We conducted a retrospective study in the anesthesiology and critical care department of The Military Hospital of Tunis. We analyzed the patients’ database of Meddeb et al. study [15]. This study was conducted from August 2011 to May 2014.
After approval of the hospital ethics committee, the patient was included according to the emergency procedure or after receiving his family agreement. In all cases, we asked for the final consent of the patient or his family subsequently before the definitive inclusion.
We included in the study, all patients over the age of 16 years in septic shock defined by the Bone criteria of the consensus conference of the American College of Chest Physicians / Society of Critical Care Medicine (ACCP/ SCCM) of 1991 [16].
We carried out fluid challenges until reaching a total volume of 20 to 40 ml/kg of crystalloids or colloids. The therapeutic objectives were a MAP greater than 65 mm Hg, a diuresis greater than 0.5 ml/kg/h, and a lactatemia less than 2 mmol/l. In case of failure, we used norepinephrine to optimize the hemodynamic state. To be included, patients must be sedated, intubated and under mechanical ventilation support.
After inclusion, the patients were randomized using a randomization table in three groups:
- Dobutamine group: each patient received continuous perfusion of dobutamine using an electric syringe pump at a rate of 5 μg/kg per minute for at least three days.
- Levosimendan group: each patient received continuous perfusion of levosimendan using an electric syringe pump at a rate of 0.2 μg/kg per minute for 24 hours then continuous perfusion with an electric syringe pump of normal saline solution for 48 hours.
- Control group: 10 patients who were not on inotropes.
We did not include moribund patients (age <18 years, pregnant women, uncontrolled hemorrhage, history of cardiopathy, and/or severe heart failure). The exclusion criterion was a decision to limit care for the first 48 hours. The principle of the study was to analyze the effect of vasopressors on microcirculation in patients with septic shock resuscitated and stabilized according to the recommendations of Surviving Sepsis Campaign (SSC) 2002 [17].
Sedation was conducted with continuous perfusions of midazolam and remifentanil. The goal was a Ramsay score of 4 to 5. Firstly, all the patients were under a probabilistic intravenous antibiotic therapy that was adapted secondly, to the infectious site and the eventual germs found. We performed veno-venous hemofiltration or conventional dialysis in patients with acute oligo-anuric renal failure. Attending physician did not have any information on the results of the microdialysis until study end. We collected the data on a notebook that bears the patient’s name, the patient’s history, the starting point of sepsis, the responsible germs, and the Sequential Organ Failure Assessment (SOFA) score, the Simplified Acute Physiology Score (SAPS II), the duration of septic shock, the duration of mechanical ventilation, the discharge status from the intensive care unit (whether alive or dead).
Every six hours, we measured hemoglobin and blood glucose levels every six hours and collected arterial samples to analyze blood gases and lactates. We performed hemodynamic and classic parameters monitoring by a Swan-Ganz catheter. We collected macrocirculation data, the dose of norepinephrine required, and the corresponding microcirculation data (lactate and pyruvate concentration in the interstitial tissue) every six hours during the first five days, by muscle microdialysis. We used for the analysis a microdialysis machine (CMA600, CMA Microdialysis AB, Stockholm, Sweden). The lactate/pyruvate ratio was calculated automatically by the machine.
The primary endpoint was the correlation between the administered dose of norepinephrine and the microcirculatory metabolic parameters (lactate, pyruvate, and muscle lactate pyruvate ratio) measured by microdialysis.
We expressed qualitative variables in number and percentage. We applied Pearson-Fischer Chi-square tests to these variables. We used the Kolmogorov-Smirnov test to evaluate the normality of the distribution of continuous variables. We presented continuous variables as mean and standard deviations of either the mean or the median (interquartile range), depending on the normality of the distribution. We used the ANOVA test for analysis of variance when appropriate. We assessed correlations between quantitative variables with Pearson’s r test or Spearman test based on the normality of variables.
The correlation was assessed according to the significance threshold (p) and the Pearson’s r coefficient or Spearman’s rho in:
- Strong correlation: [- 1, - 0.75] [0.75, 1]
- Average correlation:] –0.75, - 0.5] [0.5, 0.75 [
- Weak correlation:]–0.5, 0 [] 0, 0.5 [
- Zero correlation: r or rho = 0
These tests were two-tailed with a significance threshold (p) set at 0.05.