Xavier Monnet

Passive leg raising predicts fluid responsiveness in the critically ill.

Objective:Passive leg raising (PLR) represents a “self-volume challenge” that could predict fluid response and might be useful when the respiratory variation of stroke volume cannot be used for that purpose. We hypothesized that the hemodynamic response to PLR predicts fluid responsiveness in mechanically ventilated patients. Design:Prospective study. Setting:Medical intensive care unit of a university hospital. Patients:We investigated 71 mechanically ventilated patients considered for volume expansion. Thirty-one patients had spontaneous breathing activity and/or arrhythmias. Interventions:We assessed hemodynamic status at baseline, after PLR, and after volume expansion (500 mL NaCl 0.9% infusion over 10 mins). Measurements and Main Results:We recorded aortic blood flow using esophageal Doppler and arterial pulse pressure. We calculated the respiratory variation of pulse pressure in patients without arrhythmias. In 37 patients (responders), aortic blood flow increased by ≥15% after fluid infusion. A PLR increase of aortic blood flow ≥10% predicted fluid responsiveness with a sensitivity of 97% and a specificity of 94%. A PLR increase of pulse pressure ≥12% predicted volume responsiveness with significantly lower sensitivity (60%) and specificity (85%). In 30 patients without arrhythmias or spontaneous breathing, a respiratory variation in pulse pressure ≥12% was of similar predictive value as was PLR increases in aortic blood flow (sensitivity of 88% and specificity of 93%). In patients with spontaneous breathing activity, the specificity of respiratory variations in pulse pressure was poor (46%). Conclusions:The changes in aortic blood flow induced by PLR predict preload responsiveness in ventilated patients, whereas with arrhythmias and spontaneous breathing activity, respiratory variations of arterial pulse pressure poorly predict preload responsiveness.

Cardiac filling pressures are not appropriate to predict hemodynamic response to volume challenge.

Objective: Values of central venous pressure of 8–12 mm Hg and of pulmonary artery occlusion pressure of 12–15 mm Hg have been proposed as volume resuscitation targets in recent international guidelines on management of severe sepsis. By analyzing a large number of volume challenges, our aim was to test the significance of the recommended target values in terms of prediction of volume responsiveness. Design: Retrospective study. Setting: A 24‐bed medical intensive care unit. Patients: All consecutive septic patients monitored with a pulmonary artery catheter who underwent a volume challenge between 2001 and 2004. Intervention: None. Measurements and Main Results: A total of 150 volume challenges in 96 patients were reviewed. In 65 instances, the volume challenge resulted in an increase in cardiac index of ≥15% (responders). The pre‐infusion central venous pressure was similar in responders and nonresponders (8 ± 4 vs. 9 ± 4 mm Hg). The pre‐infusion pulmonary artery occlusion pressure was slightly lower in responders (10 ± 4 vs. 11 ± 4 mm Hg, p < .05). However, the significance of pulmonary artery occlusion pressure to predict fluid responsiveness was poor and similar to that of central venous pressure, as indicated by low values of areas under the receiver operating characteristic curves (0.58 and 0.63, respectively). A central venous pressure of <8 mm Hg and a pulmonary artery occlusion pressure of <12 mm Hg predicted volume responsiveness with a positive predictive value of only 47% and 54%, respectively. With the knowledge of a low stroke volume index (<30 mL·m−2), their positive predictive values were still unsatisfactory: 61% and 69%, respectively. When the combination of central venous pressure and pulmonary artery occlusion pressure was considered instead of either pressure alone, the degree of prediction of volume responsiveness was not improved. Conclusion: Our study demonstrates that cardiac filling pressures are poor predictors of fluid responsiveness in septic patients. Therefore, their use as targets for volume resuscitation must be discouraged, at least after the early phase of sepsis has concluded.

Predicting volume responsiveness by using the end-expiratory occlusion in mechanically ventilated intensive care unit patients

Objective:During mechanical ventilation, inspiration cyclically decreases the left cardiac preload. Thus, an end-expiratory occlusion may prevent the cyclic impediment in left cardiac preload and may act like a fluid challenge. We tested whether this could serve as a functional test for fluid responsiveness in patients with circulatory failure. Design:Prospective study. Setting:Medical intensive care unit. Patients:Thirty-four mechanically ventilated patients with shock in whom volume expansion was planned. Intervention:A 15-second end-expiratory occlusion followed by a 500 mL saline infusion. Measurements:Arterial pressure and pulse contour-derived cardiac index (PiCCOplus) at baseline, during passive leg raising (PLR), during the 5-last seconds of the end-expiratory occlusion, and after volume expansion. Main Results:Volume expansion increased cardiac index by >15% (2.4 ± 1.0 to 3.3 ± 1.2 L/min/m2, p < 0.05) in 23 patients (“responders”). Before volume expansion, the end-expiratory occlusion significantly increased arterial pulse pressure by 15% ± 15% and cardiac index by 12% ± 11% in responders whereas arterial pulse pressure and cardiac index did not change significantly in nonresponders. Fluid responsiveness was predicted by an increase in pulse pressure ≥5% during the end-expiratory occlusion with a sensitivity and a specificity of 87% and 100%, respectively, and by an increase in cardiac index ≥5% during the end-expiratory occlusion with a sensitivity and a specificity of 91% and 100%, respectively. The response of pulse pressure and cardiac index to the end-expiratory occlusion predicted fluid responsiveness with an accuracy that was similar to the response of cardiac index to PLR and that was significantly better than the response of pulse pressure to PLR (receiver operating characteristic curves area 0.957 [95% confidence interval {CI:} 0.825–0.994], 0.972 [95% CI: 0.849–0.995], 0.937 [95% CI: 0.797–0.990], and 0.675 [95% CI: 0.497–0.829], respectively). Conclusions:The hemodynamic response to an end-expiratory occlusion can predict volume responsiveness in mechanically ventilated patients.

Effects of changes in vascular tone on the agreement between pulse contour and transpulmonary thermodilution cardiac output measurements within an up to 6-hour calibration-free period*

Objectives: To examine whether the agreement between pulse contour and transpulmonary thermodilution cardiac index (CI) measurements is altered by changes in vascular tone within an up to 6-hr calibration-free period. Design: Observational study. Setting: Medical intensive care unit of a university hospital. Patients: Fifty-nine critically ill patients. Interventions: None. Measurements and Main Results: Data from 59 critically ill patients equipped with a PiCCO device were retrospectively analyzed. The database contained the transpulmonary thermodilution CI (CIT) value obtained at each time point the device was calibrated and the pulse contour CI (CIPC) value recorded immediately before this time point. Seven subsets of CI pairs were defined according to intervals of time elapsed from the previous calibration (within the first 30 mins, between 30 mins and 1 hr, and every hour up to 6 hrs). In the whole set of 400 CI pairs, CIPC correlated with CIT (r2 = .68, p < .001). The bias ± sd was 0.12 ± 0.61 L/min/m2, and the percentage error was 35%. Among the seven time-interval subsets, the percentage error was <30% only in the two first ones (27% and 26%, respectively). When changes in systemic vascular resistance by >15% occurred (129 times), CIPC correlated with CIT (r2 = .64), the bias ± sd was 0.12 ± 0.62 L/min/m2, and the percentage error was 36%. In the subset of CI pairs recorded within the 1-hr calibration-free period while vascular resistance changed by >15% (n = 32), the bias ± sd was 0.04 ± 0.47 L/min/m2 and the percentage error was 29%. Conclusions: Our study in critically ill patients suggests that the agreement between pulse contour cardiac output and transpulmonary thermodilution cardiac output was not significantly influenced by changes in vascular tone. However, after a 1-hr calibration-free period, recalibration may be encouraged. Such a procedure provides helpful information drawn from other thermodilution-derived variables.

Extravascular lung water is an independent prognostic factor in patients with acute respiratory distress syndrome.

Objective:Acute respiratory distress syndrome might be associated with an increase in extravascular lung water index and pulmonary vascular permeability index, which can be measured by transpulmonary thermodilution. We tested whether extravascular lung water index and pulmonary vascular permeability index are independent prognostic factors in patients with acute respiratory distress syndrome. Design:Retrospective study. Setting:Medical intensive care unit. Patients:Two hundred consecutive acute respiratory distress syndrome patients (age = 57 ± 17, Simplified Acute Physiology Score II = 57 ± 20, overall day-28 mortality = 54%). Measurements:Extravascular lung water index and pulmonary vascular permeability index were collected (PiCCO device, Pulsion Medical Systems) at each day of the acute respiratory distress syndrome episode. Main Results:The maximum values of extravascular lung water index and pulmonary vascular permeability index recorded during the acute respiratory distress syndrome episode (maximum value of extravascular lung water index and maximum value of pulmonary vascular permeability index, respectively) were significantly higher in nonsurvivors than in survivors at day-28 (mean ± SD: 24 ± 10 mL/kg vs. 19 ± 7 mL/kg of predicted body weight, p < 0.001 [t-test] for maximum value of extravascular lung water index and median [interquartile range]: 4.4 [3.3–6.1] vs. 3.5 [2.8–4.4], p = 0.001 for maximum value of pulmonary vascular permeability index, Wilcoxon’s test). In multivariate analyses, maximum value of extravascular lung water index or maximum value of pulmonary vascular permeability index, Simplified Acute Physiology Score II, maximum blood lactate, mean positive end-expiratory pressure, mean cumulative fluid balance, and the minimal ratio of arterial oxygen pressure over the inspired oxygen fraction were all independently associated with day-28 mortality. A maximum value of extravascular lung water index >21 mL/kg predicted day-28 mortality with a sensitivity of (mean [95% confidence interval]) 54% (44–63)% and a specificity of 73% (63–82)%. The mortality rate was 70% in patients with a maximum value of extravascular lung water index >21 mL/kg and 43% in the remaining patients (p = 0.0003). A maximum value of pulmonary vascular permeability index >3.8 predicted day-28 mortality with a sensitivity of (mean [95% confidence interval]) 67% (57–76)% and a specificity of 65% (54–75)%. The mortality rate was 69% in patients with a maximum value of pulmonary vascular permeability index >3.8 and 37% in the group with a maximum value of pulmonary vascular permeability index ⩽3.8 (p < 0.0001). Conclusions:Extravascular lung water index and pulmonary vascular permeability index measured by transpulmonary thermodilution are independent risk factors of day-28 mortality in patients with acute respiratory distress syndrome.

Passive leg-raising and end-expiratory occlusion tests perform better than pulse pressure variation in patients with low respiratory system compliance.

Objectives:We tested whether the poor ability of pulse pressure variation to predict fluid responsiveness in cases of acute respiratory distress syndrome was related to low lung compliance. We also tested whether the changes in cardiac index induced by passive leg-raising and by an end-expiratory occlusion test were better than pulse pressure variation at predicting fluid responsiveness in acute respiratory distress syndrome patients. Design:Prospective study. Setting:Medical intensive care unit. Patients:We included 54 patients with circulatory shock (63 ± 13 yrs; Simplified Acute Physiology Score II, 63 ± 24). Twenty-seven patients had acute respiratory distress syndrome (compliance of the respiratory system, 22 ± 3 mL/cm H2O). In nonacute respiratory distress syndrome patients, the compliance of the respiratory system was 45 ± 9 mL/cm H2O. Measurements and Main Results:We measured the response of cardiac index (transpulmonary thermodilution) to fluid administration (500 mL saline). Before fluid administration, we recorded pulse pressure variation and the changes in pulse contour analysis-derived cardiac index induced by passive leg-raising and end-expiratory occlusion. Fluid increased cardiac index ≥15% (44% ± 39%) in 30 “responders.” Pulse pressure variation was significantly correlated with compliance of the respiratory system (r = .58), but not with tidal volume. The higher the compliance of the respiratory system, the better the prediction of fluid responsiveness by pulse pressure variation. A compliance of the respiratory system of 30 mL/cm H2O was the best cut-off for discriminating patients regarding the ability of pulse pressure variation to predict fluid responsiveness. If compliance of the respiratory system was >30 mL/cm H2O, then the area under the receiver-operating characteristics curve for predicting fluid responsiveness was not different for pulse pressure variation and the passive leg-raising and end-expiratory occlusion tests (0.98 ± 0.03, 0.91 ± 0.06, and 0.97 ± 0.03, respectively). By contrast, if compliance of the respiratory system was ⩽30 mL/cm H2O, then the area under the receiver-operating characteristics curve was significantly lower for pulse pressure variation than for the passive leg-raising and end-expiratory occlusion tests (0.69 ± 0.10, 0.94 ± 0.05, and 0.93 ± 0.05, respectively). Conclusions:The ability of pulse pressure variation to predict fluid responsiveness was inversely related to compliance of the respiratory system. If compliance of the respiratory system was ⩽30 mL/cm H2O, then pulse pressure variation became less accurate for predicting fluid responsiveness. However, the passive leg-raising and end-expiratory occlusion tests remained valuable in such cases.

Arterial pressure-based cardiac output in septic patients: different accuracy of pulse contour and uncalibrated pressure waveform devices

IntroductionWe compared the ability of two devices estimating cardiac output from arterial pressure-curve analysis to track the changes in cardiac output measured with transpulmonary thermodilution induced by volume expansion and norepinephrine in sepsis patients.MethodsIn 80 patients with septic circulatory failure, we administered volume expansion (40 patients) or introduced/increased norepinephrine (40 patients). We measured the pulse contour-derived cardiac index (CI) provided by the PiCCO device (CIpc), the arterial pressure waveform-derived CI provided by the Vigileo device (CIpw), and the transpulmonary thermodilution CI (CItd) before and after therapeutic interventions.ResultsThe changes in CIpc accurately tracked the changes in CItd induced by volume expansion (bias, -0.20 ± 0.63 L/min/m2) as well as by norepinephrine (bias, -0.05 ± 0.74 L/min/m2). The changes in CIpc accurately detected an increase in CItd ≥ 15% induced by volume expansion and norepinephrine introduction/increase (area under ROC curves, 0.878 (0.736 to 0.960) and 0.924 (0.795 to 0.983), respectively; P < 0.05 versus 0.500 for both). The changes in CIpw were less reliable for tracking the volume-induced changes in CItd (bias, -0.23 ± 0.95 L/min/m2) and norepinephrine-induced changes in CItd (bias, -0.01 ± 1.75 L/min/m2). The changes in CIpw were unable to detect an increase in CItd ≥ 15% induced by volume expansion and norepinephrine introduction/increase (area under ROC curves, 0.564 (0.398 to 0.720) and 0.541 (0.377 to 0.700, respectively, both not significantly different from versus 0.500).ConclusionsThe CIpc was reliable and accurate for assessing the CI changes induced by volume expansion and norepinephrine. By contrast, the CIpw poorly tracked the trends in CI induced by those therapeutic interventions.

Critical care management and outcome of severe Pneumocystis pneumonia in patients with and without HIV infection

BackgroundLittle is known about the most severe forms of Pneumocystis jiroveci pneumonia (PCP) in HIV-negative as compared with HIV-positive patients. Improved knowledge about the differential characteristics and management modalities could guide treatment based on HIV status.MethodsWe retrospectively compared 72 patients (73 cases, 46 HIV-positive) admitted for PCP from 1993 to 2006 in the intensive care unit (ICU) of a university hospital.ResultsThe yearly incidence of ICU admissions for PCP in HIV-negative patients increased from 1993 (0%) to 2006 (6.5%). At admission, all but one non-HIV patient were receiving corticosteroids. Twenty-three (85%) HIV-negative patients were receiving an additional immunosuppressive treatment. At admission, HIV-negative patients were significantly older than HIV-positive patients (64 [18 to 82] versus 37 [28 to 56] years old) and had a significantly higher Simplified Acute Physiology Score (SAPS) II (38 [13 to 90] versus 27 [11 to 112]) but had a similar PaO2/FiO2 (arterial partial pressure of oxygen/fraction of inspired oxygen) ratio (160 [61 to 322] versus 183 [38 to 380] mm Hg). Ventilatory support was required in a similar proportion of HIV-negative and HIV-positive cases (78% versus 61%), with a similar proportion of first-line non-invasive ventilation (NIV) (67% versus 54%). NIV failed in 71% of HIV-negative and in 13% of HIV-positive patients (p < 0.01). Mortality was significantly higher in HIV-negative than HIV-positive cases (48% versus 17%). The HIV-negative status (odds ratio 3.73, 95% confidence interval 1.10 to 12.60) and SAPS II (odds ratio 1.07, 95% confidence interval 1.02 to 1.12) were independently associated with mortality at multivariate analysis.ConclusionThe yearly incidence of ICU admissions for PCP in HIV-negative patients in our unit increased from 1993 to 2006. The course of the disease and the outcome were worse in HIV-negative patients. NIV often failed in HIV-negative cases, suggesting that NIV must be watched closely in this population.

Lactate and Venoarterial Carbon Dioxide Difference/Arterial-Venous Oxygen Difference Ratio, but Not Central Venous Oxygen Saturation, Predict Increase in Oxygen Consumption in Fluid Responders*

Objectives:During circulatory failure, the ultimate goal of treatments that increase cardiac output is to reduce tissue hypoxia. This can only occur if oxygen consumption depends on oxygen delivery. We compared the ability of central venous oxygen saturation and markers of anaerobic metabolism to predict whether a fluid-induced increase in oxygen delivery results in an increase in oxygen consumption. Design:Prospective study. Setting:ICU. Patients:Fifty-one patients with an acute circulatory failure (78% of septic origin). Measurements:Before and after a volume expansion (500mL of saline), we measured cardiac index, o2- and Co2-derived variables and lactate. Main Results:Volume expansion increased cardiac index ≥15% in 49% of patients (“volume-responders”). Oxygen delivery significantly increased in these 25 patients (+32% ± 16%, p < 0.0001). An increase in oxygen consumption ≥15% concomitantly occurred in 56% of these 25 volume-responders (+38% ± 28%). Compared with the volume-responders in whom oxygen consumption did not increase, the volume-responders in whom oxygen consumption increased ≥15% were characterized by a higher lactate (2.3 ± 1.1 mmol/L vs. 5.5 ± 4.0 mmol/L, respectively) and a higher ratio of the veno-arterial carbon dioxide tension difference (P(v − a)Co2) over the arteriovenous oxygen content difference (C(a − v)o2). A fluid-induced increase in oxygen consumption greater than or equal to 15% was not predicted by baseline central venous oxygen saturation but by high baseline lactate and (P(v − a)Co2/C(a − v)o2 ratio (areas under the receiving operating characteristics curves: 0.68 ± 0.11, 0.94 ± 0.05, and 0.91 ± 0.06). In volume-nonresponders, volume expansion did not significantly change cardiac index, but the oxygen delivery decreased due to a hemodilution-induced decrease in hematocrit. Conclusions:In volume-responders, unlike markers of anaerobic metabolism, central venous oxygen saturation did not allow the prediction of whether a fluid-induced increase in oxygen delivery would result in an increase in oxygen consumption. This suggests that along with indicators of volume-responsiveness, the indicators of anaerobic metabolism should be considered instead of central venous oxygen saturation for starting hemodynamic resuscitation.

Prediction of volume responsiveness in critically ill patients with spontaneous breathing activity.

Purpose of reviewPredicting volume responsiveness in patients with spontaneous breathing activity is a difficult challenge in the emergency room as well as in the intensive care unit because heart–lung interactions indices cannot be reliably used as they can be in mechanically ventilated patients fully adapted to their ventilator. The aim of this review is to summarize the different tools that have been proposed to predict the hemodynamic response to fluid infusion in the presence of spontaneous breathing activity. Recent findingsClinical studies recently demonstrated that neither indicators of cardiac preload (filling pressures and end-diastolic ventricular dimensions) nor arterial pulse pressure respiratory variation was an accurate predictor of volume responsiveness in patients with spontaneous breathing activity with or without mechanical support. In contrast, performing a passive leg-raising test has been proved as valuable for this purpose. SummaryThe passive leg-raising test is the only method that has been repeatedly shown to be reliable for predicting volume responsiveness in patients who experience spontaneous breathing. The appropriate utilization of this test requires a real-time assessment of its effects on systemic blood flow.