Final 08 ssc guidelines


Published on

Published in: Health & Medicine, Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Final 08 ssc guidelines

  1. 1. Special ArticleSurviving Sepsis Campaign: International guidelines formanagement of severe sepsis and septic shock: 2008*R. Phillip Dellinger, MD; Mitchell M. Levy, MD; Jean M. Carlet, MD; Julian Bion, MD; Margaret M. Parker, MD; Roman Jaeschke, MD;Konrad Reinhart, MD; Derek C. Angus, MD, MPH; Christian Brun-Buisson, MD; Richard Beale, MD; Thierry Calandra, MD, PhD;Jean-Francois Dhainaut, MD; Herwig Gerlach, MD; Maurene Harvey, RN; John J. Marini, MD; John Marshall, MD; Marco Ranieri, MD;Graham Ramsay, MD; Jonathan Sevransky, MD; B. Taylor Thompson, MD; Sean Townsend, MD; Jeffrey S. Vender, MD;Janice L. Zimmerman, MD; Jean-Louis Vincent, MD, PhD; for the International Surviving Sepsis Campaign Guidelines Committee Objective: To provide an update to the original Surviving Sepsis Campaign sor therapy (1C); stress-dose steroid therapy given only in septic shock after bloodclinical management guidelines, “Surviving Sepsis Campaign Guidelines for Man- pressure is identified to be poorly responsive to fluid and vasopressor therapyagement of Severe Sepsis and Septic Shock,” published in 2004. (2C); recombinant activated protein C in patients with severe sepsis and clinical Design: Modified Delphi method with a consensus conference of 55 interna- assessment of high risk for death (2B except 2C for postoperative patients). In thetional experts, several subsequent meetings of subgroups and key individuals, absence of tissue hypoperfusion, coronary artery disease, or acute hemorrhage,teleconferences, and electronic-based discussion among subgroups and among target a hemoglobin of 7–9 g/dL (1B); a low tidal volume (1B) and limitation ofthe entire committee. This process was conducted independently of any industry inspiratory plateau pressure strategy (1C) for acute lung injury (ALI)/acute respi-funding. ratory distress syndrome (ARDS); application of at least a minimal amount of Methods: We used the Grades of Recommendation, Assessment, Development positive end-expiratory pressure in acute lung injury (1C); head of bed elevation inand Evaluation (GRADE) system to guide assessment of quality of evidence from mechanically ventilated patients unless contraindicated (1B); avoiding routine usehigh (A) to very low (D) and to determine the strength of recommendations. A of pulmonary artery catheters in ALI/ARDS (1A); to decrease days of mechanicalstrong recommendation (1) indicates that an intervention’s desirable effects ventilation and ICU length of stay, a conservative fluid strategy for patients withclearly outweigh its undesirable effects (risk, burden, cost) or clearly do not. Weak established ALI/ARDS who are not in shock (1C); protocols for weaning andrecommendations (2) indicate that the tradeoff between desirable and undesirable sedation/analgesia (1B); using either intermittent bolus sedation or continuouseffects is less clear. The grade of strong or weak is considered of greater clinical infusion sedation with daily interruptions or lightening (1B); avoidance of neuro-importance than a difference in letter level of quality of evidence. In areas without muscular blockers, if at all possible (1B); institution of glycemic control (1B),complete agreement, a formal process of resolution was developed and applied. targeting a blood glucose <150 mg/dL after initial stabilization (2C); equivalencyRecommendations are grouped into those directly targeting severe sepsis, rec- of continuous veno-veno hemofiltration or intermittent hemodialysis (2B); prophy-ommendations targeting general care of the critically ill patient that are consid- laxis for deep vein thrombosis (1A); use of stress ulcer prophylaxis to preventered high priority in severe sepsis, and pediatric considerations. upper gastrointestinal bleeding using H2 blockers (1A) or proton pump inhibitors Results: Key recommendations, listed by category, include early goal-directed (1B); and consideration of limitation of support where appropriate (1D). Recom-resuscitation of the septic patient during the first 6 hrs after recognition (1C); mendations specific to pediatric severe sepsis include greater use of physicalblood cultures before antibiotic therapy (1C); imaging studies performed promptly examination therapeutic end points (2C); dopamine as the first drug of choice forto confirm potential source of infection (1C); administration of broad-spectrum hypotension (2C); steroids only in children with suspected or proven adrenalantibiotic therapy within 1 hr of diagnosis of septic shock (1B) and severe sepsis insufficiency (2C); and a recommendation against the use of recombinant acti-without septic shock (1D); reassessment of antibiotic therapy with microbiology vated protein C in children (1B).and clinical data to narrow coverage, when appropriate (1C); a usual 7–10 days Conclusions: There was strong agreement among a large cohort of interna-of antibiotic therapy guided by clinical response (1D); source control with atten- tional experts regarding many level 1 recommendations for the best current caretion to the balance of risks and benefits of the chosen method (1C); administration of patients with severe sepsis. Evidenced-based recommendations regarding theof either crystalloid or colloid fluid resuscitation (1B); fluid challenge to restore acute management of sepsis and septic shock are the first step toward improvedmean circulating filling pressure (1C); reduction in rate of fluid administration with outcomes for this important group of critically ill patients.rising filing pressures and no improvement in tissue perfusion (1D); vasopressor KEY WORDS: sepsis; severe sepsis; septic shock; sepsis syndrome; infection;preference for norepinephrine or dopamine to maintain an initial target of mean Grades of Recommendation, Assessment, Development and Evaluation criteria;arterial pressure >65 mm Hg (1C); dobutamine inotropic therapy when cardiac GRADE; guidelines; evidence-based medicine; Surviving Sepsis Campaign; sepsisoutput remains low despite fluid resuscitation and combined inotropic/vasopres- bundles *The correct citation for this article is as follows. (JFD); Vivantes-Klinikum Neukoelin, Berlin, Germany (HG); tional Sepsis Forum,* Japanese Association for Acute Dellinger RP, Levy MM, Carlet, JM, et al: Surviving Consultants in Critical Care, Inc, Glenbrook, NV (MH); Univer- Medicine, Japanese Society of Intensive Care Medicine;Sepsis Campaign: International guidelines for manage- sity of Minnesota, St. Paul, MN (JJM); St. Michael’s Hospital, Society of Critical Care Medicine,* Society of Hospital Toronto, Ontario, Canada (JM); Università di Torino, Torino, Medicine,** Surgical Infection Society,* World Federationment of severe sepsis and septic shock: 2008 [pub- of Critical Care Nurses,** World Federation of Societies oflished correction appears in Crit Care Med 2008; 36: Italy (MR); West Hertfordshire Health Trust, Hemel Hemp- Intensive and Critical Care Medicine.** Participation and1394 –1396]. Crit Care Med 2008; 36:296 –327. stead, UK (GR); The Johns Hopkins University School of endorsement by the German Sepsis Society and the Latin From Cooper University Hospital, Camden, NJ (RPD); Medicine, Baltimore, MD (JS); Massachusetts General Hos- American Sepsis Institute. *Sponsor of 2004 guidelines.Rhode Island Hospital, Providence, RI (MML); Hospital Saint- pital, Boston, MA (BTT); Rhode Island Hospital, Providence, RI **Sponsors of 2008 guidelines who did not participateJoseph, Paris, France (JMC); Birmingham University, Bir- (ST); Evanston Northwestern Healthcare, Evanston, IL (JSV); formally in revision process. Members of the 2008 SSC The Methodist Hospital, Houston, TX (JLZ); Erasme University Guidelines Committee are listed in Appendix I. Appendix Jmingham, UK (JB); SUNY at Stony Brook, Stony Brook, NY provides author disclosure information.(MMP); McMaster University, Hamilton, Ontario, Canada (RJ); Hospital, Brussels, Belgium (JLV). Also published in Intensive Care Medicine (JanuaryFriedrich-Schiller-University of Jena, Jena, Germany (KR); Sponsoring organizations: American Association of 2008).University of Pittsburgh, Pittsburgh, PA (DCA); Hopital Henri Critical-Care Nurses,* American College of Chest Physi- For information regarding this article, E-mail: cians,* American College of Emergency Physicians,* Ca- Dellinger-Phil@CooperHealth.eduMondor, Créteil, France (CBB); Guy’s and St Thomas’ Hos- nadian Critical Care Society, European Society of Clinicalpital Trust, London, UK (RB); Centre Hospitalier Universitaire Copyright © 2007 by the Society of Critical Care Microbiology and Infectious Diseases,* European Society MedicineVaudois, Lausanne, Switzerland (TC); French Agency for of Intensive Care Medicine,* European Respiratory Soci-Evaluation of Research and Higher Education, Paris, France ety,* Indian Society of Critical Care Medicine,** Interna- DOI: 10.1097/01.CCM.0000298158.12101.41Crit Care Med 2008 Reprint 1
  2. 2. S evere sepsis (acute organ dys- Scheme 1. Diagnostic criteria for sepsis function secondary to infec- Infection, documented or suspected, and some of the following: tion) and septic shock (severe General variables sepsis plus hypotension not re- Fever (Ͼ38.3°C)versed with fluid resuscitation) are major Hypothermia (core temperature Ͻ36°C)healthcare problems, affecting millions of Heart rate Ͼ90 minϪ1 or Ͼ2 SD above the normal value for ageindividuals around the world each year, Tachypnea Altered mental statuskilling one in four (and often more), and Significant edema or positive fluid balance (Ͼ20 mL/kg over 24 hrs)increasing in incidence (1–5). Similar to Hyperglycemia (plasma glucose Ͼ140 mg/dL or 7.7 mmol/L) in the absence of diabetespolytrauma, acute myocardial infarction, Inflammatory variablesor stroke, the speed and appropriateness Leukocytosis (WBC count Ͼ12,000 ␮LϪ1) Leukopenia (WBC count Ͻ4000 ␮LϪ1)of therapy administered in the initial Normal WBC count with Ͼ10% immature formshours after severe sepsis develops are Plasma C-reactive protein Ͼ2 SD above the normal valuelikely to influence outcome. In 2004, an Plasma procalcitonin Ͼ2 SD above the normal valueinternational group of experts in the di- Hemodynamic variables Arterial hypotension (SBP Ͻ90 mm Hg; MAP Ͻ70 mm Hg; or an SBP decrease Ͼ40 mm Hgagnosis and management of infection and in adults or Ͻ2 SD below normal for age)sepsis, representing 11 organizations, Organ dysfunction variablespublished the first internationally ac- Arterial hypoxemia (PaO2/FIO2 Ͻ300)cepted guidelines that the bedside clini- Acute oliguria (urine output Ͻ0.5 mL/Kg hr or 45 mmol/L for at least 2 hrs, despite adequatecian could use to improve outcomes in fluid resuscitation)severe sepsis and septic shock (6, 7). Creatinine increase Ͼ0.5 mg/dL or 44.2 ␮mol/L Coagulation abnormalities (INR Ͼ1.5 or a PTT Ͼ60 secs)These guidelines represented phase II of Ileus (absent bowel sounds)the Surviving Sepsis Campaign (SSC), an Thrombocytopenia (platelet count, Ͻ100,000 ␮LϪ1)international effort to increase awareness Hyperbilirubinemia (plasma total bilirubin Ͼ4 mg/dL or 70 ␮mol/L)and improve outcomes in severe sepsis. Tissue perfusion variables Hyperlactatemia (Ͼ upper limit of lab normal)Joined by additional organizations, the Decreased capillary refill or mottlinggroup met again in 2006 and 2007 to Diagnostic criteria for sepsis in the pediatric population are signs and symptoms of inflammationupdate the guidelines document using a plus infection with hyper- or hypothermia (rectal temperature Ͼ38.5°C or Ͻ35°C),new evidence-based methodology system tachycardia (may be absent in hypothermic patients), and at least one of the followingfor assessing quality of evidence and indications of altered organ function: altered mental status, hypoxemia, increased serumstrength of recommendations (8 –11). lactate level, or bounding pulses. These recommendations are intended WBC, white blood cell; SBP, systolic blood pressure; MAP, mean arterial blood pressure; INR,to provide guidance for the clinician car- international normalized ratio; a PTT, activated partial thromboplastin for a patient with severe sepsis or Adapted from Levy MM, Fink MP, Marshall JC, et al: 2001 SCCM/ESICM/ACCP/ATS/SIS Interna-septic shock. Recommendations from tional Sepsis Definitions Conference. Crit Care Med 2003; 31:1250 –1256these guidelines cannot replace the clini-cian’s decision-making capability when Scheme 2.he or she is provided with a patient’sunique set of clinical variables. Most of Severe sepsis ϭ sepsis-induced tissue hypoperfusion or organ dysfunction (any of the followingthese recommendations are appropriate thought to be due to the infection) Sepsis-induced hypotensionfor the severe sepsis patient in both the Lactate greater than the upper limits of normal laboratory resultsintensive care unit (ICU) and non-ICU Urine output Ͻ0.5 mL/kg hr for Ͼ2 hrs, despite adequate fluid resuscitationsettings. In fact, the committee believes ALI with PaO2/FIO2 Ͻ250 in the absence of pneumonia as infection sourcethat currently, the greatest outcome im- ALI with PaO2/FIO2 Ͻ200 in the presence of pneumonia as infection sourceprovement can be made through educa- Creatinine Ͼ2.0 mg/dL (176.8 ␮mol/L) Bilirubin Ͼ2 mg/dL (34.2 ␮mol/L)tion and process change for those caring Platelet count Ͻ100,000for severe sepsis patients in the non-ICU Coagulopathy (INR Ͼ1.5)setting and across the spectrum of acutecare. It should also be noted that re- ALI, acute lung injury; INR, international normalized ratio.source limitations in some institutions Adapted from Levy MM, Fink MP, Marshall JC, et al: 2001 SCCM/ESICM/ACCP/ATS/SIS Interna-and countries may prevent physicians tional Sepsis Definitions Conference. Crit Care Med 2003; 31:1250 –1256. ACCP/SCCM Consensus Conference Committee: American College of Chest Physicians/Society of Critical Care Medicinefrom accomplishing particular recom- Consensus Conference: Definitions for sepsis and organ failure and guidelines for the use of innovativemendations. therapies in sepsis. Crit Care Med 1992; 20:864 – 874METHODS threshold for this dysfunction has varied blood pressure (SBP) Ͻ90 mm Hg or mean Sepsis is defined as infection plus sys- somewhat from one severe sepsis research arterial pressure Ͻ70 mm Hg or a SBPtemic manifestations of infection study to another. An example of typical decrease Ͼ40 mm Hg or Ͻ2 SD below nor-(Scheme 1) (12). Severe sepsis is defined thresholds identification of severe sepsis is mal for age in the absence of other causesas sepsis plus sepsis-induced organ dys- shown in Scheme 2 (12, 13). Sepsis- of hypotension. Septic shock is defined asfunction or tissue hypoperfusion. The induced hypotension is defined as a systolic sepsis-induced hypotension persisting de-2 Crit Care Med 2008 Reprint
  3. 3. spite adequate fluid resuscitation. Sepsis- ests; however, consensus as to thresh- (composed of a subset of the committeeinduced tissue hypoperfusion is defined as old for exclusion could not be reached. members) at the 2007 SCCM (Orlando,either septic shock, an elevated lactate, or Alternatively, the committee agreed to FL) and 2007 International Symposiumoliguria. ensure full disclosure and transparency on Intensive Care and Emergency Medi- The current clinical practice guidelines of all committee members’ potential cine (Brussels) meetings with recircula-build on the first and second editions from conflicts at time of publication. (See tion of deliberations and decisions to the2001 (discussed subsequently) and 2004 (6, disclosures at the end of this docu- entire group for comment or approval. At7, 14). The 2001 publication incorporated a ment.) the discretion of the chair and followingMEDLINE search for clinical trials in the The guidelines process included a adequate discussion, competing propos-preceding 10 yrs, supplemented by a man- modified Delphi method, a consensus als for wording of recommendations orual search of other relevant journals (14). conference, several subsequent meetings assigning strength of evidence were re-The 2004 publication incorporated the ev- of subgroups and key individuals, tele- solved by formal voting. On occasions,idence available through the end of 2003. conferences and electronic-based discus- voting was performed to give the com-The current publication is based on an up- sions among subgroups and members of mittee a sense of distribution of opinionsdated search into 2007 (see following meth- the entire committee, and two follow-up to facilitate additional discussion. Theods and rules). nominal group meetings in 2007. manuscript was edited for style and form The 2001 guidelines were coordinated Subgroups were formed, each charged by the writing committee with final ap-by the International Sepsis Forum; the with updating recommendations in spe- proval by section leads for their respec-2004 guidelines were funded by unre- cific areas, including corticosteroids, tive group assignment and then by thestricted educational grants from industry blood products, activated protein C, renal entire committee.and administered through the Society of replacement therapy, antibiotics, source The development of guidelines andCritical Care Medicine (SCCM), the Eu- control, and glucose control. Each sub- grading of recommendations for the 2004ropean Society of Intensive Care Medi- group was responsible for updating the guideline development process werecine (ESICM), and the International Sep- evidence (into 2007, with major addi- based on a system proposed by Sackettsis Forum. Two of the SSC administering tional elements of information incorpo- (15) in 1989, during one of the firstorganizations receive unrestricted indus- rated into the evolving manuscript American College of Chest Physicianstry funding to support SSC activities (ES- throughout 2006 and 2007). A separate (ACCP) conferences on the use of anti-ICM and SCCM), but none of this funding search was performed for each clearly de- thrombotic therapies. The revised guide-was used to support the 2006/2007 com- fined question. The committee chair lines recommendations are based on themittee meetings. worked with subgroup heads to identify GRADE system, a structured system for It is important to distinguish between pertinent search terms that always in- rating quality of evidence and gradingthe process of guidelines revision and the cluded, at a minimum, sepsis, severe sep- strength of recommendation in clinicalSSC. The SSC is partially funded by un- sis, septic shock, and sepsis syndrome practice (8 –11). The SSC Steering Com-restricted educational industry grants, crossed against the general topic area of mittee and individual authors collabo-including those from Edwards Life- the subgroup as well as pertinent key rated with GRADE representatives to ap-Sciences, Eli Lilly and Company, and words of the specific question posed. All ply the GRADE system to the SSCPhilips Medical Systems. SSC also re- questions of the previous guidelines pub- guidelines revision process. The mem-ceived funding from the Coalition for lications were searched, as were pertinent bers of GRADE group were directly in-Critical Care Excellence of the Society of new questions generated by general top- volved, either in person or via e-mail, inCritical Care Medicine. The great major- ic-related search or recent trials. Quality all discussions and deliberations amongity of industry funding has come from Eli of evidence was judged by predefined the guidelines committee members as toLilly and Company. Grades of Recommendation, Assessment, grading decisions. Subsequently, the SSC Current industry funding for the SSC Development and Evaluation (GRADE) authors used written material preparedis directed to the performance improve- criteria (discussed subsequently). Signif- by the GRADE group and conferred withment initiative. No industry funding was icant education of committee members GRADE group members who were avail-used in the guidelines revision process. on the GRADE approach was performed able at the first committee meeting and For both the 2004 and the 2006/2007 via e-mail before the first committee subsequent nominal group meetings.efforts, there were no members of the meeting and at the first meeting. Rules GRADE representatives were also used ascommittee from industry, no industry were distributed concerning assessing a resource throughout subgroup deliber-input into guidelines development, and the body of evidence, and GRADE experts industry presence at any of the were available for questions throughout The GRADE system is based on a se-meetings. Industry awareness or com- the process. Subgroups agreed electroni- quential assessment of the quality of ev-ment on the recommendations was not cally on draft proposals that were pre- idence, followed by assessment of the bal-allowed. No member of the guideline sented to committee meetings for general ance between benefits vs. risks, burden,committee received any honoraria for discussion. In January 2006, the entire and cost and, based on the preceding, de-any role in the 2004 or 2006/2007 group met during the 35th SCCM Critical velopment and grading of a managementguidelines process. The committee con- Care Congress in San Francisco, Califor- recommendations (9 –11). Keeping the rat-sidered the issue of recusement of indi- nia. The results of that discussion were ing of quality of evidence and strength ofvidual committee members during de- incorporated into the next version of rec- recommendation explicitly separate consti-liberation and decision making in areas ommendations and again discussed using tutes a crucial and defining feature of thewhere committee members had either electronic mail. Recommendations were GRADE approach. This system classifiesfinancial or academic competing inter- finalized during nominal group meetings quality of evidence as high (grade A), mod-Crit Care Med 2008 Reprint 3
  4. 4. erate (grade B), low (grade C), or very low Table 1. Determination of the quality of evidence(grade D). Randomized trials begin as high- ● Underlying methodologyquality evidence but may be downgraded A. RCTdue to limitations in implementation, in- B. Downgraded RCT or upgraded observational studiesconsistency or imprecision of the results, C. Well-done observational studiesindirectness of the evidence, and possible D. Case series or expert opinionreporting bias (Table 1). Examples of indi- ● Factors that may decrease the strength of evidence 1. Poor quality of planning and implementation of available RCTs, suggesting high likelihood ofrectness of the evidence include population biasstudied, interventions used, outcomes mea- 2. Inconsistency of results (including problems with subgroup analyses)sured, and how these relate to the question 3. Indirectness of evidence (differing population, intervention, control, outcomes, comparison)of interest. Observational (nonrandomized) 4. Imprecision of resultsstudies begin as low-quality evidence, but 5. High likelihood of reporting bias ● Main factors that may increase the strength of evidencethe quality level may be upgraded on the 1. Large magnitude of effect (direct evidence, RR Ͼ2 with no plausible confounders)basis of large magnitude of effect. An exam- 2. Very large magnitude of effect with RR Ͼ5 and no threats to validity (by two levels)ple of this is the quality of evidence for early 3. Dose-response gradientadministration of antibiotics. The GRADE system classifies recom- RCT, randomized controlled trial; RR, relative risk.mendations as strong (grade 1) or weak(grade 2). The grade of strong or weak is Table 2. Factors determining strong vs. weak recommendationconsidered of greater clinical importancethan a difference in letter level of quality What Should Be Considered Recommended Processof evidence. The committee assessed Quality of evidence The lower the quality of evidence, the less likely a strongwhether the desirable effects of adherence recommendationwill outweigh the undesirable effects, and Relative importance of the If values and preferences vary widely, a strongthe strength of a recommendation reflects outcomes recommendation becomes less likelythe group’s degree of confidence in that Baseline risks of outcomes The higher the risk, the greater the magnitude of benefitassessment (Table 2). A strong recommen- Magnitude of relative risk, Larger relative risk reductions or larger increases indation in favor of an intervention reflects including benefits, harms, and relative risk of harm make a strong recommendation burden more or less likely, respectivelythat the desirable effects of adherence to Absolute magnitude of the effect The larger the absolute benefits and harms, the greater ora recommendation (beneficial health out- lesser likelihood, respectively, of a strongcomes, less burden on staff and patients, recommendationand cost savings) will clearly outweigh Precision of the estimates of the The greater the precision, the more likely a strongthe undesirable effects (harms, more bur- effects recommendationden, and greater costs). A weak recom- Costs The higher the cost of treatment, the less likely a strongmendation in favor of an intervention in- recommendationdicates that the desirable effects ofadherence to a recommendation probablywill outweigh the undesirable effects, but make the recommendation less applicable. a recommendation was as follows: 1) tothe panel is not confident about these Being a strong recommendation does not give a recommendation a direction (for ortradeoffs— either because some of the ev- automatically imply standard of care. For against the given action), a majority ofidence is low quality (and thus there re- example, the strong recommendation for votes were to be in favor of that direction,mains uncertainty regarding the benefits administering antibiotics within 1 hr of the with Յ20% preferring the opposite direc-and risks) or the benefits and downsides diagnosis of severe sepsis, although desir- tion (there was a neutral vote allowed asare closely balanced. While the degree of able, is not currently standard of care as well); 2) to call a given recommendationconfidence is a continuum and there is verified by current practice (M Levy, per- strong rather than weak, Ն70% “strong”no precise threshold between a strong sonal communication, from first 8,000 pa- votes were required; 3) if Ͻ70% of votesand a weak recommendation, the pres- tients entered internationally into the SSC indicated “strong” preference, the recom-ence of important concerns about one or performance improvement database). The mendation was assigned a weak category ofmore of the preceding factors makes a implication of a weak recommendation is strength. We used a combination of modi-weak recommendation more likely. A that although a majority of well-informed fied Delphi process and nominal (expert)strong recommendation is worded as “we patients would accept it (but a substantial group techniques to ensure both depth andrecommend” and a weak recommenda- proportion would not), clinicians should breadth of review. The entire review grouption as “we suggest.” consider its use according to particular cir- (together with their parent organizations The implications of calling a recom- cumstance. as required) participated in the larger, iter-mendation strong are that most well- Differences of opinion among commit- ative, modified Delphi process. The smallerinformed patients would accept that inter- tee members about interpretation of evi- working group meetings, which took placevention and that most clinicians should use dence, wording of proposals, or strength of in person, functioned as the nominalit in most situations. There may be circum- recommendations were resolved using a groups. If a clear consensus could not bestances in which a strong recommendation specifically developed set of rules. We will obtained by polling within the nominalcannot or should not be followed for an describe this process in detail in a separate group meetings, the larger group was spe-individual patient because of that patient’s publication. In summary, the main ap- cifically asked to use the polling process.preferences or clinical characteristics that proach for converting diverse opinions into This was only required for corticosteroids4 Crit Care Med 2008 Reprint
  5. 5. and glycemic control. The larger group had Table 3. Initial resuscitation and infection issuesthe opportunity to review all outputs. In Strength of recommendation and quality of evidence have been assessed using the GRADE criteria,this way the entire review combined in- presented in parentheses after each guidelinetense focused discussion (nominal group) ● Indicates a strong recommendation, or “we recommend”with broader review and monitoring using ⅙ Indicates a weak recommendation, or “we suggest”the Delphi process. Initial resuscitation (first 6 hrs) Note: Refer to Tables 3–5 for con- ● Begin resuscitation immediately in patients with hypotension or elevated serum lactate Ͼ4densed adult recommendations. mmol/L; do not delay pending ICU admission (1C) ● Resuscitation goals (1C) CVP 8–12 mm HgaI. MANAGEMENT OF SEVERE Mean arterial pressure Ն 65 mm Hg Urine output Ն0.5 mL⅐kgϪ1⅐hrϪ1SEPSIS Central venous (superior vena cava) oxygen saturation Ն70% or mixed venous Ն65% ⅙ If venous oxygen saturation target is not achieved (2C) Consider further fluidA. Initial Resuscitation Transfuse packed red blood cells if required to hematocrit of Ն30% and/or Start dobutamine infusion, maximum 20 ␮g⅐kgϪ1⅐minϪ11. We recommend the protocolized re- Diagnosis suscitation of a patient with sepsis- ● Obtain appropriate cultures before starting antibiotics provided this does not significantly induced shock, defined as tissue hypo- delay antimicrobial administration (1C) perfusion (hypotension persisting Obtain two or more BCs One or more BCs should be percutaneous after initial fluid challenge or blood One BC from each vascular access device in place Ͼ48 hrs lactate concentration Ն4 mmol/L). Culture other sites as clinically indicated This protocol should be initiated as ● Perform imaging studies promptly to confirm and sample any source of infection, if safe to do so (1C) soon as hypoperfusion is recognized Antibiotic therapy and should not be delayed pending ● Begin intravenous antibiotics as early as possible and always within the first hour of recognizing severe sepsis (1D) and septic shock (1B) ICU admission. During the first 6 hrs ● Broad-spectrum: one or more agents active against likely bacterial/fungal pathogens and with of resuscitation, the goals of initial good penetration into presumed source (1B) resuscitation of sepsis-induced hypo- ● Reassess antimicrobial regimen daily to optimize efficacy, prevent resistance, avoid toxicity, perfusion should include all of the fol- and minimize costs (1C) lowing as one part of a treatment pro- ⅙ Consider combination therapy in Pseudomonas infections (2D) tocol: ⅙ Consider combination empiric therapy in neutropenic patients (2D) ⅙ Combination therapy Յ3–5 days and de-escalation following susceptibilities (2D) Central venous pressure 8 –12 mm ● Duration of therapy typically limited to 7–10 days; longer if response is slow or there are Hg undrainable foci of infection or immunologic deficiencies (1D) Mean arterial pressure (MAP) Ն65 ● Stop antimicrobial therapy if cause is found to be noninfectious (1D) mm Hg Source identification and control Urine output Ն0.5 mL·kgϪ1·hr Ϫ1 ● A specific anatomic site of infection should be established as rapidly as possible (1C) and within first 6 hrs of presentation (1D) Central venous (superior vena ● Formally evaluate patient for a focus of infection amenable to source control measures (e.g. cava) or mixed venous oxygen sat- abscess drainage, tissue debridement) (1C) uration Ն70% or Ն65%, respec- ● Implement source control measures as soon as possible following successful initial tively (grade 1C) resuscitation (1C) (exception: infected pancreatic necrosis, where surgical intervention is best delayed) (2B) Rationale. Early goal-directed resusci- ● Choose source control measure with maximum efficacy and minimal physiologic upset (1D)tation has been shown to improve sur- ● Remove intravascular access devices if potentially infected (1C)vival for emergency department patientspresenting with septic shock in a ran- GRADE, Grades of Recommendation, Assessment, Development and Evaluation; ICU, intensivedomized, controlled, single-center study care unit; CVP, central venous pressure; BC, blood culture. a(16). Resuscitation directed toward the A higher target CVP of 12–15 mm Hg is recommended in the presence of mechanical ventilationpreviously mentioned goals for the initial or preexisting decreased ventricular compliance.6-hr period of the resuscitation was ableto reduce 28-day mortality rate. The con- come in septic shock and MAP Ն65 mmsensus panel judged use of central venous iment to filling (17). Similar consider- Hg as well as central venous oxygen sat-and mixed venous oxygen saturation tar- ation may be warranted in circumstances uration (ScvO2, measured in superiorgets to be equivalent. Either intermittent of increased abdominal pressure or dia- vena cava, either intermittently or con-or continuous measurements of oxygen stolic dysfunction (18). Elevated central tinuously) of Ն70% (19). Many recentsaturation were judged to be acceptable. venous pressures may also be seen with studies support the value of early proto-Although blood lactate concentration preexisting clinically significant pulmo- colized resuscitation in severe sepsis andmay lack precision as a measure of tissue nary artery hypertension. Although the sepsis-induced tissue hypoperfusion (20 –metabolic status, elevated levels in sepsis cause of tachycardia in septic patients 25). Studies of patients with shock indi-support aggressive resuscitation. In me- may be multifactorial, a decrease in ele- cate that mixed venous oxygen saturationchanically ventilated patients or patients vated pulse rate with fluid resuscitation is ¯ (SVO2) runs 5–7% lower than central ve-with known preexisting decreased ven- often a useful marker of improving intra- nous oxygen saturation (ScvO2) (26) andtricular compliance, a higher target cen- vascular filling. Recently published ob- that an early goal-directed resuscitationtral venous pressure of 12–15 mm Hg is servational studies have demonstrated an protocol can be established in a nonre-recommended to account for the imped- association between good clinical out- search general practice venue (27).Crit Care Med 2008 Reprint 5
  6. 6. Table 4. Hemodynamic support and adjunctive therapy Rationale. The protocol used in the study cited previously targeted an increaseStrength of recommendation and quality of evidence have been assessed using the GRADE criteria, in ScvO2 to Ն70% (16). This was achieved presented in parentheses after each guideline. ● Indicates a strong recommendation, or “we recommend” by sequential institution of initial fluid re- ⅙ Indicates a weak recommendation, or “we suggest” suscitation, packed red blood cells, andFluid therapy then dobutamine. This protocol was asso- ● Fluid-resuscitate using crystalloids or colloids (1B) ciated with an improvement in survival. ● Target a CVP of Ն8 mm Hg (Ն12 mm Hg if mechanically ventilated) (1C) Based on bedside clinical assessment and ● Use a fluid challenge technique while associated with a hemodynamic improvement (1D) ● Give fluid challenges of 1000 mL of crystalloids or 300–500 mL of colloids over 30 mins. More personal preference, a clinician may deem rapid and larger volumes may be required in sepsis-induced tissue hypoperfusion (1D) either blood transfusion (if hematocrit is ● Rate of fluid administration should be reduced if cardiac filling pressures increase without Ͻ30%) or dobutamine the best initial concurrent hemodynamic improvement (1D) choice to increase oxygen delivery andVasopressors thereby elevate ScvO2, when fluid resusci- ● Maintain MAP Ն65 mm Hg (1C) ● Norepinephrine and dopamine centrally administered are the initial vasopressors of choice (1C) tation is believed to be already adequate. ⅙ Epinephrine, phenylephrine, or vasopressin should not be administered as the initial The design of the aforementioned trial did vasopressor in septic shock (2C). Vasopressin 0.03 units/min may be subsequently added to not allow assessment of the relative contri- norepinephrine with anticipation of an effect equivalent to norepinephrine alone bution of these two components (i.e., in- ⅙ Use epinephrine as the first alternative agent in septic shock when blood pressure is poorly creasing oxygen content or increasing car- responsive to norepinephrine or dopamine (2B). diac output) of the protocol on ● Do not use low-dose dopamine for renal protection (1A) ● In patients requiring vasopressors, insert an arterial catheter as soon as practical (1D) achievement of improved outcome.Inotropic therapy ● Use dobutamine in patients with myocardial dysfunction as supported by elevated cardiac B. Diagnosis filling pressures and low cardiac output (1C) ● Do not increase cardiac index to predetermined supranormal levels (1B) 1. We recommend obtaining appropri-Steroids ate cultures before antimicrobial ⅙ Consider intravenous hydrocortisone for adult septic shock when hypotension responds poorly therapy is initiated if such cultures to adequate fluid resuscitation and vasopressors (2C) do not cause significant delay in an- ⅙ ACTH stimulation test is not recommended to identify the subset of adults with septic shock who should receive hydrocortisone (2B) tibiotic administration. To optimize ⅙ Hydrocortisone is preferred to dexamethasone (2B) identification of causative organisms, ⅙ Fludrocortisone (50 ␮g orally once a day) may be included if an alternative to hydrocortisone we recommend at least two blood is being used that lacks significant mineralocorticoid activity. Fludrocortisone if optional if cultures be obtained before antibiot- hydrocortisone is used (2C) ics with at least one drawn percuta- ⅙ Steroid therapy may be weaned once vasopressors are no longer required (2D) neously and one drawn through each ● Hydrocortisone dose should be Յ300 mg/day (1A) ● Do not use corticosteroids to treat sepsis in the absence of shock unless the patient’s vascular access device, unless the de- endocrine or corticosteroid history warrants it (1D) vice was recently (Ͻ48 hrs) inserted.Recombinant human activated protein C Cultures of other sites (preferably ⅙ Consider rhAPC in adult patients with sepsis-induced organ dysfunction with clinical quantitative where appropriate), such assessment of high risk of death (typically APACHE II Ն25 or multiple organ failure) if there as urine, cerebrospinal fluid, wounds, are no contraindications (2B, 2C for postoperative patients). respiratory secretions, or other body ● Adult patients with severe sepsis and low risk of death (typically, APACHE II Ͻ20 or one organ failure) should not receive rhAPC (1A) fluids that may be the source of in- fection should also be obtained be- GRADE, Grades of Recommendation, Assessment, Development and Evaluation; CVP, central fore antibiotic therapy if not associ-venous pressure; MAP, mean arterial pressure; ACTH, adrenocorticotropic hormone; rhAPC, recom- ated with significant delay inbinant human activated protein C; APACHE, Acute Physiology and Chronic Health Evaluation. antibiotic administration (grade 1C). Rationale. Although sampling should not delay timely administration of antibi- otics in patients with severe sepsis (e.g., There are recognized limitations to riod and research to validate utility. lumbar puncture in suspected meningi-ventricular filling pressure estimates as These technologies are already available tis), obtaining appropriate cultures be-surrogates for fluid resuscitation (28, for early ICU resuscitation. fore administration of antibiotics is es-29). However, measurement of central sential to confirm infection and the 2. We suggest that during the first 6 hrs ofvenous pressure is currently the most responsible pathogens and to allow de- resuscitation of severe sepsis or septicreadily obtainable target for fluid resus- ¯ escalation of antibiotic therapy after re-citation. There may be advantages to shock, if ScvO2 or SVO2 of 70% or 65%, ceipt of the susceptibility profile. Samplestargeting fluid resuscitation to flow and respectively, is not achieved with fluid can be refrigerated or frozen if processingperhaps to volumetric indices (and even resuscitation to the central venous pres- cannot be performed immediately. Im-to microcirculation changes) (30 –33). sure target, then transfusion of packed mediate transport to a microbiologicalTechnologies currently exist that allow red blood cells to achieve a hematocrit lab is necessary. Because rapid steriliza-measurement of flow at the bedside (34, of Ն30% and/or administration of a do- tion of blood cultures can occur within a35). Future goals should be making butamine infusion (up to a maximum of few hours after the first antibiotic dose,these technologies more accessible dur- 20 ␮g·kgϪ1·minϪ1) be used to achieve obtaining those cultures before startinging the critical early resuscitation pe- this goal (grade 2C). therapy is essential if the causative organ-6 Crit Care Med 2008 Reprint
  7. 7. Table 5. Other supportive therapy of severe sepsisStrength of recommendation and quality of evidence have been assessed using the GRADE criteria, presented in parentheses after each guideline ● Indicates a strong recommendation, or “we recommend” ⅙ Indicates a weak recommendation, or “we suggest”Blood product administration ● Give red blood cells when hemoglobin decreases to Ͻ7.0 g/dL (Ͻ70 g/L) to target a hemoglobin of 7.0–9.0 g/dL in adults (1B). A higher hemoglobin level may be required in special circumstances (e.g., myocardial ischaemia, severe hypoxemia, acute hemorrhage, cyanotic heart disease, or lactic acidosis) ● Do not use erythropoietin to treat sepsis-related anemia. Erythropoietin may be used for other accepted reasons (1B) ⅙ Do not use fresh frozen plasma to correct laboratory clotting abnormalities unless there is bleeding or planned invasive procedures (2D) ● Do not use antithrombin therapy (1B) ⅙ Administer platelets when (2D) Counts are Ͻ5000/mm3 (5 ϫ 109/L) regardless of bleeding Counts are 5000–30,000/mm3 (5–30 ϫ 109/L) and there is significant bleeding risk Higher platelet counts (Ն50,000/mm3 [50 ϫ 109/L]) are required for surgery or invasive proceduresMechanical ventilation of sepsis-induced ALI/ARDS ● Target a tidal volume of 6 mL/kg (predicted) body weight in patients with ALI/ARDS (1B) ● Target an initial upper limit plateau pressure Յ30 cm H2O. Consider chest wall compliance when assessing plateau pressure (1C) ● Allow PaCO2 to increase above normal, if needed, to minimize plateau pressures and tidal volumes (1C) ● Set PEEP to avoid extensive lung collapse at end-expiration (1C) ⅙ Consider using the prone position for ARDS patients requiring potentially injurious levels of FIO2 or plateau pressure, provided they are not put at risk from positional changes (2C) ● Maintain mechanically ventilated patients in a semirecumbent position (head of the bed raised to 45°) unless contraindicated (1B), between 30° and 45° (2C) ⅙ Noninvasive ventilation may be considered in the minority of ALI/ARDS patients with mild to moderate hypoxemic respiratory failure. The patients need to be hemodynamically stable, comfortable, easily arousable, able to protect/clear their airway, and expected to recover rapidly (2B) ● Use a weaning protocol and an SBT regularly to evaluate the potential for discontinuing mechanical ventilation (1A) ● SBT options include a low level of pressure support with continuous positive airway pressure 5 cm H2O or a T piece ● Before the SBT, patients should be arousable be hemodynamically stable without vasopressors have no new potentially serious conditions have low ventilatory and end-expiratory pressure requirement require FIO2 levels that can be safely delivered with a face mask or nasal cannula ● Do not use a pulmonary artery catheter for the routine monitoring of patients with ALI/ARDS (1A) ● Use a conservative fluid strategy for patients with established ALI who do not have evidence of tissue hypoperfusion (1C)Sedation, analgesia, and neuromuscular blockade in sepsis ● Use sedation protocols with a sedation goal for critically ill mechanically ventilated patients (1B) ● Use either intermittent bolus sedation or continuous infusion sedation to predetermined end points (sedation scales), with daily interruption/lightening to produce awakening. Re-titrate if necessary (1B) ● Avoid neuromuscular blockers where possible. Monitor depth of block with train-of-four when using continuous infusions (1B)Glucose control ● Use intravenous insulin to control hyperglycemia in patients with severe sepsis following stabilization in the ICU (1B) ⅙ Aim to keep blood glucose Ͻ150 mg/dL (8.3 mmol/L) using a validated protocol for insulin dose adjustment (2C) ● Provide a glucose calorie source and monitor blood glucose values every 1–2 hrs (4 hrs when stable) in patients receiving intravenous insulin (1C) ● Interpret with caution low glucose levels obtained with point of care testing, as these techniques may overestimate arterial blood or plasma glucose values (1B)Renal replacement ⅙ Intermittent hemodialysis and CVVH are considered equivalent (2B) ⅙ CVVH offers easier management in hemodynamically unstable patients (2D)Bicarbonate therapy ● Do not use bicarbonate therapy for the purpose of improving hemodynamics or reducing vasopressor requirements when treating hypoperfusion- induced lactic acidemia with pH Ն7.15 (1B)Deep vein thrombosis prophylaxis ● Use either low-dose UFH or LMWH, unless contraindicated (1A) ● Use a mechanical prophylactic device, such as compression stockings or an intermittent compression device, when heparin is contraindicated (1A) ⅙ Use a combination of pharmacologic and mechanical therapy for patients who are at very high risk for deep vein thrombosis (2C) ⅙ In patients at very high risk, LMWH should be used rather than UFH (2C)Stress ulcer prophylaxis ● Provide stress ulcer prophylaxis using H2 blocker (1A) or proton pump inhibitor (1B). Benefits of prevention of upper gastrointestinal bleed must be weighed against the potential for development of ventilator-acquired pneumoniaConsideration for limitation of support ● Discuss advance care planning with patients and families. Describe likely outcomes and set realistic expectations (1D) GRADE, Grades of Recommendation, Assessment, Development and Evaluation; ALI, acute lung injury; ARDS, acute respiratory distress syndrome;PEEP, positive end-expiratory pressure; SBT, spontaneous breathing trial; ICU, intensive care unit; CVVH, continuous veno-venous hemofiltration; UFH,unfractionated heparin; LMWH, low-molecular weight heparin.ism is to be identified. Two or more blood cular access device. Obtaining blood that the organism is causing the severecultures are recommended (36). In pa- cultures peripherally and through a sepsis is enhanced. In addition, if thetients with indwelling catheters (for Ͼ48 vascular access device is an important culture drawn through the vascular ac-hrs), at least one blood culture should be strategy. If the same organism is recov- cess device is positive much earlier thandrawn through each lumen of each vas- ered from both cultures, the likelihood the peripheral blood culture (i.e., Ͼ2 hrsCrit Care Med 2008 Reprint 7
  8. 8. earlier), the data support the concept that obtained before initiating antibiotic gen when choosing initial therapy. Whenthe vascular access device is the source of therapy but should not prevent deemed warranted, the selection of empir-the infection (37). Quantitative cultures prompt administration of antimicro- ical antifungal therapy (e.g., fluconazole,of catheter and peripheral blood are also bial therapy (grade 1D). amphotericin B, or echinocandin) will beuseful for determining whether the cath- Rationale. Establishing vascular ac- tailored to the local pattern of the mosteter is the source of infection. Volume of prevalent Candida species and any prior cess and initiating aggressive fluid resus-blood drawn with the culture tube should administration of azoles drugs (44). Risk citation are the first priority when man-be Ն10 mL (38). Quantitative (or semi- factors for candidemia should also be con- aging patients with severe sepsis or septicquantitative) cultures of respiratory tract sidered when choosing initial therapy. shock. However, prompt infusion of anti-secretions are recommended for the di- Because patients with severe sepsis microbial agents should also be a priorityagnosis of ventilator-associated pneumo- or septic shock have little margin for and may require additional vascular ac-nia (39). Gram-negative stain can be use- error in the choice of therapy, the ini- cess ports (42, 43). In the presence offul, in particular for respiratory tract tial selection of antimicrobial therapy septic shock, each hour delay in achiev-specimens, to help decide the micro- should be broad enough to cover all ing administration of effective antibioticsorganisms to be targeted. The potential likely pathogens. There is ample evi-role of biomarkers for diagnosis of in- is associated with a measurable increase in mortality (42). If antimicrobial agents dence that failure to initiate appropri-fection in patients presenting with se- ate therapy (i.e., therapy with activityvere sepsis remains undefined. The pro- cannot be mixed and delivered promptly from the pharmacy, establishing a supply against the pathogen that is subse-calcitonin level, although often useful, quently identified as the causativeis problematic in patients with an acute of premixed antibiotics for such urgent situations is an appropriate strategy for agent) correlates with increased mor-inflammatory pattern from other ensuring prompt administration. In bidity and mortality (45– 48).causes (e.g., postoperative, shock) (40). choosing the antimicrobial regimen, cli- Patients with severe sepsis or septicIn the near future, rapid diagnostic nicians should be aware that some anti- shock warrant broad-spectrum therapymethods (polymerase chain reaction, microbial agents have the advantage of until the causative organism and its an-micro-arrays) might prove extremely bolus administration, while others re- tibiotic susceptibilities are defined. Re-helpful for a quicker identification of quire a lengthy infusion. Thus, if vascular striction of antibiotics as a strategy topathogens and major antimicrobial re-sistance determinants (41). access is limited and many different reduce the development of antimicrobial agents must be infused, bolus drugs may resistance or to reduce cost is not an2. We recommend that imaging studies appropriate initial strategy in this patient offer an advantage. be performed promptly in attempts population. to confirm a potential source of in- 2a. We recommend that initial empirical All patients should receive a full load- fection. Sampling of potential anti-infective therapy include one or ing dose of each antimicrobial. However, sources of infection should occur as more drugs that have activity against patients with sepsis or septic shock often they are identified; however, some all likely pathogens (bacterial and/or have abnormal renal or hepatic function patients may be too unstable to war- fungal) and that penetrate in ade- and may have abnormal volumes of dis- rant certain invasive procedures or quate concentrations into the pre- tribution due to aggressive fluid resusci- transport outside of the ICU. Bedside sumed source of sepsis (grade 1B). tation. Drug serum concentration moni- studies, such as ultrasound, are use- Rationale. The choice of empirical an- toring can be useful in an ICU setting for ful in these circumstances (grade tibiotics depends on complex issues re- those drugs that can be measured 1C). lated to the patient’s history, including promptly. An experienced physician or Rationale. Diagnostic studies may drug intolerances, underlying disease, clinical pharmacist should be consultedidentify a source of infection that re- the clinical syndrome, and susceptibility to ensure that serum concentrations arequires removal of a foreign body or drain- patterns of pathogens in the community, attained that maximize efficacy and min-age to maximize the likelihood of a sat- in the hospital, and that previously have imize toxicity (49 –52).isfactory response to therapy. However, been documented to colonize or infecteven in the most organized and well- 2b. We recommend that the antimicro- the patient. There is an especially widestaffed healthcare facilities, transport of bial regimen be reassessed daily to range of potential pathogens for neutro-patients can be dangerous, as can placing optimize activity, to prevent the de- penic patients.patients in outside-unit imaging devices velopment of resistance, to reduce Recently used antibiotics should gener-that are difficult to access and monitor. ally be avoided. When choosing empirical toxicity, and to reduce costs (gradeBalancing risk and benefit is therefore therapy, clinicians should be cognizant of 1C).mandatory in those settings. the virulence and growing prevalence of Rationale. Although restriction of an- oxacillin (methicillin)-resistant Staphylo- tibiotics as a strategy to reduce the devel-C. Antibiotic Therapy coccus aureus (ORSA or MRSA) in some opment of antimicrobial resistance or to communities and healthcare settings (espe- reduce cost is not an appropriate initial 1. We recommend that intravenous an- cially in the United States). If the preva- strategy in this patient population, once tibiotic therapy be started as early as lence is significant, and in consideration of the causative pathogen has been identi- possible and within the first hour of the virulence of this organism, empirical fied, it may become apparent that none of recognition of septic shock (1B) and therapy adequate for this pathogen would the empirical drugs offers optimal ther- severe sepsis without septic shock be warranted. Clinicians should also con- apy; that is, there may be another drug (1D). Appropriate cultures should be sider whether candidemia is a likely patho- proven to produce superior clinical out-8 Crit Care Med 2008 Reprint
  9. 9. come that should therefore replace em- patient will become infected with an clude a rapid diagnosis of the specific sitepirical agents. antibiotic-resistant pathogen or will of infection and identification of a focus Narrowing the spectrum of antibiotic develop a drug-related adverse effect on infection amenable to source controlcoverage and reducing the duration of (grade 1D). measures (specifically the drainage of anantibiotic therapy will reduce the likelihood Rationale. Clinicians should be cogni- abscess, debridement of infected necroticthat the patient will develop superinfection zant that blood cultures will be negative tissue, removal of a potentially infectedwith pathogenic or resistant organisms, in Ͼ50% of cases of severe sepsis or sep- device, and definitive control of a sourcesuch as Candida species, Clostridium diffi- tic shock, yet many of these cases are very of ongoing microbial contamination)cile, or vancomycin-resistant Enterococcus likely caused by bacteria or fungi. Thus, (58). Foci of infection readily amenable tofaecium. However, the desire to minimize the decisions to continue, narrow, or stop source control measures include an in-superinfections and other complications antimicrobial therapy must be made on tra-abdominal abscess or gastrointestinalshould not take precedence over the need the basis of clinician judgment and clin- perforation, cholangitis or pyelonephri-to give the patient an adequate course of ical information. tis, intestinal ischemia or necrotizing softtherapy to cure the infection that caused tissue infection, and other deep space in-the severe sepsis or septic shock. fection, such as an empyema or septic D. Source Control arthritis. Such infectious foci should be2c. We suggest combination therapy for controlled as soon as possible following patients with known or suspected 1a. We recommend that a specific ana- successful initial resuscitation (59), ac- Pseudomonas infections as a cause of tomical diagnosis of infection requir- complishing the source control objective severe sepsis (grade 2D). ing consideration for emergent with the least physiologic upset possible2d. We suggest combination empirical source control (e.g., necrotizing fas- (e.g., percutaneous rather than surgical therapy for neutropenic patients with ciitis, diffuse peritonitis, cholangitis, drainage of an abscess [60], endoscopic severe sepsis (grade 2D). intestinal infarction) be sought and rather than surgical drainage of biliary2e. When used empirically in patients diagnosed or excluded as rapidly as tree), and removing intravascular access with severe sepsis, we suggest that possible (grade 1C) and within the devices that are potentially the source of combination therapy should not be first 6 hrs following presentation severe sepsis or septic shock promptly administered for Ͼ3–5 days. De- (grade 1D). after establishing other vascular access escalation to the most appropriate 1b. We further recommend that all pa- (61, 62). A randomized, controlled trial single therapy should be performed tients presenting with severe sepsis comparing early vs. delayed surgical in- as soon as the susceptibility profile is be evaluated for the presence of a tervention for peripancreatic necrosis known (grade 2D). focus on infection amenable to showed better outcomes with a delayed source control measures, specifically approach (63). However, areas of uncer- Rationale. Although no study or meta- the drainage of an abscess or local tainty exist, such as definitive docu-analysis has convincingly demonstrated focus on infection, the debridementthat combination therapy produces a supe- mentation of infection and appropriate of infected necrotic tissue, the re- length of delay. The selection of optimalrior clinical outcome for individual patho- moval of a potentially infected device,gens in a particular patient group, combi- source control methods must weigh or the definitive control of a source benefits and risks of the specific inter-nation therapies do produce in vitro of ongoing microbial contaminationsynergy against pathogens in some models vention as well as risks of transfer (64). (grade 1C). (Appendix A provides ex- Source control interventions may cause(although such synergy is difficult to define amples of potential sites needingand predict). In some clinical scenarios, further complications, such as bleed- source control.) ing, fistulas, or inadvertent organ in-such as the two preceding, combination 2. We suggest that when infectedtherapies are biologically plausible and are jury. Surgical intervention should be peripancreatic necrosis is identified as considered when lesser interventionallikely clinically useful even if evidence has a potential source of infection, defini-not demonstrated improved clinical out- approaches are inadequate or when di- tive intervention is best delayed until agnostic uncertainty persists despite ra-come (53–56). Combination therapy for adequate demarcation of viable andsuspected known Pseudomonas pending diologic evaluation. Specific clinical sit- nonviable tissues has occurred (grade uations require consideration ofsensitivities increases the likelihood that at 2B). available choices, patient’s preferences,least one drug is effective against that 3. We recommend that when source con- and clinician’s expertise.strain and positively affects outcome (57). trol is required, the effective intervention 3. We recommend that the duration of associated with the least physiologic in- E. Fluid Therapy therapy typically be 7–10 days; longer sult be employed (e.g., percutaneous courses may be appropriate in pa- rather than surgical drainage of an 1. We recommend fluid resuscitation tients who have a slow clinical re- abscess (grade 1D). with either natural/artificial colloids sponse, undrainable foci of infection, 4. We recommend that when intravas- or crystalloids. There is no evidence- or immunologic deficiencies, includ- cular access devices are a possible based support for one type of fluid ing neutropenia (grade 1D). source of severe sepsis or septic over another (grade 1B). 4. We recommend that if the presenting shock, they be promptly removed af- Rationale. The SAFE study indicated clinical syndrome is determined to be ter other vascular access has been that albumin administration was safe and due to a noninfectious cause, antimi- established (grade 1C). equally as effective as crystalloid (65). There crobial therapy be stopped promptly Rationale. The principals of source was an insignificant decrease in mortality to minimize the likelihood that the control in the management of sepsis in- rates with the use of colloid in a subsetCrit Care Med 2008 Reprint 9
  10. 10. analysis of septic patients (p ϭ .09). Previ- F. Vasopressors Rationale. There is no high-qualityous meta-analyses of small studies of ICU primary evidence to recommend one cat-patients had demonstrated no difference 1. We recommend that mean arterial echolamine over another. Much litera-between crystalloid and colloid fluid resus- pressure (MAP) be maintained Ն65 ture exists that contrasts the physiologiccitation (66 – 68). Although administration mm Hg (grade 1C). effects of choice of vasopressor and com-of hydroxyethyl starch may increase the Rationale. Vasopressor therapy is re- bined inotrope/vasopressors in septicrisk of acute renal failure in patients with quired to sustain life and maintain perfu- shock (73– 85). Human and animal stud-sepsis, variable findings preclude definitive sion in the face of life-threatening hypo- ies suggest some advantages of norepi-recommendations (69, 70). As the volume tension, even when hypovolemia has not nephrine and dopamine over epinephrineof distribution is much larger for crystal- yet been resolved. Below a certain mean (the latter with the potential for tachy-loids than for colloids, resuscitation with arterial pressure, autoregulation in vari- cardia as well as disadvantageous effectscrystalloids requires more fluid to achieve ous vascular beds can be lost, and perfu- on splanchnic circulation and hyper-the same end points and results in more sion can become linearly dependent on lactemia) and phenylephrine (decrease inedema. Crystalloids are less expensive. pressure. Thus, some patients may re- stroke volume). There is, however, no quire vasopressor therapy to achieve a clinical evidence that epinephrine results 2. We recommend that fluid resuscita- minimal perfusion pressure and maintain in worse outcomes, and it should be the tion initially target a central venous first chosen alternative to dopamine or pressure of Ն8 mm Hg (12 mm Hg in adequate flow (71, 72). The titration of norepinephrine to as low as MAP 65 mm norepinephrine. Phenylephrine is the ad- mechanically ventilated patients). Hg has been shown to preserve tissue renergic agent least likely to produce Further fluid therapy is often re- perfusion (72). In addition, preexisting tachycardia but as a pure vasopressor quired (grade 1C). comorbidities should be considered as to would be expected to decrease stroke vol-3a. We recommend that a fluid challenge most appropriate MAP target. For exam- ume. Dopamine increases mean arterial technique be applied wherein fluid ple, a MAP of 65 mm Hg might be too low pressure and cardiac output, primarily administration is continued as long in a patient with severe uncontrolled hy- due to an increase in stroke volume and as the hemodynamic improvement pertension, and in a young previously heart rate. Norepinephrine increases (e.g., arterial pressure, heart rate, normotensive, a lower MAP might be mean arterial pressure due to its vasocon- urine output) continues (grade 1D). adequate. Supplementing end points, strictive effects, with little change in3b. We recommend that fluid challenge heart rate and less increase in stroke vol- in patients with suspected hypovole- such as blood pressure, with assess- ume compared with dopamine. Either mia be started with Ն1000 mL of ment of regional and global perfusion, may be used as a first-line agent to correct crystalloids or 300 –500 mL of col- such as blood lactate concentrations hypotension in sepsis. Norepinephrine is loids over 30 mins. More rapid ad- and urine output, is important. Ade- more potent than dopamine and may be ministration and greater amounts of quate fluid resuscitation is a fundamen- more effective at reversing hypotension in fluid may be needed in patients with tal aspect of the hemodynamic manage- patients with septic shock. Dopamine may sepsis-induced tissue hypoperfusion ment of patients with septic shock and be particularly useful in patients with com- (see Initial Resuscitation recommen- should ideally be achieved before vaso- promised systolic function but causes more dations) (grade 1D). pressors and inotropes are used, but tachycardia and may be more arrhythmo-3c. We recommend that the rate of fluid using vasopressors early as an emer- genic (86). It may also influence the endo- administration be reduced substan- gency measure in patients with severe crine response via the hypothalamic- tially when cardiac filling pressures shock is frequently necessary. When pituitary axis and have immunosuppressive (central venous pressure or pulmo- that occurs, great effort should be di- effects. nary artery balloon-occluded pres- rected to weaning vasopressors with Vasopressin levels in septic shock have sure) increase without concurrent continuing fluid resuscitation. been reported to be lower than antici- hemodynamic improvement (grade 2. We recommend either norepineph- pated for a shock state (87). Low doses of 1D). rine or dopamine as the first choice vasopressin may be effective in raising Rationale. Fluid challenge must be vasopressor agent to correct hypo- blood pressure in patients refractory toclearly separated from simple fluid ad- tension in septic shock (administered other vasopressors and may have otherministration; it is a technique in which through a central catheter as soon as potential physiologic benefits (88 –93).large amounts of fluids are administered one is available) (grade 1C). Terlipressin has similar effects but is longover a limited period of time under close 3a. We suggest that epinephrine, phenyl- lasting (94). Studies show that vasopres-monitoring to evaluate the patient’s re- ephrine, or vasopressin should not be sin concentrations are elevated in earlysponse and avoid the development of pul- administered as the initial vasopres- septic shock, but with continued shockmonary edema. The degree of intravascular sor in septic shock (grade 2C). Vaso- the concentration decreases to normalvolume deficit in patients with severe sepsis pressin 0.03 units/min may be added range in the majority of patients betweenvaries. With venodilation and ongoing cap- to norepinephrine subsequently with 24 and 48 hrs (95). This has been calledillary leak, most patients require continu- anticipation of an effect equivalent to relative vasopressin deficiency because ining aggressive fluid resuscitation during that of norepinephrine alone. the presence of hypotension, vasopressinthe first 24 hrs of management. Input is 3b. We suggest that epinephrine be the would be expected to be elevated. Thetypically much greater than output, and first chosen alternative agent in sep- significance of this finding is unknown.input/output ratio is of no utility to judge tic shock that is poorly responsive to The recent VASST trial, a randomized,fluid resuscitation needs during this time norepinephrine or dopamine (grade controlled trial comparing norepineph-period. 2B). rine alone to norepinephrine plus vaso-10 Crit Care Med 2008 Reprint
  11. 11. pressin at 0.03 units/min, showed no dif- predetermined supranormal levels CORTICUS did show a faster resolution ofference in outcome in the intent to treat (grade 1B). septic shock in patients who received ste-population. An a priori defined subgroup Rationale. Dobutamine is the first- roids. The use of the ACTH test (respondersanalysis showed that the survival of pa- choice inotrope for patients with mea- and nonresponders) did not predict thetients receiving Ͻ15 ␮g/min norepineph- sured or suspected low cardiac output in faster resolution of shock. Importantly, un-rine at the time of randomization was the presence of adequate left ventricular like the French trial, which only enrolledbetter with vasopressin. However, the filling pressure (or clinical assessment of shock patients with blood pressure unre-pretrial rationale for this stratification adequate fluid resuscitation) and ade- sponsive to vasopressor therapy, the COR-was based on exploring potential benefit quate mean arterial pressure. Septic pa- TICUS study included patients with septicin the Ն15 ␮g norepinephrine require- tients who remain hypotensive after fluid shock, regardless of how the blood pressurement population. Higher doses of vaso- resuscitation may have low, normal, or responded to vasopressors. Although corti-pressin have been associated with car- increased cardiac outputs. Therefore, costeroids do appear to promote shock re-diac, digital, and splanchnic ischemia and treatment with a combined inotrope/ versal, the lack of a clear improvement inshould be reserved for situations where vasopressor, such as norepinephrine or mortality— coupled with known side ef-alternative vasopressors have failed (96). dopamine, is recommended if cardiac fects of steroids, such as increased risk ofCardiac output measurement to allow output is not measured. When the capa- infection and myopathy— generally tem-maintenance of a normal or elevated flow bility exists for monitoring cardiac out- pered enthusiasm for their broad use. Thus,is desirable when these pure vasopressors put in addition to blood pressure, a vaso- there was broad agreement that the recom-are instituted. pressor, such as norepinephrine, may be mendation should be downgraded from the 5. We recommend that low-dose dopa- used separately to target specific levels of previous guidelines (Appendix B). There mine not be used for renal protection mean arterial pressure and cardiac out- was considerable discussion and consider- (grade 1A). put. Two large prospective clinical trials ation by the committee on the option of that included critically ill ICU patients encouraging use in those patients whose Rationale. A large randomized trial blood pressure was unresponsive to fluidsand meta-analysis comparing low-dose who had severe sepsis failed to demon- strate benefit from increasing oxygen de- and vasopressors, while strongly discourag-dopamine to placebo found no difference ing use in subjects whose shock respondedin either primary outcomes (peak serum livery to supranormal targets by use of dobutamine (99, 100). These studies did well to fluids and pressors. However, thiscreatinine, need for renal replacement, more complex set of recommendations wasurine output, time to recovery of normal not specifically target patients with se- vere sepsis and did not target the first 6 rejected in favor of the preceding singlerenal function) or secondary outcomes recommendation (Appendix B).(survival to either ICU or hospital dis- hrs of resuscitation. The first 6 hrs ofcharge, ICU stay, hospital stay, arrhyth- resuscitation of sepsis-induced hypoper- 2. We suggest that the ACTH stimulationmias) (97, 98). Thus, the available data do fusion need to be treated separately from test not be used to identify the subsetnot support administration of low doses the later stages of severe sepsis (see Ini- of adults with septic shock who shouldof dopamine solely to maintain renal tial Resuscitation recommendations). receive hydrocortisone (grade 2B).function. Rationale. Although one study sug- 6. We recommend that all patients re- H. Corticosteroids gested those who did not respond to quiring vasopressors have an arterial ACTH with a brisk surge in cortisol (fail- 1. We suggest that intravenous hydro- ure to achieve or Ͼ9 ␮g/dL increase in catheter placed as soon as practical if cortisone be given only to adult septic cortisol 30 – 60 mins after ACTH admin- resources are available (grade 1D). shock patients after it has been con- istration) were more likely to benefit Rationale. In shock states, estimation firmed that their blood pressure is from steroids than those who did re-of blood pressure using a cuff is com- poorly responsive to fluid resuscita- spond, the overall trial population ap-monly inaccurate; use of an arterial can- tion and vasopressor therapy (grade peared to benefit regardless of ACTH re-nula provides a more appropriate and re- 2C). sult, and the observation of a potentialproducible measurement of arterial Rationale. One French multicenter, interaction between steroid use andpressure. These catheters also allow con- randomized controlled trial (RCT) of pa- ACTH test was not statistically significanttinuous analysis so that decisions regard- tients in vasopressor-unresponsive septic (101). Furthermore, there was no evi-ing therapy can be based on immediate shock (hypotension despite fluid resuscita- dence of this distinction between re-and reproducible blood pressure informa- tion and vasopressors) showed a significant sponders and nonresponders in a recenttion. shock reversal and reduction of mortality multicenter trial (104). Commonly used rate in patients with relative adrenal insuf- cortisol immunoassays measure totalG. Inotropic Therapy ficiency (defined as postadrenocortico- cortisol (protein-bound and free) while tropic hormone [ACTH] cortisol increase free cortisol is the pertinent measure-1. We recommend that a dobutamine in- Յ9 ␮g/dL) (101). Two additional smaller ment. The relationship between free and fusion be administered in the presence RCTs also showed significant effects on total cortisol varies with serum protein of myocardial dysfunction as sug- shock reversal with steroid therapy (102, concentration. When compared with a gested by elevated cardiac filling pres- 103). However, a recent large, European reference method (mass spectrometry), sures and low cardiac output (grade multicenter trial (CORTICUS), which has cortisol immunoassays may over- or un- 1C). been presented in abstract form but not yet derestimate the actual cortisol level, af-2. We recommend against the use of a published, failed to show a mortality benefit fecting the assignment of patients to re- strategy to increase cardiac index to with steroid therapy of septic shock (104). sponders or nonresponders (105).Crit Care Med 2008 Reprint 11