Sepsis is part of every critical care practice. Care for septic patients is provided by internists, anesthesiologists, surgeons, pediatricians, neurologists, neurosurgeons, emergency physicians, nurses, respiratory therapists, and pharmacists. The disorder effects patients of all ages—babies, children, adolescents, young adults, older adults, and geriatric patients. It presents in every type of intensive care unit (ICU)—medical ICUs (MICUs), surgical ICUs (SICUs), pediatric ICUs (PICUs), trauma ICUs, coronary care and cardiac surgical ICUs, mixed (Med/Surg) ICUs, and neonatal ICUs (NICUs). Caring for septic patients is expensive. In 2011, managing septic patients in hospitals in the United States cost more than $20 billion, 5.2% of total hospital costs. Sepsis is common, but it is difficult to determine just how common. Four large U.S. population studies reported that the incidence of sepsis rose an average of 13% per year between 2004 and 2009. Iwashyna et al. determined that, over a 12-year period beginning in 1996, the incidence of sepsis increased threefold among patients receiving Medicare; by 2008, there were nearly 1,000,000 new cases each year in the United States and untold numbers worldwide. However, Gaieski et al. reported that, depending on the method of database abstraction used, the incidence of sepsis varied 3.5-fold, from 300 to 1031/100,000 population, and in-hospital mortality varied twofold, from 14.7% to 29.9%. Nonetheless, even conservative estimates indicate that sepsis is a major cause of mortality and morbidity worldwide. In the United States, the annual number of sepsis-associated deaths likely rivals that for coronary heart disease (375,000) and may exceed the mortality attributable to the four deadliest forms of cancer combined. It is also now recognized that sepsis survivors are plagued by functional, cognitive, and emotional disabilities that constitute an additional burden on the health care system. In short, “sepsis” identifies a syndrome that constitutes an increasingly important public health concern.
If asked to define sepsis, however, most ICU providers would struggle. They would likely invoke some combination of infection and inflammatory markers—temperature, heart rate, respiratory rate, and white blood cell count—the so-called SIRS (systemic inflammatory response syndrome) criteria. Most would also acknowledge that this definition, first articulated in a 1992 paper, identifies a large number of patients who are not septic. A more detailed discussion is clearly in order.
The term sepsis is generically used to describe a set of clinical, pathologic, and biochemical changes that may accompany infection. The root of the word is derived from the ancient Greek for “to decay” or “to putrefy.” Over the years, sepsis has been used to describe a wide and often bewildering array of medical conditions. Consensus conferences convened in 1991 and 2001 to provide structure and clarity focused on the then-accepted view that sepsis represented a generalized inflammatory host response to infection. The emphasis of the initial consensus conference on deriving a clinically useful construct resulted in the creation of SIRS. Sepsis was defined as “the systemic response to infection,” that is, SIRS in a patient with suspected, presumed, or identified infection. Sepsis could progress until patients had organ dysfunction or “organ failure.” When organ dysfunction was present, the syndrome was termed severe sepsis . Severe sepsis could continue to progress across a continuum to septic shock, defined as “sepsis-induced hypotension persisting despite adequate fluid resuscitation.” Finally, patients with sepsis could progress to develop a disorder where most organ systems function abnormally, the multiple organ dysfunction syndrome (MODS). Issues with these definitions, in particular with the nonspecific nature of SIRS, arose almost immediately. Indeed, soon after their publication, the primary architect of the 1991 definitions, Roger Bone, proposed the existence of a compensatory anti-inflammatory response syndrome (CARS), while others noted that changes in organ function corrected with survival, a finding not thought to be consistent with organ failure. Ongoing concerns ultimately led to the 2001 consensus conference. The participants noted that SIRS was present in the great majority of critically ill patients—indeed, 93% fulfill criteria on admission as part of the host response to any critical illness. The identified limitations in SIRS were addressed by expanding the list of defining criteria used. The determinants of severe sepsis were left unchanged, whereas criteria to characterize the “state of acute circulatory failure” of septic shock were expanded incrementally. Despite a number of developments suggesting that reassessment was needed, no formal attempt to revisit the definitions was subsequently undertaken. In effect, the definitions of sepsis, septic shock, and organ dysfunction remained largely unchanged for two decades.
Because of these inherent concerns, a Task Force of intensivists, infectious disease experts, and pulmonologists was convened in January 2014. The group was provided unrestricted support by the Society of Critical Care Medicine (SCCM) and the European Society of Intensive Care Medicine (ESICM), and was charged to reexamine and, as deemed appropriate, revise existing definitions of sepsis, severe sepsis, septic shock, and organ dysfunction. Proposed changes were to reflect a deeper understanding of the pathophysiology of sepsis and the availability of large electronic health records, clinical databases, and patient registries. Endorsement of the document from major international societies focused on intensive care and other relevant disciplines was then obtained, and the document was published in 2016.
Why Is Reassessing the Definitions of Sepsis and Sepsis-Related Conditions So Essential—and So Difficult?
Variability in Terminology
Sepsis. A patient is currently defined as having sepsis if she or he has presumed or demonstrated infection in the presence of two or more SIRS criteria. This approach is problematic. SIRS reflects inflammation, a normal response to “danger,” and thus is neither pathologic nor necessarily infection driven. Further, the current reliance on SIRS has led to inconsistencies in hospital reporting and in accruing epidemiologic data. As noted, Giaeiski et al. found a 3.5-fold variation in the incidence of sepsis and a twofold variation in hospital mortality when different abstraction methods were applied to the same patient population. In addition, a recent study showed that a significant number of patients admitted to critical care units with infection and organ failure did not meet SIRS criteria. Thus, current definitions impede diagnosis, assessment of outcome, and the gathering of epidemiologic data.
Organ Dysfunction . Although identifying sepsis-induced organ-specific abnormalities as “failure” has become infrequent, several issues arise when considering organ dysfunction. Perhaps most important is a lack of reliable criteria to identify “dysfunction” that is evident when one examines the selected clinical findings, laboratory data, or therapeutic interventions in current use. Many abnormalities are nonspecific, and basing assessment of severity on intervention is subject to a host of concerns, especially when therapeutic approaches are changing and their application is not subject to uniform guidelines. It has also become common practice to combine organ-based abnormalities into scoring systems. Perhaps the most frequently used system is the Sequential Organ Failure Assessment Score (SOFA), which was originally called the Sepsis-related Organ Failure Assessment Score. SOFA, however, illustrates many of the concerns with current approaches to organ dysfunction. For example, an elevation in serum bilirubin levels, which in SOFA is used to demonstrate hepatic dysfunction, can arise from hemolysis. Coagulation abnormalities are noted by a decrease in platelet counts, which is just as likely to reflect an effect on bone marrow, especially when noted in conjunction with anemia or neutropenia. The progression of cardiovascular dysfunction is based on escalating doses of vasopressors, which may be managed in different ways according to local custom. Indeed, the reliance of the SOFA cardiovascular score is based on the dose of dopamine, a drug that is no longer routinely administered. Finally, most indices, and SOFA in particular, were not derived based on data but rather reflected consensus, especially with regard to variable selection and cutoff values. That said, a higher SOFA score does seem to indicate an increasing probability of mortality.
Septic Shock . Problems with the existing definitions are particularly evident when septic shock is examined. As part of the Task Force effort, Shankar-Hari et al. undertook a systemic review of the literature. The results indicate that septic shock is defined by the presence of infection (presumed or confirmed) in conjunction with combinations of terms that are themselves defined with distressing variability, namely:
- •
Hypotension (SBP [systolic blood pressure] <90 mm Hg or MAP [mean arterial pressure] <60 or <70 mm Hg or fall in SAP pressure >40 mm Hg from baseline or >2 standard deviations from the norm for age despite “adequate fluid resuscitation”)
- •
The presence of abnormal biochemical variables (e.g., lactate >2 or >4 mmol/L or base deficit >5 mmol/L)
- •
The use of inotropes or vasopressors (not necessarily above a prespecified dose)
- •
New onset organ dysfunction (defined variably with various scoring systems such as APACHE [Acute Physiology and Chronic Health Evaluation] II, APACHE III, or the cardiovascular component of the SOFA score)
Further complicating matters are the following:
- •
Variable endpoints of adequacy of fluid resuscitation (rarely defined or reported)
- •
Variable durations of hypotension or vasopressor therapy
- •
Failure to account for the underlying blood pressure of the patient or other comorbidities
- •
Failure to account for the hypotensive effect of cointerventions such as vasodilating and/or cardiodepressant sedative agents.
In effect, issues with terminology are even more problematic for septic shock than for sepsis.
Improved Understanding of Sepsis Pathobiology
There are inherent challenges in defining sepsis. First and foremost, the term sepsis describes a complex and poorly understood process. There are no simple or unambiguous clinical criteria or biological characteristics that differentiate patients who are “septic” from those who are not. That is, clinical features or current animal models poorly delineate the complexity, variability, and time dependence of the sepsis phenotype.
The original conceptualization of sepsis as infection with SIRS focused solely on immune excess. Recent advances have demonstrated that sepsis involves early activation of both proinflammatory and anti-inflammatory responses, of nonimmunologic pathways such as cardiovascular, neuronal, autonomic, hormonal, bioenergetic, metabolic, and coagulation, all of which carry prognostic significance. Modulation of septic response involves not only the immune system but also the endocrine and central nervous systems. Organ dysfunction, even when severe, is not necessarily associated with significant cell damage, whereas a sepsislike biological response may be triggered by noninfectious host factors (damage-associated molecular patterns). In short, criteria currently in use are no longer consistent with what we understand about the pathobiologic of sepsis.
A Need for Sepsis Definitions for the Lay Public and for Health-care Practitioners
Despite its prevalence, associated morbidity, contribution to the rising cost of health care, worldwide importance, and recent high-profile cases, public awareness of sepsis is poor, and limited resources are directed toward sepsis-associated research. Furthermore, recognition of clinical sepsis is difficult for trained medical personnel and can be entirely obscure to the populace at large. There is thus a glaring need for a description of sepsis that can be appreciated by the nonmedical public.
In addition, the diagnosis of sepsis is problematic even for experienced practitioners. A recent high-profile death of a 12-year-old boy (Rory Staunton) from sepsis at least in part reflects this difficulty—sepsis is a condition that can confound the most experienced practitioners, especially at times of the year when other conditions that present with similar protean signs and symptoms are common (e.g., during flu season). Thus, health-care practitioners would benefit greatly from a simple, validated set of bedside criteria that directs them to consider sepsis when presented with infected patients.
Availability of Patient Data
Although there have been a number of important clinical trials regarding therapy for sepsis, a great deal of the material on which the definitions of sepsis are based reflects findings from small studies or expert opinion. By their nature, it is difficult to validate such studies. However, there are now in existence a number of large electronic health record (EHR) databases and patient registries that either relate directly to sepsis itself (e.g., the Surviving Sepsis Campaign [SSC] database ) or contain general information that could be leveraged to study patients with sepsis. These datasets enable the derivation and validation of variables that better capture the incidence, severity, and trajectory of sepsis.
Nature of the Problem: What Is a “Definition”?
Per the Merriam-Webster dictionary, a definition is “a statement expressing the essential nature of something.” In effect, a definition is the “gold standard” to be compared with any other descriptions or collection of signs and symptoms used to identify something. There is no such gold standard for sepsis. In contrast to “cancer,” there is no tissue specimen that, if examined under a microscope or in some other manner, can be unambiguously identified as sepsis. Unlike cystic fibrosis, there is no specific, characteristic genetic abnormality. Sepsis even differs from infection, where a culture can identify the offending microorganism. Thus, by its very nature, any current definition of sepsis cannot be validated in the clinical realm. The best that can be hoped for is to delineate characteristics or criteria to identify patients with some proxy for sepsis, and who are thus highly likely to have sepsis.
Related posts:
What Lessons Have Intensivists Learned During the Evidence-Based Medicine Era?
What Is the Role of Invasive Hemodynamic Monitoring in Critical Care?
Should Fever Be Treated?
Are Computerized Algorithms Useful in Managing the Critically Ill Patient?
Inhaled Vasodilators in ARDS: Do They Make a Difference?
Has Outcome in Sepsis Improved? What Has Worked? What Has Not Worked?
Full access? Get Clinical Tree
