Evaluation and management of suspected sepsis and septic shock in adults

INTRODUCTION — Sepsis is a clinical syndrome characterized by systemic inflammation due to infection. There is a continuum of severity ranging from sepsis to septic shock. Although wide-ranging and dependent upon the population studied, mortality has been estimated to be ≥10 percent and ≥40 percent when shock is present [1,2].

In this topic review, the management of sepsis and septic shock is discussed. Our approach is consistent for the most-part with 2021 guidelines issued by the Surviving Sepsis Campaign [3,4].

While we use the Society of Critical Care Medicine (SCCM)/European Society of Intensive Care Medicine (ESICM) definitions, such definitions are not unanimously accepted. For example, the Center for Medicare and Medicaid Services (CMS) still continues to support the previous definition of systemic inflammatory response syndrome, sepsis, and severe sepsis. In addition, the Infectious Diseases Society of America (IDSA) has pointed out that use of such definitions, while lifesaving for those with shock, may lead to overtreatment with broad-spectrum antibiotics for those with milder variants of sepsis [5].

Definitions, diagnosis, pathophysiology, and investigational therapies for sepsis, as well as management of sepsis in the asplenic patient are reviewed separately. (See "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis" and "Pathophysiology of sepsis" and "Investigational and ineffective pharmacologic therapies for sepsis" and "Clinical features, evaluation, and management of fever in patients with impaired splenic function".)

IMMEDIATE EVALUATION AND MANAGEMENT — Securing the airway (if indicated) and correcting hypoxemia, and establishing venous access for the early administration of fluids and antibiotics are priorities in the management of patients with sepsis and septic shock [3,4]. The following table summarizes emergency management of the patient with severe sepsis during the first hour (table 1).

Stabilize respiration — Supplemental oxygen should be supplied to all patients with sepsis who have indications for oxygenation, and oxygenation should be monitored continuously with pulse oximetry. Ideal target values for peripheral saturation are unknown, but we typically target values between 90 and 96 percent. Intubation and mechanical ventilation may be required to support the increased work of breathing that frequently accompanies sepsis or for airway protection since encephalopathy and a depressed level of consciousness frequently complicate sepsis [6,7]. Target values for oxygenation, techniques and sedative and induction agents for intubation are discussed separately. (See "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Fraction of inspired oxygen' and "Induction agents for rapid sequence intubation in adults for emergency medicine and critical care" and "Overview of advanced airway management in adults for emergency medicine and critical care" and "Rapid sequence intubation in adults for emergency medicine and critical care" and "The decision to intubate" and "Direct laryngoscopy and endotracheal intubation in adults".)

Establish venous access — Venous access should be established as soon as possible in patients with suspected sepsis. While peripheral venous or intraosseous access may be sufficient in some patients, particularly for initial resuscitation, the majority will require central venous access at some point during their course. However, the insertion of a central line should not delay the administration of resuscitative fluids and antibiotics. A central venous catheter (CVC) can be used to infuse intravenous fluids, medications (particularly vasopressors), and blood products, as well as to draw blood for frequent laboratory studies. While a CVC can be used to monitor the therapeutic response by measuring the central venous pressure (CVP) and the central venous oxyhemoglobin saturation (ScvO₂), evidence from randomized trials suggest that their value is limited [8-13]. (See "Central venous catheters: Overview of complications and prevention in adults" and 'Monitor response' below.)

Initial investigations — An initial brief history and examination, as well as laboratory, microbiologic (including blood cultures), and imaging studies are often obtained simultaneously while access is being established and the airway stabilized. This brief assessment yields clues to the suspected source and complications of sepsis, and therefore, helps guide empiric therapy and additional testing (table 2). (See "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis", section on 'Clinical presentation' and 'Empiric antibiotic therapy (first hour)' below.)

Quickly obtaining the following is preferable (within 45 minutes of presentation) but should not delay the administration of fluids and antibiotics:

●Complete blood counts with differential, chemistries, liver function tests, and coagulation studies including D-dimer level. Results from these studies may support the diagnosis, indicate the severity of sepsis, and provide baseline to follow the therapeutic response.

●Serum lactate – An elevated serum lactate (eg, >2 mmol/L or greater than the laboratory upper limit of normal) may indicate the severity of sepsis and is used to follow the therapeutic response [3,4,14-16].

●Peripheral blood cultures (aerobic and anaerobic cultures from at least two different sites), urinalysis, and microbiologic cultures from suspected sources (eg, sputum, urine, intravascular catheter, wound or surgical site, body fluids, rapid antigen or polymerase chain reaction tests) from readily accessible sites. Drawing blood for cultures through an indwelling or central intravascular catheter should be avoided whenever possible since ports are frequently colonized with skin flora, thereby increasing the likelihood of a false-positive blood culture. If blood cultures are drawn from an intravenous line, a second specimen should be drawn from a peripheral venipuncture site.

The importance of early blood cultures was best illustrated in a multicenter randomized trial of 325 patients with a presumed or confirmed source of infection and hypotension or elevated lactate >4 mmol/L [17]. All patients had two sets of blood cultures drawn from two separate sites before antimicrobial administration and a second set of blood cultures was obtained from zero to four hours after antimicrobial administration. Pre-antimicrobial cultures were positive in 31.4 percent compared with 19.4 percent post-antimicrobial administration. When pre-antimicrobial cultures were considered the reference gold standard, the sensitivity of post-antimicrobial blood cultures was 53 percent. When other cultures were included with post-antimicrobial blood cultures, pathogens were identified in approximately two-thirds of patients. Although some methodologic issues (eg, in some patients only one blood culture or one venipuncture was obtained instead of two), this study nonetheless highlights the importance of taking blood cultures prior to antimicrobial administration. Importantly, both the collection of cultures and the initiation of antimicrobial therapy should be prompt in those with the signs of severe sepsis.

●Arterial blood gas (ABG) analysis – ABGs may reveal acidosis, hypoxemia, or hypercapnia.

●Imaging targeted at the suspected site of infection is warranted (eg, chest radiography, computed tomography of chest and/or abdomen).

●Procalcitonin – While the diagnostic value of procalcitonin in patients with sepsis is poorly supported by evidence, its value in de-escalating antibiotic therapy is well established in populations, in particular, those with community acquired pneumonia and respiratory tract infections; measurement of procalcitonin to guide duration of antibiotic use is appropriate in those populations with sepsis, while its role in other groups is unclear. While one meta-analysis of 5158 critically ill patients reported a mortality benefit associated with procalcitonin use a survival benefit was not noted in the subset of patients with sepsis [18]. In contrast, a randomized trial of 266 patients with sepsis from lower respiratory tract infections, acute pyelonephritis, or primary bloodstream infection, procalcitonin guided therapy was associated with a reduction in 28-day mortality (28 versus 15 percent) and length of antibiotic therapy (5 versus 10 days) [19]. While encouraging, additional data from well-conducted randomized trials in patients with sepsis are required before we can routinely recommend procalcitonin use all patients with sepsis. Detailed evidence to support the use of procalcitonin is provided separately. (See "Procalcitonin use in lower respiratory tract infections" and "Clinical evaluation and diagnostic testing for community-acquired pneumonia in adults", section on 'Serum biomarkers'.)

INITIAL RESUSCITATIVE THERAPY — The cornerstone of initial resuscitation is the rapid restoration of perfusion and the early administration of antibiotics. The following table summarizes emergency management of the patient with severe sepsis during the first hour (table 1).

●Tissue perfusion is predominantly achieved by the aggressive administration of intravenous fluids (IVF), usually crystalloids (balanced crystalloids or normal saline) given at 30 mL/kg (actual body weight), started by one hour and completed within the first three hours following presentation.

●Empiric antibiotic therapy is targeted at the suspected organism(s) and site(s) of infection and preferably administered within the first hour.

Our approach is based upon several major randomized trials that used a protocol-based approach (ie, early goal-directed therapy [EGDT]) to treating sepsis [8-13]. Components of the protocols usually included the early administration of fluids and antibiotics (within one to six hours) using the following targets to measure the response: central venous oxyhemoglobin saturation (ScvO₂) ≥70 percent, central venous pressure (CVP) 8 to 12 mmHg, mean arterial pressure (MAP) ≥65 mmHg, and urine output ≥0.5 mL/kg/hour. Although all trials [9-11] (except for one [8]) did not show a mortality benefit to EGDT, it is thought that the lack of benefit was explained by an overall improved outcome in both control and treatment groups and to improved clinical performance by trained clinicians in academic centers during an era that followed an aggressive sepsis education and management campaign. In support of this hypothesis is that central line placement was common (>50 percent) in control groups so it is likely that CVP, ScvO₂, and/or lactate clearance were targeted in these patients. Furthermore, the mortality in studies that did not report a benefit to EGDT [9-11] approximated that of the treatment arm in the only study that reported benefit [8].

●One single center randomized trial of 263 patients with suspected sepsis reported a lower mortality in patients when ScvO₂, CVP, MAP, and urine output were used to direct therapy compared with those in whom only CVP, MAP, and urine output were targeted (31 versus 47 percent) [8]. Both groups initiated therapy, including antibiotics, within six hours of presentation. There was a heavy emphasis on the use of red cell transfusion (for a hematocrit >30) and dobutamine to reach the ScvO₂ target in this trial.

●Three subsequent multicenter randomized trials of patients with septic shock, ProCESS [9], ARISE [10], and ProMISE [11] and two meta-analyses [12,13] all reported no mortality benefit (mortality ranged from 20 to 30 percent), associated with an identical protocol compared with protocols that used some of these targets or usual care. In contrast, one meta-analysis of 13 trials reported a mortality benefit from early-goal directed therapy within the first six hours [20].

●A lack of benefit of resuscitation protocols has also been reported in resource-limited settings. As an example, in a randomized trial of 212 patients with sepsis (defined as suspected infection plus two systemic inflammatory response syndrome criteria) and hypotension (systolic blood pressure ≤90 mmHg or MAP <65 mmHg) in Zambia, a protocolized approach of aggressive fluid resuscitation, monitoring, blood, and vasopressor transfusion within the first six hours of presentation resulted in a higher rate of death (48 versus 33 percent) when compared with usual care [21]. However, several flaws including crude measurements of monitoring, lower than usual rates of lactate elevation, larger than typical volumes of fluid resuscitation, and use of dopamine (as opposed to norepinephrine) in a population with a high percentage of patients with human immunodeficiency virus may have biased the results.

●Another analysis from a cohort of 1871 Canadian patients reported that prehospital administration of fluids by paramedics to patients with hypotension from sepsis may be of benefit, although it was associated with increased prehospital time [22].

●A post-hoc analysis of the ANROMEDA-SHOCK trial [23] suggested that peripheral perfusion-targeted resuscitation (using capillary refill time [CRT]) may be associated with a lower mortality and faster resolution of organ dysfunction than lactate-targeted resuscitation [24]. Additional studies suggest a poor correlation between MAP and CRT [25]. Further study is required before clinicians should use CRT routinely as a marker of resuscitation.

The importance of timely treatment, particularly with antibiotics, was illustrated in a database study of nearly 50,000 patients with sepsis and septic shock who were treated with various types of protocolized treatment bundles (that included fluids and antibiotics, blood cultures, and serum lactate measurements) [26]. Compared with those in whom a three-hour bundle (blood cultures before broad spectrum antibiotics, serum lactate level) was completed within the three-hour time frame, a higher in-hospital mortality was reported when a three-hour bundle was completed later than three hours (odds ratio [OR] 1.04 per hour). Increased mortality was associated with the delayed administration of antibiotics but not with a longer time to completion of a fluid bolus (as part of a six hour bundle) (OR 1.04 per hour versus 1.10 per hour).

Intravenous fluids (first three hours) — In patients with sepsis, intravascular hypovolemia is typical and may be severe, requiring rapid fluid resuscitation. (See "Treatment of severe hypovolemia or hypovolemic shock in adults".)

Volume — Intravascular hypovolemia is typical and may be severe in sepsis. Rapid, large volume infusions of IVF (30 mL/kg) are indicated as initial therapy for severe sepsis or septic shock, unless there is convincing evidence of significant pulmonary edema. This approach is based upon several randomized trials that reported no difference in mortality when mean infusion volumes of 2 to 3 liters were administered in the first three hours [9-11] compared with larger volumes of three to five liters, which was considered standard therapy at the time [8]. However, some patients may require higher than recommended volumes, particularly those who demonstrate clinical and/or hemodynamic indicators of fluid-responsiveness. (See 'Monitor response' below.)

Fluid therapy should be administered in well-defined (eg, 500 mL), rapidly infused boluses. The clinical and hemodynamic response and the presence or absence of pulmonary edema must be assessed before and after each bolus. Intravenous fluid challenges can be repeated until blood pressure and tissue perfusion are acceptable, pulmonary edema ensues, or fluid fails to augment perfusion.

In one trial, over 1500 patients with sepsis-induced hypotension refractory to initial resuscitation with 1 to 3 L of IVF were randomized to receive either a restrictive fluid regimen (prioritizing vasopressors and lower intravenous fluid volumes) or liberal fluid regimen (prioritizing higher volumes of intravenous fluids before vasopressor use). The restrictive strategy entailed early administration of vasopressors after infusion of up to 2 L of fluid including prerandomization fluid, if needed. The liberal regimen prioritized infusion of an additional 2 L at randomization in addition to prerandomized fluid and additional boluses as needed [27]. Despite more fluid administration in the liberal group compared with the restrictive group (3400 versus 1267mL at 24 hours), there was no difference in 90-day mortality (15 versus 14 percent) or any other outcome measured. Vasopressors were administered earlier and for longer periods in those who received the restrictive regimen but did not appear to be associated with excess adverse effects. However, results may have been impacted by greater than intended fluid volumes administered in some patients in the restrictive group and lower than intended volumes in some patients in the liberal group, perhaps limiting the ability of the study to detect a meaningful difference between the interventions. In addition, the median volume administered in the liberal group is reasonably close to that which is widely practiced and both arms of this trial led to administration of significantly less fluid than would have been typical a decade ago; this suggests that adopting a more aggressive restrictive approach than currently practiced may not be associated with additional benefit.

Choice of fluid — Evidence from randomized trials and meta-analyses have found no convincing difference between using albumin solutions and crystalloid solutions (eg, Ringer's lactate, normal saline) in the treatment of sepsis or septic shock, but they have identified potential harm from using pentastarch or hydroxyethyl starch [28-37]. There is no role for hypertonic saline [38].

In our practice, we generally use a balanced crystalloid solution, and less commonly, normal saline, instead of an albumin solution because of the lack of clear benefit and higher cost of albumin. Balanced crystalloid rather than normal saline is preferred if there is a perceived need to avoid or treat the hyperchloremia that occurs when large volumes of nonbuffered crystalloid (eg, normal saline) are administered, although the data to support this practice are weak (and discussed separately). (See "Treatment of severe hypovolemia or hypovolemic shock in adults", section on 'Choice of replacement fluid'.)

Data discussing IVF choice among patients with sepsis include the following:

●Crystalloid versus albumin – Among patients with sepsis, several randomized trials and meta-analyses have reported no difference in mortality when albumin was compared with crystalloids, although one meta-analysis suggested benefit in those with septic shock [29,36,37]. In the Saline versus Albumin Fluid Evaluation (SAFE) trial performed in critically ill patients, there was no benefit to albumin compared with saline even in the subgroup with severe sepsis, who comprised 18 percent of the total group [28]. Among the crystalloids, there are no guidelines to suggest that one form is more beneficial than the other.

●Crystalloid versus hydroxyethyl starch (HES) – In the Scandinavian Starch for Severe Sepsis and Septic Shock (6S) trial, compared with Ringer's acetate, use of HES resulted in increased mortality (51 versus 43 percent) and renal replacement therapy (22 versus 16 percent) [30]. Similar results were found in additional trials of patients without sepsis.

●Crystalloid versus pentastarch – The Efficacy of Volume Substitution and Insulin Therapy in Severe Sepsis (VISEP) trial compared pentastarch to modified Ringer's lactate in patients with severe sepsis and found no difference in 28-day mortality [31]. The trial was stopped early because there was a trend toward increased 90-day mortality among patients who received pentastarch.

●Balanced salt solutions – Evidence to support the use of a balanced crystalloid solution or 0.9 percent saline in patients with sepsis is indirect and mostly derived from studies performed in a mixed population of critically ill patients. Further details are provided separately. (See "Treatment of severe hypovolemia or hypovolemic shock in adults", section on 'Choosing between 0.9 percent saline and buffered crystalloid'.)

Treating metabolic acidosis — Whether metabolic acidosis associated with sepsis should be treated with bicarbonate is discussed separately. (See "Bicarbonate therapy in lactic acidosis".)

Empiric antibiotic therapy (first hour) — Prompt identification and treatment of the site(s) of infection is the primary therapeutic intervention, with most other interventions being purely supportive.

Identification of suspected source — Empiric antibiotics should be targeted at the suspected source(s) of infection which is typically identified from the initial history, physical examination, and preliminary laboratory findings and imaging (table 2) (see 'Initial investigations' above). However, additional diagnostic testing or interventions may be required to identify the anatomic site(s) of infection. In particular, in addition to antibiotics, closed-space infections should be promptly drained or debrided (eg, empyema, abscess) for effective source control. (See 'Septic focus identification and source control' below.)

Timing — Once a presumed diagnosis of sepsis or septic shock has been made, optimal doses of appropriate intravenous antibiotic therapy should be initiated, preferably within one hour of presentation and after cultures have been obtained (see 'Initial investigations' above). The early administration of antimicrobials is often challenging because various patient- and institutional-related factors may lead to delays in timely treatment [39]. Institutional protocols should address timeliness as a quality improvement measure [40]. However, several clinician groups including the Infectious Diseases Society of America (IDSA) criticized the Society of Critical Care Medicine (SCCM) for promoting standards defining rigid time frames for initiation of antibiotics as it is often difficult to determine the actual onset of sepsis in individual patients; in addition, the one hour time frame could lead to overuse and inappropriate administration of unwarranted antimicrobials [5,41]. The IDSA favors removal of a recommendation for specific minimum time frames for initiating antibiotic therapy and instead advocates replacing these recommendations with the statement that prompt administration of antibiotics is recommended once a presumed diagnosis of sepsis or shock has been made by the treating clinician [5].

Although the feasibility of a one hour target for initiating antibiotics has not been assessed, the rationale for choosing it is based upon observational studies that report poor outcomes with delayed (even beyond one hour), inadequately dosed, or inappropriate (ie, treatment with antibiotics to which the pathogen was later shown to be resistant in vitro) antimicrobial therapy [42-52].

●In a retrospective analysis of over 17,000 patient with sepsis and septic shock, delay in first antibiotic administration was associated with increased in-hospital mortality with a linear increase in the risk of mortality for each hour delay in antibiotic administration [50]. Similar results were reported in an emergency department cohort of 35,000 patients [52].

●A prospective cohort study of 2124 patients demonstrated that inappropriate antibiotic selection was surprisingly common (32 percent) [46]. Mortality was markedly increased in these patients compared with those who had received appropriate antibiotics (34 versus 18 percent).

Choosing a regimen — The choice of antimicrobials can be complex and should consider the patient's history (eg, recent antibiotics received, previous organisms), comorbidities (eg, diabetes, organ failures), immune defects (eg, human immune deficiency virus), clinical context (eg, community- or hospital-acquired), suspected site of infection, presence of invasive devices, Gram stain data, and local prevalence and resistance patterns [53-57]. The general principles and examples of potential empiric regimens are given in this section but antimicrobial choice should be tailored to each individual.

For most patients with sepsis without shock, we recommend empiric broad spectrum therapy with one or more antimicrobials to cover all likely pathogens. Coverage should be directed against both gram-positive and gram-negative bacteria and, if indicated, against fungi (eg, Candida) and occasionally viruses (eg, influenza, coronavirus disease 2019 [COVID-19]). Broad spectrum is defined as therapeutic agent(s) with sufficient activity to cover a range of gram-negative and gram-positive organisms (eg, carbapenem, piperacillin-tazobactam). In order to ensure treatment with an effective antibiotic, many patients with septic shock suspected to be due to gram-negative organisms may require initial therapy with two antimicrobials from two different classes (ie, combination therapy), although this practice depends upon the organisms that are considered likely pathogens and local antibiotic susceptibilities.

Empiric therapy for patients with sepsis should be directed at the most common organisms causing sepsis in specific patient populations. Among organisms isolated from patients with sepsis, the most common include Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, and Streptococcus pneumoniae, such that coverage of these organisms should be kept in mind when choosing an agent [58].

However, when the organism is unknown, the clinician should be mindful of other potential pathogens when risk factors are present and consider the following:

●Methicillin-resistant S. aureus – Methicillin-resistant S. aureus (MRSA) is a cause of sepsis not only in hospitalized patients but also in community dwelling individuals without recent hospitalization [59,60]. For these reasons, we suggest empiric intravenous vancomycin (adjusted for renal function) be added to empiric regimens, particularly in those with shock or those at risk for MRSA. Potential alternative agents to vancomycin (eg, daptomycin for nonpulmonary MRSA, linezolid) should be considered for patients with refractory or virulent MRSA or with a contraindication to vancomycin. (See "Methicillin-resistant Staphylococcus aureus (MRSA) in adults: Treatment of bacteremia" and "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Risk factors for MRSA'.)

In our practice, if Pseudomonas is an unlikely pathogen, we favor combining vancomycin with one of the following:

•A third generation (eg, ceftriaxone or cefotaxime) or fourth generation cephalosporin (cefepime), or

•A beta-lactam/beta-lactamase inhibitor (eg, piperacillin-tazobactam), or

•A carbapenem (eg, imipenem or meropenem)

●Pseudomonas – Alternatively, if Pseudomonas is a likely pathogen, we favor combining vancomycin with one to two of the following, depending on local antibiotic susceptibility patterns (see "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections"):

•Antipseudomonal cephalosporin (eg, ceftazidime, cefepime), or

•Antipseudomonal carbapenem (eg, imipenem, meropenem), or

•Antipseudomonal beta-lactam/beta-lactamase inhibitor (eg, piperacillin-tazobactam), or

•Fluoroquinolone with good anti-pseudomonal activity (eg, ciprofloxacin), or

•Aminoglycoside (eg, gentamicin, amikacin), or

•Monobactam (eg, aztreonam)

Choosing among the preceding options should be individualized. Cefepime and piperacillin-tazobactam are probably the most commonly prescribed antipseudomonal agents. Piperacillin-tazobactam has been hypothesized to cause acute kidney injury, and cefepime has been hypothesized to cause neurologic dysfunction. Use of these two agents were directly compared in a randomized trial of 2511 patients with suspected infection who needed empiric treatment with an antipseudomonal agent [61]. Rates of acute kidney injury or death by day 14 were similar between the two groups. In contrast, patients in the cefepime group experienced fewer days alive and free of delirium and coma (11.9 versus 12.2 days; odds ratio 0.79, 95% CI 0.65-0.95), although the difference, while significant, was small. Interpretation is limited since the study was an open-label trial with significant crossover between the groups and longer-term outcomes beyond 14 days are not known. In addition, the dosing regimen was lower than typical for piperacillin-tazobactam and cefepime was administered as an intravenous push, which may have increased the risk of neurotoxicity. Lastly, the role of other antibiotics, such as anaerobic coverage with metronidazole, was not examined in this study. Further details regarding the adverse effects of beta-lactam antibiotics are provided elsewhere. (See "Beta-lactam antibiotics: Mechanisms of action and resistance and adverse effects", section on 'Adverse effects'.)

●Non pseudomonal gram-negative organisms (eg, E. coli, K. pneumoniae) – In the past, gram-negative pathogens routinely were covered with two agents from different antibiotic classes. However, several clinical trials and two meta-analyses failed to demonstrate superior overall efficacy of combination therapy compared to monotherapy with a third generation cephalosporin or a carbapenem [46,62-66]. Furthermore, one meta-analysis found double coverage that included an aminoglycoside was associated with an increased incidence of adverse events (nephrotoxicity) [65,66]. For this reason, in patients with suspected gram-negative pathogens, we recommend use of a single agent with proven efficacy and the least possible toxicity, except in patients who are either neutropenic or whose sepsis is due to a known or suspected Pseudomonas infection, where combination therapy can be considered [64]. (See "Pseudomonas aeruginosa bacteremia and endocarditis" and "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections".)

●Invasive fungal infections – The routine administration of empirical antifungal therapy is not generally warranted in nonneutropenic critically ill patients. Invasive fungal infections occasionally complicate the course of critical illness, especially when the following risk factors are present: surgery, parenteral nutrition, prolonged antimicrobial treatment or hospitalization (especially in the intensive care unit [ICU]), chemotherapy, transplant, chronic liver or renal failure, diabetes, major abdominal surgery, vascular devices, septic shock or multisite colonization with Candida spp. However, studies do not support the routine use of empiric antifungals in this population:

•In a meta-analysis of 22 studies (most often comparing fluconazole to placebo, but also using ketoconazole, anidulafungin, caspofungin, micafungin, and amphotericin B), untargeted empiric antifungal therapy possibly reduced fungal colonization and the risk of invasive fungal infection but did not reduce all-cause mortality [67].

•In a study of critically ill patients ventilated at least five days, empiric antifungal treatment (mostly fluconazole) was not associated with a decreased risk of mortality or occurrence of invasive candidiasis [68].

•In a multicenter randomized trial (EMPIRICUS) of 260 nonneutropenic critically ill patients with Candida colonization (at multiple sites), multiple organ failure, and ICU-acquired sepsis, empiric treatment for 14 days with micafungin did not result in improved infection-free survival at 28 days but did decrease the rate of new fungal infection [69].

However, if Candida or Aspergillus is strongly suspected or if neutropenia is present, an echinocandin (for Candida) or voriconazole (for Aspergillus) are often appropriate. (See "Treatment and prevention of invasive aspergillosis" and "Management of candidemia and invasive candidiasis in adults".)

●Other – Other regimens should consider the inclusion of agents for specific organisms such as Legionella (macrolide or fluoroquinolone) or difficult to treat organisms (eg, Stenotrophomonas), or for specific conditions (eg, neutropenic bacteremia)

Dosing — Clinicians should pay attention to maximizing the dose in patients with sepsis and septic shock using a full "high-end" loading dose where possible. This strategy is based upon the known increased volume of distribution that can occur in patients with sepsis due to the administration of fluid [70-72] and that higher clinical success rates have been reported in patients with higher peak concentrations of antimicrobials [73-75].

Continuous infusions of antibiotics as compared with intermittent dosing regimens remains investigational at this time [76]. Continuous infusions of beta lactam antimicrobials have been shown to provide higher and sustained drug levels above the minimum inhibitory concentration compared with the peaks and trough of intermittent dosing [77-79]. In a double-blind randomized trial of 607 patients prescribed meropenem for sepsis, outcomes were compared when meropenem was administered as a continuous infusion or intermittent infusion [80]. The mode of administration did not affect the composite outcome of 28-day mortality and the emergence of pandrug-resistant bacteria (47 versus 49 percent). In addition, there also was no difference in the length of ICU or hospital stay, days alive and free from the ICU, or 90-day mortality. However, therapeutic monitoring was not performed and the prescription of other microbials was common, both of which may limit the ability to detect a difference between the groups. We believe that further study is needed before continuous infusions can be recommended, although some experts favor continuous infusions given study biases, lack of harm, and possibility of benefit. Further details on prolonged infusions of antimicrobials are provided separately. (See "Prolonged infusions of beta-lactam antibiotics".)

Location of admission — Whether patients should be admitted to an ICU or ward is unclear and likely varies with the individual presenting characteristics as well as available institutional services and policy, which also may vary from state to state and country to country. For example, patients with septic shock who require mechanical ventilation and vasopressors clearly require ICU admission while those without shock who quickly respond to fluid and antibiotics may be safely transferred to the floor. For those in between these extremes, close observation and a low threshold to admit to the ICU is prudent.

Use of a systematic approach to ICU admission has been studied. One study of 3037 critically ill French patients aged 75 years or older, randomized patients to hospitals that promoted a systematic approach to ICU admission (interventional group) or to hospitals that did not use this approach (usual care) [81]. Despite a doubling of the admission rate to the ICU and an increased risk of in-hospital death, there was no difference in mortality at six months after adjustment for age, illness severity, initial clinical diagnosis, seniority of the emergency department clinician, time of ICU admission, baseline functional status, living situation, and type of home support. However, several flaws including, higher severity of illness in the intervention group, lack of blinding, and a strategy that was underpowered to detect a mortality difference may have influenced these results. In addition, international differences in the care of patients with sepsis may also explain an opposing outcome reported by a United States cohort [82].

MONITOR RESPONSE — After fluids and empiric antibiotics have been administered, the therapeutic response should be assessed frequently. We suggest that clinical, hemodynamic, and laboratory parameters be followed as outlined in the sections below. In our experience, most patients respond within the first 6 to 24 hours to initial fluid therapy, however, resolution can be protracted and take days or weeks. The response mostly influences further fluid management but can also affect antimicrobial therapy and source control.

Monitoring catheters — For many patients, a central venous catheter (CVC) and an arterial catheter are placed, although they are not always necessary. For example, an arterial catheter may be inserted if blood pressure is labile, sphygmomanometer readings are unreliable, restoration of perfusion is expected to be protracted (especially when vasopressors are administered), or dynamic measures of fluid responsiveness are selected to follow the hemodynamic response. A CVC may be placed if the infusion of large volumes of fluids or vasopressors are anticipated, peripheral access is poor, or the central venous pressure (CVP) or the central venous oxyhemoglobin saturation (ScvO₂) are chosen as methods of monitoring the hemodynamic response. (See "Intra-arterial catheterization for invasive monitoring: Indications, insertion techniques, and interpretation" and "Novel tools for hemodynamic monitoring in critically ill patients with shock" and "Central venous access in adults: General principles".)

We believe that pulmonary artery catheters (PACs) should not be used in the routine management of patients with sepsis or septic shock since they have not been shown to improve outcome [83-85]. PACs can measure the pulmonary artery occlusion pressure (PAOP) and mixed venous oxyhemoglobin saturation (SvO₂). However, the PAOP has proven to be a poor predictor of fluid responsiveness in sepsis and the SvO₂ is similar to the ScvO₂, which can be obtained from a CVC [86,87]. (See "Pulmonary artery catheterization: Indications, contraindications, and complications in adults".)

Clinical — All patients should be followed clinically for improved mean arterial pressure (MAP), urine output, heart rate, respiratory rate, skin color, temperature, pulse oximetry, and mental status. Among these, a MAP ≥65 mmHg (MAP = [(2 x diastolic) + systolic]/3) (calculator 1), and urine output ≥0.5 mL/kg per hour are common targets used in clinical practice. They have not been compared to each other nor have they been proven to be superior to any other target or to clinical assessment. Data supporting the use of these clinical parameters are discussed above. (See 'Initial resuscitative therapy' above.)

Most clinicians target a MAP ≥65 mmHg based upon data from large randomized trials that demonstrated benefit when using this target MAP (see 'Initial resuscitative therapy' above). However, the ideal target for MAP, is unknown. Furthermore, data since then suggest that higher MAPs (eg, ≥70 mmHg) may be harmful, while targeting lower MAPs (eg, 60 to 65 mmHg) may be appropriate. Thus, a reasonable goal may be to individualize targets within a range (eg, 60 to 70 mmHg) rather than targeting one specific numeric goal. Further trials are pending that should help elucidate an optimal range for a target MAP for patients with hypotension from sepsis.

●One trial that randomized patients to a target MAP of 65 to 70 mmHg (low target MAP) or 80 to 85 mmHg (high target MAP) reported no mortality benefit to targeting a higher MAP [88]. Patients with a higher MAP had a greater incidence of atrial fibrillation (7 versus 3 percent), suggesting that targeting a MAP >80 mmHg is potentially harmful.

●A pilot randomized trial reported that among patients aged 75 years or older, a higher MAP target (75 to 80 mmHg) was associated with increased hospital mortality compared with a lower MAP target (60 to 65 mmHg; 60 versus 13 percent) [89]. The same group subsequently reported outcomes in a nonblinded randomized trial of 2600 volume-repleted patients with vasodilatory shock who were older than 65 years and 80 percent of whom had sepsis (the "65 trial") [90]. Patients in whom vasopressors were used to target a MAP 60 to 65 mmHg ("permissive hypotension"; median mean MAP 66.7 mmHg) were compared with patients who received "usual care" (MAP at the discretion of the treating clinician; median MAP 72.6 mmHg). Fluid balance, rates of corticosteroid use (roughly one-third), and urine output were similar among the groups. Although the 90 day mortality was no different, the point estimate favored the permissive hypotension treatment strategy (41 versus 44 percent; adjusted odds ratio 0.82, 95% CI 0.68-0.98) and patients in the permissive hypotension group received lower doses of vasopressors for shorter duration. Importantly, there were no differences in the rates of cognitive dysfunction, cardiac arrhythmias, or acute renal failure. Although the prespecified outcome of superiority was not met, this trial suggests that at minimum, permissive hypotension was not harmful and supports the use of lower than usual target MAPs. It also suggests that similar to observations in other trials, clinicians tend to "overshoot" the target when a target MAP is set. However, several flaws including bias due to lack of blinding and a higher than usual mortality in the usual care group may limit interpretation of this study.

●Two meta-analyses that did not include the "65 trial" above [90] reported increased mortality in patients in whom a higher MAP was targeted with vasopressor use for greater than six hours [91] and a greater risk of supraventricular cardiac arrhythmias [92]. Another meta-analysis found no difference in mortality or the need for renal replacement therapy when higher versus lower MAP targets were used [93].

Hemodynamic — Static or dynamic predictors of fluid responsiveness should be employed in order to determine further fluid management. Guidelines state a preference for dynamic measures [3] since they are more accurate than static measures (eg, CVP) at predicting fluid responsiveness. However whether the use of dynamic predictors improve clinically impactful outcomes such as mortality remains unproven.

●Static – Traditionally, in addition to MAP, the following static CVC measurements were used to determine adequate fluid management:

•CVP at a target of 8 to 12 mmHg

•ScvO₂ ≥70 percent (≥65 percent if sample is drawn off a PAC)

While one early trial of patients with septic shock reported a mortality benefit to these static parameters in a protocol-based therapy, trials published since then (ProCESS, ARISE, ProMISe) have reported no mortality benefit in association with their use [8-11]. (See 'Initial resuscitative therapy' above.)

●Dynamic – Respiratory changes in the vena caval diameter, radial artery pulse pressure, aortic blood flow peak velocity, left ventricular outflow tract velocity-time integral, and brachial artery blood flow velocity are considered dynamic measures of fluid responsiveness. There is increasing evidence that dynamic measures are more accurate predictors of fluid responsiveness than static measures, as long as the patients are in sinus rhythm and passively ventilated with a sufficient tidal volume. For actively breathing patients or those with irregular cardiac rhythms, an increase in the cardiac output in response to a passive leg-raising maneuver (measured by echocardiography, arterial pulse waveform analysis, or pulmonary artery catheterization) also predicts fluid responsiveness. Choosing among these is dependent upon availability and technical expertise, but a passive leg raising maneuver may be the most accurate and broadly available. Future studies that report improved outcomes (eg, mortality, ventilator free days) in association with their use are needed. Further details are provided separately. (See "Novel tools for hemodynamic monitoring in critically ill patients with shock".)

Laboratory

●Lactate clearance – Although the optimal frequency of measuring serum lactate is unknown, we follow serum lactate (eg, every six hours) in patients with sepsis until the lactate value has clearly fallen. While guidelines promote normalization of lactate [3], lactate-guided resuscitation has not been convincingly associated with improved outcomes.

The lactate clearance is defined by the equation [(initial lactate – lactate >2 hours later)/initial lactate] x 100. The lactate clearance and interval change in lactate over the first 12 hours of resuscitation has been evaluated as a potential marker for effective resuscitation [14,94-99]. One meta-analysis of five low quality trials reported that lactate–guided resuscitation resulted in a reduction in mortality compared with resuscitation without lactate [3]. Other meta-analyses reported modest mortality benefit when lactate clearance strategies were used compared with usual care or ScvO₂ normalization [97,98]. However, many of the trials included in these meta-analyses studied heterogeneous populations and used varying definitions of lactate clearance as well as additional variables that potentially affected the outcome.

In addition, lactate is a poor marker of tissue perfusion after the restoration of perfusion [100]. As a result, lactate values are generally unhelpful following restoration of perfusion, with one exception: a rising serum lactate level should prompt reevaluation of the adequacy of perfusion. (See "Venous blood gases and other alternatives to arterial blood gases".)

Devices that allow measurement of serum lactate levels at the bedside are now available and their use may increase the practicality and utility of serial monitoring of serum lactate levels [101-103].

●Routine laboratories – Follow up laboratory studies, in particular platelet count, serum chemistries, and liver function tests are often performed (eg, every six hours) until values have reached normal or baseline. Hyperchloremia should be avoided, but if it is occurs, switching to low chloride-containing (ie, buffered) solutions may be indicated. (See "Treatment of severe hypovolemia or hypovolemic shock in adults", section on 'Buffered crystalloid'.)

●Microbiology – Follow up indices of infection are also indicated, including complete blood count and additional cultures. Results should prompt alteration of antibiotic choice if a better and safer regimen can be substituted and/or investigations directed toward source control. (See 'Septic focus identification and source control' below.)

●Arterial blood gases – It is prudent to follow worsening or resolution of gas exchange abnormalities, as well as the severity and type of acidosis (eg, resolution of metabolic acidosis and development of hyperchloremic acidosis). Worsening gas exchange may be a clue to the presence of pulmonary edema from excessive fluid resuscitation and also help detect other complications including pneumothorax from central catheter placement, acute respiratory distress syndrome, or venous thromboembolism.

SEPTIC FOCUS IDENTIFICATION AND SOURCE CONTROL — In our experience, a focused history and physical examination is the most valuable method for source detection. Following initial investigations and empiric antimicrobial therapy, further efforts aimed at identifying and controlling the source(s) of infection should be performed in all patients with sepsis. In addition, for those who fail despite therapy or those who fail having initially responded to therapy, further investigations aimed at adequacy of the antimicrobial regimen or nosocomial super infection should be considered.

●Identification – Additional investigations targeted at the suspected source(s) should be considered in patients with sepsis as promptly as is feasible (eg, within the first 12 hours). This investigation may include imaging (eg, computed tomography, ultrasonography) and sample acquisition (cultures or further diagnostic samples [eg, bronchoalveolar lavage, aspirating fluid collections or joints]); performing additional investigations depends upon the risk, if an intervention is involved, and patient stability. If invasive Candida or Aspergillus infection is suspected, serologic assays for 1,3 beta-D-glucan, galactomannan, and anti-mannan antibodies, if available, may provide early evidence of these fungal infections. These assays are discussed separately. (See "Clinical manifestations and diagnosis of candidemia and invasive candidiasis in adults", section on 'Nonculture methods' and "Diagnosis of invasive aspergillosis", section on 'Galactomannan antigen detection' and "Diagnosis of invasive aspergillosis", section on 'Beta-D-glucan assay'.)

●Source control – Source control (ie, physical measures to eradicate a focus of infection and eliminate or treat microbial proliferation and infection) should be undertaken in timely manner when they feasible since undrained foci of infection may not respond to antibiotics alone (table 3). As examples, potentially infected vascular access devices should be removed (after other vascular access has been established). Other examples include removing other infected implantable devices/hardware, when feasible, abscess drainage (including thoracic empyema and joint), percutaneous nephrostomy, soft tissue debridement or amputation, colectomy (eg, for fulminant Clostridium difficile-associated colitis), and cholecystostomy.

The optimal timing of source control is unknown but guidelines suggest no more than 6 to 12 hours after diagnosis since survival is negatively impacted by inadequate source control [3]. Although the general rule of thumb is that source control should occur as soon as possible [104-106], this is not always practical or feasible. In addition, decisions about the type and timing of source control should take into consideration the risk of a specific intervention and its potential risk of complications (eg, death, fistula formation) and the likelihood of success, particularly when there is diagnostic uncertainty regarding the source.

PATIENTS WHO FAIL INITIAL THERAPY — Patients having persistent hypoperfusion despite adequate fluid resuscitation and antimicrobial treatment should be reassessed for fluid responsiveness (see 'Hemodynamic' above) adequacy of the antimicrobial regimen and septic focus control (see 'Septic focus identification and source control' above) as well as the accuracy of the diagnosis of sepsis and/or its source and the possibility that unexpected complications or coexisting problems have occurred (eg, pneumothorax following central venous catheter insertion) (see "Evaluation of and initial approach to the adult patient with undifferentiated hypotension and shock"). Other options for treatment of persistent hypoperfusion such as the use of vasopressors, glucocorticoids, inotropic therapy, and blood transfusion are discussed in this section.

Vasopressors — Intravenous vasopressors are useful in patients who remain hypotensive despite adequate fluid resuscitation or who develop cardiogenic pulmonary edema. Based upon meta-analyses of small randomized trials and observational studies, a paradigm shift in practice has occurred such that most experts prefer to avoid dopamine in this population and favor norepinephrine as the first-choice agent (table 4 and table 5). Although guidelines suggest additional agents including vasopressin (up to 0.03 units/minute to reduce the dose of norepinephrine) or epinephrine (for refractory hypotension), practice varies considerably. Guidelines state a preference for central venous and arterial access especially when vasopressor administration is prolonged or high dose, or multiple vasopressors are administered through the same catheter [3]; while this is appropriate, waiting for placement should not delay their administration via a peripheral catheter.

●First agent – Data that support norepinephrine as the first-line single agent in septic shock are derived from numerous trials that compared the use of one vasopressor to another [107-113]. These trials studied norepinephrine versus phenylephrine [114], norepinephrine versus vasopressin [115-118], norepinephrine versus terlipressin [119-121], norepinephrine versus epinephrine [122], and vasopressin versus terlipressin [123]. While some of the comparisons found no convincing difference in mortality, length of stay in the intensive care unit or hospital, or incidence of kidney failure [118,124], two meta-analyses reported increased mortality among patients who received dopamine during septic shock compared with those who received norepinephrine (53 to 54 versus 48 to 49 percent) [110,125]. Although the causes of death in the two groups were not directly compared, both meta-analyses identified arrhythmic events about twice as often with dopamine than with norepinephrine.

However, we believe the initial choice of vasopressor in patients with sepsis is often individualized and determined by additional factors including the presence of coexistent conditions contributing shock (eg, heart failure), arrhythmias, organ ischemia, or agent availability. For example, in patients with significant tachycardia (eg, fast atrial fibrillation, sinus tachycardia >160/minute), agents that completely lack beta adrenergic effects (eg, vasopressin) may be preferred if it is believed that worsening tachycardia may prompt further decompensation. Similarly, dopamine (DA) may be acceptable in those with significant bradycardia; but low dose DA should not be used for the purposes of "renal protection."

The impact of agent availability was highlighted by one study of nearly 28,000 patients from 26 hospitals, which reported that during periods of norepinephrine shortages, phenylephrine was the most frequent alternative agent chosen by intensivists (use rose from 36 to 54 percent) [126]. During the same period, mortality rates from septic shock rose from 36 to 40 percent. Whether this was directly related to phenylephrine use remains unknown.

●Additional agents – The addition of a second or third agent to norepinephrine may be required (eg, epinephrine, dobutamine, or vasopressin).

•For patients with distributive shock from sepsis, vasopressin may be added. In a meta-analysis of 23 trials, the addition of vasopressin to catecholamine vasopressors (eg, epinephrine, norepinephrine) resulted in a lower rate of atrial fibrillation (relative risk 0.77, 95% CI 0.67-0.88) [127]. However, when including only studies at low risk of bias, no mortality benefit, reduced requirement for renal replacement therapy, or rate of myocardial injury, stroke, ventricular arrhythmias or length of hospital stay was reported. Although not studied, this effect is likely due to a reduced need for catecholamines which increase the risk of cardiac arrhythmias. This analysis is consistent with other trials and meta-analyses that have demonstrated no mortality benefit from vasopressin and selepressin in patients with septic shock [128-132].

•For patients with refractory septic shock associated with a low cardiac output, addition of an inotropic agent may be useful. In a retrospective series of 234 patients with septic shock, among several vasopressor agents added to norepinephrine (dobutamine, dopamine, phenylephrine, vasopressin), inotropic support with dobutamine was associated with a survival advantage (epinephrine was not studied) [133]. (See "Use of vasopressors and inotropes", section on 'Epinephrine' and "Use of vasopressors and inotropes", section on 'Dobutamine'.)

Whether lower vasopressor use is associated with lower mortality is unclear [134].

Additional information regarding vasopressor use including angiotensin II is provided separately. (See "Use of vasopressors and inotropes".)

Additional therapies — Most clinicians agree that additional therapies such as glucocorticoids, inotropic agents, or red blood cell (RBC) transfusion are not routinely warranted in patients with sepsis or septic shock but the use of these therapies maybe useful in refractory cases of septic shock or in special circumstances.

Glucocorticoids — Guidelines recommend against the routine use of glucocorticoids in patients with sepsis. However, corticosteroid therapy may be appropriate in patients with septic shock that is refractory to adequate fluid resuscitation and vasopressor administration. This topic is discussed in detail separately. (See "Glucocorticoid therapy in septic shock in adults".)

Inotropic therapy — A trial of inotropic therapy may be warranted in patients who fail to respond to adequate fluids and vasopressors, particularly those who also have diminished cardiac output (table 5) [8,135-137]. Inotropic therapy should not be used to increase the cardiac index to supranormal levels [138]. Dobutamine is a suitable first-choice agent; epinephrine is a suitable alternative. (See "Use of vasopressors and inotropes", section on 'Dobutamine'.)

Red blood cell transfusions — Based upon clinical experience, randomized studies, and guidelines on transfusion of blood products in critically ill patients, we typically reserve red blood cell transfusion for patients with a hemoglobin level ≤7 g/dL. Exceptions include suspicion of concurrent hemorrhagic shock or active myocardial ischemia.

Support for a restrictive transfusion strategy (goal hemoglobin >7 g/dL) is derived from direct and indirect evidence from randomized studies of patients with septic shock:

●One multicenter randomized study of 998 patients with septic shock reported no difference in 28-day mortality between patients who were transfused when the hemoglobin was ≤7 g/dL (restrictive strategy) and patients who were transfused when the hemoglobin was ≤9 g/dL (liberal strategy) [139]. The restrictive strategy resulted in 50 percent fewer red blood cell transfusions (1545 versus 3088 transfusions) and did not have any adverse effect on the rate of ischemic events (7 versus 8 percent).

●One randomized trial initially reported a mortality benefit from a protocol that included transfusing patients to a hematocrit goal >30 (hemoglobin level 10 g/dL) [8]. However, similarly designed studies published since then reported no benefit to this strategy [9-11]. These studies are discussed below.

In further support of a restrictive approach to transfusion in patients with septic shock is the consensus among experts that transfusing to a goal of >7 g/dL is also preferred in critically ill patients without sepsis [140-142], the details of which are provided separately. (See "Use of blood products in the critically ill", section on 'Red blood cells'.)

PATIENTS WHO RESPOND TO THERAPY — Once patients have demonstrated a response to therapy, attention should be directed towards continuing to control the septic focus, and de-escalation of fluids and antibiotics, as appropriate. This may occur within hours or days, depending upon the indicators of response and the individual patient. (See 'Clinical' above and 'Hemodynamic' above and 'Laboratory' above.)

Identification and control of the septic focus — Further efforts aimed at identifying and controlling the source of infection should be done if the initial evaluation and investigations fail to identify a source. (See 'Septic focus identification and source control' above.)

De-escalation fluids — In patients who respond to initial fluid therapy (ie, clinical hemodynamic and laboratory targets are met; usually hours to one to two days), we reduce the rate of or stop fluids, wean vasopressor support, and, if necessary, administer diuretics. While early fluid therapy is appropriate in sepsis, fluids may be unhelpful or harmful when the circulation is no longer fluid responsive. Careful and frequent monitoring is essential because patients with sepsis may develop cardiogenic and noncardiogenic pulmonary edema (ie, acute respiratory distress syndrome [ARDS]).

Small retrospective studies have reported that fluid overload is common in patients with sepsis and is associated with the increased performance of medical interventions (eg, diuresis, thoracentesis); the effect of fluid overload and such interventions on mortality and functional recovery in sepsis is unclear [143-145]. Data that support fluid restriction in this population include the following:

●In patients with ARDS or sepsis, a restrictive approach to intravenous fluid administration has been shown to decrease the duration of mechanical ventilation and ICU stay, compared with a more liberal approach [146-148].

●Another trial has shown no difference in mortality when fluid restriction was compared with a more liberal approach [148]. In this trial of 1554 patients with sepsis who had received at least 1 liter of fluid and were within 12 hours of the onset of shock, patients were randomized to receive either restricted intravenous fluid (ie, infusion stopped, small boluses given when needed for organ perfusion, low urine output, or insensible losses) or standard intravenous fluid therapy. There was no difference in the 90-day mortality or adverse effects. Cumulatively, at 90 days, patients in the restrictive group received approximately 2 liters less of fluids than patients in the standard group. While these data are encouraging and support safety of a restrictive approach to fluid de-escalation, both groups had received a median of 3 liters of fluid before randomization, the intervention was not blinded, and there were more violations in the restrictive group (22 versus 13 percent). Interestingly, the standard fluids group received less fluid than patients in some of the original sepsis trials [8], suggesting that practice has swung in favor of implementing a fluid-restricted approach to sepsis resuscitation. More studies are needed to guide de-escalation in patients with sepsis. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults", section on 'Conservative fluid management'.)

De-escalation and duration of antibiotics — It is appropriate that de-escalation and duration of antimicrobial agents be assessed daily [149,150]. When uncertain, it is also appropriate to obtain an infectious diseases consultation to facilitate good antimicrobial stewardship.

●De-escalation – After culture and susceptibility results return and/or after patients clinically improve, we recommend that antimicrobial therapy be narrowed (typically a few days). When possible, antimicrobial therapy should also be pathogen and susceptibility directed (also known as targeted/definitive therapy). However, since no pathogen is identified in approximately 50 percent of patients, de-escalation of empiric therapy requires a component of clinical judgement. For example, vancomycin typically is discontinued if no methicillin-resistant Staphylococcus is cultured.

While there is no consensus on de-escalation criteria, most experts use follow-up clinical (improved vital signs), laboratory and imaging data, and a fixed course of broad-spectrum therapy (eg, three to five days).

There are no high quality trials testing safety of de-escalation of antibiotic therapy in adult patients with sepsis or septic shock [151-155]. However, most observational trials report equivalent or improved outcomes with these fixed strategies of de-escalation.

●Duration – The duration of antibiotics should be individualized. For most patients, the duration of therapy is typically three to eight days [156-159]. However, longer courses are appropriate in patients who have a slow clinical response, an undrainable focus of infection, bacteremia with S. aureus, some fungal (eg, deep Candida infections) or viral infections (eg, herpes or cytomegalovirus), endocarditis, osteomyelitis, large abscesses, highly resistant gram-negative pathogens with marginal or limited sensitivities, neutropenia, or immunologic deficiencies [160-165]. Similarly, shorter courses may be acceptable in patients with negative cultures and rapid resolution of sepsis and laboratory studies. In patients who are neutropenic, antibiotic treatment should continue until the neutropenia has resolved or the planned antibiotic course is complete, whichever is longer. In nonneutropenic patients in whom infection is thoroughly excluded, antibiotics should be discontinued as early as is feasible to minimize colonization or infection with drug-resistant microorganisms and superinfection with other pathogens. Occasionally, shorter courses may be appropriate (eg, patients with pyelonephritis, urinary sepsis, or peritonitis who have rapid resolution of source control) [166-169].

●Role of procalcitonin – Although many institutions and guidelines support the use of procalcitonin to limit antibiotic (empiric or therapeutic) use in critically ill patients with suspected infection or documented infection, the evidence to support this practice is limited. While one randomized open-label trial of critically ill patients with infection reported a mortality benefit when the duration of antibiotic use was guided by normalization of procalcitonin levels [170], several randomized trials and meta-analyses found that using procalcitonin-guided algorithms to guide antimicrobial de-escalation did not result in any mortality benefit [171-177]. However, most trials report a reduction in the duration of antibiotic therapy (on average one day). One retrospective analysis suggested that use of procalcitonin was associated with lower hospital and ICU length of stay, but no clinically meaningful outcomes were measured in this study [178]. Other studies suggest that procalcitonin may distinguish infectious from noninfectious conditions and may therefore facilitate the decision to de-escalate empiric therapy [171,179-181]. However, procalcitonin's greatest utility is in guiding antibiotic discontinuation in patients with known community-acquired pneumonia and acute bronchitis; thus measuring procalcitonin in these populations is appropriate. (See "Procalcitonin use in lower respiratory tract infections".)

SUPPORTIVE THERAPIES — Details regarding supportive therapies needed for the care of critically ill patients, including those with sepsis are provided separately:

●Blood product infusion (see "Use of blood products in the critically ill")

●Nutrition (see "Nutrition support in intubated critically ill adult patients: Initial evaluation and prescription")

●Stress ulcer prophylaxis (see "Stress ulcers in the intensive care unit: Diagnosis, management, and prevention")

●Neuromuscular blocking agents (see "Neuromuscular blocking agents in critically ill patients: Use, agent selection, administration, and adverse effects")

●Venous thromboembolism prophylaxis (see "Prevention of venous thromboembolic disease in acutely ill hospitalized medical adults")

●Intensive insulin therapy (see "Glycemic control in critically ill adult and pediatric patients")

●External cooling or antipyretics (see "Fever in the intensive care unit", section on 'Management')

●Mechanical ventilation, sedation, weaning (see "Acute respiratory distress syndrome: Ventilator management strategies for adults" and "Sedative-analgesia in ventilated adults: Management strategies, agent selection, monitoring, and withdrawal" and "Initial weaning strategy in mechanically ventilated adults")

●Investigational therapies for sepsis and acute respiratory distress syndrome (eg, intravenous immune globulin, antithrombin, thrombomodulin, heparin, cytokine and toxin inactivators, as well as hemofiltration, statins, beta-2 agonists, beta blockade, and vitamin C) (see "Investigational and ineffective pharmacologic therapies for sepsis" and "Acute respiratory distress syndrome: Investigational or ineffective therapies in adults")

PREGNANCY — The optimal way to manage sepsis in pregnancy is unknown but most experts use the same principles as outlined in this topic being cognizant of the altered hemodynamics of pregnancy. Guidelines have been proposed but have not been validated [182]. Further details regarding the management of critically ill pregnant patients are provided separately. (See "Critical illness during pregnancy and the peripartum period".)

POSTSEPSIS CARE — For survivors of sepsis, attention should be paid to follow-up care and the recognition of post-intensive care syndrome (PICS). Further details regarding PICS and prognosis of patients with sepsis are provided separately. (See "Post-intensive care syndrome (PICS) in adults: Clinical features and diagnostic evaluation" and "Sepsis syndromes in adults: Epidemiology, definitions, clinical presentation, diagnosis, and prognosis", section on 'Prognosis'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sepsis in children and adults".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Sepsis in adults (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Initial evaluation – For patients with sepsis and septic shock, therapeutic priorities include securing the airway, correcting hypoxemia, and establishing appropriate vascular access for the early administration of fluids and antibiotics (table 1). Simultaneously obtaining the following is preferable (within 45 minutes) but should not delay the administration of fluids and antibiotics (see 'Immediate evaluation and management' above):

•Routine laboratory studies

•Serum lactate

•Arterial blood gases

•Blood cultures (aerobic and anaerobic) from two distinct venipuncture sites and from all indwelling vascular access devices; it is preferable that blood cultures be drawn before the initiation of antibiotics

•Cultures from easily accessible sites (eg, sputum, urine)

•Imaging of suspected sources

●Initial resuscitation – For patients with sepsis and septic shock, we suggest the infusion of intravenous fluids (30 mL/kg), commencing within the first hour and completed within the first three hours of presentation, rather than vasopressors, inotropes, or red blood cell transfusions (Grade 2B). (See 'Initial resuscitative therapy' above.)

•Intravenous fluids – Fluid boluses are the preferred method of administration and should be repeated until blood pressure and tissue perfusion are acceptable, pulmonary edema ensues, or there is no further response. (See 'Intravenous fluids (first three hours)' above.)

Crystalloid solutions (eg, normal saline or Ringer's lactate) are our preferred resuscitation fluid. Balanced crystalloid may be preferred if there is a perceived need to avoid or treat the hyperchloremia that occurs when large volumes of nonbuffered crystalloid (eg, normal saline) are administered. We recommend that a hyperoncotic starch solution not be administered (Grade 1A). (See 'Choice of fluid' above.)

•Antibiotics – For patients with sepsis, we recommend that optimal doses of empiric broad spectrum intravenous therapy with one or more antimicrobials be administered in a prompt fashion (eg, within one hour) of presentation (Grade 1B). Broad spectrum is defined as therapeutic agent(s) with sufficient activity to cover a broad range of gram-negative and positive organisms, and, if suspected, against fungi and viruses. (See 'Empiric antibiotic therapy (first hour)' above and 'Initial resuscitative therapy' above.)

For patients with septic shock associated with likely gram-negative sepsis, we suggest consideration of the use of two antibiotics from different classes to ensure effective treatment of resistant organisms.

Agent selection depends upon patient's history, comorbidities, immune defects, clinical context, suspected site of infection, presence of invasive devices, Gram stain data, and local prevalence and resistance patterns. The routine administration of antifungal therapy is not warranted in nonneutropenic patients.

●Monitoring – For most patients with sepsis and septic shock, we recommend that fluid management be guided using clinical targets including mean arterial pressure 60 to 70 mmHg (calculator 1) and urine output ≥0.5 mL/kg/hour (Grade 1B). (See 'Monitor response' above and 'Clinical' above.)

•Hemodynamics – In addition, while dynamic measures of fluid responsiveness (eg, respiratory changes in the radial artery pulse pressure) are preferred, static measures of determining adequacy of fluid administration (eg, central venous pressure 8 to 12 mmHg or central venous oxygen saturation ≥70 percent) may be more readily available. (See 'Monitoring catheters' above and 'Hemodynamic' above.)

•Laboratory – Serum lactate should be followed (eg, every six hours) until there is a definitive clinical response. It is prudent that other measures of the overall response to infection also be followed (eg, routine laboratory studies, arterial blood gases, microbiology studies). (See 'Laboratory' above.)

•Source control – Following initial investigations and empiric antimicrobial therapy, further efforts aimed at identifying and controlling the source(s) of infection (ideally within 6 to 12 hours) should be performed in all patients with sepsis (table 3 and table 2). In addition, for those who fail despite therapy or those who fail having initially responded to therapy, further investigations aimed at removal of devices suspected to be infected, adequacy of the antimicrobial regimen, or nosocomial super infection should be considered. (See 'Septic focus identification and source control' above.)

●Patients who fail initial therapy – For patients with sepsis who remain hypotensive despite adequate fluid resuscitation (eg, 3 L in first three hours), we recommend vasopressors (Grade 1B); the preferred initial agent is norepinephrine (table 5). For patients who are refractory to intravenous fluid and vasopressor therapy, additional therapies, such as glucocorticoids, inotropic therapy, and blood transfusions, can be administered on an individual basis. We typically reserve red blood cell transfusion for patients with a hemoglobin level <7 g/dL. (See 'Additional therapies' above and "Use of vasopressors and inotropes", section on 'Choice of agent in septic shock'.)

●Patients who respond to therapy – For patients with sepsis who have demonstrated a response to therapy, we suggest that the rate of fluid administration should be reduced or stopped, vasopressor support weaned, and, if necessary, diuretics administered. We also recommend that antimicrobial therapy be narrowed once pathogen identification and susceptibility data return. Antimicrobial therapy should be pathogen and susceptibility directed for a total duration of 7 to 10 days, although shorter or longer courses are appropriate for select patients. (See 'Patients who respond to therapy' above.)