Romain Pirracchio

Monocytic HLA-DR expression in intensive care patients: interest for prognosis and secondary infection prediction.

Objectives: To test early measurement of human leukocyte antigen‐DR expression on circulating monocytes (mHLA‐DR) as prognostic marker, and the trend of mHLA‐DR recovery for the prediction of late secondary infection risk in a large intensive care unit population. Design: Prospective, observational study over 16 mos. Setting: Intensive care unit in a tertiary teaching hospital. Inclusion criteria: Simplified Acute Physiology Score II >15, age >18 yrs. Measurements and Main Results: The mHLA‐DR was measured by flow cytometry within the first 3 days and twice a week until discharge. We used a logistic regression model for outcome prediction, and a competing risk approach to test the relationship between mHLA‐DR recovery (log (mHLA‐DR) slope) and incidence of secondary infection. A total of 283 consecutive patients suffering from various pathologies were monitored (Simplified Acute Physiology Score II = 39, Sepsis‐related Organ Failure Assessment of 5 on day 0). Early mHLA‐DR was decreased in the whole population, however, more deeply in sepsis (p < .0001). Low mHLA‐DR was associated with mortality in the whole population (p = .003), as in subgroups (nonseptic, neurologic, and septic), but not when adjusted on Simplified Acute Physiology Score II. In patients with a length of stay of >7 days (n = 70), the lower the slope of mHLA‐DR recovery, the higher the incidence of the first secondary infection (adjusted on early mHLA‐DR, p = .04). Conclusions: For a given severity, mHLA‐DR proved not to a predictive marker of outcome, but a weak trend of mHLA‐DR recovery was associated with an increased risk of secondary infection. Monitoring immune functions through mHLA‐DR in intensive care unit patients therefore could be useful to identify a high risk of secondary infection.

The Eldicus prospective, observational study of triage decision making in European intensive care units. Part II: intensive care benefit for the elderly.

Rationale:Life and death triage decisions are made daily by intensive care unit physicians. Admission to an intensive care unit is denied when intensive care unit resources are constrained, especially for the elderly. Objective:To determine the effect of intensive care unit triage decisions on mortality and intensive care unit benefit, specifically for elderly patients. Design:Prospective, observational study of triage decisions from September 2003 until March 2005. Setting:Eleven intensive care units in seven European countries. Patients:All patients >18 yrs with an explicit request for intensive care unit admission. Interventions:Admission or rejection to intensive care unit. Measurements and Main Results:Demographic, clinical, hospital, physiologic variables, and 28-day mortality were obtained on consecutive patients. There were 8,472 triages in 6,796 patients, 5,602 (82%) were accepted to the intensive care unit, 1,194 (18%) rejected; 3,795 (49%) were ≥65 yrs. Refusal rate increased with increasing patient age (18–44: 11%; 45–64: 15%; 65–74: 18%; 75–84: 23%; >84: 36%). Mortality was higher for older patients (18–44: 11%; 45–64: 21%; 65–74: 29%; 75–84: 37%; >84: 48%). Differences between mortalities of accepted vs. rejected patients, however, were greatest for older patients (18–44: 10.2% vs. 12.5%; 45–64: 21.2% vs. 22.3%; 65–74: 27.9% vs. 34.6%; 75–84: 35.5% vs. 40.4%; >84: 41.5% vs. 58.5%). Logistic regression showed a greater mortality reduction for accepted vs. rejected patients corrected for disease severity for elderly patients (age >65 [odds ratio 0.65, 95% confidence interval 0.55–0.78, p < .0001]) than younger patients (age <65 [odds ratio 0.74, 95% confidence interval 0.57–0.97, p = .01]). Conclusions:Despite the fact that elderly patients have more intensive care unit rejections than younger patients and have a higher mortality when admitted, the mortality benefit appears greater for the elderly. Physicians should consider changing their intensive care unit triage practices for the elderly. (Crit Care Med 2012; 40:132–138)

Fluid Resuscitation Does Not Improve Renal Oxygenation during Hemorrhagic Shock in Rats

Background:The resuscitation strategy for hemorrhagic shock remains controversial, with the kidney being especially prone to hypoxia. Methods:The authors used a three-phase hemorrhagic shock model to investigate the effects of fluid resuscitation on renal oxygenation. After a 1-h shock phase, rats were randomized into four groups to receive either normal saline or hypertonic saline targeting a mean arterial pressure (MAP) of either 40 or 80 mmHg. After such resuscitation, rats were transfused with the shed blood. Renal macro- and microcirculation were monitored with cortical and outer-medullary microvascular oxygen pressure, renal oxygen delivery, and renal oxygen consumption measured using oxygen-dependent quenching of phosphorescence. Results:Hemorrhagic shock was characterized by a drop of aortic blood flow, MAP, renal blood flow, renal oxygen delivery, renal oxygen consumption, and renal microvascular PO2. During the fluid resuscitation phase, normal saline targeting a MAP = 80 mmHg was the sole strategy able to restore aortic blood flow, renal blood flow, and renal oxygen consumption, although without improving renal oxygen delivery. However, none of the strategies using either normal saline or hypertonic saline or targeting a high MAP could restore the renal microvascular Po2. Blood transfusion increased microvascular Po2 but was unable to totally restore renal microvascular oxygenation to baseline values. Conclusions:This experimental rat study shows that (1) high MAP-directed fluid resuscitation (80 mmHg) does not lead to higher renal microvascular Po2 compared with fluid resuscitation targeted to MAP (40 mmHg); (2) hypertonic saline is not superior to normal saline regarding renal oxygenation; and (3) decreased renal oxygenation persists after blood transfusion.

The Eldicus prospective, observational study of triage decision making in European intensive care units: Part I—European Intensive Care Admission Triage Scores*

Objective:Life and death triage decisions are made daily by intensive care unit physicians. Scoring systems have been developed for prognosticating intensive care unit mortality but none for intensive care unit triage. The objective of this study was to develop an intensive care unit triage decision rule based on 28-day mortality rates of admitted and refused patients. Design:Prospective, observational study of triage decisions from September 2003 until March 2005. Setting:Eleven intensive care units in seven European countries. Patients:All patients >18 yrs with a request for intensive care unit admission. Interventions:Admission or rejection to an intensive care unit. Measurements and Main Results:Clinical, laboratory, and physiological variables and data from severity scores were collected. Separate scores for accepted and rejected patients with 28-day mortality end point were built. Values for variables were grouped into categories determined by the locally weighted least squares graphical method applied to the logit of the mortality and by univariate logistic regressions for reducing candidates for the score. Multivariate logistic regression was used to construct the final score. Cutoff values for 99.5% specificity were determined. Of 6796 patients, 5602 were admitted and 1194 rejected. The initial refusal score included age, diagnosis, systolic blood pressure, pulse, respirations, creatinine, bilirubin, PaO2, bicarbonate, albumin, use of vasopressors, Glasgow Coma Scale score, Karnofsky Scale, operative status and chronic disorder, and the initial refusal receiver operating characteristics were area under the curve 0.77 (95% confidence interval 0.76–0.79). The final triage score included age, diagnosis, creatinine, white blood cells, platelets, albumin, use of vasopressors, Glasgow Coma Scale score, Karnofsky Scale, operative status and chronic disorder, and the final score receiver operating characteristics were area under the curve 0.83 (95% confidence interval 0.80–0.86). Patients with initial refusal scores >173.5 or final triage scores = 0 should be rejected. Conclusions:The initial refusal score and final triage score provide objective data for rejecting patients that will die even if admitted to the intensive care unit and survive if refused intensive care unit admission. (Crit Care Med 2012; 40:125–131)

Impaired plasma B-type natriuretic peptide clearance in human septic shock.

Introduction:High B-type natriuretic peptide (BNP) levels are reported in the context of septic shock. We hypothesized that high BNP levels might be related to an alteration in BNP clearance pathway, namely neutral endopeptidase (NEP) 24.11. NEP 24.11 activity was measured in septic shock and in cardiogenic shock patients. We further evaluated whether baseline plasma BNP can predict fluid responsiveness and whether BNP can still be released in plasma despite high initial BNP levels, in response to overloading. Material and Methods:Prospective observational study. Patients in severe sepsis (S) or in septic shock (SS) needing a fluid challenge were included. Stroke volume (SV) and BNP were measured before (SV1, BNP1) and 45 mins after (SV2, BNP2) a standardized fluid challenge. DeltaBNP was defined as the difference between BNP2 and BNP1. NEP 24.11 activity was determined by fluorometry in 12 SS and 4 S patients before fluid challenge and in 5 cardiogenic shock patients. Results:Twenty-three patients (61 ± 18 years old, Simplified Acute Physiology Score II: 54 ± 21; 19 SS, 4 S; BNP1: 1371 ± 1434 pg/mL) were studied. BNP1 concentrations were significantly higher in SS than in S (1643 ± 1437 vs. 80 ± 35 pg/mL; p = 0.002). There was no correlation between baseline BNP and fluid responsiveness. Nine of the 11 patients with BNP1 >1000 pg/mL were fluid responders. DeltaBNP was greater in fluid nonresponders than in fluid responders (22 ± 27% vs. 6 ± 11%, p = 0.028). Plasma BNP was higher in SS than in cardiogenic shock patients (1367 ± 1438 vs. 750 ± 346 respectively; p = 0.027). NEP 24.11 activity was lower in SS than in S patients (0.10 ± 0.06 nmole/mL/min vs. 0.50 ± 0.22 nmole/mL/min, p <0.0001) cardiogenic shock patients (0.10 ± 0.06 nmole/mL/min vs. 0.58 ± 0.19 nmole/mL/min; p = 0.002). Conclusion:High levels of BNP might be related to an alteration in BNP clearance. During sepsis, high BNP levels are not predictive of fluid nonresponsiveness. Nevertheless, in fluid nonresponders, acute ventricular stretching can result in further BNP release.

Evaluation of the Propensity score methods for estimating marginal odds ratios in case of small sample size

BackgroundPropensity score (PS) methods are increasingly used, even when sample sizes are small or treatments are seldom used. However, the relative performance of the two mainly recommended PS methods, namely PS-matching or inverse probability of treatment weighting (IPTW), have not been studied in the context of small sample sizes.MethodsWe conducted a series of Monte Carlo simulations to evaluate the influence of sample size, prevalence of treatment exposure, and strength of the association between the variables and the outcome and/or the treatment exposure, on the performance of these two methods.ResultsDecreasing the sample size from 1,000 to 40 subjects did not substantially alter the Type I error rate, and led to relative biases below 10%. The IPTW method performed better than the PS-matching down to 60 subjects. When N was set at 40, the PS matching estimators were either similarly or even less biased than the IPTW estimators. Including variables unrelated to the exposure but related to the outcome in the PS model decreased the bias and the variance as compared to models omitting such variables. Excluding the true confounder from the PS model resulted, whatever the method used, in a significantly biased estimation of treatment effect. These results were illustrated in a real dataset.ConclusionEven in case of small study samples or low prevalence of treatment, PS-matching and IPTW can yield correct estimations of treatment effect unless the true confounders and the variables related only to the outcome are not included in the PS model.

Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study

BACKGROUND Improved mortality prediction for patients in intensive care units is a big challenge. Many severity scores have been proposed, but findings of validation studies have shown that they are not adequately calibrated. The Super ICU Learner Algorithm (SICULA), an ensemble machine learning technique that uses multiple learning algorithms to obtain better prediction performance, does at least as well as the best member of its library. We aimed to assess whether the Super Learner could provide a new mortality prediction algorithm for patients in intensive care units, and to assess its performance compared with other scoring systems. METHODS From January, 2001, to December, 2008, we used the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) database (version 26) including all patients admitted to an intensive care unit at the Beth Israel Deaconess Medical Centre, Boston, MA, USA. We assessed the calibration, discrimination, and risk classification of predicted hospital mortality based on Super Learner compared with SAPS-II, APACHE-II, and SOFA. We calculated performance measures with cross-validation to avoid making biased assessments. Our proposed score was then externally validated on a dataset of 200 randomly selected patients admitted at the intensive care unit of Hôpital Européen Georges-Pompidou, Paris, France, between Sept 1, 2013, and June, 30, 2014. The primary outcome was hospital mortality. The explanatory variables were the same as those included in the SAPS II score. FINDINGS 24,508 patients were included, with median SAPS-II of 38 (IQR 27-51) and median SOFA of 5 (IQR 2-8). 3002 of 24,508 (12%) patients died in the Beth Israel Deaconess Medical Centre. We produced two sets of predictions based on the Super Learner; the first based on the 17 variables as they appear in the SAPS-II score (SL1), and the second, on the original, untransformed variables (SL2). The two versions yielded average predicted probabilities of death of 0·12 (IQR 0·02-0·16) and 0·13 (0·01-0·19), whereas the corresponding value for SOFA was 0·12 (0·05-0·15) and for SAPS-II 0·30 (0·08-0·48). The cross-validated area under the receiver operating characteristic curve (AUROC) for SAPS-II was 0·78 (95% CI 0·77-0·78) and 0·71 (0·70-0·72) for SOFA. Super Learner had an AUROC of 0·85 (0·84-0·85) when the explanatory variables were categorised as in SAPS-II, and of 0·88 (0·87-0·89) when the same explanatory variables were included without any transformation. Additionally, Super Learner showed better calibration properties than previous score systems. On the external validation dataset, the AUROC was 0·94 (0·90-0·98) and calibration properties were good. INTERPRETATION Compared with conventional severity scores, Super Learner offers improved performance for predicting hospital mortality in patients in intensive care units. A user-friendly implementation is available online and should be useful for clinicians seeking to validate our score. FUNDING Fulbright Foundation, Assistance Publique-Hôpitaux de Paris, Doris Duke Clinical Scientist Development Award, and the NIH.

Implications of ICU triage decisions on patient mortality: a cost-effectiveness analysis

IntroductionIntensive care is generally regarded as expensive, and as a result beds are limited. This has raised serious questions about rationing when there are insufficient beds for all those referred. However, the evidence for the cost effectiveness of intensive care is weak and the work that does exist usually assumes that those who are not admitted do not survive, which is not always the case. Randomised studies of the effectiveness of intensive care are difficult to justify on ethical grounds; therefore, this observational study examined the cost effectiveness of ICU admission by comparing patients who were accepted into ICU after ICU triage to those who were not accepted, while attempting to adjust such comparison for confounding factors.MethodsThis multi-centre observational cohort study involved 11 hospitals in 7 EU countries and was designed to assess the cost effectiveness of admission to intensive care after ICU triage. A total of 7,659 consecutive patients referred to the intensive care unit (ICU) were divided into those accepted for admission and those not accepted. The two groups were compared in terms of cost and mortality using multilevel regression models to account for differences across centres, and after adjusting for age, Karnofsky score and indication for ICU admission. The analyses were also stratified by categories of Simplified Acute Physiology Score (SAPS) II predicted mortality (< 5%, 5% to 40% and >40%). Cost effectiveness was evaluated as cost per life saved and cost per life-year saved.ResultsAdmission to ICU produced a relative reduction in mortality risk, expressed as odds ratio, of 0.70 (0.52 to 0.94) at 28 days. When stratified by predicted mortality, the odds ratio was 1.49 (0.79 to 2.81), 0.7 (0.51 to 0.97) and 0.55 (0.37 to 0.83) for <5%, 5% to 40% and >40% predicted mortality, respectively. Average cost per life saved for all patients was

Effect of Levosimendan on Low Cardiac Output Syndrome in Patients With Low Ejection Fraction Undergoing Coronary Artery Bypass Grafting With Cardiopulmonary Bypass: The LICORN Randomized Clinical Trial

103,771 (€82,358) and cost per life-year saved was

Extracorporeal life support (ECLS) for refractory cardiac arrest after drowning: an 11-year experience.

7,065 (€5,607). These figures decreased substantially for patients with predicted mortality higher than 40%,