R Marcelle

The use of low doses of dopamine in intensive care medicine

The dopamine alpha- and beta-adrenoceptor dose-response curves are investigated in four patients who are exempt from cardiovascular disease. A dose-related increase in CO, HR and SV is observed with infusion rates of up to 3 micrograms kg-1 min-1. With concentrations greater than 10 micrograms kg-1 min-1, both BP and SVR increase. Low-dose dopamine infusion less than 3 micrograms kg-1 min-1 is investigated in ten other patients. With this infusion rate, a selective renal vasodilation is induced without peripheral or cardiac beta-adrenoceptor activation. Dopamine is responsible for an increase in diuresis FENa, GFR and RBF. These properties are indicated in renal failure, and when haemodynamic support is required in cardiac failure, if an infusion rate of up to 10 micrograms kg-1 min-1 is able to reverse cardiac insufficiency.

Analysis of endotoxin effects on the intact pulmonary circulation

OBJECTIVE The mechanism of sustained alterations in pulmonary hemodynamics during endotoxin shock remains unclear. To gain more detailed knowledge we used the four-element windkessel model as a descriptor of the pulmonary circuit. METHODS Consecutive changes in characteristic resistance (R1), vascular compliance (C), input resistance (R2) and inductance (L) were continuously assessed following injection of endotoxin in 6 anaesthetised pigs, and were compared with the corresponding values measured in a similar group of sham-operated animals. RESULTS Endotoxin challenge resulted in a biphasic pulmonary artery pressure response. Blood flow decreased progressively from 2.8 +/- 0.2 l/min to 2 +/- 0.2 l/min. Ohmic pulmonary vascular resistance (PVR) increased gradually from 0.2 +/- 0.04 to 0.76 +/- 0.1 mm Hg s ml-1. The early increase in PAP (from 14 +/- 2 to 27 +/- 4 mm Hg) was mediated by changes in both R1 (from 0.04 +/- 0.01 to 0.06 +/- 0.01 mm Hg s ml-1) and R2 (from 0.16 +/- 0.04 to 0.61 +/- 0.2 mm Hg s ml-1). These responses, in turn, altered the proximal vascular compliance. A subsequent increase in PAP (from 27 +/- 2 to 32 +/- 3 mm Hg) paralleled the specific decline in distal pulmonary vasculature compliance from 0.84 +/- 0.1 to 0.65 +/- 0.1 ml/mmHg. Analysis of the time course of PVR did not allow us to distinguish between vasoconstriction and stiffening of the vascular tree as mechanisms accounting for PAP changes. CONCLUSIONS Endotoxemia leads to pulmonary hypertension, which is a result of constriction of proximal pulmonary arteries during the early phase, whereas the late phase is characterised by a decline in distal pulmonary vasculature compliance.

Accuracy in Early Prediction of Prognosis of Patients with Septic Shock by Analysis of Simple Indices: Prospective Study

In 26 consecutive septic shock patients, we analyzed the clinical, hemodynamic, and metabolic data before and during volume infusion to test their circulatory reserve in response to fluid repletion. These patients were investigated to identify early variables that could predict outcome. There were 15 survivors (group A) and 11 nonsurvivors (group B). As a mean, group A patients were hemodynamically evaluated 2.3 h after onset of the sepsis syndrome, whereas group B patients underwent cardiac catheterization after a 12-h interval. At the initial evaluation, both groups demonstrated similarly decreased mean arterial pressure, mean heart rate, and mean cardiac filling pressure. Only group A patients evidenced elevated cardiac index (CI) (>4 L/min·m2) associated with low systemic vascular resistance index (<7400 dyne·sec/cm5·m2), which is generally recognized as hyperdynamic cardiac state. However, none of the initial cardiovascular variables could serve as a predictor for survival. Fluid challenge increased left ventricular preload from 6 to 12.4 and from 7.8 to 12.7 mm Hg in group A and group B, respectively. The increases were associated with significant increases in CI from 4.4 to 6.9 and from 3 to 3.8 L/min·m2. However, at the end of fluid challenge, only group A patients exhibited normal cardiac response, as evidenced by the change in left ventricular stroke work index (LVSWI) for a given increase in the pulmonary capillary wedge pressure (WP) that was referred to as left cardiac preload. Contrary to poor predictive value, on an individual basis, of the initial variables to assess the outcome of septic shock, the multivariate analysis that simultaneously accounted for four variables (LVSWI/WP, WBC count, Po2, and Hct) could separate with accuracy the two groups of patients.

Effects of Escherichia coli endotoxin on pulmonary vascular resistance in intact dogs

The effects of endotoxin on pulmonary hemodynamics were studied in seven intact dogs. The distribution of pulmonary vascular resistance was estimated by the effective pulmonary capillary pressure, which was derived from the pressure transient recorded while the pulmonary artery catheter was rapidly wedged. After the injection of endotoxin, cardiac output and aortic pressure consistently fell. Pulmonary artery occlusion (wedge) pressure also decreased, but not significantly. Although pulmonary artery pressure did not necessarily rise, total pulmonary vascular resistance increased in every dog. The absolute increase in pulmonary artery resistance was greater (142 mm Hg/L min/kg) than in venous resistance (111 mm Hg/L min/kg); however, the relative increase in venous resistance was higher (410% for venous resistance vs. 220% for pulmonary artery resistance). As a result of venoconstriction, there was a consistent increase in effective pulmonary capillary pressure (from 2.5 to 6.3 mm Hg). Our data indicate that the pulmonary vascular response to endotoxin injection is characterized by constriction of both pulmonary arteries and pulmonary veins. The capillary wedge pressure did not reflect the pulmonary microvascular pressure, since it varied in the opposite direction to the effective capillary pressure.

Lung Fluid Dynamics and Supply Dependency of Oxygen Uptake During Experimental Endotoxic Shock and Volume Resuscitation

Background and MethodsWe studied the effect of volume resuscitation on lung fluid balance and systemic oxygen extraction during septic shock in eight anesthetized dogs. Sepsis was induced using a 2-hr continuous infusion of Escherichia coli endotoxin at 0.25 μg/min-kg. Relationships between oxygen uptake (Vo2) and oxygen supply (Do2) were performed acutely during step wise controlled decrements in cardiac output by progressive inflation of an intracardiac balloon. At each stage, Do2 and corresponding Vo2 were measured independently and the individual critical Do2 level was referred to as the point below which the relationship held. The slope of such a constructed relationship was defined as the maximal oxygen extraction ratio. Lung fluid balance was assessed by measurements of extravascular lung water. All values were studied at baseline, after endotoxin insult, and after reversing hypotension by a 10% dextran infusion. ResultsEndotoxin infusion led to a shock state that associated hypotension (from 135 to 63 mm Hg) with increases in blood lactate (from 0.53 to 3.9 mmol/L). The mean critical Do2 and maximal oxygen extraction ratio were significantly altered from 7.9 to 17.8 mL/min-kg and from 0.81 to 0.38, respectively. After reversing hypotension by 28 mL/kg colloid infusion, the critical Do2 (11.4 mL/min-kg) and maximal oxygen extraction ratio (0.48) were significantly improved. However, restoration of normal values required a state of fluid overload by further dextran infusion (8 mL/kg). At the end of the fluid challenge, extravascular lung water significantly increased from 6.4 to 17.4 mL/kg. ConclusionsThese data suggest that volume loading may reverse endotoxin-induced peripheral perfusion abnormalities. However, substantial pulmonary edema may occur, possibly jeopardizing the beneficial effects of fluid expansion. (Crit Care Med 1991; 19:955)

Correction of pressure waveforms recorded by fluid‐filled catheter recording systems: A new method using a transfer equation

Background: Pressure measuring systems using fluid‐filled catheters can result in the recording of distorted pressure waveforms. It results in phase delay, overestimation of systolic and, to a lesser extent, of diastolic pressure. We designed and evaluated a method to correct this pressure waveform distortion using an appropriate transfer equation obtained from the dynamic response of the fluid‐filled catheter. This transfer equation is based on the principle that a fluid‐filled catheter recording system is considered as an underdamped dynamic system fully characterized by its natural frequency (ωn and damping ratio (ζ).

Pulmonary impedance and right ventricular-vascular coupling in endotoxin shock

OBJECTIVE We tested the hypothesis that right heart failure during endotoxin shock may result from altered ventriculovascular coupling responsible for impeding power transfer to the pulmonary circulation. METHODS The changes in vascular pulmonary input impedance and right ventricular contractility produced by low-dose endotoxin infusion were studied in 6 intact anesthetized dogs. RESULTS Endotoxin insult resulted in pulmonary hypertension (from 22 +/- 2 to 33 +/- 3 mmHg) associated with significant decreases in stroke volume (from 26.9 +/- 4 to 20.2 +/- 3 ml) and right ventricular ejection fraction (from 41 +/- 3 to 32 +/- 2%). The first minimum of input impedance spectrum and zero phase were shifted towards higher frequencies. Input resistance and characteristic resistance were dramatically increased. The latter change contributed to a significant increase in the pulsatile component of total right ventricular power output from 13 to 21%, indicating a reduction in the hydraulic right ventricle power output delivered into the main pulmonary artery. Overall changes in input pulmonary impedance were indicative of increased afterload facing the right ventricle leading to depressed performance. In contrast, right ventricular systolic elastance was simultaneously increased from 0.56 to 0.93 mmHg/ml indicating an increase in right heart contractility. CONCLUSION These data suggest that pulmonary hypertension in the setting of experimental endotoxin shock is accompanied by deleterious changes in the pulmonary impedance spectrum, which are responsible for a mismatch of increased contractile state of the right ventricle to the varying hydraulic load ultimately leading to ventricular-vascular uncoupling.

Unusual complication after percutaneous dilatational tracheostomy: pneumoperitoneum with abdominal compartment syndrome.

Sir: We read with interest the recently published paper by Walz and Schmidt, who reported anterior tracheal lesion caused by percutaneous dilatational tracheostomy (PDT) in post-mortem examination [1]. We report the first case of posterior tracheal lesion with pneumoperitoneum after PDT. A 59-year-old woman was admitted to the ICU for Guillain-BarrØ syndrome requiring mechanical ventilation. Eleven days later, PDT was performed using Ciaglias technique (Cook, Bloomington,Ind.). The patient was ventilated via laryngeal mask without any problem [2]. After progressive dilatation of the trachea with dilators of increasing size, a Shiley-8 tracheal cannula was inserted but seemed to be obstructed and it had to be removed. A second attempt to place the tracheal cannula was successful. Twenty-four hours later, chest X-ray showed a pneumomediastinum (Fig. 1) but CT scan did not reveal any cause for this. Tracheo-bronchoscopy was also normal. The tracheal tube cuff was left in place and fully inflated. The patient improved until day 4 when diuresis and total pulmonary static compliance abruptly dropped with concurrent abdominal air distention. An increase in intra-abdominal pressure (20 mmHg) and tension pneumoperitoneum was confirmed on abdominal CT (Fig.2). As a rigid esophago-tracheobronchoscopy demonstrated a leak in the posterior tracheal wall, 4.5 cm below the tracheostomy stoma, a reinforced tracheal tube was inserted just above the carina in order to cover this leak. The pneumoperitoneum was evacuated using a 16 gauge catheter and pulmonary compliance and diuresis returned to normal values. The tracheal tube was left in place for 10 days until the leak closed spontaneously as demonstrated by tracheoscopy. A tracheal cannula was then inserted. No related infectious complication occurred. The patient was transferred from the ICU, but died 1 month later from cardiac arrest on a general ward. Although subcutaneous, mediastinal emphysema and pneumothorax have already been reported as a perioperative complication of PDT, this is the first report of pneumoperitoneum as a complication of PDT [2]. One case of pneumoperitoneum has been reported in the literature following surgical tracheostomy, but considered as a consequence of barotrauma [3]. In the present case, the origin of the pneumoperitoneum was a breach probably occurring during the last insertion of the cannula over its dilator. Barotrauma related to mechanical ventilation is unlikely to have been responsible for the pneumoperitoneum since intrathoracic pressure was kept in the low range during the procedure. The use of a laryngeal mask could also have contributed to this complication owing to a minimum fixation of the trachea. Indeed, Kaloud et al.[4] also reported a case of full length rupture of the membranous wall after a PDT with laryngeal mask. The narrower diameter of the female trachea should be considered in the occurrence of the lesion. Although the use of fiberoptic bronchoscope for guidance and control of puncture was mandatory, this case seems to demonstrate that it is not sufficient to prevent all posterior wall lesions. Reviewing the chest X-ray and clinical evolution revealed no preliminary sign and we suspected that the interval of time between the pneumomediastinum and the pneumoperitoneum occurrence was explained by an almost complete closing of the hole by the cuff. However no satisfactory explanation was found for this prolonged time delay. This time interval could 1334

Effects of intravascular volume expansion on lung fluid balance in a canine model of septic shock

We tested the early effects of endotoxin on both the permeability of capillary membranes and microvascular pressure. One group of dogs (n = 8) were fluid loaded (30 ml/kg dextran-40) after having been subjected to a 2-h Escherichia coli endotoxin infusion (0.25 micrograms/kg X min). A second control group of animals (n = 6) was submitted to a similar (25 ml/kg) volume loading over an equivalent 30-min period. We estimated extravascular lung water (EVLW), calculated the effective pulmonary capillary pressure, and determined the alveolar-capillary filtration coefficient (Kf) after volume loading. Only the septic animals consistently showed elevated EVLW values consistent with pulmonary edema. The results showed, however, that the Kf calculated for the dogs that received endotoxin was no different from that of control group (Kf = 0.005 ml/kg X min X mm Hg). Instead, endotoxin constricted the pulmonary veins which led to a considerable rise in microvascular hydrostatic pressure above the level at which the lungs could not resist edema formation. We conclude that acute pulmonary edema that follows endotoxin insult and subsequent therapeutic volume replacement is due to an increased filtration force instead of an alteration in the microvascular permeability.

Pressure-Flow Relationships of the Pulmonary Circulation During Endotoxin Infusion in Intact Dogs

Background and MethodsWe aimed to characterize the effects of an endotoxin insult (Escherichia coli 0127:B8) on the relationships between pulmonary vascular pressure and flow in intact dogs. To achieve this goal, multipoint plots of total pressure gradient, arterial pressure gradient, and venous pressure gradient vs. flow were generated by graded inflation of a right atrial balloon, which was used to vary flow. The partitioning of the total pressure decrease across the pulmonary vasculature (total pressure gradient = pulmonary arterial pressure – pulmonary artery occlusion pressure [PAOP]) into gradients across pulmonary arterial (arterial pressure gradient = pulmonary arterial pressure – effective capillary pressure) and pulmonary venous (venous pressure gradient = effective capillary pressure – PAOP) regions was assessed by a waveform mathematical analysis of the pulmonary arterial pressure profile during arterial occlusion, with computation of both PAOP and effective pulmonary capillary pressures. Slopes and extrapolated pressure intercepts from linear regression fits to the pulmonary vascular pressure/flow plots were determined in seven dogs after a 2-hr endotoxic infusion interval and were compared with the corresponding values that characterized a similar group of sham-operated dogs. ResultsUnder normal conditions, the extrapolated pressure intercept for pulmonary arterial pressure gradient was virtually 0 mm Hg; for total pulmonary arterial pressure gradient and pulmonary venous pressure gradient, the mean extrapolated pressure intercepts were substantially positive: 2.4 ± 0.2 and 2.1 ± 0.3 mm Hg, respectively. Endotoxin infusion at 0.25 μg/kg/min significantly increased the pressure intercepts from 2.4 to 8.7 and from 2.1 to 8.3 mm Hg of total pressure gradient and venous pressure gradient vs. flow, respectively. This infusion produced a minor, nonsignificant change in the intercept of arterial pressure gradient vs. flow, whereas it increased its slope significantly (p < .05) from 0.036 to 0.081 mm Hg/mlVmin/kg. ConclusionsThese data suggest that endotoxins effects on vascular resistance are exerted at two different loci such that these effects are additive. These endotoxin-induced effects consisted of increased vascular resistance of the arterial segment and appearance of a Starling resistor at the venous side of the pulmonary circulation, which acted as the relevant back-pressure to flow.