany) prior to induction of anesthesia. A central venous catheter and a pulmonary artery catheter (Thermodilution Catheter, Arrow, Reading, PA, USA) were inserted via the right internal jugular vein.
The standardized CPB priming consisted of 600 ml of crystalloid fluid, 500 ml of 10% hydroxyethylstarch solution and a total dose of 50,000 KIU aprotinin per kg bodyweight prior to and during CPB. Pump flow was adjusted to maintain a mean arterial pressure (MAP) of 55 to 60 mmHg and an oxygen saturation >75% during CPB. Norepinephrine was used during CPB when MAP could not be maintained within the targeted range by adjusting the pump flow. During bypass an arterial partial pressure of oxygen of 150 to 250 mmHg was maintained. Body temperature was kept between 35.5 and 36°C during CPB. CPB technique was normothermic using intermittent antegrade warm blood cardioplegia as described by Calafiore and colleagues [19]. Before institution of CPB, isovolemic hemodilution using a hydroxyethylstarch solution 130/0.4 (Voluven®, Fresenius-Kabi, Bad Homburg, Germany) was performed to reduce the Hct to a level of 5 ± 1% above the target Hct level of 20 ± 1% or 25 ± 1%. For all measurements of Hct and blood lactate, a blood gas analyzer (ABL-700 series, Radiometer, Copenhagen, Denmark) was used.
Measurements of Hct were done 2 and 5 minutes after initiation of CPB and every 15 minutes when the target hematocrit was stable. In cases where the initial Hct exceeded the targeted Hct range, crystalloid fluid was substituted, and if the hematocrit was below the target range, autologous blood was transfused.
In the intensive care unit (ICU), patients were extubated as soon as possible after an observation period of 6 hours if they fulfilled the following criteria: paO2 >60 mmHg with an inspired oxygen fraction (FiO2) of 40%, adequate neurological reaction and sufficient muscle strength. Packed red blood cells were substituted according to the following protocol: Hct <23% or lactic acidosis or ST-segment elevation or secondary organ failure attributed to hemodilutional anemia. In the ICU, the patients were not infused with crystalloids according to a standard protocol. Hydroxyethylstarch solution (6% HES 130/0.4) was given when patients showed clinical symptoms of hypovolemia.
Discharge from ICU to the intermediate care unit was feasible when patients were in a stable clinical condition, that is, awake without neurological deficit or agitation, = 5 μg/kg/minute dopamine or no inotrope support, paO2 >60 mmHg with an oxygen insufflation of 4 l per minute and a normal partial pressure of carbon dioxide in arterial blood (paCO2) and no need for continuous loop diuretics to maintain urinary output or renal replacement therapy.
Outcome measures
Primary outcome measures of this trial were calculated whole body oxygen delivery, oxygen consumption and mixed venous blood lactate during CPB, at the end of surgery and in the ICU (one, six and 18 hours after admission). Secondary outcome measures were: drainage loss, transfusion utilization and Hct in the ICU, incidence of secondary organ failure, hemodynamic parameters and stay in ICU.
Oxygen delivery and oxygen consumption were calculated using standard formulae (see Additional file 1). Hemodynamic parameters such as MAP, central venous pressure, mean pulmonary artery pressure and pulmonary artery occlusion pressure were measured before hemodilution, at the end of surgery after retransfusion of autologous blood, one and six hours after admission to the ICU and before discharge from the ICU. Cardiac index, systemic vascular resistance and pulmonary vascular resistance were calculated using standard formulas whereas extravascular lung water and intrathoracic blood volume index were calculated at the same time points using the PiCCO plus-monitor (Pulsion, Munich, Germany). During CPB, pump-driven cardiac index, body temperature and blood lactate were recorded 15 minutes after institution and at the end of CPB whereas the cumulative amount of norepinephrine and urine volume were taken at the end of CPB. The incidence of acute cardiac failure, defined as need for epinephrine and/or enoximone, the need for intraaortic balloon counterpulsation for separation from CPB, and the dosage of dopamine for weaning from CPB were recorded.
In the ICU, the following indicators for secondary organ failure were assessed: neurological complications defined as transitory ischemic attack, agitated arousal reaction or palsy of extremities or hemiplegia. Myocardial infarction was determined by electrocardiogram (new Q-wave, ST-elevations >2 mm) and a ratio of creatine kinase and myocardial subtype of creatine kinase >10%. Acute cardiac failure was defined as the need for inotrope support (epinephrine, norepinephrine or phosphodiesterase inhibitors), respiratory failure as the need for reintubation due to respiratory failure, prolonged respiratory support (>24 hours) or the need for continuous positive airway pressure breathing. Renal insufficiency was assumed when patients required renal replacement therapy, continuous intravenous loop diuretics or increase of creatinine >2.0 mg/dl. Additionally, chest drainage loss, transfusion requirements and Hct, cumulative urine volume, creatinine, ICU stay in hours and mortality were recorded.
Statistical methods
Because of the limited sample sizes and/or non-symmetrically distributed observations we applied only nonparametric statistics. Results were expressed as median and interquartile range in the case of continuous variables. Absolute and relative frequencies were used for categorical and dichotomous variables. The effect of hemodilution regarding primary and secondary outcomes was analyzed using χ2 or Fisher's exact test for categorical and dichotomous variables, respectively. In the case of continuous variables, we applied the Mann-Whitney U test for inter-group analysis. A non-parametric multivariate analysis of variance (nonparametric MANOVA) for repeated measurements and small sample sizes in two and three-factorial designs [20], respectively, was performed in order to take the whole time courses into consideration simultaneously. Multiple tests for differences between the groups in question have been regarded as exploratory ones and were not adjusted for multiplicity. A two-tailed p < 0.05 was considered statistically significant. The statistical analysis was performed using SPSS for Windows, 11.0 (SPSS, Inc., Chicago, IL, USA).