In a randomized factorial, double-blinded, multicenter trial, consecutive adults admitted to an intensive care unit (ICU) and receiving mechanical ventilation were enrolled. Additionally, patients were required to have two or more organ failures related to their critical illness. Subjects were randomized to one of four groups: placebo, glutamine, antioxidants (selenium, beta carotene, vitamins E and C), or antioxidants plus glutamine. The primary outcome was 28-day mortality; several secondary outcomes were also studied. 

There were 1,223 patients randomized to one of four groups. All were followed appropriately, and an intention-to-treat analysis was conducted. No statistically significant baseline differences were detected among the four groups. Overall 28-day mortality was 29.8%.  Patients who received glutamine had a trend towards worse mortality at 28 days compared to patients who did not receive glutamine (32.4% vs. 27.2%; adjusted odds ratio [AOR], 1.28; 95% confidence interval [CI], 1.00 to 1.64; P=0.05). The combination of glutamine with antioxidants made no significant difference in survival (AOR, 1.09; 95% CI, 0.86 to 1.40; P=0.48). In-hospital mortality and mortality at 6 months were significantly higher among patients who received glutamine; they also had increases in ICU length of stay, length of mechanical ventilation, and overall hospital length of stay. Antioxidant supplementation had no effect on secondary outcomes.

The findings from this large, robust randomized study are in opposition to those of previous smaller studies. In this study, a majority of patients received enteral nutrition support and were in multiple organ failure. Moreover, only 1.3% to 3.0 % of patients were not in some form of shock. Thus, it is possible that glutamine and antioxidants may not be as efficacious in a population of seriously ill patients. The lack of benefit with antioxidants may also be explained by the patient population; these agents may be more efficacious in areas with a higher prevalence of antioxidant deficiency (e.g., selenium). For critically ill ICU patients in multiple organ failure, this work suggests that glutamine administration, with or without antioxidants, does not confer a therapeutic benefit and may cause harm.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Samples were taken within the first two weeks of life, either during the transitional period (days 1-3) or post-transitional period (days 4-14). In all, there were 103 paired blood gases and echocardiograms from 21 subjects in the transitional period and 44 paired studies from 15 subjects in the post-transitional period. Each echocardiographic examination included measurements for shortening fraction (which is a load-dependent measure of contractility), stress-velocity index (which is a load-independent measure of contractility), left and right ventricular output, assessment of the ductus arteriosus, myocardial performance index, and SVR.

The results of the study demonstrated differences in the way the older infants responded to acidosis and hypercapnia. During the transitional period, pH and Paco₂ had no effect on any of the hemodynamic measures, including shortening fraction, stress velocity index, and SVR. However, during the post-transitional period, the infants’ hemodynamics resembled those seen in adult patients. In this set of subjects, SVR dropped as the infants became more acidotic and had higher Pco₂ levels. Unlike in the adult population, no relationship was demonstrated between the shortening fraction and stress velocity index and pH or Pco₂.

This study has some limitations. Echocardiography in itself is limited in its ability to assess SVR, and the authors assumed a right atrial pressure of 4 mm HG in performing the calculations. Also, as the authors noted, distribution of subjects in the two groups was unbalanced. Finally, all subjects were hemodynamically stable, so the results may not apply to infants in shock.

Importantly, the authors raise a question that is clinically relevant, as it is still common to have buffers rapidly administered to infants who are acidotic and have myocardial dysfunction. This study would suggest that, other than providing more fluid, this therapy is not warranted. In conclusion, more studies are needed to demonstrate the relationship between acidosis (both metabolic and respiratory) and hemodynamics in children.

This Concise Critical Appraisal is authored by SCCM member Daniel E. Sloniewsky, MD. Each installment highlights journal articles most relevant to the critical care practitioner.

This prospective, cluster-randomized controlled trial was performed in the prehospital environment in France from 2009 to 2011. At least one first-degree relative per patient was included. Control groups consisted of emergency medical teams that did not routinely give the option for family members to be present during CPR. The experimental group used a communication guide to invite the relative to the resuscitation scene. Ninety days after resuscitation, a trained psychologist conducted a series of assessments for anxiety, depression and PTSD-related symptoms. The primary end point was the proportion of relatives with PTSD-related symptoms, while secondary end points included the effect of family presence on medical efforts at resuscitation, the well-being of the healthcare team, and the filing of medicolegal claims.
 
Of the 570 family members enrolled in an intention-to-treat analysis, 79% of the experimental group witnessed resuscitation versus 43% in the control group. Seventeen percent (n=95) of family members did not complete the 90-day post-event psychological assessment. There were no significant differences between the two groups in the characteristics of the resuscitation procedure, survival or characteristics of patients or enrolled family members. Twenty-seven percent of family members in the experimental group had PTSD-related symptoms versus 41% in the control group (P=0.01). Anxiety was significantly higher in the control group (24%) than in the intervention group (16%; P<0.001). Less than 1% of family members in either group interfered with the resuscitation effort.

One significant limitation to this work was the prehospital environment; the results from this trial deserve replication in an intensive care unit. It is possible that social and cultural norms and religious differences may pose a threat to external generalizability in other countries. The study results, which demonstrate several beneficial effects when family members are allowed to be present during CPR, require further investigation in the intensive care unit.  

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

The authors used propensity score analysis to limit the confounding by indication or treatment selection bias that occurs when one attempts to distinguish a risk factor or a risk marker. In his accompanying editorial, Lacroix comments that the relationships between anemia, transfusion and severity of illness are so intertwined that only a randomized trial can determine the causal nature of transfusion and increased morbidity. However, the first step in determining whether a variable is the cause of morbidity involves the completion of a descriptive study that can establish significant associations.

Kneyber and colleagues retrospectively analyzed data from 335 children between the ages of 0 months to 18 years who underwent cardiac surgery between 2007 and 2010. Of these subjects, 111 were transfused, 86 of them within the first 48 hours of admission to the pediatric intensive care unit. The authors then compared the outcomes of these 86 patients to those of the 249 who were not transfused within 48 hours. Without adjusting for severity of illness, the patients who received pRBCs within the first 48 hours had a longer duration of ventilation and inotropic support, longer ICU stay, and a higher rate of ventilator-associated pneumonia. However, after adjusting for the severity of illness, the transfusion of pRBCs within the first 48 hours remained independently associated with only the prolonged duration of mechanical ventilation.

This study was not designed to explain the mechanisms behind these findings, though the immunomodulatory effects of transfused blood have been reported in the past and are cited in this work. For example, Cholette et al recently provided evidence of this in their prospective trial demonstrating increased interleukin-6 levels (a pro-inflammatory cytokine) in pediatric patients who underwent cardiopulmonary bypass and received unwashed pRBCs.

This retrospective, single-center observational study demonstrating the association between increased ventilator days and early transfusion in pediatric cardiac patients provides more evidence against arbitrary decisions to transfuse blood products. Prospective studies are necessary to determine which transfusion triggers will provide more benefit than cost.

This Concise Critical Appraisal is authored by SCCM member Daniel E. Sloniewsky, MD. Each installment highlights journal articles most relevant to the critical care practitioner.

A systematic review of the literature was conducted using several major data sources. No language restrictions were applied. All studies had to include lower respiratory tract sampling before any clinical suspicion of VAP. Studies were assessed according to current Quality Assessment of Studies of Diagnostic Accuracy Included in Systematic Reviews (QUADAS) guidelines. A mixed-effect regression model was calculated to perform a meta-analysis. A Bayes nomogram was constructed, and heterogeneity was assessed with forest plot analyses and calculation of the I² statistic.

Five hundred twenty-five studies were retrieved; 14 studies were ultimately included in the meta-analysis, representing 791 VAP episodes in 688 patients. The incidence of VAP was only reported in three studies, and ranged from 10.4/1000 to 55/1000. In nine studies, the prevalence of VAP was reported and ranged between 7.3% and 44.6%. Eight studies used endotracheal aspirates, five used protected specimen brushes, and bronchoalveolar lavage (BAL) was used in ten studies. High accuracy was observed (area under the curve [AUC] =0.92; 95% confidence interval [CI], 0.89-0.94) when all variables (i.e., true positives/negatives, false positives/negatives, etc.) were included for the prediction of bacterial pathogens. The pooled sensitivity for all studies was 75% (95% CI, 65% to 83%) and overall specificity was 92% (95% CI, 85% to 96%). The ability to predict MDR organisms was good (AUC 0.95), but only four studies could be used to calculate the AUC, and heterogeneity was significant (I²=63%). Twice-a-week sampling resulted in better sensitivity and specificity compared to less frequent sampling.

This study is the first to provide coherent evidence for the accuracy of surveillance cultures to predict VAP. The absence of MDR organisms on surveillance cultures may be reassuring in some cases as most studies reported moderate to high specificity and accuracy.  Some studies may have been missed because the authors did not examine other databases, such as EMBASE or grey literature. Additionally, the incidence and prevalence of VAP was not reported consistently in all studies, and the range was wide. Various techniques for lower respiratory tract sampling were employed, possibly biasing the sensitivity and specificity of the cultures. Heterogeneity was high for the accuracy of VAP prediction, with I² values ranging from 0% to 92%. These differences in case mix may affect the external generalizability of the findings for specific patient populations. The interstudy variation, limited number of studies and cited limitations preclude formulating recommendations regarding routine lower respiratory surveillance cultures for the prediction of VAP. 

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

The authors conducted a systematic review that included critically ill adults in both medical and surgical ICUs. The primary outcome of interest was bleeding, with secondary outcomes including nosocomial pneumonia, all-cause ICU mortality, ICU length of stay, and Clostridium difficile infection. A comprehensive search strategy and rigorous Cochrane-based risk of bias assessment were executed. Data were analyzed with a random effects model, and pooled relative risks were calculated.  Several subgroup and sensitivity analyses were performed.

After screening 1,215 titles and abstracts, fourteen randomized trials were found, including a total of 1,720 patients. PPIs were associated with a lower risk of clinically important bleeding compared to H2RAs (risk ratio [RR] 0.36; 95% confidence interval [CI], 0.19-0.68, P=0.002; I2=0%). The number needed for prophylaxis was estimated to be 78 using a control event rate of 2%. PPIs were also associated with a lower risk of overt bleeding compared to H2RAs, with a number needed for prophylaxis of 30 using a control event rate of 5%. No statistically significant differences between PPIs and H2RAs were found for the outcomes of nosocomial pneumonia, mortality or ICU length of stay. No studies reported the incidence of C. difficile infection. A sensitivity analysis that excluded abstracts yielded results that were consistent with the primary analysis.

The overall methodology employed for this study was outstanding, and this study should serve as a model of how to properly conduct and report a systematic review and meta-analysis. A few limitations were openly acknowledged by the authors. First, some degree of publication bias was evident; small negative trials appeared to be missing. Second, the overall quality as assessed by the Grading of Recommendations Assessment, Development and Evaluation approach was low to moderate for the primary and secondary outcomes, with serious risk of bias for all outcomes. Third, the treatment effect appeared to be smaller in trials of higher quality.  Although all the included studies were randomized, elements of the randomized design, such as blinding, were missing in several of the trials, thus potentially inflating the observed benefit of PPIs. Nevertheless, based on this study, the highest quality available evidence suggests that PPIs may lower the risk of both clinically important and overt gastrointestinal bleeding in ICU patients without increasing the risk of nosocomial pneumonia.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

A native of Chicago, Illinois, Shoemaker graduated from the University of California, San Francisco Medical School. After three years in the U.S. Navy, he completed his general surgical residency at Hahnemann Medical College in Philadelphia, followed by a fellowship in surgery and biochemistry at Harvard Medical School.

His career took him from Boston to Chicago, where he instituted and directed the trauma service at Cook County Hospital and was a Professor of Surgery at the University of Illinois. He then moved to the Mount Sinai School of Medicine, where he was director of the Division of Surgical Metabolism and instituted and directed the multiprofessional intensive care unit as well as the Shock and Trauma Service at Elmhurst Municipal Hospital.

By the 1960s, Shoemaker has already established himself internationally as a leading force in the field of multidisciplinary critical care. He was a major contributor to the advancements in the science of critical illness, including clinical research into hemodynamic abnormalities, from mechanism of injury to an organized approach of treatment. He researched the pathophysiology of trauma and the patient’s response to optimal treatment, initially trying to establish normal physiology in the hope of better outcomes. In addition to his scientific accomplishments, he was also a key voice in the effort to have critical care considered a specialized area of medicine, and a staunch believer that optimal critical care was delivered by a multidisciplinary team.

Shoemaker’s capacity for conceptual thinking no doubt helped with the formation of the Society of Critical Care Medicine. His wife, Norma J. Shoemaker, RN, MN, FCCM, became the organization’s first executive director, running the SCCM “office” from a desk in their Anaheim, California, home. The couple played a vital role in the growth of the Society and the lives of others.

One of Shoemaker’s most impressive undertakings was serving as the first editor of Critical Care Medicine. In 1972, with a subscription base of only 100 SCCM members, he was instrumental in working with a publisher to found the journal and solicit articles for the then bi-monthly publication. He remained the editor until 1991, by which time the journal was published monthly as one of the most well respected resources in the field. He is well regarded as a scientist, an organizer, and an educator. He is respected and admired by all who meet him.

Please take a moment to wish Dr. Shoemaker a happy birthday!

Patients were enrolled from multiple centers in Canada, Saudi Arabia, Chile, India, and the United Statesfrom 2009 to 2012. They were eligible if they had acute ARDS with an Fio₂ ≥0.5. All were started on pressure-controlled ventilation with tidal volumes of 6 mL/kg, Fio₂ of 0.60, and positive end-expiratory pressure of 10 cm H₂O. After 30 minutes, if the Pao₂:Fio₂ ratio remained at 200 or lower, patients underwent randomization to one of two groups. Those in the control group were treated with conventional ventilation targeting a tidal volume of 6 mL/kg and plateau pressure <35 cm H₂O. In the intervention group, patients were treated with HFOV after a recruitment maneuver, targeting the highest possible frequency to maintain the Pao₂ >55 mm Hg and pH above 7.25. The primary outcome was in-hospital mortality. Logistic regression was used to compare outcomes while adjusting for pre-specified baseline variables. A survival analysis was performed to compare the two groups with respect to the time of death.

After 548 patients had been enrolled, the study was terminated due to a higher rate of death in the HFOV group;  47% of patients in the HFOV group died compared with 35% in the control group (relative risk [RR] of death with HFOV, 1.33; 95% confidence interval [CI], 1.09-1.64; P=0.005). Over the course of the study, patients in the HFOV group required more vasopressors compared to the control group (91% vs. 84%, P<0.01) and more neuromuscular blockers (83% vs. 68%, P<0.001). Rates of co-interventions, such as steroids, prone positioning and dialysis, were similar in the two groups.

The authors concluded that among patients with moderate to severe ARDS, early application of HFOV was associated with higher mortality compared to a ventilation strategy that used small tidal volumes and high positive end-expiratory pressure. Previous positive HFOV trials have been scrutinized for the use of inferior and outdated ventilation modes in the comparator arms, as well as small sample sizes. In this trial, HFOV appeared to be associated with hemodynamic compromise, possibly due to the higher mean airway pressures associated with this modality. HFOV strategies that use lower mean airway pressures, different inspiratory-to-expiratory ratios, and lower oscillatory frequencies might lead to different results. Moreover, the sedation strategy used for HFOV may partially explain the disparate outcomes results. A very small proportion of trauma patients (n=15) were enrolled, and it is possible that HFOV may benefit other specific patient populations or patients with different risk factors for ARDS. It is not clear whether expertise with HFOV was uniform among the numerous international research sites. Based on the results of this study, HFOV may be best reserved for patients with refractory hypoxemia. Reliance on an established lung-protective strategy (i.e., ARDSNet-based) was demonstrated to be superior in this trial for the early treatment of moderate to severe ARDS.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

This multicenter, randomized, noninferiority trial included mechanically ventilated adults admitted to one of nine ICUs in France. Patients in the intervention group were fed enterally via a nasogastric tube, and nutrition was stopped only when vomiting occurred. The control group consisted of patients fed via a nasogastric tube, with cessation of feeding whenever either vomiting or GRV exceeded 250 mL. The primary endpoint was the incidence of VAP. Secondary endpoints included evaluations of whether absence of GRV monitoring affected enteral nutrition delivery. Additional patient outcomes were also assessed. Intent-to-treat (ITT) and per-protocol population analyses were conducted, and the number of VAP episodes was evaluated using negative binomial regression. 

Four hundred forty-nine patients were included in the primary ITT analysis and 423 were included in the per-protocol analysis. In the ITT population, 16.7% in the intervention group developed VAP compared to 15.8% in the controls. The cumulative incidence of VAP in both groups was not significantly different (P=0.80). More patients vomited in the intervention group compared to the control group (odds ratio of vomiting, 1.86; 90% confidence interval, 1.32-2.61; P=0.003). Patients in the intervention group had a lower cumulative calorie deficit from day 0 to day 7 compared to the control group. There were no significant differences in mortality, lengths of stay, rates of diarrhea, and ICU-acquired infections between groups.

The results of this study challenge the practice of routinely checking GRVs in all mechanically ventilated patients receiving gastric feeds. A GRV cutoff of 250 mL was used; several other studies have evaluated higher GRVs (up to 500 mL). It is possible that more complications might occur at higher GRVs, but previous work has not shown VAP rates to be greater. Furthermore, the lack of blinding in this study might have caused nurses to over-report vomiting, even though patients not monitored clearly received a larger volume of enteral nutrition. Additionally, the majority of patients in this study had a medical diagnosis at admission (>90%), and patients who had abdominal surgery were excluded. Hence, the results are not necessarily externally generalizable to all ICU populations, including surgical and trauma populations. Based on the results of this study, the routine stopping of enteral nutrition for GRVs < 250 mL should be reconsidered in medical ICU patients receiving gastric feeds.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

From 2003 to 2009, a total of 889 adult patients with acute, but not massive upper GI hemorrhage were randomized to either a restrictive (hgb <7 g/dL) or liberal (hgb <9 g/dL) transfusion strategy threshold group. Block randomization was performed and the analysis proceeded according to the intention-to-treat principle. The study was not blinded. The primary outcome was death from any cause within the first 45 days after bleeding. Secondary outcomes included complications and the rate of further bleeding. A Kaplan-Meier plot was constructed and a Cox proportional hazards model was used to compare the two groups with respect to the endpoints, controlling for baseline risk factors. 

Four hundred forty-four patients were randomized to the restrictive group and 445 were randomized to the liberal group. There was no statistically significant difference between the two groups in terms of baseline characteristics. Unadjusted mortality was statistically higher for the liberal group (9%) compared to the restrictive group (5%; P=0.02). The survival benefit for the restrictive group remained after adjustment for baseline risk factors for death (hazard ratio [HR], 0.55; 95% confidence interval [CI], 0.33 to 0.92).  The rate of further bleeding was lower in the restrictive group (10%) compared to the liberal group (16%; P=0.01), and effect that persisted after adjustment for baseline risk factors (HR 0.68; 95% CI, 0.47 to 0.98). The overall rate of complications was higher in the liberal group (48%) compared to the restrictive group (40%; P=0.02). Transfusion reactions and cardiac events occurred more commonly in the liberal group. The authors concluded that a restrictive strategy was associated with improved outcomes for patients with acute upper GI bleeding.

Based on the unadjusted control and experimental event rates, 24 patients would need to be treated with the restrictive strategy to save one life (absolute risk reduction, 4%; 95% CI, 0.6% to 7.5%). Strengths of this study include its randomized design and an appropriate statistical analysis. The results are not externally generalizable for patients with massive GI bleeding, and the use of a lower transfusion threshold in patients with significant cardiovascular disease is not supported by these data. Indeed, for patients with hypovolemia and shock secondary to massive GI hemorrhage, transfusion should not be withheld. Existing guidelines already endorse a restrictive strategy (hgb <7 g/dL); this work supports such an approach for most patients. 

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

This multicenter, parallel-group trial, randomized assignment of subjects to ICP monitoring (experimental) or imaging and clinical examination alone (control) in patients with severe TBI in four Bolivian and two Ecuadorian hospitals from 2008 to 2011.  Patients in the experimental group were treated to maintain an ICP <20 mm Hg. The primary outcome was a composite of 21 components, including measures of survival, impairment of consciousness and functional status. Clinicians and study personnel were not blinded to the experimental intervention. The blocked Wilcoxon test, logistic proportional odds models and Cox models were used to test the primary hypothesis by adjusting for factors such as study site, age and severity of injury.

The experimental group included 157 patients while the control group included 167 patients. There were no significant differences in prognostic factors between the groups.  Patients in the control group were statistically significantly more likely to receive hypertonic saline and hyperventilation; there were no other differences in co-interventions or other treatments such as neurosurgical procedures (i.e., craniectomy).  No significant differences existed between the groups in terms of the primary outcome (proportional odds ratio [OR], 1.09; 95% confidence interval [CI], 0.74 to 1.58; P=0.49). Cumulative mortality at six months was 39% in the experimental group and 41% in the control group (P=0.60). Patients in the control group were more likely to have a higher median intensive care unit length of stay (4.8 days [interquartile range (IQR), 2.3-7.4] vs. 3.4 days [IQR, 1.1-7.0]; P=0.002). The authors concluded that care focused on maintaining ICP <20 mm Hg was not shown to be superior to care guided by clinical examination and repeat neuroimaging. 

As ICP monitoring is presently endorsed as the standard of care for management of TBI, this study is bound to generate discussion within the neurocritical care community. Early criticisms include lack of external generalizability, since less than half of the enrolled patients (45%) were treated by prehospital emergency medical services. Patients might have had different degrees of severe TBI compared to patients in higher-income countries. Moreover, none of the subjects received rehabilitation or advanced medical care after discharge. ICP monitoring did not include placement of ventricular catheters, which may have additional therapeutic and diagnostic benefits in TBI patients. The therapeutic protocols in the experimental and control groups were similar, but the protocols alone may account for variable efficacy. The results from this study do not refute the tenet that aggressive treatment is important for patients with severe TBI, but the means for guiding targeted therapies deserves reassessment.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Prospectively collected data from January 2005 to January 2010 were examined. The primary outcome of interest was change in compliance with sepsis bundle targets over time. Secondary outcomes included hospital mortality, hospital length of stay and intensive care unit (ICU) length of stay. A random effects logistic regression model was used to estimate hospital mortality and a random-effects linear regression model was used to find the relationship between lengths of stay and geographic region.

In the study, 18,766 patients from the United States and 6,609 patients from Europe met the inclusion criteria for analysis. Compliance with all bundle elements was higher in the United States (21.6% vs. 18.4%, P<0.0001), although compliance for four of six individual bundle elements was statistically significantly higher in Europe. Patients in Europe had a higher incidence of multiple organ dysfunction and nosocomial infections, and a higher unadjusted mortality rate compared to the United States. Unadjusted ICU mortality in Europe was higher (odds ratio [OR] 1.61; 95% confidence interval [CI], 1.32 to 1.96; P<0.0001); however, after adjustment for multiple variables associated with patient illness severity — such as type of organ failure, resuscitation performance and other patient-related variables — the mortality rate difference between the two regions disappeared (OR 1.19; 95% CI, 0.96 to 1.47; P=0.11). Adjusted ICU length of stay was longer in Europe than in the United States (P<0.0001). Further, a sensitivity analysis using propensity score matching confirmed the results from the logistic regression analysis.

The Surviving Sepsis Campaign database has several limitations. Patients are self-selected, and it is possible that the results from this study may not be generalizable for other regions of the world. The true mortality rate is unknown since some patients with severe sepsis might have been admitted to a general ward and discharged without an ICU admission. No other studies have directly compared sepsis outcomes in Europe and the United States. The results of this study raise important questions about different models of care, outcomes and compliance with accepted management bundles for severe sepsis and septic shock.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

The GWTG-Resuscitation registry included 113,514 adults with cardiac arrest at 553 hospitals during the period 2000-2009. After arrests in the emergency department, operating room and procedural areas were excluded, 84,625 patients at 374 hospitals constituted the final study population. The primary outcome was survival to discharge; secondary outcomes included neurologic disability. Advanced multivariable regression techniques were used to analyze the cohort. Regression models adjusted for age, sex, race, coexisting conditions, therapeutic interventions at the time of arrest, and hospital characteristics.

Of all arrest patients, 67,125 (79.3%) had asystole or pulseless electrical activity, and 17,490 (20.7%) patients had ventricular fibrillation or pulseless ventricular tachycardia.  Overall survival to discharge was 17%. Survival increased from 13.7% in 2000 to 22.3% in 2009 (adjusted rate ratio per year, 1.04; 95% confidence interval, 1.03 to 1.06; P<0.001 for trend). Rates of acute resuscitation survival improved from 42.7% in 2000 to 54.1% in 2009 (P<0.001 for trend). Clinically significant neurologic disability rates declined over time from 32.9% in 2000 to 28.1% in 2009 (P=0.02 for trend). The authors concluded that both survival and neurological outcomes improved substantially from 2000 to 2009.

The results from this study may be attributed to improved protocols that emphasize high-quality chest compressions with fewer interruptions, as well as other post-resuscitation care such as therapeutic hypothermia and early cardiac catheterization. However, as with all large observational studies, the results may be explained by a decrease in the baseline risk over time, reporting bias from hospitals involved with the GWTG-Resuscitation registry and other residual confounders. While improved survival for in-hospital cardiac arrest is encouraging, confirmatory studies are indicated to identify the specific aspects responsible for this positive trend.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

A variety of databases were utilized to locate 2,122 articles for possible inclusion. Articles were required to evaluate two types of ventilation in patients without ARDS. Protective lung ventilation was defined as a tidal volume set for 5 mL/kg to 8 mL/kg ideal body weight (one study included 9 mL/kg compared to 12 mL/kg), and conventional ventilation was defined as a tidal volume of 9mL/kg to 12 mL/kg ideal body weight. Randomized trials, cohort studies, cross-sectional studies, and before/after studies were included. The Grades of Recommendation Assessment, Development and Evaluation (GRADE) approach was used to summarize the quality of evidence for each outcome. Pooled estimates (relative risk) were calculated, while sensitivity analyses were carried out for different subgroups of studies.

Twenty articles were identified for inclusion. All but five studies were randomized controlled studies. In fifteen studies, the reason for intubation was surgery; in the other five, the reason was mixed. The mean tidal volume in the protective group was 6.5 (standard deviation [SD] 1.1) and 10.6 (SD 1.1) in the conventional group. In the protective ventilation group, 4.2% of patients developed lung injury versus 12.7% of patients in the conventional group (risk ratio [RR], 0.33; 95% confidence interval [CI], 0.23-0.47, number needed to treat, 11). When only randomized studies were analyzed, patients in the protective ventilation group had a lower relative risk of developing acute lung injury (RR, 0.26; 95% CI, 0.10-0.66, number needed to treat, 10). Both overall mortality and hospital stay (but not ICU length of stay) was lower in the protective ventilation group.

Strengths of this study include application of the GRADE validity criteria, sensitivity analyses, and assessments for publication bias. Although it is unlikely that most relevant studies were missed with the search strategy, the strategy did not utilize additional databases, such as EMBASE, or a search of the gray literature. While homogeneity was relatively low in the meta-analysis, the included studies consisted of disparate populations, treatment locations (e.g., operating room versus intensive care unit), and tidal volumes. Finally, patients had varying lengths of follow-up, ranging from as short as 3 hours to as long as 672 hours. Despite these and other limitations, the results regarding the primary endpoint (acute lung injury) yielded large and consistent estimates of treatment effect, with narrow confidence intervals, favoring a protective lung strategy for several outcomes of interest. The results from this trial should motivate further prospective clinical trials investigating the effect of lower tidal volumes in patients without ARDS.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Four hundred twenty-three patients were randomized either to protocolized sedation alone (control group) or protocolized sedation plus daily interruption of sedation (experimental group). None of the participants or study personnel was blinded. The two groups were similar in terms of baseline characteristics, and patients from surgical and trauma intensive care units (ICUs) were included. The primary outcome was time to successful extubation. Patients were analyzed according to the intention-to-treat principle, and a Cox proportional hazards model was used to estimate a hazard ratio for median time to extubation.

When adjusted for severity, age and other patient characteristics, median time to extubation was similar in both groups (hazard ratio, 1.04; 95% confidence interval, 0.83-1.31). Patients in the experimental group had significantly higher mean daily benzodiazepine doses (P=0.04) and higher daily opioid doses (1780 vs. 1070 μg/day fentanyl equivalents; P<0.001). Secondary outcomes, such as ICU mortality, hospital days and ICU-acquired organ failures, were not significantly different between the two groups. There were no significant differences when medical patients were analyzed separately from surgical and trauma patients.

The results from this study confirmed the alternative hypothesis that a combination of both interruption and protocolization for sedation was no better than protocolized sedation alone. Potential limitations of this work include lack of adherence with daily interruption in more than one-quarter of randomized patients, lack of blinding of the investigators, and absence of the use of shorter acting sedatives — such as propofol and dexmedetomidine. Nevertheless, this study was conducted with a pragmatic design, incorporating existing ICU bedside assignments that utilized clinicians rather than research staff, and included a broad mix of patients. The results suggest that daily interruption may increase — not decrease — the amount of analgesics and sedatives required for critically ill patients, while also increasing staff workload. Protocolized sedation alone may be sufficient for impacting the outcomes of interest in patients with a significant requirement for analgesia and sedation in the ICU.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

The IABP-SHOCK II trial was a prospective, multicenter, randomized trial conducted over a three-yr period at 37 centers in Germany. AMI patients were considered to be in cardiogenic shock if a systolic blood pressure of 90 mm Hg was observed for more than 30 minutes, with other signs of shock or need for catecholamine infusion. Patients were randomized to the IABP vs. no IABP in a 1:1 ratio. Randomization was performed centrally, all patients were accounted for in an intention-to-treat analysis, and clinical outcomes assessors were blinded to group assignments. There were no significant differences between the groups in terms of baseline characteristics, nor clinical course before randomization. Patients who crossed over from the control group to the IABP group were excluded. The major outcome of interest was 30-day, all-cause mortality; additional secondary endpoints were established a priori.

Three hundred patients were included in the IABP group and 298 in the control group.  At 30 days, no statistically significant difference in relative risk for death was found between the groups (relative risk [RR] of death with IABP, 0.96; 95% confidence interval [CI], 0.79-1.17; p=0.69). There were no significant differences in secondary endpoints (i.e., lactate, C-reactive protein levels, renal function, Simplified Acute Physiology Score II). In a multivariate analysis, the relative risk remained non-significant for the primary endpoint of death. The authors concluded that IABP counterpulsation did not reduce 30-day mortality in patients with AMI and cardiogenic shock.

AMI complicated by cardiogenic shock has a disturbingly high (40%) attendant mortality. This randomized controlled study was appropriately powered and rigorously conducted; however, the results may not be generalizable to all populations. For example, the procedure used for revascularization was primary percutaneous coronary interventionin over 95% of all patients. In other countries, immediate bypass surgery may be more commonly performed. In the group aged <50, the IABP was associated with a lower risk of death (RR 0.44, 95% CI, 0.21-0.95), and this population might deserve further study. In populations with higher mortality (> 40%), it is unclear if the IABP might be of some benefit. In view of these limitations, this study provides evidence that challenges current high-grade recommendations for the use of IABP counterpulsation in AMI patients with cardiogenic shock.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Patients in the NICE-SUGAR trial were randomly assigned to either intensive blood glucose control, targeting a glucose range of 81-108 mg/dL (4.5-6.0 mmol/L), or conventional glucose control, targeting a glucose <180 mg/dL (<10.0 mmol/L). Severe hypoglycemia was defined as a glucose value <40 mg/dL (2.2 mmol/L); values between 41 and 70 mg/dL (2.3-3.9 mmol/L) were considered moderate hypoglycemia. Each group contained 3,013 patients. Cox regression was used to calculate hazard ratios for death, adjusted for treatment assignment, baseline characteristics and time-dependent factors.

Forty-five percent of patients had moderate hypoglycemia, including 74.2% of patients in the intensive glucose control group. In the intensive glucose control group, 6.9% of patients had severe hypoglycemia (n=208/3013). Patients with moderate hypoglycemia were more likely to have severe sepsis, trauma, diabetes, and cardiovascular failure. When adjusted for treatment assignment, patients in the intensive glucose control group had a statistically significantly higher risk of death when they had severe hypoglycemia (hazard ratio 3.21; 95% confidence interval, 2.49-4.15; P<0.001). The risk of death remained statistically increased when adjustments for baseline characteristics and post-randomization factors were taken into account; both moderate and severe hypoglycemia was associated with an increased risk of death. Patients with hypoglycemia for more than 1 day were found to have an increased risk of death (>1 day vs. 1 day, P=0.01).

This post hoc analysis offers further insight into the mechanisms of increased mortality when ICU patients are treated with intensive glucose control. Because the original NICE-SUGAR trial included a wide range of ICU patients (i.e., postsurgical vs. medical), it is unclear if the association between hypoglycemia and death is valid across all patient populations. As the authors correctly point out, the results from this study cannot demonstrate a casual relationship between hypoglycemia and death. Instead, the authors rightly surmise that hypoglycemia may be a marker of severe underlying disease processes. Based on the results of this trial and additional observational data, the current American Diabetes Association guidelines, which stress avoidance of hypoglycemia and targeting of glucose values in the range of 140-180 mg/dL (8–10 mmol/L), appear reasonable for critically ill patients.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

They reviewed 356 patients with SAH in a retrospective trial at a level I trauma and stroke referral center in Seattle, Washington. Traumatic and nonaneuyrsmal SAH patients were excluded. Two groups were examined: a positive fluid balance group (n=221) and a negative fluid balance group (n=135). The patients in the positive fluid balance group had a positive fluid balance by intensive care unit day 3; those in the negative fluid balance group had a zero or negative cumulative fluid balance. All patients were managed with daily transcranial Doppler exams (TCD) according to an institutional fluid protocol for managing SAH. Adjusted and unadjusted regression models were used to assess the relationship of various known confounders and outcomes of interest.

SAH patients with a positive fluid balance were more likely to have a lower Glasgow Coma Scale score (8.7 vs. 10.0, P<0.01) and higher score on Simplified Acute Physiology Score II (39.6 vs. 31.9, P<0.01). Positive fluid balance patients also required statistically significantly more vasopressor use, more fludrocortisone use and more blood product use. In a model adjusting for variables such as age, Hunt-Hess score, Fisher grade, troponin elevation, and respiratory failure, patients with a positive fluid balance were more likely to have a longer length of stay (odds ratio [OR] 1.26; 95% confidence interval [CI], 1.05-1.51, P value not provided), but were not more likely to die in the hospital or suffer a new stroke. Patients with a positive fluid balance had a significantly higher degree of vasospasm (OR 2.25, 95% CI, 1.37-3.71) as detected by TCD exams.

This study adds to a growing body of evidence that has shown the “hypervolemia” arm of triple-H therapy to be of limited utility, and possibly harmful, for patients with vasospasm from SAH.  As an observational trial, this study has several limitations. Although an institutional protocol was used, the authors did not report compliance with this protocol or additional co-interventions. Individual practice patterns were not described, follow-up was not complete (not all patients had TCD exams), and TCD exams, not angiography, were used to confirm vasospasm. The authors’ conclusions remained guarded in view of these limitations, but based on the findings from this study, intensive care management of SAH patients should include careful intravenous fluid management (perhaps tailored to individual patient needs), keeping in mind the deleterious effects associated with the accumulation of a positive fluid balance.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Of the 316 consecutive patients included, 125 (40%) received steroids at the time ICU-acquired pneumonia was diagnosed. A wide variety of statistical techniques, including Cox regression with propensity scores for steroid use, Kaplan-Meier analysis and several sensitivity analyses, were used to evaluate the primary outcome of 28-day survival, as well as additional secondary outcomes. The American Thoracic Society guidelines, the Infectious Diseases Society of America guidelines, and the Clinical Pulmonary Infectious Score (CPIS) were used to establish the diagnosis of ICU-acquired pneumonia.

In the steroid group, unadjusted mortality was 39% (n=49) compared to 28% in the group that did not receive steroids. After adjustment for multiple independent variables, steroid treatment was statistically significantly associated with decreased 28-day survival (adjusted hazard ratio, 2.503; 95% confidence interval [CI], 1.17-5.33; p=0.017). Kaplan-Meier analysis also revealed decreased survival in patients with ICU pneumonia at 28 days when treated with steroids (log-rank test, 4.006; p=0.045). Steroid-treated patients had lower levels of inflammatory markers such as C-reactive protein, procalcitonin and interleukin-6. Patients treated with steroids had a higher rate of microbiological late-onset pneumonia (74% vs. 58%; p=0.020).

One major strength of this study is the use of multiple robust methods to control for known confounders given the nonrandomized design, but several limitations are noted. The authors did not present a power calculation to ensure that the study was adequately powered to make conclusions about mortality. Although the effect estimate for increased risk of death with steroid treatment was significant, the confidence interval was wide. Additional limitations included the use of different types of corticosteroids and the single-center study design.

The results of Ranzani et al appear to conflict with those reported by Roquilly et al last year in The Journal of the American Medical Association. In the Hydrocortisone Polytraumatise (HYPOLYTE) randomized controlled study, hydrocortisone therapy was found to significantly decrease the risk of pneumonia in intubated trauma patients; however, the HYPOLYTE study was not adequately powered to detect differences in mortality. Ranzani et al included only a minority of multiple trauma patients (n=21). Based on the present results, caution should be exercised until larger studies are conducted to define the effects of steroid use on mortality. Attenuation of the host inflammatory response by systemic corticosteroids may not be beneficial for all critically ill patients with ICU-acquired pneumonia.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.

Ten studies, consisting of eight cohort studies and two RCTs, were included with data from 17,376 catheters. There was no significant difference in the risk of CRBI between the femoral compared to the subclavian site (risk ratio [RR] 1.75; 95% confidence interval [CI], 0.80-3.8; P = 0.16).  The IJ site was associated with an overall lower risk of CRBI compared to the femoral site (RR 1.9; 95% CI, 1.21-2.97; P=0.005); however, when two cohort studies were excluded in a sensitivity analysis, there was no significant difference (RR 1.35; 95% CI, 0.84-2.19; P=0.20).  Meta-regression showed a significant interaction between the risk of infection and the year of publication, with earlier studies favoring the IJ site. The authors concluded that recent evidence demonstrates no difference in the risk of CRBI between the femoral, subclavian and IJ sites.

With the advent of maximum barrier precautions, the use of ultrasound and vigilant management of catheter exit sites, the CRBI rate in many intensive care units has decreased in recent years, perhaps independently of catheter insertion site. Only two RCTs were included in this review, but both supported the author’s conclusion that there is no greater risk of a CRBI when the femoral site is used. It is unclear how the cohort studies were assessed for risk of bias, but several sensitivity analyses were conducted when significant heterogeneity was found. Based on the results of this systematic review and meta-analysis, recent studies do not appear to support the notion that the femoral site is associated with significantly more CRBIs.

Concise Critical Appraisal is a regular feature authored by SCCM member Samuel M. Galvagno Jr., DO, PhD. Each installment highlights journal articles most relevant to the critical care practitioner.