Ebook Year book - Year book of critical care medicine 2013 (1st edition): Part 2

Part 2 book Year book - Year book of critical care medicine 2013 presents the following contents: Sepsis/septic shock, metabolism/gastrointestinal/nutrition/hematology-oncology, neurologic - Traumatic and non traumatic, renal, trauma and overdose, ethics/socioeconomic/administrative/issues, pharmacology/sedation analgesia.

Acad Emerg Med 19:983-985, 2012

Objectives.dPrevious studies have confirmed the prognostic significance

of lactate concentrations categorized into groups (low, intermediate, high)among emergency department (ED) patients with suspected infection.Although the relationship between lactate concentrations categorized intogroups and mortality appears to be linear, the relationship between lactate

as a continuous measurement and mortality is uncertain This study sought

to evaluate the association between blood lactate concentrations along

an incremental continuum up to a maximum value of 20 mmol/L andmortality

Methods.dThis was a retrospective cohort analysis of adult ED patientswith suspected infection from a large urban ED during 2007e2010 Inclu-sion criteria were suspected infection evidenced by administration of anti-biotics in the ED and measurement of whole blood lactate in the ED Theprimary outcome was in-hospital mortality Logistic and polynomialregression were used to model the relationship between lactate concentra-tion and mortality

Results.dA total of 2,596 patients met inclusion criteria and wereanalyzed The initial median lactate concentration was 2.1 mmol/L (inter-quartile range [IQR]¼ 1.3 to 3.3 mmol/L) and the overall mortality ratewas 14.4% In the cohort, 459 patients (17.6%) had initial lactatelevels > 4 mmol/L Mortality continued to rise across the continuum of incre-mental elevations, from 6% for lactate < 1.0 mmol/L up to 39% for lactate19e20 mmol/L Polynomial regression analysis showed a strong curvilinearcorrelation between lactate and mortality (R¼ 0.72, p < 0.0001)

Conclusions.dIn ED patients with suspected infection, we found acurvilinear relationship between incremental elevations in lactate concen-tration and mortality These data support the use of lactate as a continuousvariable rather than a categorical variable for prognostic purposes

Previous studies have shown that increased serum lactate correlates withincreased mortality However, these studies all used categorical lactates (eg,normal, low, intermediate, high) This study questioned whether lactate as acontinuous variable (it is after all) correlated with mortality in patients with

135

suspected infection Patients admitted to the hospital from the emergencydepartment were included if they had suspected infection (antibiotics given inthe emergency department) and a lactate measurement obtained More than

2500 patients were enrolled and they seemed to represent a typical sepsis cohortbased on length of stay, comorbidities, and mortality The authors found definitecorrelations between incremental increases of lactate and incremental increases

in mortality They developed a regression model/equation that fit their data quitewell In other words, with a given lactate, their equation could predict mortality.This study was well done by a group of researchers with a good track record Payattention to lactate in patients with suspected infection: the higher it is, the morelikely the patient in front of you is going to die

Intensive Care Med 38:422-428, 2012

Purpose.dResponse to fluid challenge is often defined as an increase incardiac index (CI) of more than 10e15% However, in clinical practice CIvalues are often not available We evaluated whether changes in meanarterial pressure (MAP) correlate with changes in CI after fluid challenge

in patients with septic shock

Methods.dThis was an observational study in which we reviewedprospectively collected data from 51 septic shock patients in whomcomplete hemodynamic measurements had been obtained before andafter a fluid challenge with 1,000 ml crystalloid (Hartman’s solution) or

500 ml colloid (hydroxyethyl starch 6%) CI was measured using dilution Patients were divided into two groups (responders and non-responders) according to their change in CI (responders: %CI > 10%)after the fluid challenge Statistical analysis was performed using a two-way analysis of variance test followed by a Student’s t test with adjustmentfor multiple comparisons Pearson’s correlation and receiver operatingcharacteristic curve analysis were also used

thermo-Results.dMean patient age was 67 ± 17 years and mean SequentialOrgan Failure Assessment (SOFA) upon admittance to the intensive careunit was 10 ± 3 In the 25 responders, MAP increased from 69 ± 9 to

77 ± 9 mmHg, pulse pressure (PP) increased from 59 ± 15 to 67 ± 16,and CI increased from 2.8 ± 0.8 to 3.4 ± 0.9 L/min/m2 (all p < 0.001).There were no significant correlations between the changes in MAP, PP,and CI

Conclusions.dChanges in MAP do not reliably track changes in CIafter fluid challenge in patients with septic shock and, consequently,

should be interpreted carefully when evaluating the response to fluid lenge in such patients.

Accurate, dynamic markers of fluid responsiveness in shock are highly soughtafter but often remain illusive In a prospective observational study, Pierrakos et alassessed whether changes in arterial pressure could predict fluid responsiveness

of 51 septic shock patients by determining the percentage of change in cardiacindex (CI) via the thermodilution method Half of the patients were determined

to be fluid responders as defined by a greater than 10% increase in CI after fluidchallenge with 1 L crystalloid or 500 mL colloid bolus While pulse pressure andmean arterial pressure (MAP) significantly increased in responders when chal-lenged with a fluid bolus, there was no significant correlation with CI The lack

of correlation between changes in arterial pressure and cardiac index shouldcaution the intensivist that an increase in MAP after volume expansion doesnot necessarily improve hemodynamic status In other words, augmentation ofMAP with a fluid challenge does not accurately identify fluid responsiveness

in septic shock patients

Methods.dWe conducted a prospective observational study comparingresults for 30 trauma and emergency surgery patients to 20 burn patients.Whole-blood samples collected with routine blood cultures (BCs) weretested using a new multiplex, PCR-based, pathogen detection system.PCR results were compared to culture data

Results.dPCR detected rapidly more pathogens than culture methods.Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequen-tial Organ Failure Assessment (SOFA), and Multiple Organ Dysfunction(MODS) scores were greater in PCR-positive versus PCR-negative traumaand emergency surgery patients (P#.033) Negative PCR results (oddsratio, 0.194; 95% confidence interval, 0.045e0.840; P ¼.028) acted as

an independent predictor of survival for the combined surgical patientpopulation

Conclusion.dPCR detected the presence of pathogens more frequentlythan blood culture These PCR results were reported faster than bloodculture results Severity scores were significantly greater in PCR-positive

Chapter 6eSepsis/Septic Shock / 137

trauma and emergency surgery patients The lack of pathogen DNA as mined by PCR served as a significant predictor of survival in the combinedpatient population PCR testing independent of traditional prompts forculturing may have clinical value in burn patients These results warrantfurther investigation through interventional trials.

Trauma, emergency, and burn surgery patients are at a particularly high risk ofsepticemia secondary to their disease states As a result, immediate empiric anti-biotic treatment is necessary to improve mortality.1However, with increasingresistant pathogens, aggressive de-escalation is important as well

In this prospective, observational study, patients displaying signs and toms of sepsis had blood cultures and blood samples run through polymerasechain reaction (PCR) Arbitrated case reviews were performed to determine ifappropriate antibiotic regimens were dispensed, given first blood culture resultsalone and then PCR results PCR had a markedly increased turnaround time(5.9 h vs 25.3 h) Additionally, PCR was positive in 11 cases, whereas bloodcultures were either negative or failed to identify pathogen species Arbitratedcase review found antimicrobial therapy was inadequate relative to PCR results

symp-in 29% PCR-positive patients PCR was found to have more pathogen detectionevents, even in serial samples while patients were on concurrent antimicrobialtherapy A negative PCR was found to be an independent predictor of survival.While this practice is cost prohibitive and not approved by the US Food andDrug Administration, PCR can be a very powerful tool in early pathogen detec-tion in sepsis It can also serve as a predictor for mortality According to theauthors, it does have its limitations: it is dependent on the assay to detect specificpathogens, genetic polymorphisms, and whole blood sample matrix effects

Advances in Mesenchymal Stem Cell Research in Sepsis

Wannemuehler TJ, Manukyan MC, Brewster BD, et al (Indiana Univ School ofMedicine, Indianapolis)

J Surg Res 173:113-126, 2012

Background.dSepsis remains a source of morbidity and mortality in thepostoperative patient despite appropriate resuscitative and antimicrobialapproaches Recent research has focused upon additional interventionssuch as exogenous cell-based therapy Mesenchymal stem cells (MSCs)exhibit multiple beneficial properties through their capacity for homing,attenuating the inflammatory response, modulating immune cells, and

promoting tissue healing Recent animal trials have provided evidence thatMSCs may be useful therapeutic adjuncts.

Materials and Methods.dA directed search of recent medical literaturewas performed utilizing PubMed to examine the pathophysiology of sepsis,mechanisms of mesenchymal stem cell interaction with host cells, sepsisanimal models, and recent trials utilizing stem cells in sepsis

Results.dMSCs continue to show promise in the treatment of sepsis bytheir intrinsic ability to home to injured tissue, secrete paracrine signals

to limit systemic and local inflammation, decrease apoptosis in threatenedtissues, stimulate neoangiogenesis, activate resident stem cells, beneficiallymodulate immune cells, and exhibit direct antimicrobial activity Theseeffects are associated with reduced organ dysfunction and improvedsurvival in animal models

Conclusion.dResearch utilizing animal models of sepsis has provided agreater understanding of the beneficial properties of MSCs Their capacity

to home to sites of injury and use paracrine mechanisms to change the localenvironment to ultimately improve organ function and survival makeMSCs attractive in the treatment of sepsis Future studies are needed tofurther evaluate the complex interactions between MSCs and host tissues

Sepsis holds a mortality rate of 28.6% and costs $16.7 billion nationally Assuch, immediate, effective treatment for a septic patient is vital Interest inmesenchymal stem cells (derived from bone marrow, adipose, placenta, andumbilical cord) has been applied in the treatment of sepsis for its multiple abil-ities, which is reviewed in this article

Mesenchymal stem cells have the ability to home to injured tissues, partlyfrom interacting with host cytokines They also have paracrine signaling effects

to promote tissue regeneration, prevent tissue loss, and improve tissue functionvia its anti-inflammatory, antiapoptotic, nonangiogenic activation of residentstem cell and immunomodulatory capabilities Although in vivo studies mesen-chymal cells also had antimicrobial effects, vitro models revealed they couldnot Thus, they likely require host cell stimulation and signaling to do so Inanimal studies, stem cell treatment of sepsis has shown reduced levels of proin-flammatory cytokines (interleukin-1, tumor necrosis factor-a, interleukin-6),decreased organ injury, and increased organ function Additionally, stem cellscan improve bacterial clearance through increased macrophage phagocyticactivity More importantly, septic animals treated with stem cells had a signifi-cantly improved survival rate

In conclusion, mesenchymal stem cells administration is a promising ment for sepsis through its multiple effects on the host’s immune system,organ tissue, and function Although there are no clinical trials as of yet,continued research in the animal model is crucial

treat-S Zanotti, MDChapter 6eSepsis/Septic Shock / 139

A multicenter trial to compare blood culture with polymerase chainreaction in severe human sepsis

Bloos F, Hinder F, Becker K, et al (Univ Hosp Jena, Germany; Univ HospMu¨nster, Germany)

Intensive Care Med 36:241-247, 2010

Objective.dTo assess the presence of microbial DNA in the blood bypolymerase chain reaction (PCR) and its association with disease severityand markers of inflammation in severe sepsis and to compare the perfor-mance of PCR with blood culture (BC)

Design.dProspective multicentric controlled observational study.Setting.dThree surgical intensive care units in university centers andlarge teaching hospitals

Patients.dOne hundred forty-two patients with severe sepsis and 63surgical controls

Interventions.dPresence of microbial DNA was assessed by multiplexPCR upon enrollment, and each time a BC was obtained

Measurements and Main Results.dControls had both approximately4% positive PCRs and BCs In severe sepsis, 34.7% of PCRs were positivecompared to 16.5% of BCs (P < 0.001) Consistently, 70.3% of BCs had acorresponding PCR result, while only 21.4% of PCR results wereconfirmed by BC Compared to patients with negative PCRs at enrollment,those testing positive had higher organ dysfunction scores [SOFA, median(25the75th percentile) 12 (7e15) vs 9 (7e11); P ¼ 0.023] and a trendtoward higher mortality (PCR negative 25.3%; PCR positive 39.1%;

P¼ 0.115)

Conclusions.dIn septic patients, concordance between BC and PCR ismoderate However, PCR-based pathogen detection correlated withdisease severity even if the BC remained negative, suggesting that presence

of microbial DNA in the bloodstream is a significant event The clinicalutility to facilitate treatment decisions warrants investigation

Sepsis is one of the most important causes of morbidity and mortality in cally ill patients The administration of early appropriate antibiotics is a corner-stone of sepsis therapy Studies have shown that mortality is increased withdelays in initiating antibiotics that cover the culprit pathogen in sepsis Thishas led to an increase in the use of initial empiric antibiotic regimens with broadcoverage Without proper de-escalation, the increased use of broad-spectrumantibiotics will lead to increased antibiotic resistance Unfortunately, negativeblood cultures are a common occurrence in sepsis This makes modification ofantibiotic regimes and diagnosis of infection difficult

criti-Many experts have proposed that culture-independent molecular biologyebased diagnostic tests, such as real-time polymerase chain reaction (PCR),could overcome some of the limitations cultures present and be more useful inthe management of sepsis In this intriguing prospective multicenter study, theauthors assessed the diagnostic utility of culture-independent PCR-based de-tection of pathogens compared with blood cultures (BC) in sepsis The study

basically found that concordance between BC and PCR was moderate in septicpatients There were more positive PCRs with negative BCs However, negativePCRs were still common, making it unlikely that at present a negative PCR could

be used to rule out infection and/or stop antibiotics An additional finding ofgreat interest was the correlation of a positive PCR (presence of microbialDNA in the bloodstream) with increased severity and poor outcomes Thisfinding suggests that the presence of bacteria DNA (even with negative bloodcultures) is an important clinical finding and may warrant further investigationregarding its value to guide therapeutic interventions

S Zanotti, MD

Diagnostic value of positron emission tomography combined withcomputed tomography for evaluating patients with septic shock ofunknown origin

Kluge S, Braune S, Nierhaus A, et al (Univ Med Ctr Hamburg-Eppendorf,Germany)

J Crit Care 27:316.e1-316.e7, 2012

Purpose.d18F-fluorodeoxyglucose (FDG) positron emission raphy (PET) combined with computed tomography (CT) is a promisingnew tool for the identification of infectious foci The aim of our workwas to evaluate the diagnostic value of FDG-PET/CT in critically ill patientswith septic shock of unknown origin

tomog-Methods.dWe performed a single-center, 6-year retrospective tion of the value of FDG-PET/CT in critically ill patients with severe sepsis

evalua-or septic shock of unknown evalua-origin

Results.dEighteen patients underwent FDG-PET/CT Microbiologicaltests (blood culture, urine, and respiratory secretions), chest x-rays, CTscans, and transesophageal echocardiography were performed on allpatients before FDG-PET/CT scanning Pathologic FDG accumulationcould be demonstrated in 14 of 18 FDG-PET/CT scans On a per-patientbasis, 11 were “true positive,” 3 were “false positive,” 4 were true negative,and there were no false negatives In 6 cases, the results of the PET/CT scanhad direct therapeutic consequences (surgery, 2; pacemaker removal, 2;initiation of antibiotic therapy, 1; and prolonged antibiotic therapy, 1);

12 (66%) of the 18 patients survived to hospital discharge

Conclusions.dThe FDG-PET/CT is a valuable tool for the localization

of infectious foci in critically ill patients with severe sepsis/septic shock inwhom conventional diagnostic methods fail to detect these foci Prospec-tive studies with more patients are warranted to further evaluate the diag-nostic accuracy and feasibility of this diagnostic tool in critically illpatients with severe sepsis

Positron emission tomography/computed tomography (PET/CT) is wellestablished in the arena of oncologic imaging and evaluation of chronic infec-tions Its potential role in evaluating acute infections and patients with severe

Chapter 6eSepsis/Septic Shock / 141

sepsis and septic shock is undetermined Simons et al (2010) previously tigated the role of PET/CT in mechanically ventilated intensive care unit (ICU)patients.1This study presents a novel application of an established modality toevaluate for acute infections in the setting of severe sepsis and septic shock.This was a retrospective observational study in all adult ICU patients over a6-year period with severe sepsis or septic shock with an unknown sourcewho underwent PET/CT evaluation The results of PET/CT were comparedwith a final diagnosis that was made using all clinical information excludingthe PET/CT findings In 5 cases (27%), PET/CT was essential for diagnosis

inves-in the followinves-ing cases: pseudomembranous colitis, inves-infected bypass graft,infected inferior vena cava thrombus, infected pacemaker, and cervical abscess

In 6 cases (33%), PET/CT results altered patient management The survival ratefor this group of patients was 66% Most of the patients in this study were eval-uated for persistent fever or suspected septic emboli This study did not specifysevere sepsis or septic shock as indications In 14% of patients, the PET/CTresults altered management The limitations of the current study were the retro-spective design and the small study population A major drawback of PET/CT inthe critical care setting is the time required for the exam, the expense, and avail-ability as an inpatient PET/CT evaluation may serve as an important adjunct fordiagnosis of sources of severe sepsis or septic shock when other modalities(clinical, microbiologic, and anatomic imaging) have proven unrevealing.Further validation of this modality in sepsis in a prospective randomized trialwould determine if a survival benefit exists

A F Miller, MD

S Zanotti, MDReference

1 Simons KS, Pickkers P, Bleeker-Rovers CP, Oyen WJ, van der Hoeven JG fluorodeoxyglucose positron emission tomography combined with CT in critically ill patients with suspected infection Intensive Care Med 2010;36:504-511.

F-18-Early goal-directed therapy (EGDT) for severe sepsis/septic shock: whichcomponents of treatment are more difficult to implement in acommunity-based emergency department?

O’Neill R, Morales J, Jule M (Genesys Regional Med Ctr, Grand Blanc, MI)

J Emerg Med 42:503-510, 2012

Background.dEarly goal-directed therapy (EGDT) has been shown toreduce mortality in patients with severe sepsis/septic shock, however,implementation of this protocol in the emergency department (ED) issometimes difficult

Objectives.dWe evaluated our sepsis protocol to determine whichEGDT elements were more difficult to implement in our community-based ED

Methods.dThis was a non-concurrent cohort study of adult patientsentered into a sepsis protocol at a single community hospital from July

2008 to March 2009 Charts were reviewed for the following processmeasures: a predefined crystalloid bolus, antibiotic administration, centralvenous catheter insertion, central venous pressure measurement, arterialline insertion, vasopressor utilization, central venous oxygen saturationmeasurement, and use of a standardized order set We also compared theindividual component adherence with survival to hospital discharge.Results.dA total of 98 patients presented over a 9-month period.Measures with the highest adherence were vasopressor administration(79%; 95% confidence interval [CI] 69e89%) and antibiotic use (78%;95% CI 68e85%) Measures with the lowest adherence included arterialline placement (42%; 95% CI 32e52%), central venous pressure measure-ment (27%; 95% CI 18e36%), and central venous oxygen saturationmeasurement (15%; 95% CI 7e23%) Fifty-seven patients survived tohospital discharge (Mortality: 33%) The only element of EDGT todemonstrate a statistical significance in patients surviving to hospitaldischarge was the crystalloid bolus (79% vs 46%) (respiratory rate[RR]¼ 1.76, 95% CI 1.11e2.58)

Conclusion.dIn our community hospital, arterial line placement,central venous pressure measurement, and central venous oxygen satura-tion measurement were the most difficult elements of EGDT to implement.Patients who survived to hospital discharge were more likely to receive thecrystalloid bolus

It has been more than a decade since the publication of the landmark earlygoal-directed therapy (EGDT) study by Rivers et al.1 This study showed asignificant improvement in mortality in patients with severe sepsis-inducedhypoperfusion treated with EGDT Despite the questions that remain unan-swered and concerns with specific aspects of the trial, EGDT has been widelyrecommended by various guidelines, such as the Surviving Sepsis Guidelines.2

In this study, the investigators sought to evaluate the adherence of specificelements of the EGDT protocol in a community emergency department (ED).The authors report that arterial line placement, central venous pressuremeasurement, and central venous oxygen saturation measurement were themost difficult elements of EGDT to implement The authors also report thatthe patients who survived to hospital discharge were more likely to receivethe crystalloid bolus (2 L in the first hour) as prescribed in their protocol.Although this study does not provide answers to which aspects of EGDTwork in decreasing survival, it does shed important light on some of the barriersthat still persist in the implementation of these protocols in community EDs.Currently, there are 3 large multicenter studies evaluating different aspects ofEGDT and resuscitation of patients with severe sepsis/septic shock The results

of these ongoing trials may help define the best way forward

S Zanotti, MDChapter 6eSepsis/Septic Shock / 143

1 Rivers E, Nguyen B, Havstad S, et al Early goal-directed therapy in the treatment

of severe sepsis and septic shock N Engl J Med 2001;345:1368-1377.

2 Dellinger RP, Levy MM, Carlet JM, et al Surviving Sepsis Campaign: tional guidelines for management of severe sepsis and septic shock: 2008 Crit Care Med 2008;36:296-327.

interna-Antibiotic strategies in severe nosocomial sepsis: Why do we not escalate more often?

de-Heenen S, Jacobs F, Vincent J-L (Erasme Hosp, Brussels, Belgium)

Crit Care Med 40:1404-1409, 2012

Objectives.dTo assess the use of antibiotic de-escalation in patientswith hospital-acquired severe sepsis in an academic setting

Design.dWe reviewed all episodes of severe sepsis treated over a 1-yrperiod in the department of intensive care Antimicrobial therapy wasconsidered as appropriate when the antimicrobial had in vitro activityagainst the causative microorganisms According to the therapeutic strategy

in the 5 days after the start of antimicrobial therapy, we classified patientsinto four groups: de-escalation (interruption of an antimicrobial agent orchange of antibiotic to one with a narrower spectrum); no change in antibio-therapy; escalation (addition of a new antimicrobial agent or change in anti-biotic to one with a broader spectrum); and mixed changes

Setting.dA 35-bed medico-surgical intensive care department in whichantibiotic strategies are reviewed by infectious disease specialists threetimes per week

Patients.dOne hundred sixty-nine patients with 216 episodes of severesepsis attributable to a hospital-acquired infection who required broad-spectrum b-lactam antibiotics alone or in association with other anti-infectious agents

Measurements and Main Results.dThe major sources of infection werethe lungs (44%) and abdomen (38%) Microbiological data were available

in 167 of the 216 episodes (77%) Initial antimicrobial therapy was propriate in 27 episodes (16% of culture-positive episodes) De-escalationwas applied in 93 episodes (43%), escalation was applied in 22 episodes(10%), mixed changes were applied in 24 (11%) episodes, and therewas no change in empirical antibiotic therapy in 77 (36%) episodes Inthese 77 episodes, the reasons given for maintaining the initial antimicro-bial therapy included the sensitivity pattern of the causative organisms andprevious antibiotic therapy The number of episodes when the chance tode-escalate may have been missed was small (4 episodes [5%])

inap-Conclusion.dEven in a highly focused environment with close ration among intensivists and infectious disease specialists, de-escalationmay actually be possible in <50% of cases.

Early appropriate antibiotics are a cornerstone in the treatment of patients withsevere sepsis and septic shock Studies have found that both delays in initiationand the use of inappropriate empiric antibiotics (defined as antibiotics that donot cover organisms recovered later in cultures) are associated with increasedmorbidity and mortality This finding has led to recommendations supportingthe use of broad-spectrum antibiotics as the initial regimen in critically ill patientswith the goal of assuring appropriate coverage However, utilization of broad-spectrum antibiotics can also increase antibiotic pressure and facilitate the emer-gence of resistant bacteria, already a significant problem in most intensive careunits (ICUs) Rapid de-escalation is recommended to decrease the risk forincreased antibiotic resistance In this very interesting study, the authorsassessed how de-escalation is applied in patients with hospital-acquired severesepsis in an academic ICU In this study, most infections were in the lungs andabdomen De-escalation was applied in 43% of the cases, and initial antibioticswere inappropriate in 16% of culture-positive cases De-escalation was notassociated with increased morbidity and mortality It is important to recognizethat the results of this study may not be applicable to ICUs with different patterns

of antibiotic resistance or different set-ups However, in an academic ment with close collaboration with infection diseases, de-escalation waspossible in less than 50% of cases of severe nosocomial sepsis Clinicians canlearn from this study and continue to push for de-escalation as soon as possible

Crit Care Med 40:2945-2953, 2012

Objective.dTo evaluate the effects of single-dose etomidate on theadrenal axis and mortality in patients with severe sepsis and septic shock.Design.dA systematic review of randomized controlled trials andobservational studies with meta-analysis

Setting.dLiterature search of EMBASE, Medline, Cochrane Database,and Evidence-Based Medical Reviews

Subjects.dSepsis patients who received etomidate for rapid sequenceintubation

Interventions.dNone

Measurements and Main Results.dWe conducted a systematic review ofrandomized controlled trials and observational studies with meta-analysisassessing the effects of etomidate on adrenal insufficiency and all-causemortality published between January 1950 and February 2012 We only

Chapter 6eSepsis/Septic Shock / 145

FIGURE 2.dA, Pooled relative risks (RRs) for all-cause mortality: all studies B, Pooled RR for mortality: randomized controlled trials C, Pooled RR for mortality: 28-day mortality rates only CI, confi- dence interval (Reprinted from Chan CM, Mitchell AL, Shorr AF Etomidate is associated with mortality and adrenal insufficiency in sepsis: a meta-analysis Crit Care Med 2012;40:2945-2953, with permission from the Society of Critical Care Medicine and Lippincott Williams and Wilkins.)

examined studies including septic patients All-cause mortality served as ourprimary end point, whereas the prevalence of adrenal insufficiency was oursecondary end point Adrenal insufficiency was determined using a cosyn-tropin stimulation test in all studies We used a random effects model foranalysis; heterogeneity was assessed with the I2statistic Publication biaswas evaluated with Begg’s test Five studies were identified that assessedmortality in those who received etomidate A total of 865 subjects wereincluded Subjects who received etomidate were more likely to die (pooledrelative risk 1.20; 95% confidence interval 1.02e1.42; Q statistic, 4.20; I2statistic, 4.9%) Seven studies addressed the development of adrenalsuppression associated with the administration of etomidate; 1,303 subjectswere included Etomidate administration increased the likelihood of devel-oping adrenal insufficiency (pooled relative risk 1.33; 95% confidenceinterval 1.22e1.46; Q statistic, 10.7; I2 statistic, 43.9%).

Conclusions.dAdministration of etomidate for rapid sequence tion is associated with higher rates of adrenal insufficiency and mortality

intuba-in patients with sepsis (Fig 2)

Etomidate was initially developed as a continuous infusion drug for sedation

It was later shown to cause prolonged adrenal insufficiency and its use for longed sedation was abandoned However, some of its characteristics, mainlyits rapid onset/offset of action and its favorable hemodynamic profile, made it

pro-an attractive cpro-andidate for rapid sequence intubation In this context, clinicipro-ansstarted using a single dose of etomidate as part of their rapid sequence intuba-tion Concerns of the potential effects on adrenal function lead to severalstudies that once again documented suppression of adrenal function withadministration of etomidate The potential perils of this effect were particularlyconcerning in patients with severe sepsis and septic shock because of theirincreased risk of adrenal dysfunction from their underlying disease The lack

of studies showing a clear association between etomidate administration toincreased mortality in sepsis led to an ongoing debate on the topic This system-atic review and meta-analysis evaluated the hypothesis that the use of etomi-date during rapid sequence intubation increases the risk of death and inducesadrenal insufficiency in patients with sepsis The study showed a clear associ-ation between the administration of a single dose of etomidate and increasedmortality (Fig 2) It also showed that etomidate effectively produces adrenalinsufficiency as measured by cosyntropin stimulation testing in patients withsepsis The results of this study should serve as a strong indicator of the needfor clinicians to carefully weigh the benefits/risks of using etomidate as a part

of their rapid sequence intubation protocol Furthermore, it serves as a greatbasis for a randomized controlled study to further examine this issue

S Zanotti, MDChapter 6eSepsis/Septic Shock / 147

Severe Sepsis and Septic Shock in Pregnancy

Barton JR, Sibai BM (Central Baptist Hosp, Lexington, KY; Univ of TexasHealth Science Ctr, Houston)

Obstet Gynecol 120:689-706, 2012

Pregnancies complicated by severe sepsis and septic shock are associatedwith increased rates of preterm labor, fetal infection, and preterm delivery.Sepsis onset in pregnancy can be insidious, and patients may appear decep-tively well before rapidly deteriorating with the development of septicshock, multiple organ dysfunction syndrome, or death The outcome andsurvivability in severe sepsis and septic shock in pregnancy are improvedwith early detection, prompt recognition of the source of infection, and tar-geted therapy This improvement can be achieved by formulating a stepwiseapproach that consists of early provision of time-sensitive interventionssuch as: aggressive hydration (20 mL/kg of normal saline over the firsthour), initiation of appropriate empiric intravenous antibiotics (genta-micin, clindamycin, and penicillin) within 1 hour of diagnosis, centralhemodynamic monitoring, and the involvement of infectious diseasespecialists and critical care specialists familiar with the physiologic changes

in pregnancy Thorough physical examination and imaging techniques orempiric exploratory laparotomy are suggested to identify the septic source.Even with appropriate antibiotic therapy, patients may continue to deterio-rate unless septic foci (ie, abscess, necrotic tissue) are surgically excised Thedecision for delivery in the setting of antepartum severe sepsis or septicshock can be challenging but must be based on gestational age, maternalstatus, and fetal status The natural inclination is to proceed with emergentdelivery for a concerning fetal status, but it is imperative to stabilize themother first, because in doing so the fetal status will likewise improve.Prevention Aggressive treatment of sepsis can be expected to reduce theprogression to severe sepsis and septic shock and prevention strategiescan include preoperative skin preparations and prophylactic antibiotictherapy as well as appropriate immunizations (Fig 4)

Sepsis is one of the leading causes of morbidity and mortality in critically illpatients Pregnancy is associated with specific risk factors predisposing pregnantpatients to develop sepsis Furthermore, when pregnant patients develop severesepsis or septic shock, they are at risk for increased morbidity and mortality.Caring for critically ill pregnant patients poses a number of challenges for theintensivist First, there is the added stress of caring for more than 1 life at thesame time (mother and fetus) Concerns for using drugs, procedures, or diag-nostic tests that are potentially harmful for the fetus are a constant in thesecases Second, pregnancy is associated with significant physiological changesthat need to be considered by the intensivist, especially when providing organsupport and evaluating hemodynamics Finally, there are certain conditionsthat are unique to pregnancy that can compound the clinical picture in a pregnantpatient with severe sepsis or septic shock In this comprehensive review, Bartonand Sibai thoroughly discuss relevant aspects of caring for pregnant patients with

severe sepsis and septic shock One of the most valuable aspects of this review isthe algorithm presented for overall management (Fig 4) The paucity or lack ofstudies in this particular population makes review articles such as this one veryvaluable for practicing physicians.

Chapter 6eSepsis/Septic Shock / 149

The effectiveness of hypertonic saline and pentoxifylline (HTSePTX)resuscitation in haemorrhagic shock and sepsis tissue injury: Comparisonwith LR, HES, and LRePTX treatments

Kim HJ, Lee KH (Bucheon Hosp of Soonchunhyang Univ, South Korea; WonjuChristian Hosp of Yonsei Univ, South Korea)

Injury 43:1271-1276, 2012

Purpose.dTo compare lung and liver injury and laboratory results inhaemorrhagic shock and sepsis models treated with combinations oflactated Ringer’s solution (LR), 7.5% hypertonic saline (HTS), hydrox-yethyl starch (HES), and pentoxifylline (PTX)

Methods.dMale Sprague-Dawley rats (200e290 g) were assignedrandomly to one of four treatment groups (n¼ 16 per group): (1) LR;(2) HES; (3) LRePTX; and (4) HTSePTX Each group was subdividedinto (1) haemorrhagic shock (n¼ 8) and (2) sepsis (n ¼ 8) model groups

A venous catheter was used to inject resuscitation fluids, and an arterialcatheter was used to withdraw blood and monitor mean arterial pressure(MAP) Lung and liver histology, bronchoalveolar lavage (BAL) fluid, andcytokine levels were evaluated

Results.dThe mean lung injury score was 1.7 At 24 h after treatment,the total leucocyte count in the BAL fluid was significantly (p < 0.05) higherwith LR treatment (10 106±0.8) than with other treatments in the sepsismodel groups (HES, 6 106±1.2; LRePTX, 5  106±1.5; HTSePTX,

5 106±0.6) The higher total leucocyte count after LR treatmentwas attributable to a greater increase in the number of neutrophils(17 ± 1.5%) compared with increases after the other treatments (HES,

6 ± 0.8%; LRePTX, 10 ± 1.3%; HTSePTX, 5 ± 0.4%) In the sepsismodel groups, the total hepatic injury score was also significantly(p < 0.05) higher with LR treatment (9.9 ± 0.5) than with the other treat-ments (HES, 6.7 ± 0.8; LRePTX, 5.6 ± 0.7; HTSePTX, 3.1 ± 0.9).This also occurred in the shock model (LR, 10.6 ± 2.1; HES, 5.8 ± 0.9;LRePTX, 7.3 ± 0.9; HTSePTX, 3.5 ± 0.9) As compared with LR treat-ment, HTSePTX resuscitation resulted in a 49% decrease in TNF-a, 29%decrease in IL-1b, and 58% decrease in IL-6 in the shock model at 24 h(p < 0.05), and the respective decreases were 45, 24, and 35% in the sepsismodel (p < 0.05)

Conclusion.dHTSePTX was superior to HES, LRePTX, and LR fortreating shock and sepsis, and LRePTX and HES gave better resultsthan LR therapy alone

Sepsis and hemorrhagic shock are both pathologies that require large tative efforts; however, even with adequate resuscitation these can result in largefluid volumes Additionally, the resuscitative fluid of choice (usually lactatedRinger’s solution) can lead to proinflammatory states, exacerbating an alreadyinflammatory pathology As such, the authors looked into a fluid that wouldreduce fluid volume as well as decrease inflammatory effects Pentoxifylline, anonspecific phosphodiesterase inhibitor, was found to decrease neutrophil

resusci-activation and decrease organ injury; thus, it would provide an anti-inflammatoryeffect opposite that of lactated Ringer’s solution alone Other fluids that havebeen studied are hydroxyethyl starch, which decreased pulmonary microvascularpermeability in one study, and hypertonic saline Hypertonic saline has thebenefit of requiring minimum fluid for increased intravascular volume expansion(therefore improved organ perfusion) as well as immunomodulatory effects viadecreases in neutrophil accumulation and cytokine release.1

Rats were assigned to 4 different resuscitative fluids: lactated Ringer’s (LR),Hydroxyethyl starch (HES), lactated Ringer’s and pentoxifylline (LR-PTX),and hypertonic saline and pentoxifylline (HTS-PTX) They underwent eitherhemorrhagic shock or sepsis through intratracheal lipopolysaccharide adminis-tration Parameters observed were histology of the lung, particularly macrophageand neutrophil accumulation, as well as alveolar hemorrhage and edema

A bronchial lavage was performed for neutrophil and total leukocyte count.Histologic hepatic pathology was examined along with alanine aminotransferaselevels Finally, inflammatory cytokines tumor necrosis factor-alpha (TNF-a),interleukin-beta (IL-1b), and IL-6 were studied between the 4 resuscitativegroups

In the septic shock and hemorrhagic shock models, HTS-PTX performedsignificantly better than lactated Ringer’s solution alone with decreased inflam-matory effects, organ damage, and edema It also appeared superior to HES inthe hemorrhagic shock model

This study has very promising implications in sepsis and hemorrhagic shockresuscitative efforts From a fluid status standpoint, the consequences of massivefluid resuscitation can be just as morbid as the diseases themselvesdforexample, abdominal compartment syndrome, and pulmonary edema Usinghypertonic saline has a 2-fold benefit of reducing fluid requirements as well asits anti-inflammatory effects Combining these attributes with pentoxifylline,which is also anti-inflammatory and reduces organ injury, can produce a moreideal resuscitative fluid

S Zanotti, MDReference

1 Oliveira RP, Velasco I, Soriano FG, Friedman G Clinical review: hypertonic saline resuscitation in sepsis Crit Care 2002;6:418-423.

Transfusion of packed red blood cells is not associated with improvedcentral venous oxygen saturation or organ function in patients withseptic shock

Fuller BM, Gajera M, Schorr C, et al (Washington Univ School of Medicine, StLouis, MO; Cooper Univ Hosp, Camden, NJ)

Study Objective.dTo evaluate whether PRBC transfusion is associatedwith improved central venous oxygen saturation (ScvO2) or organ func-tion in patients with severe sepsis and septic shock receiving early goal-directed therapy (EGDT).

Methods.dRetrospective cohort study (n ¼ 93) of patients presentingwith severe sepsis or septic shock treated with EGDT

Results.dThirty-four of 93 patients received at least one PRBC sion The ScvO2 goal > 70% was achieved in 71.9% of the PRBC groupand 66.1% of the no-PRBC group (p¼ 0.30) There was no difference inthe change in Sequential Organ Failure Assessment (SOFA) score withinthe first 24 h in the PRBC group vs the no-PRBC group (8.6e8.3 vs5.8e5.6, p ¼ 0.85), time to achievement of central venous pressure

transfu-> 8 mm Hg (732 min vs 465 min, p¼ 0.14), or the use of norepinephrine

to maintain mean arterial pressure > 65 mm Hg (81.3% vs 83.8%,

p¼ 0.77)

Conclusions.dIn this study, the transfusion of PRBC was not associatedwith improved cellular oxygenation, as demonstrated by a lack ofimproved achievement of ScvO2> 70% Also, the transfusion of PRBCwas not associated with improved organ function or improved achieve-ment of the other goals of EGDT Further studies are needed to determinethe impact of transfusion of PRBC within the context of early resuscitation

of patients with septic shock

Packed red blood cell (PRBC) transfusion is a common practice in the sive care unit Despite evidence linking PRBC transfusion to adverse clinicaloutcomes, significant controversy exists over its use in early goal-directedtherapy (EGDT) sepsis resuscitation After achieving adequate volume statusand mean arterial pressure, the EGDT protocol recommends considering PRBCtransfusion for patients with persistent central venous oxygen saturation(ScvO2) less than 70% and hematocrit less than 30% The PRBC transfusion ispart of a bundle approach to improve global tissue hypoxia In this retrospectivestudy, the effect of PRBC transfusion on ScvO2and organ function was evalu-ated Data were collected on 93 patients with severe sepsis and septic shock.All were treated with EGDT protocol The PRBC group consisted of 34 patientswho received an average of 4.56 units of PRBC transfusion per patient Theunits were ordered during the first 6 hours and administered within the first

inten-24 hours of the start of resuscitation The primary outcome of ScvO2 greaterthan 70% was achieved in 71.9% of the PRBC group and 66.1% of the non-PRBC group

The secondary outcome included improvement in Sequential Organ FailureAssessment (SOFA) scores Both results did not reach statistical significance.The EGDT bundle is a complex approach with multiple components, whichincludes PRBC transfusion What makes this study interesting is that both thetransfused arm and nontransfused arm were resuscitated using the EGDTprotocol This allowed controlling of different factors and evaluating PRBCtransfusion as a single variable between the 2 groups

The results of the study are even more interesting On one hand, they areconsistent with previous data suggesting that physiologic indicators may notnecessarily identify patients that are likely to benefit from PRBC transfusion.This is attributed to the changes occurring at the cellular level during PRBCstorage Stored PRBCs seem to lose functional and structural capability toimprove tissue oxygen deficit On the other hand, these study results are inmajor conflict with those of the Rivers et al1trial Rivers et al showed significantimprovement in ScvO2after transfusion and decreased mortality when used as apart of the EGDT protocol in the emergency department (ED) within 6 hours ofthe sepsis diagnosis This difference could be in part secondary to the timing oftransfusion Sixty-four percent of the patients in the Rivers et al trial receivedtheir transfusions in the first 6 hours in the ED In this study, although PRBCtransfusions were ordered in ED, the time of administration was not clear Thetiming of transfusion could be an important factor in determining benefit Theother limitation could be the small sample size This might have contributed tothe insignificant improvement in the ScvO2and SOFA scores.

This study raises concerns regarding basic physiologic principles that involvered cell changes and oxygen delivery and consumption It can serve as hypoth-esis for larger prospective trials that should take into account potential modifierssuch as the age of the stored PRBCs and the timing of transfusion

Crit Care Med 40:725-730, 2012

Objectives.dThere has long-been controversy about the possible riority of norepinephrine compared to dopamine in the treatment ofshock The objective was to evaluate the effects of norepinephrine anddopamine on outcome and adverse events in patients with septic shock.Data Sources.dA systematic search of the MEDLINE, Embase, Scopus,and CENTRAL databases, and of Google Scholar, up to June 30, 2011.Study Selection and Data Extraction.dAll studies providing informa-tion on the outcome of patients with septic shock treated with dopaminecompared to norepinephrine were included Observational and randomizedtrials were analyzed separately Because time of outcome assessment variedamong trials, we evaluated 28-day mortality or closest estimate Heteroge-neity among trials was assessed using the Cochrane Q homogeneity test

supe-Chapter 6eSepsis/Septic Shock / 153

A Forest plot was constructed and the aggregate relative risk of death wascomputed Potential publication bias was evaluated using funnel plots.Methods and Main Results.dWe retrieved five observational (1,360patients) and six randomized (1,408 patients) trials, totaling 2,768 patients(1,474 who received norepinephrine and 1,294 who received dopamine) Inobservational studies, among which there was significant heterogeneity(p < 001), there was no difference in mortality (relative risk, 1.09; confi-dence interval, 0.84e1.41; p ¼ 72) A sensitivity analysis identified onetrial as being responsible for the heterogeneity; after exclusion of thattrial, no heterogeneity was observed and dopamine administration wasassociated with an increased risk of death (relative risk, 1.23; confidenceinterval, 1.05e1.43; p < 01) In randomized trials, for which no heteroge-neity or publication bias was detected (p¼ 77), dopamine was associatedwith an increased risk of death (relative risk, 1.12; confidence interval,1.01e1.20; p ¼ 035) In the two trials that reported arrhythmias, thesewere more frequent with dopamine than with norepinephrine (relativerisk, 2.34; confidence interval, 1.46e3.77; p ¼ 001).

Conclusions.dIn patients with septic shock, dopamine administration

is associated with greater mortality and a higher incidence of arrhythmicevents compared to norepinephrine administration

Investigation of the optimal vasopressor for the management of septic shockhas generated considerable controversy The 2008 Surviving Sepsis Campaign(SSC) guidelines recommend the use of norepinephrine or dopamine as first-line in septic shock However, dopamine has more recently been associatedwith increased mortality in several observational trials Prior meta-analyses wereunderpowered to demonstrate superiority, but the recent publication of large,randomized, clinical trials has added a significant volume of outcome datacomparing the 2 vasopressors De Backer et al performed a meta-analysis throughJune 2011 incorporating 5 observational and 6 randomized trials assessingshort-term mortality in septic shock patients treated with norepinephrine versusdopamine Analysis of all trials found no mortality difference in the presence ofsignificant heterogeneity Exclusion of one trial eliminated heterogeneity andshowed increased mortality with use of dopamine compared with norepinephrine(relative risk, 1.23) In addition, the number of arrhythmic events was significantlygreater in patients treated with dopamine versus those treated with norepineph-rine While the results of the meta-analysis were heavily influenced by the largesize of the Sepsis Occurrence in Acutely Ill Patients randomized trial, exclusion

of this trial produces similar mortality risk with wider confidence intervals fore, the authors’ conclusions are consistent with the growing body of evidencethat dopamine is associated with greater risk of death and arrhythmias than norepi-nephrine in the treatment of septic shock Moreover, this has led to the modifica-tion of the current 2013 SSC guidelines to downgrade the recommendation fordopamine as a first-line alternative to norepinephrine, specifying its use only inselect patients at low risk of tachyarrhythmias

There-E Damuth, MD

S Zanotti, MD

Red blood cell transfusions are associated with lower mortality in patientswith severe sepsis and septic shock: A propensity-matched analysisPark DW, Chun B-C, Kwon S-S, et al (Korea Univ College of Medicine, Ansan;Korea Univ College of Medicine, Seoul; et al)

Crit Care Med 40:3140-3145, 2012

Objectives.dTo evaluate the effects of transfusions in patients withsevere sepsis and septic shock on mortality

Design.dPropensity-matched analysis of a prospective observationaldatabase (April 2005 to February 2009)

Setting.dTwenty-two medical and surgical intensive care units in 12teaching hospitals in Korea

Patients.dOne thousand fifty-four patients with community-acquiredsevere sepsis and septic shock

Interventions.dNone

Measurements and Main Results.dOf the 1,054 patients, 407 (38.6%)received a blood transfusion The mean pretransfusion hemoglobin levelwas 7.7 ± 1.2 g/dL Transfused patients had higher 28-day and in-hospital mortality rates (32.7% vs 17.3%; p < 001, 41.3% vs 20.3%;

p < 001, respectively) and a longer duration of hospital stay (21 tile range, 10e35] vs 13 [interquartile range, 8e24] days; p < 001), butwere more severely ill at admission (lower systolic blood pressure, higherAcute Physiology and Chronic Health Evaluation II score, and SequentialOrgan Failure Assessment score at admission) In 152 pairs matchedaccording to the propensity score depending on patient transfusion status,transfused patients had a lower risk of 7-day (9.2% vs 27.0%; p < 001),28-day (24.3% vs 38.8%; p¼ 007), and in-hospital mortality rates(31.6% vs 42.8%; p¼ 044) After adjusting for blood transfusion as atime-dependent variable in multivariable analysis, blood transfusion wasindependently associated with lower risk of 7-day (hazard ratio 0.42,95% confidence interval 0.19e0.50, p ¼ 026), 28-day (hazard ratio 0.43,95% confidence interval 0.29e0.62, p < 001), and in-hospital mortality(hazard ratio 0.51, 95% confidence interval 0.39e0.69, p < 001)

[interquar-Conclusions.dIn this observational study of patients with acquired severe sepsis and septic shock, red blood cell transfusions wereassociated with lower risk of mortality

bene-In this study, the authors evaluated the effects of red blood cell transfusions

on mortality in patients with severe sepsis and septic shock A multicenterobservational study was conducted evaluating 1054 patients using the Korean

Chapter 6eSepsis/Septic Shock / 155

sepsis registry from April 2005 to February 2009 A total of 407 (38.6%) ofthese patients received a blood transfusion From the demographics, the trans-fused group was sicker as noted from a higher Charlson’s index score, APACHE

II score, and Sequential Organ Failure Assessment score Transfused patientshad both higher mortality and longer duration of hospital stays The authorsperformed a propensity-matched analysis that included 152 patients fromeach group Transfused patients had lower risk of 7-day, 28-day, and in-hospital mortality The authors concluded that blood transfusions may be asso-ciated with survival benefit and can be used in the critical care setting.This conclusion should be approached with caution, as there are a few pointsworthy of discussion

1 The characteristics of the propensity-matched patients show the fused group is sicker than the total number of nontransfused patients Thetransfused patients in the propensity-matched group are less sick than thegroup of all transfused patients In addition to acuity differences among thegroups, the indication for transfusion is not reported This raises the question

nontrans-of selection bias Because the hemoglobin in the nontransfused arm was notmeasured, it is not clear whether the hemoglobin in this group reached a crit-ical level below which patients needed to be transfused

2 The median hemoglobin level for all transfused patients was less than 7.7 g/dL

It would have been interesting to see the actual hemoglobin value in thepropensity-matched arm This raises the question whether there is a lowercutoff point below which transfusion could be beneficial

3 Furthermore, outcomes in severe sepsis and septic shock have been ated with multiple time-sensitive factors Those lacking in the study includetiming of antibiotics, lactate level, mean arterial pressure less than 65, ScVO2

associ-less than 70% These variables have important clinical implications on survival

4 Another limitation is that the study looked only at mortality without ining other possible adverse effects of transfusion that could explain thelonger hospital length of stay in the group of transfused patients

exam-Data concerning the negative effects of PRBC transfusion has been mountingover many years The positive conclusion of this study, despite its limitations,highlights the need for the topic of transfusion in severe sepsis and septicshock to be revisited

Z Kobeissi, MD

Red blood cell transfusions are associated with lower mortality in patientswith severe sepsis and septic shock: A propensity-matched analysisPark DW, Chun B-C, Kwon S-S, et al (Korea Univ College of Medicine, Ansan;Korea Univ College of Medicine, Seoul; et al)

Crit Care Med 40:3140-3145, 2012

Objectives.dTo evaluate the effects of transfusions in patients withsevere sepsis and septic shock on mortality

Design.dPropensity-matched analysis of a prospective observationaldatabase (April 2005 to February 2009).

Setting.dTwenty-two medical and surgical intensive care units in 12teaching hospitals in Korea

Patients.dOne thousand fifty-four patients with community-acquiredsevere sepsis and septic shock

Interventions.dNone

Measurements and Main Results.dOf the 1,054 patients, 407 (38.6%)received a blood transfusion The mean pretransfusion hemoglobin levelwas 7.7 ± 1.2 g/dL Transfused patients had higher 28-day and in-hospital mortality rates (32.7% vs 17.3%; p < 001, 41.3% vs 20.3%;

p < 001, respectively) and a longer duration of hospital stay (21 tile range, 10e35] vs 13 [interquartile range, 8e24] days; p < 001), butwere more severely ill at admission (lower systolic blood pressure, higherAcute Physiology and Chronic Health Evaluation II score, and SequentialOrgan Failure Assessment score at admission) In 152 pairs matched accord-ing to the propensity score depending on patient transfusion status, trans-fused patients had a lower risk of 7-day (9.2% vs 27.0%; p < 001),28-day (24.3% vs 38.8%; p¼ 007), and in-hospital mortality rates(31.6% vs 42.8%; p¼ 044) After adjusting for blood transfusion as atime-dependent variable in multivariable analysis, blood transfusion wasindependently associated with lower risk of 7-day (hazard ratio 0.42,95% confidence interval 0.19e0.50, p ¼ 026), 28-day (hazard ratio 0.43,95% confidence interval 0.29e0.62, p < 001), and in-hospital mortality(hazard ratio 0.51, 95% confidence interval 0.39e0.69, p < 001)

[interquar-Conclusions.dIn this observational study of patients with acquired severe sepsis and septic shock, red blood cell transfusions wereassociated with lower risk of mortality

One possible exception was the original EGDT study by Rivers et al, in whichPRBC transfusion was included as part of the algorithm to achieve an centralvenous oxygen saturation$70% However, the individual impact of transfusion

on outcome was not reported, although subjects in the treatment arm had overall

Chapter 6eSepsis/Septic Shock / 157

better outcomes In the study reviewed here, the investigators collected data on aprospective observational basis on more than 1000 patients with severe sepsisand septic shock admitted to a large number of medical and surgical ICUs toassess outcome differences based on the impact of PRBC transfusion.

Outcomes for transfused patients and nontransfused patients were compared

in both crude and multivariate analyses, as is typical for studies of this design

In addition, 152 matched pairs from these 2 groups were compared usingpropensity-matching, a relatively recently developed statistical technique thatattempts to account for the absence of randomization In effect, it is an attempt

to factor in the likelihood that a subject would have received the treatmentbeing studied based on a variety of variables or characteristics, which areincluded in a logistic regression model used to predict the probability (or so-called “propensity”) that the patient would have received the treatment underinvestigation In both the crude and basic multivariate analyses, the patientsreceiving transfusions had worse outcomes as measured by a number of param-eters This is likely not unexpected because they were sicker as measured by anumber of parameters and had lower hemoglobins However, when the propen-sity matched pairs were evaluated, the transfused patients had lower 7- and28-day and hospital mortality than did nontransfused patients These findingsare interesting, in that they are at odds with the significant preponderance of liter-ature that has accumulated in the past 2 decades regarding the impact of PRBCtransfusion on outcomes in a wide variety of critically ill patients However, theauthors point out that there have been other studies within the past few yearsalso indicating that septic patients may benefit from PRBC transfusion.Although several potential explanations are posited, no definitive ones areprovided Perhaps the most reasonable is that with the adoption of a restrictivetransfusion strategy, relatively strict adherence may show benefit if practitionerslimit transfusion to unstable patients with hemoglobins in the range of 7 g/dL,

as appears to have been done in this study Of course, as an observationalstudy, it is limited by the nature of its design In addition, although propensitymatching is intended to improve the accuracy and validity of nonrandomizedstudies, this methodology has its detractors, who feel it has the potential tointroduce its own confounding or biases into the analysis.1

Nevertheless, these findings add an interesting bit of food for thought to thecritical care menu, and they may well open the door for a new prospective study

on transfusion, this time targeting patients specifically with severe sepsis andseptic shock

D R Gerber, DOReference

1 Nuttall GA, Houle TT Liars, damn liars, and propensity scores Anesthesiology 2008;108:3-4.

Initial resuscitation guided by the Surviving Sepsis Campaignrecommendations and early echocardiographic assessment ofhemodynamics in intensive care unit septic patients: A pilot study

Bouferrache K, Amiel J-B, Chimot L, et al (Univ Hosp Ambroise Pare´,Boulogne, France; CHU de Limoges, France)

Crit Care Med 40:2821-2827, 2012

Objective.dTo compare therapeutic interventions during initial tation derived from echocardiographic assessment of hemodynamics andfrom the Surviving Sepsis Campaign guidelines in intensive care unit septicpatients

resusci-Design and Setting.dProspective, descriptive study in two intensivecare units of teaching hospitals

Methods.dThe number of ventilated patients with septic shock whowere studied was 46 Transesophageal echocardiography was first per-formed (T1 < 3 hrs after intensive care unit admission) to adapt therapyaccording to the following predefined hemodynamic profiles: fluid loading(index of collapsibility of the superior vena cava$ 36%), inotropic support(left ventricular fractional area change < 45% without relevant index ofcollapsibility of the superior vena cava), or increased vasopressor support(right ventricular systolic dysfunction, unremarkable transesophageal echo-cardiography study consistent with sustained vasoplegia) Agreement fortreatment decision between transesophageal echocardiography andSurviving Sepsis Campaign guidelines was evaluated A second transesopha-geal echocardiography assessment (T2) was performed to validate thera-peutic interventions

Results.dAlthough transesophageal echocardiography and SurvivingSepsis Campaign approaches were concordant to manage fluid loading in

32 of 46 patients (70%), echocardiography led to the absence of bloodvolume expansion in the remaining 14 patients who all had a central venouspressure < 12 mm Hg Accordingly, the agreement was weak between trans-esophageal echocardiography and Surviving Sepsis Campaign for the deci-sion of fluid loading (k: 0.37 [0.16;0.59]) With a cut-off value < 8 mm Hgfor central venous pressure, k was 0.33 [0.03;0.69] Inotropes wereprescribed based on transesophageal echocardiography assessment in 14patients but would have been decided in only four patients according toSurviving Sepsis Campaign guidelines As a result, the agreement betweenthe two approaches for the decision of inotropic support was weak (k:0.23 [0.04;0.50]) No right ventricular dysfunction was observed Nopatient had anemia and only three patients with transesophageal echocardi-ography documented left ventricular systolic dysfunction had a centralvenous oxygen saturation < 70%

Conclusions.dA weak agreement was found in the prescription offluid loading and inotropic support derived from early transesophageal

Chapter 6eSepsis/Septic Shock / 159

echocardiography assessment of hemodynamics and Surviving SepsisCampaign guidelines in patients presenting with septic shock.

The optimal modality for hemodynamic assessment in early treatment of septicshock remains a much-debated issue Research has evaluated static and dynamicparameters for assessing volume responsiveness This article describes a prospec-tive study in which the early use of transesophageal echocardiography (TEE) inmechanically ventilated septic shock patients was compared with the SurvivingSepsis Campaign (SSC) guidelines In particular, the study evaluates earlyvolume assessment with fluid loading and the implementation of inotropicsupport In the 46 patients studied, a TEE was performed 2 separate times; early(<3 hours) in their intensive care unit (ICU) stay and then again after interven-tion The key results of the study were as follows: First, a large proportion ofpatients (30%) did not receive further volume expansion as would have been rec-ommended by SSC guidelines for central venous pressure target because of theirroutine use of superior vena cava (SVC) collapsibility index to determine volumeresponsiveness Second, an additional 10 patients were placed on inotropictherapy based on TEE evaluation of left ventricular (LV) function when comparedwith indications for starting inotropes if following the SSC guidelines This led to

an appropriate conclusion by the authors that their routine use of TEE may lead toless fluid administration and more frequent use of inotropes than if SSC guide-lines were followed

It is important to emphasize that the specific population studied was septicshock patients requiring mechanical ventilation This is a key point in any discus-sion of volume responsiveness because all dynamic parameters for determiningvolume responsiveness have only been studied in appropriately sedated patients

on mechanical ventilation with >8 mL/kg tidal volume In this article, thedynamic parameter used was an SVC collapsibility index of >36%, which has avery well-reported sensitivity and specificity in predicting volume responsiveness.Additionally, the authors used transesophageal echocardiography for their hemo-dynamic assessment Certainly, transthoracic echocardiography to aid in hemody-namic assessment in the ICU has become increasingly common, although it stillwould not be considered routine in clinical practice worldwide However, TEEwould be even less common as part of the ICU clinician’s armamentarium forhemodynamic assessment It is this issue that severely limits any major clinicalimpact from this study An often-used argument in studies using echocardiog-raphy or other bedside ultrasound for critically ill patients is that the technique

is very practitioner-dependent This argument would be relevant here and possiblyeven more so with use of TEE as opposed to transthoracic echocardiogram.Based on the design of the study, no conclusions can be made about clinicalpatient-oriented outcomes However, this study makes some very importantpoints First, dynamic parameters for predicting volume responsiveness should

be used when available in the appropriate patient as outlined previously.Second, targets for initiation of inotropic support, specifically a central venousoxygen saturation <70%, often do not correlate with LV dysfunction based onechocardiographic visualization Ultimately, whether use of echocardiography

for hemodynamic assessment leads to improved patient-oriented outcomes

Am J Respir Crit Care Med 185:1088-1095, 2012

Rationale.dFever control may improve vascular tone and decreaseoxygen consumption, but fever may contribute to combat infection.Objectives.dTo determine whether fever control by external coolingdiminishes vasopressor requirements in septic shock

Methods.dIn a multicenter randomized controlled trial, febrile patientswith septic shock requiring vasopressors, mechanical ventilation, andsedation were allocated to external cooling (n¼ 101) to achieve normo-thermia (36.5e37C) for 48 hours or no external cooling (n¼ 99) Vaso-pressors were tapered to maintain the same blood pressure target in thetwo groups The primary endpoint was the number of patients with a50% decrease in baseline vasopressor dose after 48 hours

Measurements and Main Results.dBody temperature was significantlylower in the cooling group after 2 hours of treatment (36.8 ± 0.7 vs.38.4 ± 1.1C; P < 0.01) A 50% vasopressor dose decrease was signifi-cantly more common with external cooling from 12 hours of treatment(54 vs 20%; absolute difference, 34%; 95% confidence interval [95%CI],46 to 21; P < 0.001) but not at 48 hours (72 vs 61%; absolutedifference, 11%; 95% CI,23 to 2) Shock reversal during the intensivecare unit stay was significantly more common with cooling (86 vs 73%;absolute difference,13%; 95% CI, 2 to 25; P¼ 0.021) Day-14 mortalitywas significantly lower in the cooling group (19 vs 34%; absolute differ-ence,16%; 95% CI, 28 to 4; P ¼ 0.013)

Conclusions.dIn this study, fever control using external cooling wassafe and decreased vasopressor requirements and early mortality in septicshock

Fever in the intensive care unit (ICU) has been defined by the Society of ical Care Medicine as a temperature greater than 38.3C The incidence of fever inthe ICU ranges from 23% to 70%, with approximately half of these cases beingattributable to an infectious process Fever is often considered a common finding

Crit-in patients with sepsis However, at presentation, 10% of septic patients arehypothermic and 35% are normothermic Physician and staff response to fevervaries institutionally It is common for the patient to receive either pharmacologic

or mechanical antipyretic therapy Studies evaluating the effect of fever and its

Chapter 6eSepsis/Septic Shock / 161

modulation on patient outcomes have yielded conflicting results There arestudies suggesting a beneficial effect of fever on patient outcomes in sepsis Inone study of 218 patients with gram-negative bacteremia, the presence offever was found to have a positive correlation with survival.1In this study, failure

to mount a febrile response within the first 24 hours was associated withincreased mortality Another study of prospectively collected data found thatpatients with a temperature greater than 101.3F were 59% less likely to diewhen compared with patients with a normal temperature.2With regard to fevermodulation, the Fever and Antipyretic in Critically ill patients Evaluation trialfailed to show a mortality benefit in sepsis patients with external cooling.3Other studies have suggested that aggressive fever control is associated withhigher mortality

It is still unclear whether fever is globally beneficial or harmful in sepsis andperhaps more important whether modulation of fever improves patient outcomes

In this multicenter, randomized, controlled trial, the authors attempt to shine somelight on this much-debated issue Case study patients were allocated to externalcooling with the objective of achieving normothermia (36.5C to 37C) duringthe study period of 48 hours versus no external cooling in the control group.The primary endpoint of the study was the number of patients with a 50%decrease in baseline vasopressor dose after 48 hours Both groups had similardemographics with the exception of the cumulative dose of vasopressors (higher

in the control group (P¼ 03)) The study found that the percentage of patientswith a 50% vasopressor dose decrease versus baseline between 2 groups wasnot significantly different at 48 hours (primary endpoint) Shock reversal wassignificantly more common in the cooling group than in the control group (signif-icance persisted after adjustment for severity of illness and vasopressor doseimbalance) The investigator proposed that beneficial effect of cooling wascaused by decreased oxygen consumption along with early vasopressor sparing.Furthermore, the risk of death on Day 14 was significantly lower in the coolinggroup However, this difference in mortality was not significant at ICU or hospitaldischarge The incidence of acquired infections by Day 14 was 32.6 per 1000 ICUdays in the cooling group and 23.8 per 1000 ICU days in the control group.The use of targeted temperature management (TTM) has been found to bebeneficial in other critically ill patient populations (hypothermia after cardiacarrest) This study is a preliminary evaluation of a TTM targeting normothermia

in patients with severe sepsis and septic shock Although the study results are ofgreat interest, they by no means answer all our questions regarding this topic.Current clinical consensus recommends initiating treatment when fever isgreater than 40C with a goal to achieve a temperature range of 37.5C to38.4C Further research is needed to best define the value and the way toproceed with TTM in severe sepsis and septic shock

M Gajera, MD

S Zanotti, MDReferences

1 Bryant RE, Hood AF, Hood ED, et al Factors affecting mortality of gram-negative rod bacteremia Arch Intern Med 1971;127:120-128.

2 Sanga R, Zanotti S, Schorr C, et al Relation between temperature in the initial 24 hours in patients with severe sepsis or septic shock with mortality and length of stay in the ICU Crit Care 2012;16:P57.

3 Fever and Antipyretic in Critically ill patients Evaluation (FACE) Study Group; Lee BH, Inui D, Suh GY, et al Association of body temperature and antipyretic treatments with mortality of critically ill patients with and without sepsis: multi- centered prospective observational study Crit Care 2012;16:R33.

Fever Control Using External Cooling in Septic Shock: A RandomizedControlled Trial

Schortgen F, Clabault K, Katsahian S, et al (Groupe Hospitalier Henri Mondor,Cre´teil, France; Centre Hospitalier Universitaire de Rouen, Rouen, France; et al)

Am J Respir Crit Care Med 185:1088-1095, 2012

Rationale.dFever control may improve vascular tone and decreaseoxygen consumption, but fever may contribute to combat infection.Objectives.dTo determine whether fever control by external coolingdiminishes vasopressor requirements in septic shock

Methods.dIn a multicenter randomized controlled trial, febrile patientswith septic shock requiring vasopressors, mechanical ventilation, andsedation were allocated to external cooling (n¼ 101) to achieve normo-thermia (36.5e37C) for 48 hours or no external cooling (n¼ 99) Vaso-pressors were tapered to maintain the same blood pressure target in thetwo groups The primary endpoint was the number of patients with a50% decrease in baseline vasopressor dose after 48 hours

Measurements and Main Results.dBody temperature was significantlylower in the cooling group after 2 hours of treatment (36.8 ± 0.7 vs.38.4 ± 1.1C; P < 0.01) A 50% vasopressor dose decrease was signifi-cantly more common with external cooling from 12 hours of treatment(54 vs 20%; absolute difference, 34%; 95% confidence interval [95%CI],46 to 21; P < 0.001) but not at 48 hours (72 vs 61%; absolutedifference, 11%; 95% CI,23 to 2) Shock reversal during the intensivecare unit stay was significantly more common with cooling (86 vs 73%;absolute difference, 13%; 95% CI, 2 to 25; P¼ 0.021) Day-14 mortalitywas significantly lower in the cooling group (19 vs 34%; absolute differ-ence,16%; 95% CI, 28 to 4; P ¼ 0.013)

Conclusions.dIn this study, fever control using external cooling wassafe and decreased vasopressor requirements and early mortality in septicshock

Fever is one of the defining criteria for systemic inflammatory response andthe varying degrees of sepsis.1-3 Although fever control is intuitive, this studyanswers the question about whether it is efficacious The authors demonstratereduced use of vasoactive drugs, and remarkably, with a short 48-hour inter-vention, improved outcomes (Fig 5 in the original article)

Chapter 6eSepsis/Septic Shock / 163

These data are obtained from medical rather than surgical patients In ular, neurosurgical patients are excluded This is 1 patient group in whichhyperthermia is clearly deleterious.

partic-Why did outcome improve with external cooling? It is possible that demands

on oxygen consumption were reduced.4Toxicity of vasopressor utilization wasalso decreased because these agents were required in smaller amounts.Although therapies combined with cooling are discussed, the etiology of septicshock is not revealed

Quick examination of Fig 3 from the original article reveals that although thetemperature difference was statistically significant, the “cooled” group did nothave drastic temperature reduction In summary, this work suggests that briefnormalization of a temperature curve may improve outcome in patients withfever complicating severe infection

D J Dries, MSE, MDReferences

1 Su F, Nguyen ND, Wang Z, Cai Y, Rogiers P, Vincent JL Fever control in septic shock: beneficial or harmful? Shock 2005;23:516-520.

2 Ryan M, Levy MM Clinical review: fever in intensive care unit patients Crit Care 2003;7:221-225.

3 Peres Bota D, Lopes Ferreira F, Me´lot C, Vincent JL Body temperature alterations

in the critically ill Intensive Care Med 2004;30:811-816.

4 Manthous CA, Hall JB, Olson D, et al Effect of cooling on oxygen consumption in febrile critically ill patients Am J Respir Crit Care Med 1995;151:10-14.

Physical and Mental Health in Patients and Spouses After Intensive Care ofSevere Sepsis: A Dyadic Perspective on Long-Term Sequelae Testing theActorePartner Interdependence Model

Rosendahl J, Brunkhorst FM, Jaenichen D, et al (Friedrich-Schiller Univ, Jena,Germany)

Crit Care Med 41:69-75, 2013

Objective.dTo examine the physical and mental long-term consequences

of intensive care treatment for severe sepsis in patients and their spousesunder consideration of a dyadic perspective using the ActorePartnerInterdependence Model

Design.dProspective study

Setting.dPatients and spouses who had requested advice from theGerman Sepsis Aid’s National Helpline were invited to participate.Subjects.dWe included 55 patients who survived severe sepsis and theirspouses an average of 55 months after ICU discharge

Measurements and Main Results.dThe Hospital Anxiety and sion Scale, the Short Form-12 Health Survey, the Posttraumatic StressScale-10, and the Giessen Subjective Complaints List-24 were used TheActorePartner Interdependence Model was tested using multilevelmodeling with the actor effect representing the impact of a person’s post-traumatic stress symptoms on his or her own mental health-related quality

Depres-of life and the partner effect characterized by the impact Depres-of a person’s traumatic stress symptoms on his or her partner’s mental health-relatedquality of life A significant proportion of patients and spouses (26%e42%) showed clinically relevant scores of anxiety and depression; approx-imately two thirds of both, patients and spouses, reported posttraumaticstress symptoms defined as clinically relevant Compared with normativesamples, patients reported greater anxiety, poorer mental and physicalhealth-related quality of life, and greater exhaustion; spouses had animpaired mental health-related quality of life and increased anxiety.Testing the ActorePartner Interdependence Model revealed that posttrau-matic stress symptoms were related to patients’ (b ¼ 0.71, 95% confi-

confidence interval 0.79 to 0.46) own mental health-related quality

of life Posttraumatic stress symptoms further influenced the mentalhealth-related quality of life of the respective other (b ¼ 0.18, 95% confi-dence interval0.35 to 0.003 for patients; b ¼ 0.15, 95% confidenceinterval0.32 to 0.02 for spouses)

FIGURE 2.dPatients’ and spouses’ physical and mental health in comparison to normative samples Differences expressed as standardized mean differences with 95% confidence interval Mental health- related quality of life (HRQOL) ¼ mental component summary of Short Form-12 Health Survey (SF-12); Physical HRQOL ¼ physical component summary of SF-12 (Reprinted from Rosendahl J, Brunkhorst FM, Jaenichen D, et al Physical and mental health in patients and spouses after intensive care

of severe sepsis: a dyadic perspective on long-term sequelae testing the actorepartner interdependence model Crit Care Med 2013;41:69-75, with permission from the Society of Critical Care Medicine and Lippincott Williams & Wilkins.)

Chapter 6eSepsis/Septic Shock / 165

Conclusions.dInterventions to treat posttraumatic stress symptomsafter critical illness to improve mental health-related quality of life shouldnot only include patients, but also consider spouses (Fig 2).

prob-The results of this study showed that an average of 55 months after intensivecare of severe sepsis, patients show significant impairment of physical andmental healtherelated quality of life Furthermore, spouses reported significantlydecreased mental healtherelated quality of life when compared with the normalpopulation (Fig 2) The study found a significant dyadic association betweenpatients and spouses regarding mental health Anxiety and depression scores,posttraumatic stress symptoms, and mental healtherelated qualities of lifewere significantly related between patients and spouses

Notwithstanding the limitations of this study, these results are an importantaddition to our evolving understanding of the long-term consequences ofsurviving severe sepsis-related critical illness The results of this study stronglysuggest that patients and their spouses act as emotionally interdependent enti-ties with interrelated physical and mental health Clinicians should think aboutthe long-term effects on their patients as well as their families when counselingthem in the intensive care unit

S Zanotti, MD

An evaluation of the diagnostic accuracy of the 1991 American College ofChest Physicians/Society of Critical Care Medicine and the 2001 Society ofCritical Care Medicine/European Society of Intensive Care Medicine/American College of Chest Physicians/American Thoracic Society/Surgical Infection Society sepsis definition

Zhao H, Heard SO, Mullen MT, et al (Univ of Massachusetts Med School,Worcester; et al)

Crit Care Med 40:1700-1706, 2012

Objectives.dLimited research has been conducted to compare the testcharacteristics of the 1991 and 2001 sepsis consensus definitions Thisstudy assessed the accuracy of the two sepsis consensus definitionsamong adult critically ill patients compared to sepsis case adjudication

by three senior clinicians

Design.dObservational study of patients admitted to intensive care units.Setting.dSeven intensive care units of an academic medical center.Patients.dA random sample of 960 patients from all adult intensivecare unit patients between October 2007 and December 2008.

Intervention.dNone

Measurements and Main Results.dSensitivity, specificity, and the areaunder the receiver operating characteristic curve for the two consensus defi-nitions were calculated by comparing the number of patients who met ordid not meet consensus definitions vs the number of patients who were

or were not diagnosed with sepsis by adjudication The 1991 sepsis tion had a high sensitivity of 94.6%, but a low specificity of 61.0% The

defini-2001 sepsis definition had a slightly increased sensitivity but a decreasedspecificity, which were 96.9% and 58.3%, respectively The areas underthe receiver operating characteristic curve for the two definitions werenot statistically different (0.778 and 0.776, respectively) The sensitivitiesand areas under the receiver operating characteristic curve of both defini-tions were lower at the 24-hr time window level than those of the intensivecare unit stay level, though their specificities increased slightly Fever, highwhite blood cell count or immature forms, low Glasgow coma score,edema, positive fluid balance, high cardiac index, low Pao2/Fio2 ratio,and high levels of creatinine and lactate were significantly associated withsepsis by both definitions and adjudication

Conclusions.dBoth the 1991 and the 2001 sepsis definition have a highsensitivity but low specificity; the 2001 definition has a slightly increasedsensitivity but a decreased specificity compared to the 1991 definition.The diagnostic performances of both definitions were suboptimal

A parsimonious set of significant predictors for sepsis diagnosis is likely

to improve current sepsis case definitions

Clinicians have recognized sepsis as an important disease for centuries It iscurrently considered one of the leading causes of morbidity and mortality in crit-ically ill patients Sepsis is an important area of research in critical care, withmillions of dollars invested in the search for new therapeutic interventions Yet,despite all this, we are still trying to find clear, concise definitions that can helpclinicians diagnose sepsis at the bedside In 1991, the American College ofChest Physicians and the Society of Critical Care Medicine convened a confer-ence in an attempt to provide a framework of standardized definitions for sepsis.1This consensus conference produced definitions for sepsis, severe sepsis, andseptic shock and coined the term systemic inflammatory response syndrome(SIRS) Although these definitions helped establish a common language forresearch, they often were criticized for their lack of specificity at the bedside

To this point, critics argued that SIRS was often found in critically ill patientsand was not a good discriminator for sepsis/infection In 2001, a new consensusconference was convened with the goal of revising and improving the 1991 defi-nitions The new definition expanded the SIRS criteria to include a list of general,inflammatory, hemodynamic, organ dysfunction, or tissue perfusion criteria.2

Chapter 6eSepsis/Septic Shock / 167

There has been a paucity of research evaluating the effect of the 1991 and

2001 definitions In this study, the authors assessed the test characteristics(sensitivity, specificity, and area under the receiver operating characteristic[ROC] curve) of the 1991 and 2001 consensus definitions compared with sepsiscase adjudication by 3 senior intensive care clinicians The study found that the

1991 sepsis definition had a high sensitivity (94.6%) but a low specificity (61%).The 2001 sepsis definition had a slightly increased sensitivity (96.9%) but adecreased specificity (58.3%) The areas under the ROC curve were essentiallythe same for both definitions (0.778 and 0.776, respectively) The studyutilized a logistic regression to identify specific criteria that were significantlyassociated with sepsis diagnosis both by the definitions and case adjudication.The criteria that were more likely to be associated with a final diagnosis of sepsisincluded fever, high white cell count or immature forms, low Glasgow comascore, edema, positive fluid balance, high cardiac index, low PaO2/FiO2 ratio,high level of creatinine, and high levels of lactate Despite its limitations, this is

an important study because it is the first to utilize a methodology to evaluatehow the consensus definitions for sepsis perform The bottom line is that boththe 1991 and 2001 consensus definitions are suboptimal in the clinical arena.The results of this study may be important to consider in future attempts to refineour definitions of sepsis

S Zanotti, MD

References

1 Bone RC, Balk RA, Cerra FB, et al Definitions for sepsis and organ failure and lines for the use of innovative therapies in sepsis The ACCP/SCCM Consensus Conference Committee American College of Chest Physicians/Society of Critical Care Medicine Chest 1992;101:1644-1655.

guide-2 Levy MM, Fink MP, Marshall JC, et al 2001 SCCM/ESICM/ACCP/ATS/SIS national Sepsis Definitions Conference Crit Care Med 2003;31:1250-1256.

Inter-Septic Shock Attributed to Candida Infection: Importance of EmpiricTherapy and Source Control

Kollef M, Micek S, Hampton N, et al (Washington Univ School of Medicine, StLouis, MO; Barnes-Jewish Hosp, St Louis, MO; Hosp Informatics Group, StLouis, MO; et al)

Clin Infect Dis 54:1739-1746, 2012

Background.dDelayed treatment of candidemia has previously beenshown to be an important determinant of patient outcome However, septicshock attributed to Candida infection and its determinants of outcome havenot been previously evaluated in a large patient population

Methods.dA retrospective cohort study of hospitalized patients withseptic shock and blood cultures positive for Candida species was con-ducted at Barnes-Jewish Hospital, a 1250-bed urban teaching hospital(January 2002eDecember 2010)

Results.dTwo hundred twenty-four consecutive patients with septicshock and a positive blood culture for Candida species were identified.Death during hospitalization occurred among 155 (63.5%) patients Thehospital mortality rate for patients having adequate source control andantifungal therapy administered within 24 hours of the onset of shockwas 52.8% (n¼ 142), compared to a mortality rate of 97.6% (n ¼ 82)

in patients who did not have these goals attained (P < 001) Multivariatelogistic regression analysis demonstrated that delayed antifungal treatment(adjusted odds ratio [AOR], 33.75; 95% confidence interval [CI],9.65e118.04; P ¼.005) and failure to achieve timely source control(AOR, 77.40; 95% CI, 21.52e278.38; P ¼.001) were independently asso-ciated with a greater risk of hospital mortality

Conclusions.dThe risk of death is exceptionally high among patientswith septic shock attributed to Candida infection Efforts aimed at timelysource control and antifungal treatment are likely to be associated withimproved clinical outcomes

Severe sepsis and septic shock are among the leading causes of morbidity andmortality in critically ill patients The cornerstone of treatment for septic shock isbased on rapid administration of appropriate antimicrobials, hemodynamicsupport, and source control when possible There is a clear relationship betweendelays in the initiation of appropriate antimicrobials and increased mortality.The number of cases of septic shock with positive blood cultures for Candida

is somewhere in the range of 15% Over time, cases of septic shock resultingfrom Candida seem to be increasing Septic shock due to Candida has tradition-ally been associated with very poor outcomes Many clinicians attribute this tothe fact that patients who develop septic shock from Candida are usually sickerpatients with numerous comorbid conditions Alternatively, others have pointedout that we usually recognize Candida as a potential cause of septic shock after apositive blood culture By the nature of how long it takes for these blood cultures

to be reported as positive, it is usually associated with significant delays in theinitiation of appropriate antifungals This delay in the initiation of appropriateantifungal therapy could also represent a significant factor in determining pooroutcomes in fungal septic shock

This study had 2 goals: (1) to evaluate the appropriateness of antimicrobialtherapy prescribed for patients with septic shock attributed to Candida infectionand (2) to examine the influence of appropriate antimicrobial therapy on patientoutcomes This single-center retrospective cohort study identified 224 consecu-tive patients with septic shock and a blood culture positive for Candida species.Overall hospital mortality was high at 63.5% In patients with adequate sourcecontrol and antifungal therapy administered within 24 hours of onset of septicshock, hospital mortality was 52.8% In patients who did not meet these criteria,mortality was significantly higher (97.6%) Furthermore, multivariate logisticregression analysis demonstrated that delayed antifungal treatment and failure

to achieve timely source control were independent factors associated withgreater mortality

Chapter 6eSepsis/Septic Shock / 169

This study shows the exceptionally high risk of death among patients withseptic shock resulting from Candida It also provides strong evidence of theimportance of initiation of early appropriate antifungal therapy and timely sourcecontrol Considering how most arrive at a diagnosis of Candida as the cause ofseptic shock, it is evident that we need to start appropriate therapy much earlierthan the results of blood cultures are finalized Pending new methods to identifypathogens, clinicians should consider carefully the need to initiate early empiricantifungals in appropriate patients with septic shock.

S Zanotti, MD

Effect of Bedside Ultrasonography on the Certainty of Physician ClinicalDecisionmaking for Septic Patients in the Emergency DepartmentHaydar SA, Moore ET, Higgins GL III, et al (Maine Med Ctr, Portland)Ann Emerg Med 60:346-358, 2012

Study Objective.dSepsis protocols promote aggressive patient ment, including invasive procedures After the provision of point-of-careultrasonographic markers of volume status and cardiac function, weseek to evaluate changes in emergency physician clinical decisionmakingand physician assessments about the clinical utility of the point-of-careultrasonographic data when caring for adult sepsis patients

manage-Methods.dFor this prospective before-and-after study, patients with pected sepsis received point-of-care ultrasonography to determine cardiaccontractility, inferior vena cava diameter, and inferior vena cava collaps-ibility Physician reports of treatment plans, presumed causes of observedvital sign abnormalities, and degree of certainty were compared beforeand after knowledge of point-of-care ultrasonographic findings The clin-ical utility of point-of-care ultrasonographic data was also evaluated.Results.dSeventy-four adult sepsis patients were enrolled: 27 (37%)sepsis, 30 (40%) severe sepsis, 16 (22%) septic shock, and 1 (1%) systemicinflammatory response syndrome After receipt of point-of-care ultrasono-graphic data, physicians altered the presumed primary cause of vital signabnormalities in 12 cases (17% [95% confidence interval {CI} 8% to25%]) and procedural intervention plans in 20 cases (27% [95% CI17% to 37%]) Overall treatment plans were changed in 39 cases (53%[95% CI 41% to 64%]) Certainty increased in 47 (71%) cases anddecreased in 19 (29%) Measured on a 100-mm visual analog scale, themean clinical utility score was 65 mm (SD 29; 95% CI 58 to 72), withusefulness reported in all cases

sus-Conclusion.dEmergency physicians found point-of-care graphic data about cardiac contractility, inferior vena cava diameter,and inferior vena cava collapsibility to be clinically useful in treatingadult patients with sepsis Increased certainty followed acquisition ofpoint-of-care ultrasonographic data in most instances Point-of-care

ultrasono-ultrasonography appears to be a useful modality in evaluating and treatingadult sepsis patients.

Following introduction of early goal-directed therapy and the publication ofthe Surviving Sepsis Campaign guidelines, many emergency departments initi-ated sepsis protocols to aid in early identification and treatment of severe sepsisand septic shock patients As noted in this study and others, there has been areluctance to adhere strictly to these guidelines partly due to assessment ofcentral venous pressure and central venous oxygen saturation with the relativelyinvasive procedure of central venous catheter placement Use of bedside ultra-sound has been proposed as a potential noninvasive method to guide initialmanagement in some of these patients The objectives of this prospective ques-tionnaire-based study were to determine if certain early bedside ultrasound inter-ventions in septic patients affected physician decision-making or changedphysician-perceived clinical utility of bedside ultrasound in septic patients Clin-ical utility was defined as the degree to which the ultrasound data influenced theirmanagement decisions Data were collected using a before-and-after question-naire regarding treatment plans, degree of certainty, and clinical utility, as definedpreviously The specific ultrasound interventions were assessment of inferiorvena cava (IVC) diameter and collapsibility and a single measurement to estimateleft ventricular (LV) ejection fraction The main results are summarized as follows:First, change in treatment plan occurred more than 50% of the time Second, 90%

of treating clinicians perceived the ultrasound data to be of positive clinical utility.The fundamental confounding issue with this study is whether the change intreatment plan was appropriate or not The assumption by the authors in theirdiscussion section is that the change in treatment plan was beneficial to thepatient This may have been the case but there is no way to know this based

on the study design and the recorded outcomes Simply more data, even whencollected noninvasively, are not always a helpful or even ethical approach It isquite unclear from the presented data that a 1-time evaluation of IVC diameterand collapsibility and LV function can lead to improved patient care regardless

of whether it changes decision plans of treating physicians Some examples ofpotential pitfalls would include the following: What treatment decisions can bereached after determining volume status by IVC diameter and/or collapsibility

in a patient who is not in shock or having signs of systemic hypoperfusion(elevated serum lactate)? How does the clinician react to a relatively low ejectionfraction in a severe sepsis patient who also appears to be underresuscitated based

on IVC collapsibility? With IVC diameter and collapsibility, is the clinician makingdecisions about merely volume status or predicting fluid responsiveness?

In the authors’ conclusion, they accurately state that clinicians caring forsepsis patients “considered point-of-care ultrasonographic data about cardiaccontractility, inferior vena cava diameter and collapsibility to be clinicallyuseful.” This is simply perceived clinical utility and one should not infer thatthis provides evidence for actual clinical utility of bedside ultrasound in thisspecific setting It is important to note 3 issues in regard to the assessment ofclinical utility: First, the questionnaire was apparently immediately filled outafter the ultrasound results were given to the treating clinician Thus, clinical

Chapter 6eSepsis/Septic Shock / 171

utility was assessed before seeing the results of any intervention Second, thetreating clinician did not perform the ultrasound This means that the clinicalutility was determined without taking into account the time needed to completethe procedure This is often an issue in clinical practice, where decisions to usebedside ultrasound from patient to patient are at least partially based on the timeneeded to perform the procedure Third, the time to perform the ultrasound wasnoted to be an average of 138 minutes after patient arrival Considering thesepatients required informed consent and a large portion were critically ill,completing the ultrasound just over 2 hours after arrival seems to be quite animpressive feat However, one could argue that in clinical practice, use ofbedside ultrasound in the critically ill patient is done even sooner and may beeven more clinically useful earlier on in the patient’s stay.

Crit Care Med 40:770-777, 2012

Objectives.dSystemic inflammatory response variability displays differingdegrees of organ damage and differing outcomes of sepsis C1-esterase inhib-itor, an endogenous acute-phase protein, regulates various inflammatory andanti-inflammatory pathways, including the kallikrein-kinin system and leuko-cyte activity This study assesses the influence of high-dose C1-esterase inhib-itor administration on systemic inflammatory response and survival inpatients with sepsis

Design.dOpen-label randomized controlled study

Setting.dSurgical and medical intensive care units of nine universityand city hospitals

Patients.dSixty-one patients with sepsis

Interventions.dPatients were randomized to receive either 12,000 U ofC1-esterase inhibitor infusions in addition to conventional treatment orconventional treatment only (n¼ 41 C1-esterase inhibitor, 20 controls).Blood samples for measurement of C1-esterase inhibitor, complement com-ponents C3 and C4, and C-reactive protein concentrations were drawn ondays 1, 3, 5, 7, 10, and 28

Measurements and Main Results.dQuartile analysis of C1-esteraseinhibitor activity in sepsis subjects revealed that the lowest quartile subgrouphad similar activity levels (0.7e1.2 U/L), when compared to healthy volun-teers (p > 05) These normal-level C1-esterase inhibitor sepsis patientsnevertheless displayed increased C-reactive protein (p¼ 04) productionand higher likelihoods of a more severe sepsis (p¼ 001) Overall, infusion

of C1-esterase inhibitor increased C1-esterase inhibitor (p < 005 vs control

on days 2, 3, and 5) functional activity, resulted in higher C3 levels (p < 05

vs control on days 2 and 3), followed by decreased C-reactive protein(p < 05 vs control on days 3 and 10) Simultaneously, C1-esterase inhibitorinfusion in sepsis patients was associated with reduced all-cause mortality(12% vs 45% in control, p¼ 008) as well as sepsis-related mortality(8% vs 45% in control, p¼ 001) assessed over 28 days The highest abso-lute reduction risk of 70% was achieved in sepsis patients with SimplifiedAcute Physiology Score II scores > 27

Conclusion.dIn the present study, patients in the lowest quartile of esterase inhibitor activity in combination with high C-reactive proteindemonstrated a higher risk of developing severe sepsis In general, high-dose C1-esterase inhibitor infusion down-regulated the systemic inflam-matory response and was associated with improved survival rates in sepsispatients, which could have important treatment and survival implicationsfor individuals with C1-esterase inhibitor functional deficiency (Fig 6)

With the recent withdrawal of recombinant human activated protein C fromthe market, the area of host response modulation in severe sepsis is once againorphan, with no approved treatments After decades of failed clinical trials fornovel host-modulating agents, it has become evident that patients with severesepsis and septic shock have a variable inflammatory response Therefore, it is

FIGURE 6.dA, All-cause mortality in the C1-esterase inhibitor (C1INH) (n ¼ 41) and control (n ¼ 20) groups; p ¼ 0.008 (log rank Mantel-Cox) B, Sepsis-related mortality in the C1INH (n ¼ 39) and control (n ¼ 20) groups; p ¼ 0.001 (log rank Mantel-Cox) C, Survival in the subgroup of Simplified Acute Physiology Score II >27 C1INH (n ¼ 15) group vs control group (n ¼ 10); p ¼ 0.0001 (log rank Mantel-Cox) Kaplan-Meier 28-day survival curves for the comparison of the treatment arm (C1-esterase inhibitor [C1INH] presented as solid line) and control arm (as dashed line) in subgroups stratified by the parameters of sepsis severity (Reprinted from Igonin AA, Protsenko DN, Galstyan GM, et al C1-esterase inhibitor infusion increases survival rates for patients with sepsis Crit Care Med 2012;40:770-777, with permission from the Society of Critical Care Medicine and Lippincott Williams & Wilkins.)

Chapter 6eSepsis/Septic Shock / 173

unlikely that 1 agent blocking a specific aspect of the host response will work forall patients with sepsis It seems that the future would hold treatments based onspecific biomarkers or inflammatory patterns for individual patients C1-esteraseinhibitor (C1INH) is an endogenous acute-phase protein with regulatory actions

of inflammatory and anti-inflammatory pathways C1INH has important effects

on the kallikrein-kinin system, the coagulation cascade, and leukocyte activity.These properties make C1INH an intriguing agent for treating severe sepsis.Furthermore, C1INH is already safely used to treat hereditary angioedema Inthis open-label randomized study, the investigators evaluated the use ofC1INH in the treatment of severe sepsis The study also measured various inflam-matory markers in each patient including levels of endogenous C1-esteraseinhibitor activity The most significant finding of this study was the association

of C1INH administration at high doses with improved survival in sepsis patients(Fig 6) All-cause mortality and sepsis-related mortality were both improvedsignificantly in the group of patients that received C1INH compared with thecontrol group Patients with increased severity as measured by a SimplifiedAcute Physiology Score II score of >27 had the greatest benefit in mortalityfrom treatment with C1INH Finally, when patients were classified into quartilesbased on C1-esterase inhibitor endogenous activity, the quartile with the lowestlevels responded the best to the C1INH treatment The results of this studyshould be considered preliminary Many previous novel agents showed promise

at this stage and later failed in larger confirmatory trials However, the results ofthis study may offer a different approach to the evaluation of novel agents for thetreatment of severe sepsis This new approach would involve the use ofbiomarkers, specifically endogenous C1-esterase inhibitor activity, to identify astudy group of patients with a higher likelihood of success

S Zanotti, MD

An evaluation of the diagnostic accuracy of the 1991 American College ofChest Physicians/Society of Critical Care Medicine and the 2001 Society ofCritical Care Medicine/European Society of Intensive Care Medicine/American College of Chest Physicians/American Thoracic Society/Surgical Infection Society sepsis definition

Zhao H, Heard SO, Mullen MT, et al (Univ of Massachusetts Med School,Worcester; et al)

Crit Care Med 40:1700-1706, 2012

Objectives.dLimited research has been conducted to compare the testcharacteristics of the 1991 and 2001 sepsis consensus definitions Thisstudy assessed the accuracy of the two sepsis consensus definitions amongadult critically ill patients compared to sepsis case adjudication by threesenior clinicians

Design.dObservational study of patients admitted to intensive care units.Setting.dSeven intensive care units of an academic medical center.Patients.dA random sample of 960 patients from all adult intensivecare unit patients between October 2007 and December 2008