Acute hyperventilation increases the central venous-to-arterial PCO2 difference in stable septic shock patients.

BACKGROUND: To evaluate the effects of acute hyperventilation on the central venous-to-arterial carbon dioxide tension difference (∆PCO2) in hemodynamically stable septic shock patients. METHODS: Eighteen mechanically ventilated septic shock patients were prospectively included in the study. We measured cardiac index (CI), ∆PCO2, oxygen consumption (VO2), central venous oxygen saturation (ScvO2), and blood gas parameters, before and 30 min after an increase in alveolar ventilation (increased respiratory rate by 10 breaths/min). RESULTS: Arterial pH increased significantly (from 7.35 ± 0.07 to 7.42 ± 0.09, p < 0.001) and arterial carbon dioxide tension decreased significantly (from 44.5 [41-48] to 34 [30-38] mmHg, p < 0.001) when respiratory rate was increased. A statistically significant increase in VO2 (from 93 [76-105] to 112 [95-134] mL/min/m2, p = 0.002) was observed in parallel with the increase in alveolar ventilation. While CI remained unchanged, acute hyperventilation led to a significant increase in ∆PCO2 (from 4.7 ± 1.0 to 7.0 ± 2.6 mmHg, p < 0.001) and a significant decrease in ScvO2 (from 73 ± 6 to 67 ± 8%, p < 0.001). A good correlation was found between changes in arterial pH and changes in VO2 (r = 0.67, p = 0.002). Interestingly, we found a strong association between the increase in VO2 and the increase in ∆PCO2 (r = 0.70, p = 0.001). CONCLUSIONS: Acute hyperventilation provoked a significant increase in ∆PCO2, which was the result of a significant increase in VO2 induced by hyperventilation. The clinician should be aware of the effects of acute elevation of alveolar ventilation on ∆PCO2.

Impact of tapered-cuff tracheal tube on microaspiration of gastric contents in intubated critically ill patients: a multicenter cluster-randomized cross-over controlled trial.

PURPOSE: Studies on the impact of tapered-cuff tracheal tubes on rates of microaspiration and ventilator-associated pneumonia (VAP) in intubated patients have reported conflicting results. The aim of this study was to determine the influence of this shape of tracheal cuff on abundant microaspiration of gastric contents in critically ill patients. METHODS: All patients intubated in the intensive care unit (ICU) and requiring mechanical ventilation for at least 48 h were eligible for this multicenter cluster-randomized controlled cross-over open-label study. The primary outcome was abundant microaspiration of gastric contents, defined by the presence of pepsin at significant level in >30% of tracheal aspirates. Quantitative measurement of pepsin and salivary amylase was performed in all tracheal aspirates during the 48 h following enrollment. RESULTS: A total of 326 patients were enrolled in the ten participating ICUs (162 in the PVC tapered-cuff group and 164 in the standard-cuff group). Patient characteristics were similar in the two study groups. The proportion of patients with abundant microaspiration of gastric contents was 53.5% in the tapered-cuff and 51.0% in the standard-cuff group (odds ratio 1.14, 95% CI 0.72-1.82). While abundant microaspiration of oropharyngeal secretions was not significantly different (77.4 vs 68.6%, p = 0.095), the proportion of patients with tracheobronchial colonization was significantly lower (29.6 vs 43.3%, p = 0.01) in the tapered-cuff than in the standard-cuff group. No significant difference between the two groups was found for other secondary outcomes, including ventilator-associated events and VAP. CONCLUSIONS: This trial showed no significant impact of tapered-cuff tracheal tubes on abundant microaspiration of gastric contents. TRIAL REGISTRATION: ClinicalTrials.gov, number NCT01948635.

Ratios of central venous-to-arterial carbon dioxide content or tension to arteriovenous oxygen content are better markers of global anaerobic metabolism than lactate in septic shock patients.

BACKGROUND: To evaluate the ability of the central venous-to-arterial CO2 content and tension differences to arteriovenous oxygen content difference ratios (∆ContCO2/∆ContO2 and ∆PCO2/∆ContO2, respectively), blood lactate concentration, and central venous oxygen saturation (ScvO2) to detect the presence of global anaerobic metabolism through the increase in oxygen consumption (VO2) after an acute increase in oxygen supply (DO2) induced by volume expansion (VO2/DO2 dependence). METHODS: We prospectively studied 98 critically ill mechanically ventilated patients in whom a fluid challenge was decided due to acute circulatory failure related to septic shock. Before and after volume expansion (500 mL of colloid solution), we measured cardiac index, VO2, DO2, ∆ContCO2/∆ContO2 and ∆PCO2/∆ContO2 ratios, lactate, and ScvO2. Fluid-responders were defined as a ≥15 % increase in cardiac index. Areas under the receiver operating characteristic curves (AUC) were determined for these variables. RESULTS: Fifty-one patients were fluid-responders (52 %). DO2 increased significantly (31 ± 12 %) in these patients. An increase in VO2 ≥ 15 % ("VO2-responders") concurrently occurred in 57 % of the 51 fluid-responders (45 ± 16 %). Compared with VO2-non-responders, VO2-responders were characterized by higher lactate levels and higher ∆ContCO2/∆ContO2 and ∆PCO2/∆ContO2 ratios. At baseline, lactate predicted a fluid-induced increase in VO2 ≥ 15 % with AUC of 0.745. Baseline ∆ContCO2/∆ContO2 and ∆PCO2/∆ContO2 ratios predicted an increase of VO2 ≥ 15 % with AUCs of 0.965 and 0.962, respectively. Baseline ScvO2 was not able to predict an increase of VO2 ≥ 15 % (AUC = 0.624). CONCLUSIONS: ∆ContCO2/∆ContO2 and ∆PCO2/∆ContO2 ratios are more reliable markers of global anaerobic metabolism than lactate. ScvO2 failed to predict the presence of global tissue hypoxia.

Impact of tracheal cuff shape on microaspiration of gastric contents in intubated critically ill patients: study protocol for a randomized controlled trial.

BACKGROUND: Ventilator-associated pneumonia (VAP) is the most common infection in intubated critically ill patients. Microaspiration of the contaminated gastric and oropharyngeal secretions is the main mechanism involved in the pathophysiology of VAP. Tracheal cuff plays an important role in stopping the progression of contaminated secretions into the lower respiratory tract. Previous in vitro studies suggested that conical cuff shape might be helpful in improving tracheal sealing. However, clinical studies found conflicting results. The aim of this study is to determine the impact of conical tracheal cuff shape on the microaspiration of gastric contents in critically ill patients. METHODS/DESIGN: This prospective cluster randomized controlled crossover open-label trial is currently being conducted in ten French intensive care units (ICUs). Patients are allocated to intubation with a polyvinyl chloride (PVC) standard (barrel)-shaped or a PVC conical-shaped tracheal tube. The primary objective is to determine the impact of the conical shaped tracheal cuff on abundant microaspiration of gastric contents. Secondary outcomes include the incidence of microaspiration of oropharyngeal secretions, tracheobronchial colonization, VAP and ventilator-associated events. Abundant microaspiration is defined as the presence of pepsin at significant level (>200 ng/ml) in at least 30 % of the tracheal aspirates. Pepsin and amylase are quantitatively measured in all tracheal aspirates during the 48 h following inclusion. Quantitative tracheal aspirate culture is performed at inclusion and twice weekly. We plan to recruit 312 patients in the participating ICUs. DISCUSSION: BEST Cuff is the first randomized controlled study evaluating the impact of PVC tracheal-cuff shape on gastric microaspirations in patients receiving invasive mechanical ventilation. Enrollment began in June 2014 and is expected to end in October 2015. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT01948635 (registered 31 August 2013).

Determinants of noninvasive ventilation success or failure in morbidly obese patients in acute respiratory failure.

PURPOSE: Acute respiratory failure (ARF) is a common life-threatening complication in morbidly obese patients with obesity hypoventilation syndrome (OHS). We aimed to identify the determinants of noninvasive ventilation (NIV) success or failure for this indication. METHODS: We prospectively included 76 consecutive patients with BMI>40 kg/m2 diagnosed with OHS and treated by NIV for ARF in a 15-bed ICU of a tertiary hospital. RESULTS: NIV failed to reverse ARF in only 13 patients. Factors associated with NIV failure included pneumonia (n = 12/13, 92% vs n = 9/63, 14%; p<0.0001), high SOFA (10 vs 5; p<0.0001) and SAPS2 score (63 vs 39; p<0.0001) at admission. These patients often experienced poor outcome despite early resort to endotracheal intubation (in-hospital mortality, 92.3% vs 17.5%; p<0.001). The only factor significantly associated with successful response to NIV was idiopathic decompensation of OHS (n = 30, 48% vs n = 0, 0%; p = 0.001). In the NIV success group (n = 63), 33 patients (53%) experienced a delayed response to NIV (with persistent hypercapnic acidosis during the first 6 hours). CONCLUSIONS: Multiple organ failure and pneumonia were the main factors associated with NIV failure and death in morbidly obese patients in hypoxemic ARF. On the opposite, NIV was constantly successful and could be safely pushed further in case of severe hypercapnic acute respiratory decompensation of OHS.

Effects of sitting position and applied positive end-expiratory pressure on respiratory mechanics of critically ill obese patients receiving mechanical ventilation*.

OBJECTIVE: To evaluate the extent to which sitting position and applied positive end-expiratory pressure improve respiratory mechanics of severely obese patients under mechanical ventilation. DESIGN: Prospective cohort study. SETTINGS: A 15-bed ICU of a tertiary hospital. PARTICIPANTS: Fifteen consecutive critically ill patients with a body mass index (the weight in kilograms divided by the square of the height in meters) above 35 were compared to 15 controls with body mass index less than 30. INTERVENTIONS: Respiratory mechanics was first assessed in the supine position, at zero end-expiratory pressure, and then at positive end-expiratory pressure set at the level of auto-positive endexpiratory pressure. Second, all measures were repeated in the sitting position. MEASUREMENTS AND MAIN RESULTS: Assessment of respiratory mechanics included plateau pressure, auto-positive end-expiratory pressure, and flow-limited volume during manual compression of the abdomen, expressed as percentage of tidal volume to evaluate expiratory flow limitation. In supine position at zero end-expiratory pressure, all critically ill obese patients demonstrated expiratory flow limitation (flow-limited volume, 59.4% [51.3-81.4%] vs 0% [0-0%] in controls; p < 0.0001) and greater auto-positive end-expiratory pressure (10 [5-12.5] vs 0.7 [0.4-1.25] cm H2O in controls; p < 0.0001). Applied positive end-expiratory pressure reverses expiratory flow limitation (flow-limited volume, 0% [0-21%] vs 59.4% [51-81.4%] at zero end-expiratory pressure; p < 0.001) in almost all the obese patients, without increasing plateau pressure (24 [19-25] vs 22 [18-24] cm H2O at zero end-expiratory pressure; p = 0.94). Sitting position not only reverses partially or completely expiratory flow limitation at zero end-expiratory pressure (flow-limited volume, 0% [0-58%] vs 59.4% [51-81.4%] in supine obese patients; p < 0.001) but also results in a significant drop in auto-positive end-expiratory pressure (1.2 [0.6-4] vs 10 [5-12.5] cm H2O in supine obese patients; p < 0.001) and plateau pressure (15.6 [14-17] vs 22 [18-24] cm H2O in supine obese patients; p < 0.001). CONCLUSIONS: In critically ill obese patients under mechanical ventilation, sitting position constantly and significantly relieved expiratory flow limitation and auto-positive end-expiratory pressure resulting in a dramatic drop in alveolar pressures. Combining sitting position and applied positive end-expiratory pressure provides the best strategy.

Anaemia to predict outcome in patients with acute coronary syndromes.

BACKGROUND: Owing to the heterogeneous population of patients with acute coronary syndromes (ACS), risk stratification with tools such as the GRACE risk score is recommended to guide therapeutic management and improve outcome. AIM: To evaluate whether anaemia refines the value of the GRACE risk model to predict midterm outcome after an ACS. METHODS: A prospective registry of 1064 ACS patients (63 ± 14 years; 73% men; 57% ST-segment elevation myocardial infarction [MI]) was studied. Anaemia was defined as haemoglobin less than 13 mg/dL in men or less than 12 mg/dL in women. The primary endpoint was 6-month death or rehospitalization for MI. RESULTS: The primary endpoint was reached in 132 patients, including 68 deaths. Anaemia was associated with adverse clinical outcomes (hazard ratio 3.008, 95% confidence interval 2.137-4.234; P<0.0001) in univariate analysis and remained independently associated with outcome after adjustment for the Global Registry of Acute Coronary Events (GRACE) risk score (hazard ratio 2.870, 95% confidence interval 1.815-4.538; P<0.0001). Anaemia provided additional prognostic information to the GRACE score as demonstrated by a systematic improvement in global model fit and discrimination (c-statistic increasing from 0.633 [0.571;0.696] to 0.697 [0.638;0.755]). Subsequently, adding anaemia to the GRACE score led to reclassification of 595 patients into different risk categories; 16.5% patients at low risk (≤ 5% risk of death or rehospitalization for MI) were upgraded to intermediate (>5-10%) or high risk (>10%); 79.5% patients at intermediate risk were reclassified as low (55%) or high risk (24%); and 45.5% patients at high risk were downgraded to intermediate risk. Overall, 174 patients were reclassified into a higher risk category (17.3%) and 421 into a lower risk category (41.9%). CONCLUSION: Anaemia provides independent additional prognostic information to the GRACE score. Combining anaemia with the GRACE score refines its predictive value, which often overestimates the risk.

Use of sodium-chloride difference and corrected anion gap as surrogates of Stewart variables in critically ill patients.

INTRODUCTION: To investigate whether the difference between sodium and chloride ([Na(+)] - [Cl(-)]) and anion gap corrected for albumin and lactate (AG(corr)) could be used as apparent strong ion difference (SID(app)) and strong ion gap (SIG) surrogates (respectively) in critically ill patients. METHODS: A total of 341 patients were prospectively observed; 161 were allocated to the modeling group, and 180 to the validation group. Simple regression analysis was used to construct a mathematical model between SID(app) and [Na(+)] - [Cl(-)] and between SIG and AG(corr) in the modeling group. Area under the receiver operating characteristic (ROC) curve was also measured. The mathematical models were tested in the validation group. RESULTS: in the modeling group, SID(app) and SIG were well predicted by [Na(+)] - [Cl(-)] and AG(corr) (R(2) = 0.973 and 0.96, respectively). Accuracy values of [Na(+)] - [Cl(-)] for the identification of SID(app) acidosis (<42.7 mEq/L) and alkalosis (>47.5 mEq/L) were 0.992 (95% confidence interval [CI], 0.963-1) and 0.998 (95%CI, 0.972-1), respectively. The accuracy of AG(corr) in revealing SIG acidosis (>8 mEq/L) was 0.974 (95%CI: 0.936-0.993). These results were validated by showing excellent correlations and good agreements between predicted and measured SID(app) and between predicted and measured SIG in the validation group (R(2) = 0.977; bias = 0±1.5 mEq/L and R(2) = 0.96; bias = -0.2±1.8 mEq/L, respectively). CONCLUSIONS: SID(app) and SIG can be substituted by [Na(+)] - [Cl(-)] and by AG(corr) respectively in the diagnosis and management of acid-base disorders in critically ill patients.

Impact of functional decline on outcome in elderly patients with acute coronary syndromes.

BACKGROUND: Management of acute coronary syndromes in elderly patients is poorly defined. OBJECTIVE: To assess the impact of functional decline on all-cause mortality in elderly patients with acute coronary syndromes. METHOD: Clinical data, including the Global Registry of Acute Coronary Events score and assessment of functional status obtained by using the Katz scale, were prospectively collected on 272 patients 70 years or older hospitalized for acute coronary syndromes. All-cause mortality was assessed at 6 months, and longer term outcome data were obtained. RESULTS: Mean age of the patients was 78 years (SD, 6), and 58% were men. A total of 28% had functional decline. Six months after the index hospitalization, 38 patients had died. Another 29 patients died during a median follow-up of 611 days after the initial 6 months. Functional decline was associated with both 6-month (hazards ratio, 3.63; 95% CI, 1.91-6.88; P < .001) and long-term (hazards ratio, 2.69; 95% CI, 1.28-5.64; P = .009) outcomes. Functional decline remained associated with both 6-month and long-term outcomes in multivariate analysis and was systematically selected in the most predictive multivariate models for 6-month and long-term mortality. The multivariate model including the Global Registry of Acute Coronary Events score and functional decline was predictive of 6-month mortality, but the combination of functional decline and biological data was more predictive of long-term mortality than was a model combining functional decline and the Global Registry score. CONCLUSIONS: Functional decline in elderly patients with acute coronary syndromes is predictive of poor outcomes.

Rescue therapy by switching to total face mask after failure of face mask-delivered noninvasive ventilation in do-not-intubate patients in acute respiratory failure.

OBJECTIVE: To evaluate the impact of switching to total face mask in cases where face mask-delivered noninvasive mechanical ventilation has already failed in do-not-intubate patients in acute respiratory failure. DESIGN AND SETTING: Prospective observational study in an ICU and a respiratory stepdown unit over a 12-month study period. INTERVENTION: Switching to total face mask, which covers the entire face, when noninvasive mechanical ventilation using facial mask (oronasal mask) failed to reverse acute respiratory failure. PATIENTS: Seventy-four patients with a do-not-intubate order and treated by noninvasive mechanical ventilation for acute respiratory failure. MAIN RESULTS: Failure of face mask-delivered noninvasive mechanical ventilation was associated with a three-fold increase in in-hospital mortality (36% vs. 10.5%; p = 0.009). Nevertheless, 23 out of 36 patients (64%) in whom face mask-delivered noninvasive mechanical ventilation failed to reverse acute respiratory failure and, therefore, switched to total face mask survived hospital discharge. Reasons for switching from facial mask to total face mask included refractory hypercapnic acute respiratory failure (n = 24, 66.7%), painful skin breakdown or facial mask intolerance (n = 11, 30%), and refractory hypoxemia (n = 1, 2.7%). In the 24 patients switched from facial mask to total face mask because of refractory hypercapnia, encephalopathy score (3 [3-4] vs. 2 [2-3]; p < 0.0001), PaCO2 (87 ± 25 mm Hg vs. 70 ± 17 mm Hg; p < 0.0001), and pH (7.24 ± 0.1 vs. 7.32 ± 0.09; p < 0.0001) significantly improved after 2 hrs of total face mask-delivered noninvasive ventilation. Patients switched early to total face mask (in the first 12 hrs) developed less pressure sores (n = 5, 24% vs. n = 13, 87%; p = 0.0002), despite greater length of noninvasive mechanical ventilation within the first 48 hrs (44 hrs vs. 34 hrs; p = 0.05) and less protective dressings (n = 2, 9.5% vs. n = 8, 53.3%; p = 0.007). The optimal cutoff value for face mask-delivered noninvasive mechanical ventilation duration in predicting facial pressure sores was 11 hrs (area under the receiver operating characteristic curve, 0.86 ± 0.04; 95% confidence interval 0.76-0.93; p < 0.0001; sensitivity, 84%; specificity, 71%). CONCLUSION: In patients in hypercapnic acute respiratory failure, for whom escalation to intubation is deemed inappropriate, switching to total face mask can be proposed as a last resort therapy when face mask-delivered noninvasive mechanical ventilation has already failed to reverse acute respiratory failure. This strategy is particularly adapted to provide prolonged periods of continuous noninvasive mechanical ventilation while preventing facial pressure sores.

The repeatability of Stewart's parameters and anion gap in a cohort of critically ill adult patients.

PURPOSE: To examine the repeatability of Stewart's parameters and anion gap in a cohort of critically ill patients and to determine the smallest detectable changes in individual patients. METHODS: A total of 161 patients were included prospectively. They underwent two subsequent blood samplings within 10 min of each other and samples were analyzed using the same central laboratory analyzer. Measured and calculated parameters from the two samples were compared. The repeatability was expressed as the smallest detectable difference (SDD), coefficient of variation (CV) and intraclass correlation coefficient (ICC). RESULTS: The mean differences ± SD (mEq/L) for the repeated measurements were 0.1 ± 0.76, 0.12 ± 0.68, -0.02 ± 1.02, and -0.08 ± 1.05 for the apparent strong ion difference (SID(app)), effective strong ion difference (SID(eff)), strong ion gap (SIG), and albumin-corrected anion gap (AG(corr)), respectively. The SDDs (mEq/L) for SID(app), SID(eff), SIG, and AG(corr), were ±1.49, ±1.33, ±2, and ±2.06, respectively. The CVs (%) for these variables were 1.4, 1.45, 13.3, and 4.15, respectively. The ICCs for all these variables were high, largely above 0.75. CONCLUSIONS: The repeatability of all these calculated variables was good. In repeated measurements, a change in value of these parameters exceeding 1.96√2 CV (%), the least significant change (LSC) or the SDD should be regarded as significant. Use of SDD is preferable to CV and LSC (%) because of its independence from the levels of variables and its expression in absolute units. Expressed as SDD, a SIG change value, e.g., of at least ±2 mEq/L should be significant.

Extravascular lung water indexed or not to predicted body weight is a predictor of mortality in septic shock patients.

PURPOSE: The purpose was to investigate whether extravascular lung water (EVLW) indexed to actual body weight (EVLWa) is an independent predictor of mortality in patients with septic shock, to determine the relationship between EVLWa and other markers of lung injury, and to test if indexing EVLW with predicted body weight (EVLWp) strengthens its predictive power. METHODS: Extravascular lung water, pulmonary vascular permeability index, and other markers of lung injury were measured prospectively in 55 patients with septic shock for 3 days. RESULTS: At day 1, EVLWa, EVLWp, and pulmonary vascular permeability index were not significantly different between survivors and nonsurvivors. However, in parallel to the course of septic shock, these variables decreased only in the survivors and remained elevated in the nonsurvivors, reaching intergroup difference by day 3. In multiple logistic regression analysis, both EVLWa and EVLWp (at day 3) were predictors of mortality with an odds ratio of 2 (95% confidence interval, 1.12-3.7) and 1.7 (95% confidence interval, 1.1-2.5) per SD increase, respectively. The receiver operating characteristic curve analysis showed that EVLWp did not improve the discriminative power of EVLW to predict mortality. Extravascular lung water indexed to actual body weight correlated with lung injury score and with the ratio of arterial oxygen partial pressure to inspired oxygen fraction but not with static respiratory compliance. Indexing EVLW to predicted body weight did not ameliorate these correlations. CONCLUSIONS: Extravascular lung water indexed or not to predicted body weight is an independent predictor of mortality in patients with septic shock. Repeated measurements of EVLW indexes over time, rather than a too-early measurement, seem to be more appropriate for predicting outcome.

Addition of B-type natriuretic peptide to the GRACE score to predict outcome in acute coronary syndrome: a retrospective (development) and prospective (validation) cohort-based study.

AIMS: The present study was designed to build and validate a composite score based on the Global Registry of Acute Coronary Events (GRACE) score and B-type natriuretic peptide (BNP) concentrations to predict outcome in patients with acute coronary syndromes (ACS). METHODS: The GRACE risk score and BNP concentrations were obtained in a retrospective and a prospective cohort. A composite score including the GRACE score and BNP concentrations was first developed in a retrospective cohort of 248 patients with ACS and then validated in a prospective cohort of 575 patients. The primary outcome was 6-month death or myocardial infarction. RESULTS: End points were reached in 34 patients in the retrospective cohort and in 68 patients in the prospective cohort. Both higher BNP concentration and GRACE score were independently associated with outcome in the retrospective cohort (p=0.003 and p<0.0001). The composite score could be obtained as follows: GRACE+BNP/60. The use of the composite score increased the accuracy of the GRACE score, with an increase in the C statistic from 0.810 (0.727 to 0.892) to 0.822 (0.745 to 0.902) in the retrospective cohort and from 0.724 (0.657 to 0.791) to 0.750 (0.686 to 0.813) in the prospective cohort. Finally, 7% of patients in the prospective study population were reclassified from low to high risk or from high to low risk using this composite score. CONCLUSIONS: Plasma BNP levels refine the accuracy of the GRACE score. A comprehensive risk score, which includes BNP concentration and the GRACE risk score, might improve ACS risk stratification in clinical practice.

Prognostic value of hemoglobin decline over the GRACE score in patients hospitalized for an acute coronary syndrome.

In this work, we analyzed the prognostic significance of changes in hemoglobin during intensive care unit (ICU) stay in patients with acute coronary syndromes (ACS). We prospectively enrolled 591 patients (62 ± 14 years old, 73% male, 48% ST elevated myocardial infarction) free of blood cell transfusion or bleeding events. Changes in hemoglobin between admission and ICU discharge were obtained. The primary endpoint was death or hospitalization for MI within 6 months. Hemoglobin decreased from 13.65 ± 1.77 to 13.17 ± 1.74 g/dl, p < 0.0001 in the whole population. The end point was reached in 43 patients at a mean follow-up of 180 (range 2-180 days). A decrease in hemoglobin ≥0.9 g/dl (32% of the population) was associated with adverse clinical outcomes (HR 2.37, 95% CI (1.30-4.35), p = 0.005, respectively). In multivariate analysis, age >77 year-old (p = 0.0016), Killip class ≥2 (p = 0.009), anemia (p = 0.0064), decreased estimated glomerular filtration rate (p = 0.003), and hemoglobin decline ≥0.9 g/dl (p < 0.0001) were independently associated with outcome. Hemoglobin decline and anemia both provided additional prognostic information on top of the GRACE score, as demonstrated by a systematic improvement in model global fit, discrimination, and calibration. Hemoglobin decline is frequent during ICU stay in non-bleeding ACS patients. A decline in hemoglobin ≥0.9 g/dl identifies high-risk patients. Identification of these patients refines the prognostic value of the GRACE score.