Management of Sepsis

Download 80.2 Kb.
Date conversion29.04.2016
Size80.2 Kb.
Management of Sepsis

Ian Jenkins, MD

Stefanos Parpos, MD


Sepsis is the runaway version of the healthy response to infection, with vasodilation, increased capillary permeability, etc remote from the original site of infection. The same pathophysiology can happen without infection (bad pancreatitis, or toxic shock (toxin absorbed but no bacterial invasion). 10-60% of patients die despite your best efforts, with >600 deaths a day and > 16 billion spent each year. Definitions from a 1992 consensus panel:
A) Systemic inflammatory response syndrome (SIRS): widespread inflammatory response to a variety of severe insults. > 1 of the following:
1) Temp >38ºC or <36ºC

2) HR >90/min

3) RR >20/min or PaCO2 <32 mmHg

4) WBC >12,000, <4000, or >10% bands
B) Sepsis (systemic response to infection): SIRS + definitive evidence of infection. Does NOT require bacteremia, as outcomes are the same with and without it.
C) Severe sepsis: Sepsis + organ dysfunction, hypoperfusion, or hypotension (may include, but are not limited to, lactic acidosis (in contrast to the respiratory alkalosis seen earlier in sepsis), hemostatic abnormalities, hepatic dysfunction, oliguria, or acute alteration in mental status).
D) Septic shock: Severe Sepsis with hypotension or pressor requirement despite adequate fluid resuscitation.
E) Hypotension: systolic BP <90 mmHg or 40 below baseline.
F) Multiple organ failure: altered organ function in an acutely ill patient such that homeostasis cannot be maintained without intervention.


A) USE the definitions. They help you recognize MILD sepsis when someone looks good so you can act aggressively before it becomes SEVERE sepsis, before you are called for hypotension, anuria and the like.
B) When you ARE called for hypotension or clinical decline, lots of things are on your mind from falsely low BP to cardiac (arrhythmia, low output, tamponade), pulmonary (tension PTX, PE), meds, etc. Start with history (a story of infection or of a predisposing factor) and vitals (tachypnea may be the FIRST abnormality; temp can be low or NORMAL; meds may blunt tachycardia, and a normal BP may represent a fall for some patients or the vasoconstriction reserve of previously healthy patients). Classically patients look "ill," with warm, vasodilated extremities (shock from cardiopulm sources causes vasoconstriction); do a focused exam for source (chest, belly, skin, inflamed line or IV, and you may have to ROLL them to find the decub or perirectal abscess).
C) Do appropriate studies (blood cultures x 2 sets and from any line, urine culture, sputum if making it, CBC/diff, chemistries, often LFTs, often coags and fibrinogen, ddimer, sometimes a lactate; CXR, belly imaging where needed) and TREAT aggressively.
D) How sick is the patient? ER predictors of death = terminal illness, suspected pneumonia, platelets < 150k or bands > 5%, nursing home, >65 yrs old, delta MS, and most importantly, tachypnea or hypoxia and shock.


Innumerable cytokines and interleukins interact to cause sepsis; of innumerable trials that tried to interrupt the cascade, only a few things worked. Be cautious interpreting results: 1 in 20 studies is significant at the p < .05 level just by chance. We still use everything we think might help because there is little specific therapy available.

A: FLUIDS: Don't write for an infusion and walk away; bolus, reassess, and rebolus until goals are met. While the novice is often hesitant to use much fluid, initial requirements may be >10 liters in the first day, and after capillary leak causes extravasation of much IVF, patients can end up dozens of liters up over a micu stay. The MUST protocol comes from the only hard data for fluid use—a study of protocol, goal driven therapy: 500cc NS boluses until CVP was 8-12, then pressors were added for a MAP >65, and if SVO2 was <70%, transfusion for to hct > 27 and dobutamine as needed were added to support O2 delivery. Mortality was 25.2 instead of 40.7%! The prudent MD will also consider HR, urine output, correction of hyperlactatemia, exam evidence of perfusion, or wedge pressure.

B: ANTIBIOTICS: Generally, you can go by Sanford for empiric therapy until you have culture data and sensitivities. Treat EARLY. Antibiotics for meningitis should NEVER wait for CT/LP, and yofor example. Culture and treat. Other:
1) Severe infection doesn't mean antibiotic resistance, but the cost of a mistake is higher. Therefore, it may be ok to treat mild strep pneumo pneumonia (in a healthy host with no risk factors for drug resistance) with penicillin, but if the infection is severe, be cautious—ceftriaxone has near certain coverage; some would add a second drug (respiratory quinolone) in the really sick. Have someone floridly septic from the UTI that was cultured 2 days ago? Their sensitive E coli may merit only cefazolin or ampicillin and when they're ready for discharge, bactrim.
2) Recognize the patient who needs broad spectrum coverage (nursing home, hospitalizations, ICU acquired infections putting them at risk for MRSA, dangerous GNRs). Look at prior culture data. You may find the patient failing levoflox for pneumonia had drug-resistant pseudomonas in prior sputa.
3) GNR sepsis, or what looks like it, is often double covered (best data for a B-lactam and aminoglycoside) to try to get synergistic killing and ensure at least one drug is active. The data that this is better than a 3rd generation cephalosporin or carbapenem alone is scant; in neutropenic fever for example, ceftaz or cefepime is equal to piperacillin and gent.
4) Vanco is overused. If you are worried about resistant pathogens, start it; if you have no cultures from the ET tube, or the blood, or the line after 3 days, try to stop it. Don't mistake spectrum for power. Vanco is often just static. For susceptible staph, naf and ox are stronger / cidal, cheaper, and reduce vanco resistance. That said, community acquired MRSA is becoming more common and if missed, is deadly.
5) Don't forget C diff. Even without diarrhea, and especially if there is ileus, abdominal distention or pain, high WBC, suspect it and treat it / rule it out. There are exotoxin B only C Diffs which will have (-) stool tests.
6) Some basics... CAP in the ICU can be treated with ceftriaxone + azithro. VAP needs MRSA (vanco) + GNR coverage (ceftaz / pip-tazo / carbapenem) until you have more data. Cholecystitis, cholangitis, diverticulitis and colonic perfs require gram (-), (+), and anaerobe coverage. Amp + gent + flagyl, amp-sulbactam or pip-tazo, or levo (suboptimal enteroccocus coverage) + flagyl are options.
7) Remove sources: antibiotics do not get into abscesses; drain them. Sources of toxic shock (wound packing, tampons), infected prosthetic devices and lines must be removed; a necrotic gallbladder, ischemic bowel, or colon with severe C diff may require surgery. If they have pneumonia, you can't remove it, but you can affect recurrence (or prevent) by keeping the head of bed elevated > / = 30 degrees.

C: PRESSORS AND INOTROPES: Know a few; the rest comes with practice. Don't memorize the dose yet; ask the RN, look it up. Remember, your objective is organ perfusion, NOT blood pressure. You can give a hypovolemic patient powerful pressors and get the BP up, but their toes will turn black from ischemia, and their kidneys, gut (leading to bacterial translocation), liver and other organs will suffer too. No good data to prefer one over another but there are some physiologic rationales:

1) Dobutamine: B1 (chronotrope, inotrope) and B2 (vasodilator). Increases cardiac output, but may lower BP fall, and predisposes to tachyarrythmia. In pure sepsis, the primary problem is vasodilation, not poor cardiac function. However some, and the MUST protocol, would use it to improve O2 delivery when low as suggested by a mixed venous sat <70% (suggests high O2 extraction hence insufficient cardiac output).
2) Dopamine: B1 > B2, plus alpha. Inotrope plus vasoconstricting (pressor) alpha to support BP. Less chronotrope than dobutamine. A good all-purpose pressor for both cardiac and septic shock and can be given peripherally until the central line is in. No data for “renal dose” dopa.
3) Norepinephrine (levophed): Like dopamine but predominant alpha so a more powerful pressor and less chronotrope. The pressor used in the trial of goal directed therapy, hence preferred.
4) Phenylephrine: all alpha. Very bad for cardiac shock; since there is more oxygen utilization at higher cardiac outputs, a pressor with inotrope in it may be a better first choice for sepsis—but does not predispose to SVTs like dobuta, dopa, and norepi.
5) Vasopressin: premade and stored in the pituitary. Early in sepsis, it is released by a normal host. Then supplies run out. A replacement infusion can help get patients off other pressors and has a physiologic rationale.
6) For sepsis: Bolus, bolus, bolus! Then, levophed, and if more are needed consider vasopressin infusion, or phenylephrine if SVTs are a factor.
7) Additional details on these agents at the end of this presentation

D: ACTIVATED PROTEIN C (drotrecogin or Xigris): Seems counterintuitive, but the coagulation and inflammatory cascades are intertwined, and it has antiinflammatory and anti-DIC properties. A 96 hour infusion started within 24 hours of presentation reduced mortality from 30.8 to 24.7 among ~1700 patients, but caused a trend to more bleeding (sometimes fatal). It also costs $6,000+. Thus there are strict inclusion and exclusion criteria on the web (don't memorize them!). Briefly: known or suspected infection, 3 of 4 SIRS criteria, plus hypotension after fluids, oliguria, hypoxia, thrombocytopenia or acidosis, and NOT be under 18, pregnant, have plts < 30k, bleeding or predisposition (recent GIB, head trauma, surgery, diathesis), dialysis dependence, organ transplant, advanced AIDS, merely acute pancreatitis, moribund state, chronic liver disease, and most anticoagulants or procoagulant states). Had favorable cost-benefit ratio for patients with APACHE (an illness severity measure) scores > 25 (24k per year of life gained), and poor ratio for scores < 25 ($575k per year of life gained).

E: STEROIDS: A large study of severe sepsis patients found that a week of 50mg hydrocortisone IV Q6H + 50mcg fludrocortisone PO QD reduced morality from 63 to 53% (still a horrible outcome—sicker than the APC trial patients?). This was among only those who were adrenally insufficient (<10mcg/dl response to a 250mcg cosyntropin stimulation test); the few adrenally responsive patients trended to increased mortality on steroids. So cort-stim patients with severe sepsis and treat while results are pending and if (+).
If they are sick, give some Dexamethasone ASAP then test—it doesn't cross-react with the assay, but the others do. Don't miss adrenal crisis! $10 worth of steroids can save a critically ill patient in hours. It is miraculous to watch. Clues: hyponatremia, hyperkalemia, eosinophilia, no other cause seen, predisposition (other autoimmune disorders like type 1 diabetes, or known metastatic cancer, immunosuppression (infiltrative infections of the adrenals)).

F: INSULIN and NUTRITION and OTHER: In >1500 intubated patients, intensive (gtt) insulin therapy with a goal of 80-100 instead of 180-200 reduced SICU mortality from 8 to 4.6%, hospital mortality from 10.9 to 7.2%, with less bacteremia and polyneuropathy with a cost of more hypoglycemia. Glucose control improves acute outcomes (ie, immune function, wound healing, survival in MI) so it is plausible these benefits extend to MICU and not just SICU patients, but that needs to be confirmed. Also, providing nutrition supports albumin, immune function, and wound healing, and if enteral, you avoid line complications (mechanical, infection) and might decrease bacterial translocation across the guy by keeping it healthy. The exact best regimen is undetermined. Lastly, keep HOB > 30 degrees to avoid aspiration in all unless impossible; use daily, protocol-driven vent wean-ability assessments, and pay daily attention to GI / DVT prophylaxis and use of sedation.


  • Sepsis is common, dangerous, complicated and costly

  • Early sepsis is hard to recognize and easy to treat; late sepsis is easy to recognize and hard to treat

  • Aggressive bolus fluid resuscitation until goals are met (CVP 8-12)

  • AFTER volume, use norepinephrine, then dopamine and vasopressin as needed; phenylephrine is good when there are SVTs.

  • When MAP > 70, check SVO2, and if < 70, use transfusions for hct <27 and dobutamine for >27 and reassess.

  • Appropriate and early antibiotic therapy and removal of the source when possible

  • Does the patient fit all the inclusion and none of the exclusion factors for activate protein C?

  • Test for adrenal insufficiency and treat if positive (can cover with dexamethasone while testing)

  • Aim for normoglycemia, don’t forget nutrition, and think about other details of good care (daily reconsideration of sedation, vent weaning, prophylaxis, and semirecumbent position).


Rivers, et al. Early Goal Directed Therapy in the Treatment of Severe Sepsis and Septic Shock. NEJM 345:19. Nov 2001. 1368-1377.
Van de Bergh, et al. Intensive Insulin Therapy in Critically Ill Patients. NEJM Nov 2001. 345:19 1359-1367.
Annane et al. Effect of Treatment With Low Doses of Hydrocortisone and Fludrocortisone on Mortality in Patients With Septic Shock. JAMA, Aug 2002; 288: 862 – 871
Bernard et al. Efficacy and Safety of Recombinant Human Activated Protein C for Severe Sepsis. NEJM, Mar 2001. 344: 699-709.

IV Pressors/Inotropes/Vasodilators in the ICU
Mechanism of Action of Adrenergic Receptor Stimulation:

 alpha 1: vasoconstrict peripheral arterioles

 beta  1: predominate in myocardium;cause myocardial inotropy and chronotropy
 beta  2: predominate in periphery; cause vasodilation of skeletal and mesenteric beds and bronchodilation
 dopa  1: vasodilate mesenteric and renal beds (also some in coronary and cerebral vessels)
 dopa  2: inhibit norepinephrine release
 vasop  : see attached supplement

Hemodynamic Parameters during Hypotension:

 cardiogenic:   PCWP increased; CO increased; SVR increased

 hypovolemic:   PCWP decreased; CO decreased; SVR increased
 sepsis     :   PCWP decr/norm; CO increased; SVR decreased

Overview of Rx:

 Sympathomimetic Rx:

  Dopamine (Intropin)
  Dobutamine (Dobutrex)
  Isoproterenol (Isuprel)
  Norepinephrine (Levophed)
  Epinephrine (Adrenalin)
  Phenylephrine (Neosynephrine)

 Phosphodiesterase Inhibitors:

  Amrinone (Inocor)
  Milrinone (Primacor)


  Nitroprusside (Nipride)


    A. Starting dose/dose range
       1. Renal Range: 2-5mcg/kg/min
       2. Beta Range: 5-10mcg/kg/min
       3. Alpha Range: >10-20mcg/kg/min (pressor effect)
    B. Sites of Action (as above)
    C. Indications:
           Renal perfusion, hypotension, inotropic support, pressor support
    D. Side Effects:
           tachycardia; arrythmias, ischemia, subcutaneous necrosis
    E. Comments:
           More tachycardia than Dobutamine
           Not as potent a pressor as Levophed
           Will increase renal blood flow in patients on Levophed


    A. Starting dose/dose range
       3-10 mcg/kg/min
    B. Sites of Action
           Beta 1 receptors of the heart, some Beta 2 receptors in the periphery
    C. Indications:
           inotropic support for cardiogenic shock and CHF
    D. Side Effects:
           tachycardia; arrythmias, hypotension
    E. Comments:
       NOT a pressor
         essentially a potent inotrop and mild vasodilator to increase EF
         not as potent a vasodilator as mulrinone
       more inotropic than chronotropic (as oppposed to isoproterenol)


    A. Starting dose/dose range
       1-10 mcg/min (up to 30mcg/min)
    B. Sites of Action
           Beta 1 receptors of the heart, some Beta 2 receptors in the periphery
    C. Indications:
           chronotropic incompetent states to increase heart rate (heart block, bradycardia, torsades)
    D. Side Effects:
           arrythmias, hypotension, nervousness, headaches
    E. Comments:
         great chronotrop
         reduces BP


    A. Starting dose
       2-4 mcg/kg/min
    B. Sites of Action
           Beta 1 receptors of the heart,  Beta 2 AND alpha 1 receptors in the periphery
    C. Indications:
           hypotension, inotropic support
    D. Side Effects:
           arrythmias, hypotension, decreased mesenteric perfusion, skin necrosis, reflex bradycardia
    E. Comments:
         most potent vasopressor
           less of a chronotrop than dopamine or dobutamine


    A. Starting dose
       2 mcg/kg/min
    B. Sites of Action
           Beta 1 and 2 receptors of the heart,   alpha 1 > Beta 2 receptors in the periphery
    C. Indications:
           cardiac arrest, anaphylaxis, hypotension, inotropic support
    D. Side Effects:
           arrythmias, ischemia, decreased mesenteric and renal perfusion
    E. Comments:
         may cause profound hypertension in presence of beta blocker


    A. Starting dose
       20 mcg/min  (UP TO  200mcg/min)
    B. Sites of Action
           alpha 1 receptors in the periphery
    C. Indications:
    D. Side Effects:
           headaches, reflex bradycardia, arrythmias, restlessness
    E. Comments:
         caution w/ severe MR
         no beta effects so good to use when tachycardia present


    A. Starting dose
       LOAD 50mcg/KG/min over 10 minutes
    B. Sites of Action
           beta 1 receptors of the heart and some beta 2 in periphery
    C. Indications:
           inotropic support for cardiogenic shock and CHF
    D. Side Effects:
           hypotension, arrythmias
    E. Comments:



The database is protected by copyright © 2016
send message

    Main page