Effects of the heat-moisture exchangers on dynamic hyperinflation of mechanically ventilated COPD patients

In recent years the use of devices called Heat and Moisture Exchangers (HME) has become widespread as gas conditioners for ICU patients requiring mechanical ventilation. As an important variation of the resistive properties of the HME, related to flow and duration of use, has recently been pointed out during in vitro studies, the use of these devices in COPD patients could increase the levels of auto PEEP and dynamic hyperinflation. In this study we have compared the levels of auto PEEP and difference in functional residual capacity ( FRC) in a group of COPD patients, requiring controlled mechanical ventilation (CMV), at basal conditions and after the insertion into the circuit of three HMEs (Dar Hygrobac, Pall Ultipor, Engstrom Edith) at random: the results obtained excluded a significant increase of auto PEEP and (FRC) both with new HMEs and after 12 h of continuous use.     

Transthoracic electrical impedance during extracorporeal hemodialysis in acute respiratory failure (“Shocked lung syndrome”)

The alteration (Z ₀ ) of transthoracic electrical impedance (TEI) during extracorporeal hemodialysis (EHD) was investigated in two Groups of patients with acute renal and acute respiratory failure, that differed with respect to the severity of respiratory insufficiency. Group I had moderate respiratory failure (Fi _{O 2} 0.31±0.10, Pa _{0 2} 84±14 mmHg), and Group II had severe respiratory failure (Fi _{0 2} 0.75±0.17, Pa _{O O} 77±14 mmHg). There was a significant correlation between increase in TEI (Z₀) and decrease in body weight (BW) in each individual patient, but the slope of regression lines was remarkably flattened in Group II. In Group I, TEI was 1.9±0.9 , the calculated TEI for 500 gr decrease in BW (Z_{0–500 gr}) was 0.59±0.21 , and a significant correlation existed between pooled data of Z₀ and BW. In Group II TEI increased less significantly, TEI was 0.6±0.3 (P<0.001), Z_{0–500 gr} was 0.26±0.27 (P<0.01), and there was no correlation between pooled data of Z₀ and BW. Increase of TEI in Group II could be completely attributed to increase in hematocrit. It is concluded that patients of Group I with acute renal failure and moderate respiratory failure lost intrathoracic fluid during EHD, whereas patients of Group II with severe respiratory failure did not. TEI during EHD may serve as a test for detection of fixed fluid within the pulmonary interstitium indicating a poor prognosis of the acute respiratory failure.     

Intrathoracic blood volume accurately reflects circulatory volume status in critically ill patients with mechanical ventilation

Positive pressure ventilation in patients with acute respiratory failure (ARF) may render the interpretation of central venous pressure (CVP) or pulmonary wedge pressure (PCWP) difficult as indicators of circulating volume. The preload component of cardiac (CI) and stroke index (SI) is also influenced by the increased intrathoracic pressures of positive pressure ventilation. Moreover CI and SI do not indicate volume status exclusively but also contractility and afterload. We investigated whether intrathoracic blood volume (ITBV) more accurately reflects blood volume status and the resulting oxygen transport (DO₂). CVP, PCWP, cardiac (CI) and stroke index (SI) were measured, oxygen transport index (DO₂I) and oxygen consumption index (VO₂I) were calculated in 21 ARF-patients. Ventilatory patterns were adjusted as necessary. CI, SI and intrathoracic blood volume index (ITBVI) were derived from thermal dye dilution curves which were detected with a 5 F fiberoptic thermistor femoral artery catheter and fed into a thermaldye-computer. All data were collected in intervals of 6h. There were 224 data sets obtained. Linear regression analysis was performed between absolute values as well as between the 6 h changes (prefix ). The following correlation coefficients were determined: CVP/CI and PCWP/CI 0.01 and –0.142 (p<0.05); CVP/SI and PCWP/SI –0.108 and –0.228 (p<0.01); ITBVI/CI and ITBV/SI 0.488 (p<0.01) and 0.480 (p<0.01); ITBVI and DO₂I 0.460 (p<0.01); CVP/CI and PCWP/CI –0.069 and–0.018; CVP/SI and PCWP/SI –0.083 and –0.009; ITBVI/CI and ITBVI/SI 0.715 (p<0.01) and 0.646 (p<0.01); ITBVI and DO₂I 0.707 (p<0.01). We conclude that in mechanically ventilated patients ITBV is a suitable indicator of circulating blood volume.     

Adaptive Lung Ventilation (ALV) during Anesthesia for Pulmonary Surgery: Automatic Response to Transitions to and from One-Lung Ventilation

Adaptive lung ventilation is a novel closed-loop-controlled ventilation system. Based upon instantaneous breath-to-breath analyses, the ALV controller adjusts ventilation patterns automatically to momentary respiratory mechanics. Its goal is to provide a preset alveolar ventilation (V A) and, at the same time, minimize the work of breathing. Aims of our study were (1) to investigate changes in respiratory mechanics during transition to and from one-lung ventilation (OLV), (2) to describe the automated adaptation of the ventilatory pattern. Methods. With institutional approval and informed consent, 9 patients (33–72 y, 66–88 kg) underwent ALV during total intravenous anesthesia for pulmonary surgery. The ALV controller uses a pressure controlled ventilation mode. V A is preset by the anesthesiologist. Flow, pressure, and CO₂ are continuously measured at the DLT connector. The signals were read into a IBM compatible PC and processed using a linear one-compartment model of the lung to calculate breath-by-breath resistance (R), compliance (C), respiratory time constant (TC), serial dead space (VdS) and V A. Based upon the results, the controller optimizes respiratory rate (RR) and tidal volume (VT) such as to achieve the preset V A with the minimum work of breathing. In addition to V A, only PEEP and FIO₂ settings are at the anesthesiologist's discretion. All patients were ventilated using FIO₂ = 1,0 and PEEP = 3 cm H₂O. Parameters of respiratory mechanics, ventilation, and ABG were recorded during three 5-min periods: 10 min prior to OLV (I), 20 min after onset of OLV (II), and after chest closure (III). Data analyses used nonparametric comparisons of paired samples (Wilcoxon, Friedman) with Bonferroni's correction. Significance was assumed at p < 0.05. Values are given as medians (range). Results. 20 min after onset of OLV (II), resistance had approximately doubled compared with (I), compliance had decreased from 54 (36–81) to 50 (25–70) ml/cm H₂O. TC remained stable at 1.4 (0.8–2.4) vs. 1.2 (0.9–1.6) s. Institution of OLV was followed by a reproducible response of the ALV controller. The sudden changes in respiratory mechanics caused a transient reduction in VT by 42 (8–59) %, with RR unaffected. In order to reestablish the preset V A, the controller increased inspiratory pressure in a stepwise fashion from 18 (14–23) to 27 (19–39) cm H₂O, thereby increasing VT close to baseline (7.5 (6.6–9.0) ml/kg BW vs. 7.9 (5.4–11.7) ml/kg BW). The controller was, thus, effective in maintaining V A. The minimum PaO₂ during phase II was 101 mmHg. After chest closure, respiratory mechanics had returned to baseline. Conclusions. Respiratory mechanics during transition to and from OLV are characterized by marked changes in R and C into opposite directions, leaving TC unaffected. The ALV controller manages these transitions successfully, and maintaines V A reliably without intervention by the anesthesiologist. VT during OLV was found to be consistently lower than recommended in the literature.     

Left ventricular function after acute myocardial infarction

10 patients with their first AMI were studied within the first 48 hours and again after 3 weeks. Central and peripheral haemodynamics (CI, SV, SW, TPR) were examined, including indices of contractility (dp/dlmax) and wall stiffness (P/V, relation P/V to P) of the left ventricle.In the early phase CI and SW, as well as LV dp/dt_max were depressed in accordance with symptoms of LV failure. P/V was increased. Elevation of LVEDP correlated well with ventricular gallop rhythm, but less consistently with LV functional disturbance.During convalescence CI increased uniformly, both in digitalized and non-digitalized individuals. In contrast heart rate, aortic pressure, LVEDP and dp/dt_max remained unchanged. The increase of CI, SV and SW was accompanied by a fall of TPR and P/V. LV wall stiffness was still elevatedabove normal after 3 weeks. The improvement of cardiac pumping during infarct convalescence may have been effected through a fall of TPR and LV wall stiffness. Recovery of depressed contractile performance was generally not observed, and does therefore not seem to contribute to recuperation.Herrn Prof. Dr. med. P. Schölmerich zum 60. Geburtstag.     

Pattern of lung emptying and expiratory resistance in mechanically ventilated patients with chronic obstructive pulmonary disease

Objectives To study the pattern of lung emptying and expiratory resistance in mechanically ventilated patients with chronic obstructive pulmonary disease (COPD).Design A prospective physiological study.Setting A 12-bed Intensive Care Unit.Patients Ten patients with acute exacerbation of COPD.Interventions At three levels of positive end-expiratory pressure (PEEP, 0, 5 and 10 cmH₂O) tracheal (Ptr) and airway pressures, flow (V) and volume (V) were continuously recorded during volume control ventilation and airway occlusions at different time of expiration.Measurements and results V-V curves during passive expiration were obtained, expired volume was divided into five equal volume slices and the time constant () and dynamic deflation compliance (Crs_dyn) of each slice was calculated by regression analysis of V-V and post-occlusion V-Ptr relationships, respectively. In each volume slice the existence or not of flow limitation was examined by comparing V-V curves with and without decreasing Ptr. For a given slice total expiratory resistance was calculated as /Crs_dyn, whereas expiratory resistance (Rrs) and time constant (_rs) of the respiratory system were subsequently estimated taken into consideration the presence of flow limitation. At zero PEEP, _rs increased significantly toward the end of expiration due to an increase in Rrs. PEEP significantly decreased Rrs at the end of expiration and resulted in a faster and relatively constant rate of lung emptying.Conclusions Patients with COPD exhibit a decrease in the rate of lung emptying toward the end of expiration due to an increase in Rrs. PEEP decreases Rrs, resulting in a faster and uniform rate of lung emptying.Electronic Supplementary Material Supplementary material is available in the online version of this article at An erratum to this article can be found at     

The effects of PEEP on arterial oxygenation

The effect of positive end expiratory pressure (PEEP) on arterial oxygenation depends on many factors. Two of the most important are the quality and the quantity of the physiological shunt. The quality depends on the mixed venous oxygen tension, and the quantity on the shunt fraction. Each of these factors may rise or fall with PEEP, depending on the pulmonary and circulatory state of the patient. Their ultimate balance influences the change in arterial oxygen tension.     

Effects of protective and conventional mechanical ventilation on pulmonary function and systemic cytokine release after cardiopulmonary bypass

Objective To evaluate the effects of protective and conventional ventilation with or without positive end-expiratory pressure (PEEP), on systemic tumor necrosis factor-, interleukin-6 levels and pulmonary function during open heart surgery.Design Prospective, randomized clinical study.Setting Single university hospital.Patients and participants Forty-four patients undergoing elective coronary artery bypass grafting surgery with cardiopulmonary bypass.Interventions Patients ventilated with (1) protective tidal volumes (6 ml/kg, respiratory rate: 15 breaths/min, PEEP 5 cmH₂O, n=15) group PV; (2) conventional tidal volumes (10 ml/kg, respiratory rate: 9 breaths/min, PEEP 5 cmH₂O, n=14) group CV+PEEP and (3) conventional tidal volumes (10 ml/kg, respiratory rate: 9 breaths/min, n=15) without PEEP, group CV+ZEEP. Various pulmonary parameters, systemic TNF- and IL-6 levels were determined throughout the study.Measurements and results There were no differences among the groups regarding the systemic TNF- and IL-6 levels. The plateau airway pressures of group PV were lower than those of groups CV+PEEP (p=0.02) and CV+ZEEP (p=0.001) after cardiopulmonary bypass. The shunt fraction of group PV was significantly lower than that of group CV+ZEEP 24 h after surgery (p<0.05). Oxygenation and the alveolar-arterial oxygen difference were better in both PEEP groups than in group CV+ZEEP 24 h after the operation.Conclusions We could not find any evidence that protective mechanical ventilation prevents some of the adverse effects of cardiopulmonary bypass on the lung, nor systemic cytokine levels, postoperative pulmonary function or length of hospitalization.This research is partially supported by Fresenius—Kabi and Aventis Pharma.     

Alveolar pressure during high-frequency jet ventilation

We studied the influence of ventilatory frequency (1–5 Hz), tidal volume, lung volume and body position on the end-expiratory alveolar-to-tracheal pressure difference during high-frequency jet ventilation (HFJV) in Yorkshire piglets. The animals were anesthetized and paralysed. Alveolar pressure was estimated with the clamp off method, which was performed by a computer controlled ventilator and which had been extensively tested on its feasibility. The alveolar-to-tracheal pressure difference increased with increasing frequency and with increasing tidal volume, the common determinant appearing to be the mean expiratory flow. The effects in prone and in supine position were similar. Increasing thoracic volume decreased the alveolar-to-tracheal pressure difference indicating a dependence of this pressure difference on airway resistance. We concluded that the main factors determining the alveolar-to-tracheal pressure difference (P) during HFJV are expiratory flow (V_E) and airway resistance (R), PV_E×R.     

Functional Residual Capacity Measurement during Tracheal Gas Insufflation

Objective. Tracheal gas insufflation (TGI) is considered an adjunctive method to enhance carbon dioxide elimination during permissive hypercapnia in patients with acute respiratory distress syndrome. Due to increasing tidal volume and/or expiratory resistance, TGI may cause intrinsic PEEP (PEEPi), and may lessen the advantages of permissive hypercapnia. There is no reliable method to measure PEEPi during TGI. Using an argon washout method to evaluate dynamic hyperinflation, we developed a method to measure FRC with TGI flow. Methods. We measured FRC during TGI by washing out both the ventilator and TGI circuit with 100% oxygen (O₂) previously equilibrated with 10% argon and 90% O₂. To test the accuracy of our system, we measured the volume in a model lung composed of two flasks. The FRC of the model lung was changed by varying its volume of water, to active 500, 1000, and 1500 mL. The change of FRC (FRC) of the model lung was measured at a flow of 0, 4, 8, and 12 L/min. Then the FRC of a bellows-type model lung was measured at the same TGI flow. PEEPi of the model lung was also recorded as the pressure inside the bellows at end-expiration. Results. Our FRC measurements were accurate within 10% except for that of 500 mL without TGI (12.7% ± 1.1%). As inspiratory time (T_I) and/or TGI flow increased, the FRC of the bellows-type model lung increased. PEEPi and FRC showed a positive correlation (r = 0.843, p < 0.001). The higher the TGI flow, the greater was the FRC with both continuous and expiratory-phase TGI. FRC during continuous TGI was higher than during expiratory-phase TGI especially during long T_I and high TGI flow. Conclusions. The system developed in this study can be used as a method to detect air-trapping during TGI.     

Thromboxane and prostacyclin release in adult respiratory distress syndrome

Plasma thromboxane B₂ (TXB₂) and 6-ketoprostaglandin F₁ (6-keto-PGF₁) were measured in 84 patients at risk of developing adult respiratory distress syndrome (ARDS) (44 patients following multiple trauma, 29 patients following abdominal surgery and 11 patients with acute pancreatitis). Forty-nine of these 84 patients developed an ARDS. High (>140 pg/ml plasma) TXB₂ values were found in 52/84 patients. The median values of TXB₂ were: 360 pg/ml in multiple injured, 250 pg/ml in abdominal surgery and 410 pg/ml in acute pancreatitis patients. The median TXB₂ value was 575 pg/ml in patients developing ARDS and 140 pg/ml in those without this complication: this difference was statistically significant (p<0.05). The median values of 6-keto-PGF₁ were 55pg/ml in multiple injured, 25 pg/ml in abdominal surgery and 120 pg/ml in acute pancreatitis patients. The median 6-keto-PGF₁ value was 122 pg/ml in ARDS patients and 25 pg/ml in non-ARDS patients (statistically significant: p<0.05). High TXB₂ and 6-keto-PGF₁ values were particularly related to sepsis in abdominal surgery patients (p<0.05) and in multiple injured patients (p<0.01). No relation could be established between abnormal TXB₂ or 6-keto-PGF₁ values and death. High TXB₂ values often persisted for several days and were observed particularly at the time ARDS diagnostic criteria were fulfilled. An imbalance between TXB₂ and 6-keto-PGF₁ was observed: 6-keto-PGF₁ values were always lower than TXB₂ values and did not persist for more than 24 h except in four cases. Our data demonstrate a significant production of prostanoids in ARDS patients particularly in sepsis and indicate a disturbance in balance of the prostacyclin/thromboxane axis.     

Respiratory response and inspiratory effort during pressure support ventilation in COPD patients

Objective Pressure Support Ventilation (PSV) is now widely used in the process of weaning patients from mechanical ventilation. The aim of this study was to evaluate the effects of various levels of PS on respiratory pattern and diaphragmatic efforts in patients affected by chronic obstructive pulmonary disease (COPD).Setting Intermediate intensive care unit.Patients We studied ten patients undergoing PSV and recovering from an episode of acute respiratory failure due to exacerbation of COPD.Methods Three levels of PSV were studied, starting from the lowest (PSb) one at which it was possible to obtain an adequate Vt with a pH7.32 and an SaO₂>93%. Then, PS was set at 5 cmH₂O above (PSb+5) and below (PSb-5) this starting level. Ventilatory pattern, transdiaphragmatic pressure (Pdi), the pressure-time product of the diaphragm (PTPdi), the integrated EMG of the diaphragm, static PEEP (PEEPi, stat), dynamic PEEP (PEEPi, dyn), and the static compliance and resistance of the total respiratory system were recorded.Results Minute ventilation did not significantly change with variations in the level of PS, while Vt significantly increased with PS (PS–5=6.3±0.5 ml/kg vs. PSb=10.1±0.9 [p<0.01] and vs. PS+5=11.7±0.6 [p<0.01]), producing a reduction in respiratory frequency with longer expiratory time. The best values of blood gases were obtained at PSb, while at PSb–5, PaCO₂ markedly increased. During PSb and PSb+5 and to a lesser extent during PSb–5, most of the patients made several inspiratory efforts that were not efficient enough to trigger the ventilator to inspire; thus, the PTPdi wasted during these inefficient efforts was increased, especially during PS+5. The application of an external PEEP (PEEPe) of 75% of the static intrinsic PEEP during PSb caused a significant reduction in the occurrence of these inefficient efforts (p<0.05). Minute ventilation remained constant, but Vt decreased, together with Te, leaving the blood gases unaltered. The PTPdi per breath and the dynamic PEEPi were also significantly reduced (by 59% and 31% of control, respectively,p<0.001) with the application of PEEPe.Conclusion We conclude that in COPD patients, different levels of PSV may induce different respiratory patterns and gas exchange. PS levels capable of obtaining a satisfactory equilibrium in blood gases may result in ineffective respiratory efforts if external PEEP is not applied. The addition of PEEPe, not exceeding dynamic intrinsic PEEP, may also reduce the metabolic work of the diaphragm without altering gas exchange.     

Surfactant adjunctive therapy forPneumocystis carinii pneumonitis in an infant with acute lymphoblastic leukaemia

We report successful treatment of adult respiratory distress syndrome (ARDS) with artificial surfactant (40mg/kg, Colfosceril Palmitate, Exosurf, Wellcome) in an infant with severePneumocystis carinii pneumonitis.     

Predictive value of tetrapolar body impedance measurements for hydration status in critically ill patients

Objective To design and evaluate a simple and rapid method to predict body hydration status in critically ill patients.Design Prospective, consecutive sample.Setting Medical intensive care unit of a university hospital.Patients 31 consecutive patients.Methods All patients were classified daily for hydration status by the attending physician based on clinical impression, weight changes and laboratory measurements. The hydration status was scored as dehydrated, euvolemic or edematous. The total body impedance was measured daily by a tetrapolar impedance technique.Results Resistances >700 were found in dehydrated subjects and resistances of <400 in edematous patients. Weight gain was observed in dehydrated and weight loss in edematous patients. A discriminant analysis was used to create a predictive model for hydration using the daily impedance and weight measurements. If a cutoff point of 60% for the predicted classification was used to categorize the patient's hydration as dehydrated, euvolemic and edematous, no false positive predictions were observed for the dehydrated or the edematous state.Conclusion Impedance measurements are in close agreement with the clinical impression of hydration of critically ill patients. Future investigations must elucidate the clinical importance.     

Functional residual capacity as a noninvasive indicator of optimal positive end-expiratory pressure

We hypothesized that functional residual capacity (FRC) could be used as a noninvasive indicator of optimal positive end-expiratory pressure (PEEP), the level of PEEP that results in venous admixture below 15% with an inspired oxygen fraction less than 0.5. We compared several variables for PEEP optimization—oxygen transport, total respiratory system compliance, FRC-based compliance, mixed venous oxygen saturation, end-tidal to arterial carbon dioxide tension difference, and arterial oxygen saturation—by producing four different PEEP levels, 0, 5, 10, and 15 cm H₂O, in 24 mongrel dogs in which pulmonary injury was produced. The data were regressed versus PEEP by using analysis of variance for regression. Venous admixture (F_1,23=149.3;P<0.0001), end-tidal to arterial carbon dioxide tension difference (F_1,23=64.9;P<0.0001), and oxygen transport (F_1,23=95.1;P<0.0001) decreased linearly with PEEP. FRC (F_1,23=248.1;P<0.0001) and arterial oxygen saturation (F_1,23=66.9;P<0.0001) increased linearly with PEEP. Total respiratory system compliance (F_1,23=66.6;P<0.0001) and mixed venous oxygen saturation (F_1,23=12.2;P<0.002) had a quadratic relationship with respect to PEEP with a peak at 5 cm H₂O. FRC-based compliance did not have a significant relationship to PEEP. The maximum values of total respiratory system compliance, FRC-based compliance, mixed venous oxygen saturation, and oxygen transport did not occur at PEEP levels that corresponded to a venous admixture below 15% (optimal PEEP). In this canine oleic acid lung injury model, maximizing these variables would be a poor technique for PEEP titration. FRC and arterial oxygen saturation had a strong relationship to PEEP and venous admixture, and these two would be good noninvasive variables for use in PEEP titration.This work was supported by a grant from Siemens-Elema Ventilator Division, Solna, Sweden.     

The sex killer

Thirteen sex killers were compared to 13 nonsex killers and 13 nonhomicidal sexually aggressive men on sexual history, substance abuse, history of violence, mental illness and personality, brain pathology and endocrine abnormalities. There were more similarities among the three groups than differences. Among the findings, the presence of transvestism and the early appearance of sadism differentiated the sex killers from the other two groups. Sex killers also killed by strangulation more often than nonsex killers and more often victimized a female stranger. The sex killer was more often diagnosed as antisocial personality and sadist than the other two groups. They were more often considered psychotic at the time of their offence and more often considered not guilty by reason of insanity. Directions for future research are noted.     

Thromboxane and prostacyclin release after endotoxin infusion in the rat

To determine in the rat whether pulmonary artery hypertension accompanies thromboxane release, we sequentially monitored pulmonary and systemic artery pressures and cardiac output. We measured pulmonary and aortic plasma levels of TxB₂ as well as 6-keto-prostaglandin F₁ (6-keto-PGF₁) in awake unrestrained adult male Sprague-Dawley rats given a single infusion of endotoxin at the relatively high dose commonly administered to this endotoxin-resistant species. At 40 min after endotoxin infusion, both pulmonary and aortic TxB₂ and 6-keto-PGF₁ levels increased nine-fold and seven-fold above baseline, respectively. In the pulmonary artery, 40 min after infusion, both mediator levels differed significantly from baseline (p<0.05), whereas in the aorta, because of marked variance in the response of different animals, only the 6-keto-PGF₁ levels achieved significance (p<0.05). These changes were associated with a fall in systemic blood pressure and cardiac output, but no demonstrable rise in pulmonary artery pressure (PAP). Despite ultrastructural evidence of vascular injury, these data indicate that in the rat thromboxane and prostacyclin release following a single infusion of endotoxin is not associated with pulmonary hypertension and that increased prostacyclin production may contribute to systemic hypotension.     

Elevated serum bleomycin-detectable iron concentrations in patients with sepsis syndrome

Objective To determine serum bleomycin-detectable free iron in patients with septic shock and to relate these findings to both outcome and a marker of free radical damage.Design A prospective observational study.Setting A nine-bed intensive care unit in a university teaching hospital.Patients Sixteen consecutive patients with septic shock, defined as: (1) Clinical evidence of acute infection; (2) hypo- or hyperthermia (<35.6° or >38.3°C); (3) tachypnoea (>20 breaths/min or ventilated); (4) tachycardia (>90 beats min); (5) shock (systolic pressure <90 mmHg) or on inotropes. Fourteen patients also had secondary organ dysfunction.Measurements and results Bleomycin-detectable iron concentrations were elevated in all patients (37.2±11.0 mol/l vs 5.1±3.3 mol/l in healthy subjects,P<0.0001), but there was no difference between patients who died and those who survived (39.2±9.3 and 36.2±12.3 mol/l, respectively). Thiobarbituric acid reactive substances 9an index of lipid peroxidation) were higher in those who died (3.33±2.29 mol/l) than in the surviving patients (0.99±0.14 mol/l,P<0.01) or healthy subjects (0.92±0.39 mol/l,P<0.01). Free iron did not correlate with thiobarbituric acid-reactive substances. However, a significant correlation was found between lipid peroxidation and clinical severity (APACHE II) score (r=0.54,P<0.05).Conclusions The present study provides evidence of lipid peroxidation in patients who die with septic shock. The data suggest that ironcatalysed hydroxyl radical generation does not form an important contribution to this lipid peroxidation in patients with sepsis.     

Prostacyclin counteracts the increase in capillary permeability induced by tumour necrosis factor-α

Objective To analyse how prostacyclin interferes with the short-term local circulatory effects of tumour necrosis factor- (TNF) in a skeletal muscle.Design An autoperfused sympathectomised cat gastrocnemius muscle enclosed in a plethysmograph.Interventions Arterial blood flow, total and segmental vascular resistances (large-bore arterial vessels, arterioles and veins), hydrostatic capillary pressure, tissue volume and capillary filtration coefficient were followed during local intraarterial infusion of TNF at various rates (2.5, 5.0 and 7.5 g/kg per min) and during intra-arterial infusion of prostacyclin simultaneously with the highest dose of TNF. The capillary filtration coefficient reflects the capillary surface for fluid exchange.Results Arterial infusion of TNF had no influence on vascular resistance up to 5.0 g/kg per min but induced vasodilation at 7.5 g/kg per min. No effects on the recorded hydrostatic capillary pressure were observed. The capillary filtration coefficient and the capillary filtration increased with the infusion rate of TNF the former by 55%. Simultaneous arterial infusion of prostacyclin (350 ng/kg per min) caused further vasodilation and an increase in hydrostatic capillary pressure and completely restored the capillary filtration coefficient to control. The TNF-induced filtration was partly restored.Conclusions The local circulatory effect of TNF is small apart from a graded increase in the capillary filtration coefficient, most likely reflecting an increase in capillary permeability. The prostacyclininduced decrease in capillary filtration coefficient most likely reflects a restoration of capillary permeability. The TNF-induced transcapillary filtration is not fully reduced by prostacyclin due to a simultaneous increase in hydrostatic capillary pressure.     

Infectious and inflammatory dissemination are affected by ventilation strategy in rats with unilateral pneumonia

Objective To evaluate the effect of V_T reduction and alveolar recruitment on systemic and contralateral dissemination of bacteria and inflammation during right-side pneumonia.Design Interventional animal study.Setting University hospital research laboratory.Subjects A total of 54 male Wistar rats.Interventions One day after right lung instillation of 1.4×10⁷ Pseudomonas aeruginosa, rats were left unventilated or ventilated for 2 h at low V_T (6 ml/kg) with different strategies of alveolar recruitment: no PEEP, 8 cm H₂O PEEP, 8 cm H₂O PEEP in a left lateral position, 3 cm H₂O PEEP with partial liquid ventilation, or high V_T (set such as end-inspiratory pressure was 30 cm H₂O) without PEEP (ZEEP). After ventilation the lungs, spleen and liver were cultivated for bacterial counts. Global bacterial dissemination was scored considering the percentage of positive spleen, liver and left lung cultures. TNF- was assayed in plasma before and after mechanical ventilation.Measurements and results All rats had right-side pneumonia with similar bacterial counts. All mechanical ventilation strategies, with the exception of low V_T-PEEP 8, promoted contralateral lung dissemination. Overall bacterial dissemination was less in non-ventilated controls (22%) and low V_T-PEEP 8 (22%) than in high V_T-ZEEP (67%), low V_T-PEEP 8 in left lateral position (59%) and low V_T-ZEEP (56%) (p<0.05). Partial liquid ventilation prevented systemic bacterial translocation, but at the expense of contralateral bacterial seeding. Plasma TNF- concentration increased significantly after mechanical ventilation with no PEEP at both high and low V_T.Conclusions Our results suggest that PEEP might reduce the risk of ventilation-induced bacterial and inflammatory mediator dissemination during pneumonia.Electronic Supplementary Material Supplementary material is available in the online version of this article at http://dx.doi.org/10.1007/s00134-003-2147-7This study was supported by a grant from Fondation pour la Recherche Médicale.