Severe Sepsis and Septic Shock
Cryptic Septic Shock
1. Description of the problem
Sepsis is a clinical syndrome that occurs in patients with infection (known or suspected) defined by the presence of two or more systemic inflammatory response syndrome (SIRS) criteria, which include the following:
Fever or hypothermia (temperature [T] >100.4°F or <96.8°F)
Resting respiration (RR) >20
Heart rate (HR) >90
White blood cell count (WBC) >12, <4, or 10% bands
Hyperglycemia (in nondiabetic patient)
Altered mental status
Severe sepsis is a clinical syndrome of sepsis that is associated with at least one new organ dysfunction:
Mean arterial pressure (MAP) <60 mmHg, systolic blood pressure (SBP) <90 mmHg
Lactic acidosis (2.0–4.0) or poor extremity perfusion (cool extremities, livedo reticularis)
Increased creatinine (>2.0 or increase by 50%) or oliguria
Hypoxia or high FiO2 requirement
Change in mental status, delirium
Abnormal prothrombin time/partial thromboplastin time (PT/PTT) or low platelets (<100,000)
Hyperbilirubinemia (>2.0) or liver function tests (LFTs) greater than twice the upper limit of normal
Septic shock is a subcategory of severe sepsis with severe cardiovascular dysfunction as defined by hypotension (SBP <90 mmHg or 40 mm Hg less than baseline) with evidence of hypoperfusion, despite an intravenous (IV) fluid challenge of at least 20 mL/kg.
Cryptic septic shock is a relatively new term defined as sepsis with severe lactic acidosis (4.0 mM/L or greater) despite a normal or high blood pressure.
The clinical manifestations of sepsis may include the local symptoms and signs of the inciting infection, as well as systemic symptoms and systemic signs that are manifestations of the body’s response to the infection. Although the clinical features of the localized infection are relatively specific to the anatomic site, the systemic response is not.
The clinical manifestations of the systemic response (SIRS), include both vital sign (fever, tachycardia and tachypnea) and laboratory (leukocytosis or leukopenia) abnormalities, as described above in the previous section.
These signs are sensitive but non-specific. In addition, in certain populations (immunosuppressed, end-stage liver or kidney disease), fever — the hallmark of infection — is commonly absent.
Sepsis is defined as severe when it is associated with hypoperfusion, organ dysfunction, or hypotension. The magnitude of the sepsis severity and related prognoses depend on the nature, extent, and duration of the organ dysfunction. The greater the number and duration of the organ dysfunction, the higher the likelihood of mortality. The most important characteristic that portends a poor prognosis is if organ dysfunction worsens or even fails to improve, despite appropriate interventions. Criteria for organ dysfunction have been specifically defined and are listed in a prior section.
Although many patients with severe sepsis have evidence for hypovolemia and hypoperfusion, patients may meet organ dysfunction criteria for severe sepsis without these findings, including those with hepatic, hematologic, gastrointestinal, pulmonary, or central nervous system dysfunction.
In contrast, septic patients who present with hypotension, lactic acidosis, acute kidney injury, oliguria, livedo reticularis, or cool extremities are manifesting signs of hypoperfusion. This distinction of whether there are signs of hypoperfusion impacts triage and therapeutic decisions. Those without hypoperfusion may not require aggressive fluid resuscitation, and many do not require intensive care. Other indications for ICU transfer, however, include respiratory failure and central nervous system depression.
Cryptic septic shock patients usually have clinical signs of shock on examination, accompanied by severe lactic acidosis (4.0 mM/L or greater), despite a normal or high systemic blood arterial pressure. This syndrome is important to recognize, as the prognosis and treatment is similar to those with septic shock.
Thus, in order to appropriately risk stratify and appropriately manage these patients, ALL PATIENTS with sepsis or severe sepsis should have a serum lactate measured, since an elevated value may change the prognosis and management required.
Key management points
The key management point is to promptly recognize and intervene with these patients, as the treatments are both life saving and time sensitive. Thus, one must always have a high index of suspicion for sepsis as the underlying etiology of any clinical deterioration of a hospitalized patient.
Nuances about recognition
Although patients with severe sepsis often present with organ dysfunction after or simultaneously with manifestations of their underlying infection, it is not uncommon for patients to present with an organ dysfunction as their first manifestation of severe sepsis.
For example, new onset of delirium, hyperglycemia, or oliguria in a hospitalized patient, with or without two or more SIRS criteria, may be the initial manifestation of severe sepsis, prior to the appearance of the culprit infection. This may be more common in the hospitalized setting while patients are being cared for their initial clinical problem and may not have the cognitive capacity — or a family member — to speak up on their behalf to provide the new salient historical information.
In addition, if the underlying infection is endovascular, which is more common in the hospitalized setting, the infection is typically asymptomatic, often without localizing signs such that the organ dysfunction is often the presenting and sole manifestation. Similarly, immunosuppressive medications may mask the signs of the underlying infection. Thus, clinicians need to consider severe sepsis as the diagnosis in any hospitalized patient who deteriorates clinically with any constellation of signs/symptoms/laboratory abnormalities, where the cause is not immediately apparent.
In addition, presumptive treatment for severe sepsis (intravenous fluids [IVF], antibiotics [ABx], etc) should also be initiated, unless a reasonable alternative etiology for the findings is found.
Principles about ICU Transfer
Many patients with sepsis and some subsets of severe sepsis may be managed safely on the general care floor. However, ICU transfer should be strongly considered for all patients with cardiopulmonary dysfunction. The inidications for ICU transfer depend on the patient’s comorbidities, magnitude and trajectory of the cardiopulmonary dysfunction, and goals of care. However, consensus guidelines recommend that all patients with septic shock, and those with cryptic septic shock on presentation (lactate 4.0 mM/L or greater), be triaged to the ICU. In addition, ICU transfer should also be considered for most patients with moderate to severe respiratory failure and central nervous system depression.
Septic shock is classified as a distributive shock; however, the vast majority have moderate to severe hypovolemia and 50-60% have an associated septic cardiomyopathy. Thus, the typical distributive shock phenotype — with warm and well perfused extremities — is often not manifested at presentation as these findings are only seen when there is a high cardiac output. This is characteristic of the well resuscitated septic shock state. In contrast, the hypovolemia, venodilation, and cardiomyopathy act in concert to reduce cardiac output, explaining why these patients often present with “cold shock” phenotypically similar to hypovolemic shock.
Although microcirculatory shunting and mitochondrial dysfunction are described in patients with severe sepsis, both of which would tend to diminish tissue oxygen extraction, in most patients (85-90%) oxygen extraction is elevated initially. This suggests that in the vast majority of patients, tissues oxygen extraction is preserved despite these derangements. However, in approximately 10% of patients, the micro circulatory and/or mitochondrial dysfunction overwhelms the low flow state and patients present with a normal to very high venous oxygen saturation. Not surprisingly, this phenotype has a much worse prognosis. What distinguishes these patients — with normal to high ScVo2 — from the euvolemic or well resuscitated septic shock patient, is other concomitant findings of poor perfusion such as lactic acidosis, persistent, oliguria, low BP and a narrow PP, tachycardia, and skin/extremity findings of low flow. The warm, well-perfused, septic state with a low oxygen extraction ratio (high ScvO2) typically manifests when the patient has been well resuscitated and demonstrates good perfusion systemically.
When resuscitating patients with septic shock, the goal is to restore BOTH the MAP and systemic perfusion. The latter is important because ultimately, restoration of organ perfusion is key to survival.
The MAP goal is prioritized to ensure adequate and immediate cerebral and coronary perfusion; however, a normal or even elevated MAP does not ensure tissue perfusion. The goal should be to a MAP ≥65, or 75% of the patients MAP when recently well. This should be accomplished with IV crystalloid (LR) initially but after the initial IVF challenge (~2 L) concomitant vasopressor use is encouraged. This can be titrated down when additional volume is infused. A recent RCT, called ALBIOS, showed no benefit to resuscitating to a higher MAP target than 65, except for those patients with chronic HTN, wherein a higher MAP goal was associated with was less AKI and need for CRRT.
Secondarily, one should strive to normalize peripheral perfusion to help restore organ function. This can be accomplished by serial IVF boluses following conventional bedside parameters; however, if a CVC is in place, using ScvO2 is very helpful. The recent ProCESS trial showed equivalent outcomes with or without ScvO2, but this study was conducted in academic EDs with attending critical care specialists at the bedside. Thus, in other clinical venues it is still reasonable to use the ScvO2 to guide the resuscitation.
Although the etiology for organ dysfunction in sepsis is multifactorial, including many different physiochemical pathways, tissue hypoperfusion is a key component. Regrettably, since there are currently no effective therapeutic interventions to rectify the other pathobiological derangements, aside from controlling the infection with early ABx and source control, the importance of early and rapid restoration of global oxygen delivery is tantamount. Fortunately, in many patients, organ dysfunction resolves coincident with attaining these global perfusion goals.
Importance of clinical documentation
Although not clinically relevant for individual patient care, correctly documenting severe sepsis in the medical record and correctly classifying severity is important.
Health care institutions benefit greatly if clinicians document this diagnosis accurately because this may increase substantially hospital reimbursement and more accurately portray the acuity of the patient, which directly affects performance metrics (e.g., observed to expected mortality rates). This is an important quality measure routinely used by external rating agencies (and internally) as a benchmark by which to evaluate and compare the performance of health care institutions. In the future, it is also possible that this metric could be used to assess individual physician quality/performance.
2. Emergency Management
Assess and Manage ABCs
If concerned about airway protection, intubate immediately. If the patient is in respiratory distress or respiratory failure, you may consider non-invasive positive pressure ventilation (NIPPV) but only if patient is hemodynamically stable (not in shock), is co-operative and has minimal secretions.
In patients with severe sepsis or septic shock with increased work of breathing, one should have a low threshold to intubate and mechanically ventilate to reduce or eliminate the requirement for increased blood flow to respiratory muscles and decrease total body oxygen consumption.
If there is no respiratory distress and ventilation is adequate based on ABGs, provide O2 to maintain oxygen saturation greater than >92-94%. If evidence for shock (by BP/HR, skin/extremity perfusion, lactate, urine output) place two peripheral 18 gauge IVs, or one if a CVC is already in place (PICC, PORT, Hickman etc are all fine to use) for IVF resuscitation.
Give IVF boluses (lactated ringers) in 500-1000 mL over 10 to 20 min (with a pressure bag) q 15-30 min prn until resuscitation goals are achieved, there is no evidence of fluid responsiveness, or side effects ensue. Lactated ringers is preferred, as recent observational studies suggest that high chloride solutions may be deleterious in terms of mortality and renal function. Animal studies reveal that high chloride solutions cause renal afferent vasoconstriction, which may be the underlying mechanism. Caution must be taken when LR is used in patients with hyperkalemia or cirrhosis.
If patient remains in shock (hypoperfusion), one should not continue to give IVF unless there is clear evidence of a response, as fluid overload may be associated with worse outcomes and organ function, not to mention the issues with fluid removal during recovery. If uncertain about the effectiveness of IVF boluses in this setting, we recommend objective measurements of stroke volume pre-post bolus as a guide. There are several non-invasive tools available for this purpose.
IVF bolus goals
MAP 65 mm Hg or greater (but not <25% of patient’s baseline MAP)
Eliminate signs of shock (HR, UO, mental status, skin perfusion)
If CVC is in place (for vasopressors):
Target an ScvO2 greater than 70% with IVF, and only give PRBC if HgB <7.0-7.5. Dobutamine should be considered if persistent shock and no further fluid responsiveness, although checking an echocardiogram at this point is recommended to rule out other causes for low cardiac output besides septic cardiomyopathy (tamponade, outflow tract obstruction, valvular heart disease, right heart failure)
Note: Following 6 hours of EGDT (in the original NEJM study) the mean ScvO2 was 77% +/10%; thus, if hypoperfusion persists, do not necessarily stop IVF boluses after reaching an ScvO2 target of ~70%.
Triage Patient to Appropriate Level of Care
If patient requires vasopressors for BP but there is no longer any evidence of hypoperfusion, additional IVF boluses may not be beneficial and could have adverse consequences (tissue edema, organ dysfunction, IAH). We favor vasoactive agents to achieve MAP target in the absence of hypoperfusion.
Patients with septic or cryptic septic shock require ICU admission for aggressive support and monitoring including potentially EGDT guided by a CVC (ScvO2). There are many patients with severe sepsis who can be managed on the floor, but most patients with either cardiopulmonary dysfunction or septic encephalopathy should go the ICU.
All others may be considered for floor management with heightened vigilance/monitoring for deterioration.
Send off labs (including lactate, and ScvO2 if CVC in place) and pan culture. Then administer broad spectrum antibiotics based on the presumed source (and host immune status), prior antibiotic use, and institutional antibiogram. Antibiotics should be administered within 1 hour of signs of severe sepsis, as mortality increases steadily as time passes
See “Specific Therapy” below for further management details.
It is vitally important to recognize sepsis promptly, particularly when severe, since delays will reduce the effectiveness of life-saving interventions. This may be challenging, particularly in the hospitalized patient, as the clinical presentation can vary widely and at times may be subtle, depending on the source of infection and host co-morbidities.
In sick patient populations, SIRS criteria have low diagnostic utility as they are overly sensitive and non-specific. In addition, in some particularly vulnerable populations, such as the elderly or those taking immunosuppressive medications, the clinical manifestation of both the infection and the SIRS response may be markedly attenuated despite overwhelming infection.
In these patients, among others, the signs of organ dysfunction (e.g., delirium or oliguria) may be the presenting manifestation and only clinical clue of underlying severe sepsis. This underscores the importance of maintaining a high degree of suspicion for the presence of sepsis as the underlying cause of any clinical deterioration in the critically ill patient.
Given the adverse impact that delays in recognition may have on the effectiveness of some interventions, it is prudent to initiate treatment presumptively for severe sepsis, unless or until an alternative diagnosis has been established.
Recognizing the inherent challenges in diagnosing sepsis, a consensus conference was reconvened in 2001 to review the conventional definitions of sepsis (see definition section). Although the definitions were not altered, the consensus panel added other clinical signs, associated commonly with sepsis, to the existing list of criteria that should raise the suspicion of sepsis. These include unexplained hyperglycemia, change in mental status, significant edema and a markedly positive fluid balance (>20 mL/kg over 24 h).
When a patient is suspected of having sepsis it is important to establish both the etiology of infection and assess the severity of sepsis. If a specific source of infection can be found it supports the diagnosis of sepsis over other systemic conditions, informs the appropriate choice of antibiotics, and expedites source control. The severity of sepsis dictates which interventions are indicated immediately and whether ICU transfer is indicated.
In addition to a focused history and physical examination, cultures and directed radiographic imaging should be obtained to localize the source of infection. If possible, all cultures should be obtained prior to antibiotic administration, to preserve their diagnostic sensitivity. However, even in the most acutely ill patients, pathogens will grow in blood cultures only in half the patients.
Sepsis severity is readily established by physical examination and routine laboratory studies of organ function. Although some of these patients may be cared for on the general ward, when there is cardiovascular, respiratory, or central system dysfunction, more intensive monitoring and other life-saving interventions may be are indicated.
Specifically, in patients with septic shock or cryptic septic shock (lactate ≥ 4 mM/L, with a normal or high blood pressure), early aggressive IVF resuscitation, immediate appropriate antibiotics, and prompt ICU transfer is warranted. Given the similarly high mortality of cryptic septic shock as compared to full blown septic shock, and the effectiveness of these interventions in mitigating the risks of death, a serum lactate should always be obtained promptly in all patients with suspected sepsis, regardless of whether there are any other signs of organ dysfunction.
Another reason for measuring lactate in these patients is that even marginally elevated values (i.e., ≥2 mmol/L), may be associated with increased morbidity and mortality, independent of other organ dysfunction or level of blood pressure. Thus, one may consider using a lower lactate threshold in the decision to move a patient to the ICU, particularly in frail or vulnerable patients (e.g., if significant immunosuppression).
A limited set of disorders can mimic the warm shock (low afterload) state characteristic of well-resuscitated septic shock; however, in those patients presenting in cold shock, which is more common, one must consider a broader differential including disorders associated with a low cardiac output and high afterload. This includes all the other categories of shock including hypovolemic, cardiogenic, and obstructive.
Although this includes a long list of disorders, the initial findings on history and physical examination (eg, clinical setting and absence of jugular venous distension and rales), usually allow one to readily exclude cardiogenic and obstructive shock from initial consideration. (See Table I)
4. Specific Treatment
Algorithm and principles
1 . ABs
– Consider NIPPV only if airway protected, minimal secretions, and not in shock.
– Have a low threshold to intubate patients with septic shock (ie, poor mental status, breathless, inappropriate PaCO2/O2) to improve hemodynamics but initiate volume repletion before induction (if at all possible, through 18 gauge IV) to avoid precipitating severe hypotension.
2. Cs—Circulation (Catheter)
-Catheter: Obtain adequate IV access:
-Need at least 1 peripheral IV catheter (18 gauge is fine).
-Circulation: -If hypotensive or evidence of hypoperfusion, give 500 mL to 1L fluid challenges q 15 to 30 minutes x 3 to 4.
-If lactate was 4 or greater before IVF or BP still low after IVF: place a CVC (unless one is already in place) for CVP/ScvO2 monitoring; and check results immediately and intervene.
1. General Resuscitation Goals
-Good extremity perfusion
-Also look for a greater than 10% reduction in lactate and consider the dose/side effects of vasopressors.
and prognosis from underlying co-morbidities from primary or co-attendings.
-Good urine output (non-oliguric)
-Of note, if the ScvO2 is high on presentation, before resuscitation, this suggests severe oxygen extraction failure from microcirculatory or mitochondrial derangements despite a low cardiac output.
-A high lactate and “cold shock” on examination, in this setting, is further evidence to support this pathophysiology. This atypical phenotype of septic shock patients (<10% of septic shock patients) has been described, and, as a group, has a higher mortality rate. Interestingly, if these patients respond to treatment, resuscitation results in both a decline in the ScvO2, lactate.
2 Specific resuscitation Goals
a. MAP ≥65 (or within 25% of baseline value, if baseline HTN)
b. CVP at least 8 to 12
c. ScvO2 ≥70% (Note: Mean ScvO2 after resuscitation in the Rivers N Engl J Med study ~ 85%).
3. How toResuscitate:
a.Equipment: Pressure bag/trauma tubing/ IVF warmer/PreSEP or CVC. Can use indwelling chronic line for CVP/ScvO2 (PICC, PORT) but then use peripheral line (18 gauge) for IVF boluses
b. IV Fluids: Only crystalloids. Of the choices, use Ringers lactate, unless liver failure or high K+ to avoidhyperchloremic metabolic acidosis, particularly in patients already acidotic, those with CKI, or incipient or frank respiratory failure. If liver failure or high K+, can use D5W with 3 Amps NaHCO3
c.IV Fluid Rate: 500 – 100 cc IVF bolus over 5-10 min; repeat q 15-30 minutes if improvement by vitals, exam, ScvO2, or decreasing pressor requirements and/or if ScvO2 < 70%. (Average needs ~ 5 L in 6 hrs, 6-8 L/24 hrs). Stop IVF boluses if signs of ↑ pulmonary edema (↑ hypoxia/frothy secretions/dyspnea)or ↑IAH (15->20).
d. EGDT Protocol:
After IVF boluses if CVP> 8-12, but ScvO2<70% continue IVF boluses if fluid responsive and no adverse affects until ScvO2 >70% and other goals achieved. If ScvO2 remains < 70% but patient is no longer fluid responsive or has adverse effects of IVF (increasing hypoxia/ IAH): Then: If HgB<10→ Give PRBC to goal ScVO2 or HgB ≥10. Or, If HgB ≥10 and Scv02 still 70% start dobutamine at 2.5 mcg/kg/min (but consider echo to r/o an alternative pathophysiology other than septic CM, that may not respond to inotropes e.g. outflow tract obstruction, inappropriate ventricular dilatation, tamponade or other unexpected pathophysiology).
e. Vasoactive Meds: After IVF, if CVP>8-12, but MAP<65:
Start Norepinephrine: If NE dose is > 15-20 mcg/min, also start Vasopressin at .03 Units/min and hydrocortisone at 50 mg q6 (see below under corticosteroid
Do not titrate down NE to <5 mcg/min if still on VP.
If SVT, or SVT develops, despite adequate IVF (CVP > 8-12), consider phenylephrine or VP alone rather than NE (It is also OK to use these strictly vasopressor agents if no evidence of septic cardiomyopathy by echo)
f. Corticosteroids: If SBP remains < 90 or MAP <65 despite significant (20mcg) or increasing NE doses for > 1 hour, start hydrocortisone at 50 mg q 6 hours (Rx until off pressors, then wean off over 3-5 days).
g. NaHCO3: Ineffective at ~ pH of 7.20, but if refractory severe acidosis is present i.e. pH<7.15, consider bicarbonate boluses or drip if patient can tolerate volume infusion (with goal pH greater than 7.15). If volume overload and AKI, also initiate CVVHD, if another reason for HD, or UF.
2. D’s – Diagnosis, Drug treatment, Drainage (source control)
Send all cultures immediately including 2 blood Cx (at least 1 peripheral, none from old lines), sputum, UA and UCx etc and give appropriate (Broad spectrum) antibiotics ASAP (< 1 hour from shock).
a. HOW TO CHOOSE ABx:
Choose ABX based on nature of host, suspected sources, prior ABx use, prior culture data, institutional antibiogram, and allergies. To simplify, and ensure timely administration this is best accomplished by therapeutic algorithms and use of rapid response teams:
If no prior BS antibiotic (in recent past) and no B lactam allergy:
– Cefepime, Tobramycin, Vancomycin.
**If intra-abdominal source – add metronidazole 500 mg IV and consider caspofungin 70 mg if upper GI tract perforation is suspected.
**If immunosuppressed with cytotoxic drugs and/or steroids for >3-4 weeks and a bacterial pathogen hasn’t been identified consider adding caspofungin 70 mg; use Voriconazole if nodular (or wedge shaped) pulmonary infiltrates.
If B-lactam allergy alone: Substitute, levofloxacin for cefepime
If B-lactam and fluoroquinolone allergies: Substitute aztreonam for cefepime
Prior antibiotic therapy (for > 3 days) during past 4- 8 weeks
– If no B-lactam allergy: Meropenem (1 gm), Amikacin and Vancomycin
– If B lactam allergy: Aztreonam, Tobramycin, and Vancomycin
b. HOW TO ADMINISTER ABx:
Order and rate of administration: Always first infuse drugs against gram negative pathogens: Cefepime (5 min), Amikacin (30 min), Aztreonam (15 min), Levofloxacin (90 min), Meropenem (15 min). Start Vanco (60 min for 1 gm), simultaneously. Then caspofungin (60 min).
c. Diagnostic tests: Send appropriate labs and tests: Lactate acid if not already done,ScvO2 (venous co-oximetry), ABG, Chem, LFT’s amylase lipase, CBC w/diff, coags, Ca, Phos, Mg, CPK, CXR, UA and UCx, Sputum Cx.
-If patient was an OSHtx, call OSH daily for all Micro updates.
d. Source control: Remove or drain focus of presumed infection ASAP (includes tunnelled lines-if in shock). May need additional imaging once stable for transfer. Surgery consult for abdominal or soft tissue sources.
3. E- Evaluate again to assess response to treatment:
1. Improved w/i 1-3 days, de-escalate Abx at day 3 based on cultures.
2. If no improvement in 48 hours, consider:
– Additional or altered Abx coverage:
e.g. Caspo or Vori for fungus, Linezolid or Daptomycin for VRE, Bactrim for nocardia/xanthomonas/PCP, tetracycline for rickettsial pathogens, TB drugs, Ganciclovir/Acyclovir for CMV, HSV, VZV)
– Look for an undrained focus with CT scan or US to r/o direct extension or hematogenous spread of primary focus.
-Consider an Alternative diagnosis for the syndrome (catastrophic APLS, HLH, carcinomatosis or lymphomatosis, CVD, Adrenal Insufficiency, drug reaction or OD, liver failure, non-infectious SIRS)
4. Other Important Treatments:
a. Insulin: Keep blood sugar levels 140-180, using insulin drip if necessary IV drip protocol). Verify all POC results if < 90, as you may miss occult hypoglycemia.
b. DVT prophylaxis: Sq heparin 5000 q 8 hrs. If many risk factors, add pneumoboots
c. GI bleed prophylaxis: Zantac 150 mg IV or PO q 12 – if normal Cr/UO (or PPI)
d. Prevent Pneumonia with HOB > 30 degrees
a. Determine expectations and goals of care from patients and/or family surrogates early on.
b. Inform them about high mortality rate and prognostic importance of response to therapy over next 2-3 days (extend to 4-5 days with certain comorbidities e.g. ESLDx).
c. In advising patients/family, obtain history of pre-morbid functional status.
d. Best predictor of prognosis is whether the patient is responding to therapy within a 2-4 day period. Lack of response or worsening after 4-5 days, makes it very unlikely the patient will improve.
1. Fluid Resuscitation Therapy
Fluid resuscitation should begin with lactated ringers. There is no role for albumin initially, but it is reasonable to use adjunctively in those patients with persistent signs of shock, despite an initial crystalloid resuscitation, and have anasarca with a low serum albumin (<3.0). In these circumstance, albumin (20%) can be given as a replacement fluid — targeting a serum albumin concentration of >3.0. This recommendation is based on the results of a recent RCT showing a mortality benefit in a retrospectively selected subset of patients with septic shock in the ALBIOS trial (which included all patients with severe sepsis). Additional support for the use of albumin in septic shock patients comes from two prior meta-analyses and one large RCT, the SAFE trial. Clearly, there is no role for colloids other than albumin. There have been several recent RCT’s looking at the effectiveness of various starch colloids (e.g., Hetastarch and Pentastarch), which have demonstrated worse outcomes with these agents including either an increase in mortality and/or an increase in AKI and need for CRRT.
Repeated fluid boluses—500-1000 mL of crystalloid should be administered through at least an 18 gauge IV given over 10 to 20 minutes—using a pressure bag. Fluid boluses should continue, providing there is persistent shock with or without hypotension, and evidence of circulatory improvement using routine physical exam and laboratory parameters (e.g., heart rate, MAP, extremity perfusion, urine output, lactate, and ScvO2). IVF should stop when either shock has resolved or there is evidence of harm.
After an initial 2-3 L of crystalloid are given, vasopressors should be used to reach the goal MAP, and continued IVF should be used to reach the goal of reversing hypoperfusion, with the caveats listed above about when to stop.
After the initial resuscitation goals have been achieved and shock has resolved, after a period of stability one should be encouraged to achieve a net negative fluid balance. Often times this occurs spontaneously, heralded by polyuria and negative fluid balance, however, diuretics may be required to affect this goal. Careful monitoring of organ perfusion is essential during diuretic-induced fluid mobilization, to avoid secondary organ dysfunction.
2. Vasoactive Therapy
The goals of resuscitation include maintaining an adequate mean arterial pressure, to support cerebral and coronary perfusion, and improving oxygen delivery (cardiac output) to support systemic organ perfusion. Both goals are best achieved initially through IV crystalloid resuscitation; however, vasoactive medications may be required after 2-3 L of IVF and sometime sooner if hypotension is refractory or severely reduced.
Studies show no improvement in organ perfusion when a higher MAP is targeted. However, when setting BP goals in the individual patient, it is important to consider the patient’s baseline BP; since organ perfusion may be compromised at this MAP goal in patients with baseline HTN. Regardless, the effectiveness (and adverse impact) of a given therapy should always be reassessed by following serial measurements of organ perfusion as therapy is titrated (e.g., physical exam, serum lactate measurements, mixed central venous oxygenation, urine output, and mental status).
Based on the cardiovascular derangements in septic shock, including both vasodilation and a high incidence of cardiomyopathy, it is preferable to use a vasoactive agent with both vasopressor and inotropic actions. The optimal vasopressor choice for septic shock based on RCT is norepinephrine, primarily because it is highly potent and has less arrythmogenic potential than dopamine. Nuances in patient presentation, requires that a firm grasp of the cardiovascular pharmacology of each potential vasoactive agent is known to be able to maximize the therapeutic role in a given patient (see Table II).
If BP remains refractory to 10-15 mcg/min of NE, we recommend starting vasopressin at 0.03 units/min, without titration, as well as adding stress dose corticosteroids. This is based on weak evidence favouring this combination of agents using a subset analysis of an RCT, however, given this approach has no risks compared to alternative strategies, we would favor this approach.
Finally, the optimal means to wean vasopressors is similarly unknown and requires further investigation. However, in the setting of a septic cardiomyopathy, if vasopressin has been added, one should not wean norepinephrine prior to weaning vasopressin to avoid the exclusive effects of increased afterload without inotropic support.
3. Corticosteroid Therapy
The role of low-dose (or “stress-dose”) corticosteroids in the treatment of severe sepsis and septic shock remains controversial. Annane and colleagues found a mortality benefit in a single-center randomized clinical trial when low-dose corticosteroids (hydrocortisone 50 mg IV q 6H) and fludrocortisone (50 mcg enterally daily) were administered for 7 days to patients who did not respond to an adrenocorticotropic hormone stimulation test (≤9 mg/dL).
In contrast, the more recently published Corticosteroid Therapy of Septic Shock (CORTICUS) study failed to confirm a mortality benefit, although the patients enrolled differed significantly (less ill) from those in the original Annane trial. Both trials, however, did demonstrate a significant reduction in the time to shock reversal in patients receiving low-dose corticosteroids. Thus, at present, it is recommended that low-dose corticosteroids only be considered for patients with fluid and vasopressor-refractory septic shock, defined as failing low dose vasopressors.
Table II.Cardiovascular Pharmacology of Vasoactive Agents in Septic Shock
4. Anti-Microbial Therapy
The timely administration of effective broad-spectrum anti-microbial therapy is of paramount importance in the initial management of the patient with severe sepsis. In a recent cohort study, Kumar and colleagues demonstrated that mortality increases nearly 7% for every hour that antibiotics are delayed beyond the first hour of hypotension. This has been confirmed in several other studies.
In addition, several studies have shown that if the initial antimicrobial regimen is ineffective against the pathogen, mortality increases even if the appropriate antibiotic is given subsequently based on culture data. These studies provide the rational for the recommendation that broad spectrum anti-microbial therapy should be administered empirically, and within one hour of identification of severe septis or septic shock. In most patients, this would include vancomycin, a broad spectrum anti-pseudomonal agent and an aminoglycoside. Double coverage with 2 new agents active against resistant gram negative pathogens is important, given the risk of a highly resistant strain in those patients with health care-associated septic shock. The antibiotics must be new as patients developing severe sepsis while being treated with an antibiotic for >3-4 days will not become sick from a germ that was being treated with the first antibiotic.
When choosing an empiric antibiotic regimen, clinicians must also take into account prior culture results (i.e., patient colonization), recent courses of antimicrobials (to avoid same class of agents), drug allergies, and the hospital antibiogram. After 3 days of treatment, the antibiotic regimen should be de-escalated to the narrowest spectrum possible based on available cultures, or discontinued completely if an alternative diagnosis is established.
However, if after 2-3 days the patient fails to respond to the initial antibiotics, the drug regimen should be changed (see section below on what to do if patient fails to improve), and a directed and systematic search for a localized source of infection should be initiated, e.g. CT chest and abdomen or at least ultrasound if too unstable for transport. Definitive source control may require invasive procedures, including surgical intervention.
5. Tight Glycemic Control
The therapeutic role of intensive insulin therapy, which has intuitive value given insulin’s anti-inflammatory properties and the association between hyperglycemia and poor outcomes, has recently been called into question. Several recent studies of intensive insulin therapy failed to demonstrate a mortality benefit while simultaneously exposing an increased risk of hypoglycemia. Furthermore, inaccuracies inherent in point-of-care glucose monitoring may underestimate the incidence of hypoglycemia. It is prudent to control blood glucose in moderation, with a goal of 140-180 mg/dL to avoid marked hyperglycemia while reducing the risk of hypoglycemia.
What should I do in refractory cases?
In <10% of cases, patients will die of refractory hypotension. This usually results from severe vascular paralysis or concomitant cardiogenic shock. In either case, one must be certain that the patient is not hypovolemic and does not have an unexpected reversible lesion, which should be assessed by echocardiography (e.g., dynamic LVOT obstruction). In addition, broadening antibiotics and searching for a drainable collection of pus are all important measures to take prior to deeming a patient refractory. In all such cases, patients, or their surrogates, should be informed about the high risk of death and for patients with pure vascular paralysis, they should be informed that CPR is not indicated if progressive hypotension occurs.
Aside from escalating vasopressors (high dose NE, VP, and sometimes EPI) and stress dose corticosteroids the focus should be on ensuring adequate source control and appropriate ABx.
Concomitant low cardiac output (based on physical exam and/or ScvO2 <<70%):
Here the problem may include either or both cardiogenic or obstructive shock. One should not simply increase the dose of inotropes without further diagnostic testing, as this therapy may not be appropriate and may be harmful. Thus evaluation should include EKG, CXR, ventilatory mechanics (to r/o autoPEEP), intra-abdominal pressure, and most importantly, echocardiography to determine the pathophysiology.
1. Pulmonary causes:
2. Extrapulmonary causes:
Abdominal compartment syndrome
3. Cardiogenic causes:
Dilated septic cardiomyopathy with very low ejection fraction (most common with prior cardiomyopathy)
Septic cardiomyopathy with low ventricular compliance and inadequate ventricular dilatation
Right heart failure from sepsis and/and or pulm HTN associated lung injury (or as a comorbidity
Ventricular outflow tract obstruction caused or exacerbated by beta agonist therapy
Valvular heart disease previously unappreciated or underestimated
Specific therapy must be directed against the underlying pathophysiology as determined above, while ensuring appropriate ABX and source control and continuing dialogue with the family to ensure that the level of care is meeting the goals and expectations of the patient or their surrogate.
5. Disease monitoring, follow-up and disposition
Expected response to treatment?
In general, most patients will respond to antibiotics within 2-3 days with signs of stabilization or improvement. This may take longer (e.g., 4-5 days) with some infections and in some hosts (e.g., end-stage liver disease). If patient worsens over the first 2-3 days, or fails to improve (see below), one must review and consider broadening the ABX coverage to cover pathogenic organisms that may not have been considered initially or treated empirically (e.g., Candida, aspergillus [or other molds], vancomycin-resistant enterococcus, vibrio vulnificus, nocardia, herpes simplex, influenza, and PCP.
When a patient fails to improve or worsens over 2-3 days despite treatment, one should consider several scenarios:
There is another cause (not sepsis) for the presentation
The antimicrobial treatment is ineffective against the pathogen
The infection has spread locally or systemically despite antibiotics and source control is required (e.g., by surgical or IR drainage)
A secondary complication related or unrelated to the septic event is overshadowing the clinical findings (e.g., PE, hemorrhagic shock [not yet reflected in HgB drop], intestinal ischemia, atrial fibrillation, C. Diff, or a neurologic catastrophe [e.g., ICH, or massive stroke]).
In such cases, as above, the following should be done:
Consultation with an infectious disease expert to review and consider broadening the ABX coverage to treat potential pathogenic organisms that may not have been considered initially or treated empirically e.g. candida, aspergillus (or other molds), vancomycin-resistant enterococcus, vibrio vulnificus, nocardia, herpes simplex, influenza, and PCP.
Perform a CT scan of the chest, abdomen, pelvis if stable enough for transfer, or directed ultrasound exams to evaluate for an undrained collection in the thorax or abdomen.
Perform thorough exam and diagnostic testing to assess for a concomitant disorder as listed above.
Most importantly, all culture samples should be reviewed for pathogens and their sensitivities vs current antibiotics. This includes a review of all cultures sent prior to admission if seen initially at an outside facility prior to transfer.
Follow-up labs (e.g., WBC, lactate, other organ function) should improve.
While in the ICU, it is imperative that all evidenced-based ICU protocols, that have been shown to reduce complications and improve outcomes, be strictly adhered to. These include but are not limited to the following:
DVT pr ophylaxis
Prevent in fection:
Hand-washing, HOB >30, etc
Justify all lines and tubes daily and pull them ASAP
For the care of vented patients:
Lung protective ventilation (if ARDS)
Daily screening for spontaneous breathing trial
Daily screen for sedation interruption or weaning
Prevent hyperglycemia (>180) and avoid hypoglycemia while maintaining glucose in a goal range of 140-180.
Transfuse PRBC only if HgB <7
The severity of an infection is dependent on both the pathogen’s virulence and the body’s pathophysiologic response to the infection. The toll-like receptor family plays an important and proximal role in both pathogen recognition (innate immunity) and in the initiation of the host’s inflammatory response. Peptidoglycan in the cell wall of gram-positive pathogens is recognized by TLR-2; lipopolysaccharide in the outer membrane of gram-negative pathogens is detected by TLR-4. Viral and fungal pathogens are also recognized by this family of receptors.
TLRs also play an integral role in the initiation of the “cytokine storm” of sepsis. These inflammatory cytokines stimulate neutrophils and endothelial cells, including the activation of procoagulant pathways. This adaptive immune response, through a complex, interactive relationship, augments the innate immune response to enable a more effective response to the infection. There are biologic pathways activated simultaneously, which serve to down-regulate and control this response; however, the host’s immune response may become maladaptive, resulting in organ dysfunction, circulatory shock, and death.
The specific mechanisms that regulate this response, and the genetics underlying their control, currently represent intense areas of investigation. Although the pathophysiology of organ dysfunction is multifactorial and beyond the scope of this chapter, circulatory shock plays a very important role. In addition, since shock is an important and early clinical manifestation of severe sepsis, the treatment of which may be life-saving, it is important to understand the pathophysiology and hemodynamics of septic shock.
Activated neutrophils release mediators which increase the permeability of the microvasculature. This leads to third spacing and further exacerbates the decrease in intravascular fluid volume typically present on admission, due to anorexia and external fluid losses (sweating, gastrointestinal, etc). In addition, activated endothelial cells induce nitric oxide production which leads to widespread vasodilation (low systemic vascular resistance) impairing the compensatory response to hypovolemia.
Furthermore, in up to 60% of patients with septic shock, a sepsis-induced cardiomyopathy (SIC) contributes to the circulatory derangements. Although serum troponin is a sensitive marker of SIC, and levels are proportional to both the severity of cardiac dysfunction and patient prognosis, the underlying pathophysiology does not involve coronary ischemia or infarction.
It is important to note that although widespread vasodilation (and low afterload) is a central feature of septic shock, the physical exam findings characteristic of a high cardiac output state (i.e., “warm” shock) are often absent on initial presentation. Rather, it is common for septic patients to present with findings consistent with a low cardiac output (i.e., “cold shock”) with a narrow pulse pressure, cool extremities, and mottled skin. This apparent paradox can be reconciled by dismissing the notion that vasodilation directly causes a reflex increase in cardiac output. For cardiac output to rise in response to systemic vasodilation, there must be sufficient driving pressure (i.e., the mean systemic pressure; Pms) in the venous system to promote an increase in venous return, since cardiac output must equal venous return.
Since Pms is determined by the venous blood volume and compliance of the venous blood vessels, it is often decreased on patient presentation due to profound hypovolemia and sepsis induced venodilation (which increases venous compliance), respectively. Furthermore, SIC, present in many patients, further contributes to the impairment in cardiac output. Hence the findings of poor extremity perfusion on admission, as well as a low CVP and ScvO2 (discussed below).
Nevertheless, after intravascular volume has been repleted, which may require up to 7-9 liters in the first 24 hours, the classic findings of “warm shock” become manifest in most patients, including a wide pulse pressure, and warm, well-perfused, extremities. This adaptive response to volume repletion occurs in most patients, and is essential for survival. Even with SIC, as the reduction in inotropy is compensated for by an increase in end diastolic volume (preload) due to ventricular dilation, and a decrease in afterload from arterial vasodilation.
In the United States alone, based on earlier estimates, approximately 750,000 individuals develop severe sepsis annually and more than 200,000 of these patients will die. However, more recent data suggests that these estimates may be off by 3-4 fold, as they are all based on claims data. Now sepsis is the 3rd leading cause of death in the United States, after cancer and heart disease, and accounts for nearly 20% of all admissions to ICUs, and is estimated to cost $17 billion annually.
Prognosis is dependent on the patient’s comorbidities and the severity of sepsis upon presentation as indicated by the nature and extent of the organ dysfunction. Patients with shock, respiratory failure requiring mechanical ventilation and multiorgan dysfunction syndrome have the highest mortality. However, no matter how severe the sepsis is upon presentation is, unless the underlying comorbidities are life-threatening or the patient has a terminal illness, one cannot reliably predict who will survive and thus the best predictor of patient outcome is the response to initial therapy.
In general, prognosis cannot be determined until a 2-3 day trial of intensive care. If over this time period there is worsening organ dysfunction or a lack of improvement, the chance of survival is markedly reduced. In some patients (e.g., endovascular infection or those with end stage liver disease) improvement may take a bit longer than 2-3 days.
What's the evidence?
Angus, DC, Linde-Zwirble, WT, Lidicker, J. “Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated cost of care”. Crit Care Med . vol. 29. 2001. pp. 1303-1310. (The most often cited reference of the epidemiology of severe sepsis in the United States.)
Dellinger, RP, Levy, MM, Carlet, JM. “Surviving Sepsis Campaign: International guidelines for management of severe sepsis and septic shock”. Crit Care Med . vol. 36. 2008. pp. 296-327. (Most comprehensive review of the evidence for severe sepsis interventions.)
Kumar, A, Roberts, D, Wood, KE. “Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock”. Crit Care Med. vol. 34. 2006. pp. 1589-1596. (This prospective observational study is the best evidence to date in support of the life-saving effects of early antibiotics. They found that for each hour delay in antibiotic administration beyond the first hour of hypotension mortality increase by ~ 7%.)
Levy, MM, fink, MP, Marshal, JC. “2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference”. Crit Care Med. vol. 31. 2003. pp. 1250-1256. (The second consensus conference to review and revise the prior definitions of sepsis, severe sepsis and septic shock, which were established initially by Dr. Roger Bone and colleagues in 1992.)
Mikkelsen, ME, Miltiades, AN, Gaieski, DF. “Serum lactate is associated with mortality in severe sepsis independent of organ failure and shock”. Crit Care Med . vol. 37. 2009. pp. 1670-1677. (One of several studies to show the value of serum lactate as an independent predictor of outcome in patients with severe sepsis. This study strongly supports the guideline recommendation for the routine measurement of serum lactate in all patients with sepsis.)
Rivers, E, Nguyen, B, Havstad, S. “Early goal-directed therapy in the treatment of severe sepsis and septic shock”. N Engl J med . vol. 345. 2001. pp. 1368-1377. (This is the landmark, though controversial, study that shows the value of early aggressive resuscitation in septic shock. A non-evidenced based, multi-interventional protocol was used in this study, which makes it difficult to evaluate the relative effects of each. However, the measurement, and treatment of the ScvO2 was the only variable in the protocol between the two experimental groups, suggesting that it is this parameter which is most important.)
Annane, D, Sebille, V, Carpentier, C. “Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock”. JAMA. vol. 288. 2002. pp. 862-871. (The first and only, large multi-centered RCT to show the life-saving effects of initiating stress dose steroids for patients with vasopressor-refractory septic shock.)
Caironi, P, Tognoni, G, Masson, S. “Albumin replacement in patients with severe sepsis or septic shock”. N Engl J Med . vol. 370. 2014. pp. 1412-1421. (Multi-center, RCT showing no benefit of Albumin replacement therapy in severe sepsis. However, in the septic shock group identified retrospectively, the mortality was lower.)
“The ProCESS Investigators. A randomized trial of protocol-based care for early septic shock”. N Engl J Med. vol. 370. 2014. pp. 1683-1693. (Multi-center, RCT showing that a resuscitation strategy using traditional EGDT with ScvO2 as a goal – per the Rivers Trial in 2001, resulted in similar outcomes to one that uses a protocol with traditional bedside parameters, or one that has no protocol approach (usual care).)
Asfar, P, Meziani, F, Hamel, J-F. “High versus low blood-pressure target in patients with septic shock”. N Engl J Med . vol. 370. 2014. pp. 1583-1593. (Multi-center, RCT showing that targeting a higher than traditional MAP goal for patients with septic shock was of no benefit vs 65 mm HG. However, in the prospectively established subset of patients with chronic HTN targeting a higher BP resulted in less AKI and need for HD).
Dellinger, RP, Levy, MM, Rhodes, A. “Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock, 2012”. Int Care Med . vol. 39. 2013. pp. 165-228. (A recent very comprehensive revision of the last Surviving Sepsis Campaign Guidelines, which was not supported by any external funding agencies. All the recommendations in this document are graded systematically using the GRADE system.)
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.
- 1. Description of the problem
- 2. Emergency Management
- 3. Diagnosis
- 4. Specific Treatment
- 5. Disease monitoring, follow-up and disposition
- What's the evidence?