Surviving Sepsis Campaign.

Abstract

It was the turn of the millennium. For most, sepsis was an obscure entity of importance only to intensivists, clinical academics, and basic science investigators. The molecular biology was beginning to be unraveled and industry was increasingly hopeful that sepsis might be treated with innovative biologic therapies. Consensus definitions had been created a decade earlier but had received little attention outside of clinical trialists. Then, in short order, came a series of randomized controlled trials (RCTs) supporting successful treatment of sepsis with protocolized early resuscitation, anticoagulant therapy, and steroids (1–3). This caught the attention of the frontline clinical community caring for sepsis patients as signaled by the intense interest in these trial results at scientific meetings. The stage was now set for engaging these enthused clinicians as to how sepsis might be diagnosed sooner and treated more effectively to improve outcomes. Sepsis outcomes based on analysis of control group clinical trial data were at this time disappointing (1–3). THE ORIGINS OF THE SURVIVING SEPSIS CAMPAIGN The Surviving Sepsis Campaign (SSC) was introduced at the 2002 European Society of Intensive Care Medicine (ESICM) annual meeting in Barcelona with the “Barcelona Declaration” (4). Mitchell Levy, Phil Dellinger, and Graham Ramsay presented the concept to the leadership of three professional organizations (the ESICM, the Society of Critical Care Medicine [SCCM], and the International Sepsis Forum [ISF]). The SSC was initially administered by these three organizations. The “Barcelona Declaration” committed the three organizations to strive to reduce mortality of sepsis by 25% within 5 years by improving recognition and treatment. In addition, the “Barcelona Declaration” urged governments and healthcare providers to recognize the growing burden of sepsis and to commit to providing adequate resources to combat it. A 5-point action plan was developed: Diagnosis—Facilitate early and accurate diagnosis through the adoption of one, single, clear definition of sepsis. Treatment—Ensure appropriate and timely use of treatments and interventions via consistent clinical protocols. Referral—Recognize universally acceptable referral guidelines in all countries of the world. Education—Provide leadership, support, and information to clinicians about sepsis management. Counseling—Provide a framework for improving and accelerating access to continuing post-ICU care and counseling for patients. The plan was for the SSC to be rolled out in three phases: 1) The initial declaration and commitment; 2) The development of evidence-based guidelines for sepsis management; and 3) A performance improvement initiative (on a global scale) to implement the guidelines and reduce sepsis mortality. DEVELOPMENT AND EVOLUTION OF THE GUIDELINES OVER THE YEARS The first set of SSC guidelines was published in 2004 (5). Since then, updates have been published on a roughly 4-year cycle (2008, 2012, 2016, and 2021) (6–9) (Fig. 1). Each edition has seen an evolution of the methodology for the data analysis and an evolution in the clinical guidance for optimal management. The field of sepsis management was evolving quickly and the SSC was a major driver in identifying problems and promoting advances. In consideration of the worldwide impact of the COVID-19 pandemic leading to hospitals and ICUs filled with critically ill COVID-19 patients, the SSC published separate guidelines on management of the critically ill COVID patient (10).Figure 1.: Surviving Sepsis Campaign (SSC) guidelines publication timeline and metrics. Guideline mastheads are displayed from original 2004 publication through the 2021 revision. The timeline displays and contrasts page numbers and numbers of references. To the left is the original SSC logo (2004 and 2008 editions) and the revised logo (2012 and forward).Innovative and dedicated leadership has been key to the success of the SSC. A broad coalition of sponsoring organizations has been built over the years, beginning with 11 in 2004 and with 21 in 2021 (Table 1). TABLE 1. - Organizations Endorsing the 2021 Surviving Sepsis Guidelines Society of Critical Care Medicine European Society of Intensive Care Medicine American Association of Critical Care Nurses American College of Chest Physicians American College of Emergency Physicians American Thoracic Society African Sepsis Alliance Asia and Pacific Sepsis Alliance Association De Medicina Intensiva Brasileira Australian and New Zealand Intensive Care Society Canadian Critical Care Society Chinese Society of Critical Care Medicine European Respiratory Society European Society of Clinical Microbiology and Infectious Diseases Indian Society of Critical Care Medicine Infectious Diseases Society of North America Japanese Society of Intensive Care Medicine Latin American Sepsis Institute Society for Academic Emergency Medicine Scandinavian Critical Care Trials Group Key areas where the guidelines have evolved include funding streams, the management of potential conflicts of interest (COI), the evidence-based methodology, ensuring adequate panel diversity, authorship issues, and the decision that the guidelines need to be relevant to the entire world including children. In addition, the last two sets of guidelines have introduced the public/patient voice into the process to ensure that they are relevant for our patients, who may receive care directed by them. CONFLICTS OF INTEREST From the outset, the guidelines’ leadership recognized the importance of potential COI. At no point has the panel had any members from industry and industry has had no input into the guidelines development process. In addition, no member of the guidelines panel has received any reimbursement or honoraria for their work in relation to these activities. All panelists complete a COI declaration, which has been actively managed, either through the guideline’s leadership group or, since 2012, through an independent panel made up of very senior past officers of ESICM and SCCM. In recent iterations, there has been additional focus directed toward academic (nonfinancial) COI. Based on COI, guideline panelists have been asked either to abstain from voting on recommendations with potential conflict or in certain circumstances asked not to participate in the evidence synthesis process. PANEL SELECTION/DIVERSITY Since 2012, there has been greater recognition that the panel needs to adequately represent the populations of healthcare patients and providers it serves with greater race and gender balance as well as national, geographical, and income-setting representation. The process has matured but remains a work in progress. For example, the number of women on the guidelines committee over the five renditions of the guidelines (2004, 2008, 2012, 2016, and 2021) are 4 (10%), 5 (9%), 13 (19%), 11 (19%), and 17 (28%). Another example is geographical representation outside of North America and Europe which has progressed over the five renditions of the guidelines: 2 (5%), 5 (9%), 11 (16%), 10 (17%), and 15 (25%). In addition, with the last two iterations, a target was established of at least 33% panel turnover from the previous version. It was felt that this would adequately maintain the organizational memory of the process while allowing new thinking to emerge that did not leave the product stale and old. AUTHORSHIP While initial versions of the guideline (2004–2008) had a mast head author writing committee that drafted the article for subsequent sign-off by the entire committee, since 2012, all panelists have been asked to sign the paper as a named author. This gave greater transparency over who and what was behind the development processes and gave authors more sense of responsibility for the content of the entire document. PEDIATRICS From the beginning of the development of the SSC guidelines, it was recognized that children should be considered separately from adults. The first guidelines, in 2004, as well as the updates in 2008 and 2012, included a section called “Pediatric Considerations” pointing out important differences between children and adults. The critical care community, however, recognized the limitations of this approach, and the SSC leadership formed a task force with broad international representation to develop SSC guidelines for children, first co-published in Pediatric Critical Care Medicine and Intensive Care Medicine in 2020 (11). LOW- AND MIDDLE-INCOME COUNTRIES Recent versions of the guideline have put more focus on optimizing recommendations relevancy for different parts of the world. This goal was facilitated by ensuring that low- and middle-income country (LMIC) settings are represented in each working group of the panel. The SSC guidelines have several recommendations as to therapies or tools that should not be used, which is relevant to set priorities in resource-limited settings. Other recommendations highly relevant for LMICs include the prompt administration of antibiotics with adequate dosage and de-escalation strategies. However, recommendations for lower-income countries remain challenging due to scant LMIC-specific data. Limited resources for organ support may also bear on optimal therapy, for example, aggressiveness of fluid resuscitation. The work of the SSC Sepsis in Resource-Limited Nations Workgroup includes publication of a before and after feasibility trial of increasing evidence based interventions in LMIC (12). EVOLUTION OF SSC GUIDELINE METHODOLOGY A trustworthy guideline results from rigorous methodology that minimizes personal biases, allows transparent assessment of the total body of evidence, and incorporates all essential variables to formulate recommendations. The first iteration of the SSC guideline was completed before significant advancement in guideline development methodology. The panel used an evaluation system developed by Sackett (13), classifying evidence according to study design and precision (grades A, B, and C). During that period, most guideline developers used a modified Delphi process to formulate recommendations. Guideline development methodology at that time lacked a systematic summary of the evidence, comprehensive assessment of the quality of evidence, a structured framework to transition from evidence to recommendation, and did not assign strength to recommendations. From 2008, the SSC guideline structured the questions in the Population, Intervention, Control, and Outcome (PICO) format, which provided explicit criteria for study selection and allowed judgment of the quality of the evidence at an outcome level (6). As the methodology of guideline development evolved, the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach emerged as a comprehensive, rigorous, and transparent methodology (14). Therefore, the SSC leadership adopted the GRADE approach with GRADE experts represented on the guidelines committee and charged to guide the methodology of the process. Unlike traditional approaches, the GRADE approach allowed assessment of the quality of evidence considering crucial factors that went beyond basic study design, such as inconsistency, indirectness, imprecision, and publication bias. In addition, it allowed the panel to develop recommendations considering not only quality of evidence but also balance of benefit and harm, patients’ values, cost, equity, and feasibility. The panel assigned a strength (strong vs weak) to each recommendation which reflected the confidence in the overall balance between desirable and undesirable consequences. In 2016, best practice statements were included in the guidelines. Best practice statements are recommendations the committee feels are important but not amenable for formal recommendations because the evidence is difficult to summarize or assess using GRADE methodology. These recommendations represent ungraded strong recommendations that are typically used when the benefit or harm is unequivocal. In the 2021 guidelines, the structure expanded to include librarians, systematic review experts, methodologists, and public members that collaborated with the panel. Professional medical librarians facilitated the electronic searches tailored to each PICO question. The systematic review team assisted the panel to identify and summarize the relevant evidence. Working with the methodology team, the quality of evidence was assessed and the evidence-to-decision framework was used to support the panel in formulating recommendations (15). Public members provided input on patients’ values and preferences. Table 2 shows the evolution of methodology over the five renditions of the guidelines. TABLE 2. - Surviving Sepsis Campaign Guideline Methodology Category SSC 2004 SSC 2008 SSC 2012 SSC 2016 SSC 2021 Panela 44 members 55 members 68 members 55 members 60 members Number of recommendations 46 recommendations 78 recommendations 84 recommendations 71 recommendations 75 recommendations 18 BPS 15 BPS COI managementb Financial COIs Financial COIs Financial COIs Financial COIs Financial and academic COIs Topic prioritizationc Panel discussion Panel discussion Panel discussion Panel discussion Multifaceted approachd Structured Patient/Problem, Intervention, Comparison, and Outcome questions No Yes Yes Yes Yes Outcome prioritizatione Not formally done Not formally done Not formally done Not formally done Yes Search strategyf Panel members Panel members Panel members A professional medical librarian with input from panel A team of professional medical librarians with input from panel Source of evidenceg Medline database Medline database At least one electronic database At least two electronic databases At least two electronic databases Assessment of the quality of evidenceh Approach by Sackett (13) GRADE approach GRADE approach GRADE approach GRADE approach Involvement of patients No No No No 11 public members were involvedi Formulating recommendation Modified Delphi approach GRADE approach GRADE approach GRADE approach GRADE approach and evidence to decision frameworkj Types of recommendations and statements Recommendations with no distinction of strength Level 1k Level 1k Strong recommendation Strong recommendation Level 2 Level 2 Weak recommendation Weak recommendation Ungraded statements BPSl BPSl Implications of recommendationsm No Yes Yes Yes Yes BPS = best practice statement, COI = conflicts of interest, GRADE = Grading of Recommendations, Assessment, Development, and Evaluation, SSC = Surviving Sepsis Campaign.aThe total number of panel members that participated in the guideline.bTypes of COI that were adjudicated and managed in each guideline.cUsing a systematic approach to select topics of high priority to address in the guideline.dUsing a framework based on practice variability (based on surveying 800 intensivists globally), awareness of new evidence, and panel ranking.eUsing a systematic approach to identify outcomes that are relevant most to patients for each recommendation.fThis category outlines who performed the electronic literature searches for relevant evidence.gElectronic databases that were searched for relevant studies.hAlso known as confidence or certainty in the evidence.iPublic members provided input on outcomes importance, patients’ values, and reviewed the final list of recommendations especially for the long-term outcomes subgroup.jEvidence to decision framework asks the panel to consider the following factors when formulating the recommendation: priority of topic, magnitude of benefit and harm, balance of effect, patent values, cost and resources, acceptability, and feasibility.kLevel 1 recommendation corresponds to a strong recommendation, while level 2 corresponds to a weak recommendation.lAll BPSs followed strict criteria, a BPS is an ungraded strong recommendation.mThis step involves clearly stating the implications of recommendations for stakeholders including patients, healthcare providers, and policymakers. EVOLUTION OF SPECIFIC GUIDELINE RECOMMENDATIONS OVER THE YEARS High-quality studies, adequately designed and powered to assess patient-centered outcomes, have facilitated updates of guidelines’ recommendations over time. Earlier editions of the guidelines included recommendations supported by publications reporting positive results with the testing of promising science-based interventions in sepsis (1,2). Progress, however, is often not linear. Confirmatory trials are needed and these confirmatory trials, even when they are negative, still move the field forward. The first study published (2001) that appeared to show a significant impact of a specific treatment of sepsis was that of “early goal-directed therapy (EGDT),” which supported resuscitation of sepsis-induced tissue hypoperfusion targeting first central venous pressure (CVP) and then superior vena cava oxygen saturation (Scvo2) measured with catheter insertion into the superior vena cava (1). The 2004, 2008, and 2012 SSC guidelines recommended these two EGDT targets. However, after three multicenter and multinational studies failed to confirm improved outcomes with CVP and Scvo2 targets, the 2016 guidelines no longer recommended them (16–18). Nonetheless, SSC maintained the principles of early resuscitation with clear statements that sepsis and septic shock are medical emergencies and individualized resuscitation targets remain important. Similarly, recommendations for choice of resuscitation fluids have changed throughout the years. In 2004, results of a meta-analysis in general critically ill patients supported a weak recommendation to use either crystalloids or colloids as first-line fluids for sepsis resuscitation (19). Subsequent large RCTs demonstrated worse outcomes with artificial starch colloids (20–23). In 2012, the SSC guidelines recommended the use of crystalloids as first-line fluids and recommended against the use of starches (24). Albumin was recommended in the fluid resuscitation of severe sepsis and septic shock when patients require substantial amounts of crystalloids. The 2016 guidelines supported either balanced solutions or saline for crystalloid resuscitation. This recommendation was revisited in 2021 based on the Saline against Lactated Ringer’s or Plasma-Lyte and Isotonic Solutions and Major Adverse Renal Events Trial trials and modified to preferential use of balanced crystalloid solutions (25,26). Following publication of the 2021 guidelines, two large randomized trials and a meta-analysis comparing normal saline and balanced crystalloids were published (27–29). The subgroup of septic patients in the meta-analysis including these two new studies was reported to have an estimated effect on mortality using balanced crystalloids that ranged from a 14% relative reduction to a 1% relative increase in mortality. This new evidence should be taken into consideration in the next interaction of the SSC guidelines. After a single study showed a large reduction in mortality with the use of low dose (200 mg/d) IV hydrocortisone, the 2004 SSC guideline issued a recommendation to use corticosteroids in patients with septic shock (3). However, a subsequent study in septic patients with less severe shock failed to show improved outcome and the 2008 guidelines suggested corticosteroids be given only to adults with septic shock after blood pressure is identified to be poorly responsive to fluid resuscitation and vasopressor therapy (30). The wording of the recommendation changed subtly in 2012 and 2016 to suggest against IV hydrocortisone to treat adult septic shock patients if adequate fluid resuscitation and vasopressor therapy were able to restore hemodynamic stability. Two major RCTs published in 2018, Adjunctive Corticosteroid Treatment in Critically Ill Patients with Septic Shock and Activated Protein C and Corticosteroids for Human Septic Shock, led to a weak recommendation in the 2021 guidelines in favor of corticosteroids in patients with ongoing shock following fluid resuscitation. The committee follows this recommendation with a remark that supports a specific threshold vasopressor dose for triggering administration, with steroids commenced when the dose of norepinephrine or epinephrine is greater than or equal to 0.25 µg/kg/min at least 4 hours after vasopressor initiation (31,32). Using steroids as an example, the timeline depicted in Figure 2 gives a sense of how clinical trial results of corticosteroids in septic shock influenced guidelines recommendations between 2002 and 2021.Figure 2.: Timeline of interaction of clinical trial results of corticosteroids in septic shock on guidelines recommendations between 2002 and 2021. 1 - 2004 Surviving Sepsis Campaign (SSC) - recommends IV hydrocortisone for septic shock. 2 - 2008 SSC suggests IV hydrocortisone only when blood pressure is poorly responsive to fluids and vasopressors. 3 - 2012, 2016 SSC suggests against IV hydrocortisone if fluids and vasopressors restore hemodynamic stability. 4 - 2021 SSC suggests IV hydrocortisone use with ongoing requirement for higher vasopressor therapy. ADRENAL = Adjunctive Corticosteroid Treatment in Critically Ill Patients with Septic Shock, APROCCHSS = Activated Protein C and Corticosteroids for Human Septic Shock, JAMA = Journal American Medical Association, NEJM = New England Journal of Medicine.SURVIVING SEPSIS CAMPAIGN GUIDELINES INFLUENCE Five sets of SSC guidelines were published in the journals Critical Care Medicine and Intensive Care Medicine, totaling 10 source publications. In Scopus, the guidelines have been cited 21,158 times in over 7,000 publications, 25 non-English languages, representing authors from over 125 countries. The influence of the SSC guidelines on sepsis thinking is demonstrated through two Scopus citation metrics: the Field Weighted Citation Impact (FWCI) and the Citation Benchmark. The FWCI demonstrates how well cited a publication is compared with similar publications. A FWCI greater than 1.00 signifies greater citations than expected according to the average. A FWCI of 1.48 means 48% more citations accrued than expected. The SSC guideline publication FWCI ranged from 19 to 186.51 (Table 3). The Citation Benchmark demonstrates how citations compare with the average for similar publications. Every SSC published guideline since the inaugural publication in 2004 is in the 99th percentile indicating all SSC guidelines ranked in the top 1% globally for the number of generated citations. TABLE 3. - Surviving Sepsis Campaign Guideline Publication Impact by Scopus and Year Guidelines Year Journal Title Publication Date Metric Dates Citations (n) Field Weighted Citation Index Citation Benchmark Percentile 2004 CCM 2004 2004–2019 2,562 72.93 99th 2004 ICM 2004 2004–2019 671 19 99th 2008 CCM 2008 2008–2022 3,909 119 99th 2008 ICM 2008 2008–2022 1,358 37.62 99th 2012 CCM 2013 2012–2022 4,327 186.51 99th 2012 ICM 2013 2012–2022 3,172 120.03 99th 2016 CCM 2017 2016–2022 1,755 71.31 99th 2016 ICM 2017 2016–2022 3,214 181.2 99th 2020a CCM 2020 2020–2022 430 51.63 99th 2020a ICM 2020 2020–2022 980 119.87 99th 2021 CCM 2021 2021–2022 75 30.27 99th 2021 ICM 2021 2021–2022 115 45.19 99th CCM = Critical Care Medicine, ICM = Intensive Care Medicine.aSurviving Sepsis Campaign COVID Guidelines.Guidelines year is the year during which work was completed. Publication year is the year the document was published. Scopus allows for 15 yr of continuous data from publication date. Field Weighted Citation Impact demonstrates how well cited an article is compared with like publications. It considers the year of publication, document type, and disciplines associated with its source. It is the ratio of the publication’s citations to the average number of citations received by all similar publications over a 3-yr window. An FWCI > 1 means the output is more cited than expected according to the global average (Field Weighted Citation Impact of 1.48 mean 48% more cited than expected). The Citation Benchmark demonstrates how citations received by a publication compare with the average for similar publications in an 18-mo window. It accounts for the publication year, document type, and source disciplines. The 99th percentile is high and indicates a publication is in the top 1% globally for the number of citations. Scopus access date: June 5, 2022. The Altmetric Attention Score indicates the amount of attention the work received through additional outlets such as news, blogs, and social media. All SSC guidelines publications scored in the top 5% of all research outputs ever tracked by Altmetric. When comparing articles of a similar publication period, the SSC guidelines ranked higher than 97% to 99% of its contemporaries (Table 4). Additional impact is demonstrated through several domains including related nursing publications (33,34) and a nursing complement to the SSC guidelines, which ranked in the 85th percentile for citations compared with like articles in Scopus with a FWCI of 1.82 (34). TABLE 4. - Surviving Sepsis Campaign Guideline Publication Impact by Altmetric Attention Score and Year Guidelines Year Journal Title Publication Date Altmetrics: Attention Score Altmetrics Attention Score: Percentile of All Research Outputs Ever Tracked by Altmetric Altmetrics High Attention Score: Percentile Compared With Articles of Similar Age 2004 CCM 2004 37 96 97 2004 ICM 2004 ND ND ND 2008 CCM 2008 47 96 98 2008 ICM 2008 38 96 97 2012 CCM 2013 120 98 99 2012 ICM 2013 62 97 98 2016 CCM 2017 381 99 99 2016 ICM 2017 694 66 99 2020a CCM 2020 456 99 99 2020a ICM 2020 591 99 99 2021 CCM 2021 638 99 99 2021 ICM 2021 1,284 99 99 CCM = Critical Care Medicine, ICM = Intensive Care Medicine, ND = no data.aSurviving Sepsis Campaign COVID Guidelines.Guidelines year is the year during which work was completed. Publication year is the year the document was published. The Altmetric Attention Score functions as an indicator of the amount of attention the work received through additional outlets such as news, blogs, and social media. All the Surviving Sepsis Campaign (SSC) guidelines publications scored in the top 5% of all research outputs ever tracked by Altmetric. When comparing articles of a similar publication period, the SSC guidelines ranked higher than 97% to 99% of its contemporaries. Altmetric data accessed from source journals, access date: June 5, 2022. SSC PERFORMANCE IMPROVEMENT PROGRAM After publication of the 2004 guidelines, the SSC set about creating an international quality improvement program that would facilitate hospitals in implementing evidence-based interventions. The SSC approached the Institute for Healthcare Improvement (IHI) (Boston, MA) to collaborate on designing this approach (35). The IHI had been pioneering the use of “bundle technology,” groupings of evidence-based interventions in disease management that when performed reliably have been shown to improve outcomes (36) in management of other diseases. The key to bundles is two-fold: first that bundle elements should be easy to measure and second, that elements should be linked in time. Working with IHI’s critical care leaders, the 6- and 24-hour sepsis bundles, based on the new guidelines, were developed. Other improvement strategies gleaned from IHI included the importance of measuring the change in clinical care that is associated with bundle implementation—“How do you know if the changes you are making are leading to improvement?” (37) and developing collaborative communities who learn from each other as they improve care (37). Novel at the time, the SSC commissioned the design of a database with a graphic user interface to screen patients, collect data, provide results of that data to bedside practitioners and produce improvement reports that demonstrated progress over time. The local version of the database, installed at hospitals worldwide, captured patient-level information related to sepsis bundle interventions and transmitted data back to a master database at the SCCM. To encourage site participation, a downloadable version of the database translated into seven languages was deployed in 2006. Between 2005 and 2012, the SSC conducted a number of learning collaboratives with global reach. In the United States, individual collaboratives were sponsored on the East Coast, Midwest, and West Coast comprising 59 hospitals. Similarly, successful sepsis performance improvement collaboratives were also conducted in other countries including England, Spain, and Brazil (38–40). These efforts, plus SSC list serve and website promotion, were sufficient to catalyze substantial participation in the SSC performance improvement campaign. In 2010, analyzing 15,022 patients’ data from 165 sites worldwide, Levy et al (41) reported that the adjusted odds ratio for mortality improved the longer a site was in the Campaign, resulting in an adjusted absolute drop of 0.8% per quarter and 5.4% over 2 years (95% CI, 2.5–8.4 yr). By 2015, data from 29,470 patients were available for analysis demonstrating that each 10% increase in compliance and additional quarter of participation in the SSC initiative was associated with a significant decrease in the odds ratio for hospital mortality (42). As the evidence base supporting sepsis interventions evolved, the SSC revised the sepsis bundles to reflect changes in guidelines, recommendations and contemporary practice. Following studies challenging the efficacy of tight glucose control in the medical ICU, administration of hydrocortisone for shock and use of recombinant activated protein C, the 6- and 24-hour bundles were replaced with the 3- and 6-hour bundle in 2012 (30,43,44