Management of Sepsis Bryan D. Kraft, M.D. - PowerPoint Presentation

1. Management of SepsisBryan D. Kraft, M.D.Assistant Professor of MedicinePulmonary, Allergy, & Critical Care MedicineDuke University Medical Center

2. DisclosuresNone

3. ObjectivesDefine sepsisLearn basic sepsis pathophysiologyUnderstand central venous O2 saturationReview literature on fluid resuscitationDiscuss approach to treatment

4. Definitions of SIRS and sepsisSystemic Inflammatory Response Syndrome: Requires ≥ 2 of the following:Temperature > 38°C or < 36°CRespiratory rate > 20/min, or PaCO2 < 32 mmHgHeart rate > 90 bpmWBC count >12,000 or <4,000 cells/µL, or > 10% bandsSepsis: ≥ 2 SIRS criteria + evidence of infectionACCP/SCCM Consensus Conference. Crit Care Med 1992;20(6):864–74.

5. Severe SepsisSepsis-induced tissue hypoperfusion or organ dysfunction (any of the following):Hypotension (SBP < 90 mmHg, MAP < 70 mmHg, or SBP of 40 mmHg below baseline)Elevated lactateUOP < 0.5 mL/kg/hr x 2 hours, despite adequate IVFCreatinine > 2 mg/dLPaO2/FIO2 < 250 in absence of PNA as sourcePaO2/FIO2 < 200 in presence of PNA as sourceBilirubin > 2 mg/dLPlatelets < 100,000/µLCoagulopathy (INR > 1.5)Levy MM, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003;31:1250-1256.

6. Shock in general is a state of cellular and tissue hypoxia due to:Reduced O2 delivery (e.g. anemia, hypoxemia); Outstripped O2 consumption (e.g. thyrotoxicosis); and/or Impaired utilization (e.g. mitochondrial dysfunction)Septic shock commonly manifests with #1 and #3 concurrently in a patientMost common clinical definition of septic shock is presence of severe sepsis + hypotension refractory to intravenous fluids alonei.e. the patient requires vasopressorsgenerally hypotension is mean arterial pressure (MAP) < 65 mmHgWatch out for shock in a chronically hypertensive patient with “normal BP” that is significantly below baselineSome argue presence of elevated lactate (>4 mmol/L) also indicates shock regardless of MAPSeptic shock

7. This model of sepsis suggests it is a linear spectrum of disease:Linear Model of SepsisSepsisSeptic shockSevere sepsisSIRS

8. This model of sepsis suggests it is a linear spectrum of disease:Is this really true?For example, can patients have severe sepsis without meeting SIRS criteria?Linear Model of SepsisSepsisSeptic shockSevere sepsisSIRS

9. Kaukonen K et al. N Engl J Med 2015;372:1629-1638.This study analyzed > 100,000 patients with severe sepsis and stratified by SIRS status (< 2 vs. ≥ 2 criteria) How useful are these definitions?

10. Kaukonen K et al. N Engl J Med 2015;372:1629-1638.This study analyzed > 100,000 patients with severe sepsis and stratified by SIRS status (< 2 vs. ≥ 2 criteria) How useful are these definitions?12% of patients with severe sepsis (1 in 8) met only 0 or 1 SIRS criterionThis was especially true of older patients, use of β-blockers, etc.

11. Kaukonen K et al. N Engl J Med 2015;372:1629-1638.The same study then asked whether there was anything special about meeting 2 SIRS criteria, i.e. the cut-off usedHow useful are these definitions?

12. Kaukonen K et al. N Engl J Med 2015;372:1629-1638.The same study then asked whether there was anything special about meeting 2 SIRS criteria, i.e. the cut-off usedHow useful are these definitions?They found a cut-off of “2” was not associated with a step-up in mortality, but rather there was a linear relationship between SIRS and mortality. So the sicker you are (i.e. more SIRS criteria you meet) the worse your prognosis is.

13. Singer M, et al. Sepsis-3. JAMA. 2016;315(8):801-10.In 2016, the Society of Critical Care Medicine and European Society of Intensive Care Medicine released a consensus statement redefining the sepsis syndrome.“SIRS” and “severe sepsis” were eliminated“Septic shock” was redefined as hypotension despite adequate IVF requiring vasopressors to maintain MAP > 65 mmHg, and serum lactate > 2 mmol/L.New Definitions of Sepsis?

14. Emphasis was placed on organ failure (SOFA score), especially the “quick SOFA” or qSOFA score:Presence of altered mental statusSystolic blood pressure ≤100 mmHgRespiratory rate ≥ 22/minuteA score of ≥ 2 reflects high risk for organ failure and mortality and should prompt appropriate interventionNew Definitions of Sepsis?Singer M, et al. Sepsis-3. JAMA. 2016;315(8):801-10.

15. While these definitions may prove to be clinically useful, they are not yet validated, so should be used cautiously.For example, while SIRS sacrifices specificity for sensitivity, qSOFA is likely under-sensitive, which is problematic for a screening test because it may miss septic patients.New Definitions of Sepsis?Singer M, et al. Sepsis-3. JAMA. 2016;315(8):801-10.Additional reading on Sepsis-3 definitions:Simpson SQ. New Sepsis Criteria: A Change We Should Not Make. Chest. 2016;149(5):1117-8.

16. Incidence of Sepsis is RisingBenchmarking the Incidence and Mortality of Severe Sepsis in the United StatesGaieski, D; Edwards, J; Kallan, M; Carr, B. Crit Care Med 41(5):1167-1174, 2013.SepsisSevere sepsisSeptic shockICD-9 codes:Mortality:10.6%36.9%42.2%Total cases (n)

17. Severe Sepsis—MortalityMackenzie, I. et al. BMJ 335:929-932, 2007Deaths/100,000/year05010015020025030015-2425-4445-64>65Age Age-related Mortality020406080100012345Number of Failing Organs Mortality by Organ Failure%Two of the most important factors affecting prognosis are organ failure and age: Patients with sepsis generally die from multiple organ failure. Age is a major factor, especially after age 65

18. Severe Sepsis—OutcomesYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. What are the outcomes of patients severe sepsis or shock compared to general population?

19. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime Yende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. The Y-axis is worsening functional status towards death. The X-axis is time.What are the outcomes of patients severe sepsis or shock compared to general population?

20. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime No sepsisYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. These patients are the general publicthat do not get sepsis. There is a steady decline in functional status due to aging, but it is slow.

21. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime Severe sepsisNo sepsisYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. These patients with severe sepsis get an initial, significant drop in functional status, but survive their illness and mostly recover, but not back to baseline.

22. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime RefractoryShockSevere sepsisNo sepsisYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. These patients with refractory shock typically come in with overwhelming bacteremia and die within the first few days due to shock, acidosis, and/or hypoxemia.

23. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime RefractoryShockChronic critical illness and immune suppressionSevere sepsisNo sepsisYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. These patients survive the acute illness but their organ failure never recovers and they remain life support-dependent. Older patients get trapped in ICU. Their ICU course waxes and wanes with complications such as infections due to VRE or Candida, organisms which are otherwise not very virulent.

24. Severe Sepsis—OutcomesDecliningFunctional ReserveDeathTime RefractoryShockChronic critical illness and immune suppressionSevere sepsisNo sepsisYende S, Iwashyna TJ, and Angus DC. Interplay between sepsis and chronic health Trends in Molecular Medicine, 20, 4 , 2014. These patients survive the acute illness but their organ failure never recovers and they remain life support-dependent. Older patients get trapped in ICU. Their ICU course waxes and wanes with complications such as infections due to VRE or Candida, organisms which are otherwise not very virulent.Additional reading on chronic critical illness:Nelson JE, et al. Chronic critical illness. Am J Respir Crit Care Med. 2010; 182(4):446-54.

25. So how can sepsis outcomes be improved?

26. 701. The Rapid Choice of the Right Antibiotic(s)Morell MR et al. The Management of Severe Sepsis and Septic ShockInfectious Disease Clinics of North America 23 (3): 485-501, 2009 6050403020100Mortality (%)AppropriateInappropriateKang et al.Micek et al.Khatib et al.Schramm et al.Garey et al.Tumbarello et al.Cheong et al.Early, broad-spectrum antibiotics are absolutely critical. Failure to treat up front with the right antibiotics increases mortality by 10-20 percentage points.

27. 1. The Rapid Choice of the Right AntibioticsReview of risk factors for certain infectionsE.g., recent bowel surgery should make you think about gram-negatives, anaerobes, and Candida.Identification of source(s) or suspected source(s)E.g., comprehensive work-up based on risk factors, history, physical, laboratory, and imaging studiesCorrect stratification of patients as hospital-acquired, health-care associated, or community-acquiredE.g., recent hospitalization, SNF patient, etc.Review of prior infection history and prior microbiological cultures and sensitivitiesE.g., urine culture from 2 months ago grew Zosyn-resistant E. coliEarly (within 60 minutes of diagnosis) delivery of antibioticsWhat does this look like in action?1. The Rapid Choice of the Right Antibiotic(s)

28. 1. The Rapid Choice of the Right AntibioticsThis means EARLY involvement of procedural and surgical colleagues for adequate source control:E.g. inserting chest tubes for empyemas, biliary tubes for biliary obstructions, nephrostomy tubes for obstructing kidney stones, joint wash-out for septic arthritis, ex-laps for bowel perfs, etc.One caveat – patients much be stabilized for these procedures, which means they need to be fluid resuscitated first (see next slide).It is also common for patients to transiently worsen following source control procedure (classically drainage of liver abscess) due to “stirring up infection”. Anticipate this and stay on top of it with fluids and vasopressors! Consider inadequate source control in a patient who fails to improve with fluids and antibiotics aloneAntibiotics will not effectively penetrate undrained abscesses2. Achieve Adequate Source Control

29. 3. Fluid resuscitation: Central tenet of therapy2012 Surviving Sepsis Campaign recommends specific physiologic goals for fluid resuscitation:MAP > 65 mmHgCVP 8-12 mmHg (or 12-15 during mechanical vent)Central venous O2 sat > 65% or Mixed venous > 70%Urine output > 0.5 ml/kg/hNormalization of lactateIt is important to note that these goals are for the first 6 hours of sepsis therapyWe are going to dive deeper into fluids and central venous O2 saturation (ScvO2) in a few slides

30. First-line: Norepinephrine (dose: 0.01 - 1 µg/kg/min)Increases MAP mainly by vasoconstriction, but also is a mild inotrope which is important for sepsis-mediated cardiac stunningLow doses improve cardiac output, and cerebral, renal, and splanchnic blood flowCompared to dopamine: Less tachycardia, fewer arrhythmias, lower RR of death (0.91; 0.83–0.99) Second-line: Vasopressin (dose: 0.01 – 0.04 units/min)Goal is to decrease dose of NE Reasonable to start when NE dose gets to 0.2 µg/kg/minSome say dose is “fixed” at 0.04, but in reality you can adjust.4. Maintain Blood Pressure Support

31. Third-line: Epinephrine (dose: 0.01 - 1 µg/kg/min)Same dosing regimen as norepinephrineFYI on Epinephrine:If patient is dead, the dose is 1 mg IV q3min.If patient is alive but hypotensive, the dose is 10 mcg IV (or 1 ml of 1:100,000 concentration) q1min prn. Don’t give 1 mg!Can use Phenylephrine infusion in certain situations, e.g. avoidance of beta-adrenergic activation if patient has rapid Afib.4. Maintain Blood Pressure Support

32. Let’s get back to discussing fluid resuscitation goals in sepsis and what the literature saysBy now you have heard of the 2001 NEJM paper on Early Goal Directed Therapy by Manny Rivers (next slide)A bit of background, “Early” refers to the first 6 hours (pre-ICU). The studies done pre-Rivers looked at “Goal Directed Therapy” for sepsis, which targeted physiologic endpoints once patients arrived in the ICU but not before, and did not show a difference in outcomes. We’ll come back to that concept later.Severe Sepsis – Fluid therapy

33. Severe Sepsis – Fluid therapy

34. Prospective, single center, randomized, controlled trial263 adult patients presenting to a tertiary care center ED with severe sepsis or septic shockSepsis defined as:≥ 2 SIRS criteria, ANDSBP ≤ 90 after 20-30 ml/kg (~1.5 – 2 L) IVF bolus over 30 min OR blood lactate ≥ 4 mmol/LNotable exclusion criteria included immunosuppression, active cancer, & need for urgent surgeryRivers E, et al. NEJM 2001.Rivers E et al. N Engl J Med 2001;345:1368-1377.

35. Rivers E et al. N Engl J Med 2001;345:1368-1377.Here is the enrollment and randomization schema for the studyEnrollment and Resuscitation Algorithm

36. Enrollment and Resuscitation AlgorithmHere is the enrollment and randomization schema for the studyYou will notice that the only difference between the 2 groups was the use of central venous saturation (ScvO2) as a surrogate for oxygen delivery to guide resuscitation.Rivers E et al. N Engl J Med 2001;345:1368-1377.

37. Enrollment and Resuscitation AlgorithmHere is the enrollment and randomization schema for the studyYou will notice that the only difference between the 2 groups was the use of central venous saturation (ScvO2) as a surrogate for oxygen delivery to guide resuscitation.So let’s review ScvO2 and what it means…Rivers E et al. N Engl J Med 2001;345:1368-1377.

38. Central venous vs. Mixed venous O2 satCentral venous oxyhemoglobin saturation (ScvO2) is measured hereMixed venous oxyhemoglobin saturation (MvO2) is measured here

39. Central venous vs. Mixed venous O2 satCentral venous oxyhemoglobin saturation (ScvO2) is measured hereMixed venous oxyhemoglobin saturation (MvO2) is measured hereWhat’s the difference? ScvO2 reflects tissue O2 extraction from only the upper extremities and brain, missing what the visceral organs (including coronary sinus, which empties into the RA) and lower extremities are doing. Visceral organs typically have higher extraction ratios and therefore lower venous O2 saturations. The MvO2 is therefore lower than the ScvO2 and reflects the average O2 extraction for the entire body. It is called “mixed” because the venous return from the SVC, IVC, and coronary sinus is mixed by the RV.

40. Central venous vs. Mixed venous O2 satCentral venous oxyhemoglobin saturation (ScvO2) is measured hereMixed venous oxyhemoglobin saturation (MvO2) is measured hereWhat’s the difference? ScvO2 reflects tissue O2 extraction from only the upper extremities and brain, missing what the visceral organs (including coronary sinus, which empties into the RA) and lower extremities are doing. Visceral organs typically have higher extraction ratios and therefore lower venous O2 saturations. The MvO2 is therefore lower than the ScvO2 and reflects the average O2 extraction for the entire body. It is called “mixed” because the venous return from the SVC, IVC, and coronary sinus is mixed by the RV.We tend to check ScVO2 rather than MvO2 because the latter requires a PA catheter. What are they supposed to represent? They are surrogates for oxygen delivery (DO2) and oxygen uptake (VO2). This is derived from the Fick Equation…

41. Fick PrincipleC.O.= VO2/(CaO2-CvO2)VO2 =C.O. (CaO2-CvO2)Volume of O2 inVolume of O2 outCvO2 CaO2 The Fick Principle

42. Fick PrincipleC.O.= VO2/(CaO2-CvO2)VO2 =C.O. (CaO2-CvO2)Volume of O2 inVolume of O2 outCvO2 CaO2 The Fick PrincipleVenous Oxygen Content:(Hb)(MVO2 %)(1.34) + (0.003)(PvO2)Arterial Oxygen Content:(Hb)(SaO2 %)(1.34) + (0.003)(PaO2)

43. Fick PrincipleC.O.= VO2/(CaO2-CvO2)VO2 =C.O. (CaO2-CvO2)Volume of O2 inVolume of O2 outCvO2 CaO2 The Fick PrincipleSo the concept then is that by measuring MvO2 (or ScvO2), you can make assumptions about tissue oxygenation. For example, if the MvO2 is high (e.g. > 65%), then the tissue oxygen extraction is not maxed out and therefore O2 delivery is adequate. If the MvO2 is low (e.g. < 50%), then the tissues are extracting more oxygen suggesting delivery is inadequate. The assumption, then, based on the Rivers paper, is that RBC transfusion (to increase Hb) or Dobutamine (to increase CO), will then improve oxygen uptake. (Fick equation above). There are 2 major problems with this assumption. First, while measuring the venous O2 content may be helpful, it isn’t enough because of the two other independent variables (CO and VO2). For example, VO2 may be depressed due to cytopathic hypoxia from sepsis rather than poor CO or anemia. Second, the MvO2 is a weighted average of the entire body. Each organ has a specific CO and VO2…

44. 75%20%75%20%75%20%90%10%60%10%75%20%C.O.= 5 lpmMvO2 is weighted average of entire bodyEach circle is an organ. The top number is the organ-specific SvO2% and the bottom number is organ blood flow as a % of the total cardiac output.HealthyKidneyLiverBrainHeartGutMuscle75%Mixed venousOrganSvO2% COWeighted ContributionBrain750.215Heart600.16Liver750.215Gut750.215Kidney900.19Sk. Muscle750.215MVO2 = 75%

45. 75%20%75%20%20%10%70%5%60%15%75%20%75%20%75%20%75%20%90%10%60%10%SvO2=%CO/OBF=75%20%Severe sepsisPre-resuscitationHealthyC.O.= 5 lpmC.O.= 7.5 lpmGutMixed venousWhat happens during severe sepsis?= SvO2=%CO/OBF75%Kidney61%

46. 75%20%75%20%20%10%70%5%60%15%75%20%75%20%75%20%75%20%90%10%60%10%SvO2=%CO/OBF=75%20%Severe sepsisPre-resuscitationHealthyC.O.= 5 lpmC.O.= 7.5 lpmGutMixed venousKidneyWhat happens during severe sepsis?= SvO2=%CO/OBF75%61%But blood flow to the gut and kidney has decreased. This has lowered the venous saturations due to higher extraction and caused an overall decrease in MvO2. This is why we resuscitate.What has happened in the septic patient? CO has increased due to catecholamine stimulation.

47. What happens during severe sepsis?75%20%75%20%60%20%90%10%60%15%75%15%C.O.= 10 lpmKidneyGut71%Severe sepsisPost-resuscitation= SvO2=%CO/OBFWith fluid resuscitation, blood flow has been largely restored to vital organs and CO has continued to increase. There is also a rise in MvO2 from 61% to 71%.

48. What happens during severe sepsis?75%20%75%20%60%20%90%10%60%15%75%15%C.O.= 10 lpmKidneyGut71%Severe sepsisPost-resuscitation= SvO2=%CO/OBFBut what happens if you drop out a unit? For example, the patient develops gut ischemia…With fluid resuscitation, blood flow has been largely restored to vital organs and CO has continued to increase. There is also a rise in MvO2 from 61% to 71%.

49. 75%20%75%20%0%0%90%10%60%20%75%30%VO2/Q dropout75%20%75%20%60%20%90%10%60%15%75%15%C.O.= 10 lpmC.O.= 10 lpmKidneyGutGutGut ischemia71%74%Severe sepsisPost-resuscitation= SvO2=%CO/OBF= SvO2=%CO/OBF

50. 75%20%75%20%0%0%90%10%60%20%75%30%VO2/Q dropout75%20%75%20%60%20%90%10%60%15%75%15%C.O.= 10 lpmC.O.= 10 lpmKidneyGutGutGut ischemia71%74%Severe sepsisPost-resuscitation= SvO2=%CO/OBF= SvO2=%CO/OBFHere the gut becomes under-perfused and the MvO2 actually goes up because blood flow is redistributed. So a rise in MvO2 doesn’t necessarily mean oxygen delivery has been improved. Additional reading:Roberts JK, al. Oxygen Delivery in Septic Shock. Ann Am Thorac Soc. 2015;12(6):952-5.

51. Rivers E et al. N Engl J Med 2001;345:1368-1377.Protocol for Early Goal-Directed Therapy500 ml q30 minBack to Rivers’ paper…Here was the algorithm

52. Baseline patient characteristicsVariableStandard (N=133)EGDT (N=130)DemographicsAge (yr)Sex (% male)64.4 ± 17.150.467.1 ± 17.450.8Entry criteriaTemperature (°C)Heart rate (beats/min)Systolic BP (mmHg)Respiratory rateWBC count (per µl)Lactate (mmol/l)36.6 ± 2.3114 ± 27109 ± 3430.2 ± 10.614,200 ± 9,6006.9 ± 4.535.9 ± 3.2117 ± 31106 ± 3631.8 ± 10.813,600 ± 8,3007.7 ± 4.7Baseline laboratory valuesAnion gap (mmol/l)Creatinine (mg/dl)21.4 ± 8.52.6 ± 221.7 ± 7.62.6 ± 2Diagnosis (%)PneumoniaUrosepsisOther39.527.721.938.525.623.1Features of SepsisSevere Sepsis (%)Septic Shock (%)Bacteremia (%)APACHE II score48.751.336.120.445.354.734.221.4Rivers E et al. N Engl J Med 2001;345:1368-1377.

53. Treatments AdministeredRivers E et al. N Engl J Med 2001;345:1368-1377.TreatmentHours After the Start of Therapy0 – 67 – 720 – 72Total fluids (ml) Standard therapy EGDT P value3499 ± 24384981 ± 2984<0.00110602 ± 62168625 ± 51620.0113358 ± 772913443 ± 6390NSRed-cell transfusion (%) Standard therapy EGDT P value18.564.1<0.00132.811.1<0.00144.568.4<0.001Any vasopressor (%) Standard therapy EGDT P value30.327.4NS42.929.10.0351.336.80.02Dobutamine (%) Standard therapy EGDT P value0.813.7<0.0018.414.5NS9.215.4NS

54. Treatments AdministeredRivers E et al. N Engl J Med 2001;345:1368-1377.TreatmentHours After the Start of Therapy0 – 67 – 720 – 72Total fluids (ml) Standard therapy EGDT P value3499 ± 24384981 ± 2984<0.00110602 ± 62168625 ± 51620.0113358 ± 772913443 ± 6390NSRed-cell transfusion (%) Standard therapy EGDT P value18.564.1<0.00132.811.1<0.00144.568.4<0.001Any vasopressor (%) Standard therapy EGDT P value30.327.4NS42.929.10.0351.336.80.02Dobutamine (%) Standard therapy EGDT P value0.813.7<0.0018.414.5NS9.215.4NSPatients randomized to EGDT got more fluid in the first 6 hours. The Standard therapy group then played catch up with fluid between 7-72 hours.

55. Treatments AdministeredRivers E et al. N Engl J Med 2001;345:1368-1377.TreatmentHours After the Start of Therapy0 – 67 – 720 – 72Total fluids (ml) Standard therapy EGDT P value3499 ± 24384981 ± 2984<0.00110602 ± 62168625 ± 51620.0113358 ± 772913443 ± 6390NSRed-cell transfusion (%) Standard therapy EGDT P value18.564.1<0.00132.811.1<0.00144.568.4<0.001Any vasopressor (%) Standard therapy EGDT P value30.327.4NS42.929.10.0351.336.80.02Dobutamine (%) Standard therapy EGDT P value0.813.7<0.0018.414.5NS9.215.4NSBut total fluid administration was the same at 72 hours…

56. VariableStandard (N=133)EGDT (N=130)P ValueAPACHE II score (7-72 hours)15.9 ± 6.413.0 ± 6.3<0.001SAPS II score (7-72 hours)42.6 ± 11.536.9 ± 11.3<0.001MODS score (7-72 hours)6.4 ± 45.1 ± 3.9<0.001In-hospital mortality59 (46.5%)38 (30.5%)0.00928-Day mortality61 (49.2%)40 (33.3%)0.0160-Day mortality70(56.9%)50 (44.3%)0.03OutcomesRivers E et al. N Engl J Med 2001;345:1368-1377.Even though total fluids were the same at 72 hours, there was still more death and organ failure in Standard therapy group. This suggests that fluids during the first 6 hours is critical and if you get behind you can’t make it up later.

57. Suggestion that the Standard group algorithm was not followed as well as EGDT algorithm:Control group received less fluid over first 6 hours despite same hemodynamic triggers (CVP 8 -12, MAP > 65). The CVP and MAP both lower in control group at 6 hours14% of the control group (compared to <1% of EGDT group) did not reach combined physiologic end points for CVP, MAP, and UOPCould not conclude which components of the algorithm were more most crucial for observed survival benefit~ 10% of patients in the study did not receive Abx within 6 hoursOther observations and criticisms

58. Barriers to EGDT implementationConcern of the lack of generalizability of the studyTranslation from a highly controlled research environment to diverse clinical care setting?Need for timely insertion of a central venous catheter which is cumbersome and resource heavyTime and resources required to implement the protocol in general are highCurrently EGDT is only followed strictly in ~ 10% of candidate patientsCarlbom DJ, et al. Crit Care Med 2007;35:2525-2532.Jones AE, et al. Acad Emerg Med 2007;14:1072-1078.

59. There have been 3 major follow-up studies to Rivers EGDT to address some of these concernsProCESS (NEJM 2014) – U.S.ARISE (NEJM 2014) – Australia/NZProMISe (NEJM 2015) – U.K.These studies evaluated EGDT vs. usual care in patients with early severe sepsis or septic shockFollow-up StudiesThe ProCESS Investigators. N Engl J Med 2014; 370:1683-1693.The ARISE Investigators and the ANZICS Clinical Trials Group. N Engl J Med 2014; 371:1496-1506.Mouncey PR, et al. for the ProMISe Investigators N Engl J Med 2015; 372:1301-1311.

60. Multicenter, prospective, randomized, controlled trials of patients with severe sepsis and septic shock≥ 2 SIRS criteria, ANDEvidence of hypotension or hypoperfusion (1 of the following):SBP < 90 or MAP < 65 after fluid bolus (0.5 - 1 L or 20-30 ml/kg over 30-60 min)Requiring vasopressors to achieve BP goalsLactate > 4 mmol/lProCESS, ARISE, & ProMISe trials

61. Acute CVAAcute coronary syndromeAcute pulmonary edemaStatus asthmaticusMajor cardiac arrhythmiaSeizureDrug overdoseBurn or traumaEmergent surgeryCD4 count < 50/mm2*Active bleeding*Desired limited care*CVC contraindicated*Refusal of blood*Imminent death*Pregnancy*Age < 18 yrsProCESS, ARISE, & ProMISe trialsExclusion criteria similar to Rivers study:*ARISE only

62. ProCESS, ARISE, & ProMISe trialsStudy groupsEarly goal-directed therapy compared with usual care (and protocol driven standard therapy in ProCESS trial)For Usual Care group, decisions about diagnosis, monitoring, treatment were made by treating clinical teamScvO2 measurements were not permitted during the first 6-hour intervention

63. ProCESS Trial1,341 patients randomized in ED at 1:1:1439 protocol-based EGDT, 446 protocol-based standard therapy, 456 usual care (UC)60 day mortality 21%, 18.2% and 18.9%, respectively (P=NS)No differences in 90 day or 1 year mortality or need for organ support for:EGDT vs. Standard Protocol Both Protocols vs. UC

64. ARISE Trial1,600 patients (796 EGDT, 804 UC)EGDT group received ~250 ml more IVF in first 6 h (1964 vs. 1713 ml)90 day mortality of 18.6 vs. 18.8% for EGDT vs. UC respectively (P=NS)No differences in survival time, duration of organ support, LOS

65. ProMISe Trial1,260 patients (630 EGDT, 630 UC)EGDT group received ~200 ml more IVF in first 6 h (2000 vs. 1784 ml)90 day mortality of 29.5% vs. 29.2% for EGDT vs. UC respectively (P=NS)No differences in duration of organ support, length of hospital stay

66. Clinical Characteristics Across StudiesRivers, et al.ProcESSARISEProMISeVariableEGDTN=130UCN=133EGDTN=439Stand.N=446UCN=456EGDTN=796UCN=804EGDTN=625UCN=626APACHE II21.420.420.820.620.715.415.818.718.0% Shock54.7%51.3%55.6%53.8%53.3%70%69.8%54.1%55.6%Lactate7.76.94.85.04.94.44.27.06.8% PNA38.539.531.934.133.136.532.836.533.1% Urinary25.627.722.820.220.618.720.117.318.7Bacteremia34.2%36.1%31.7%28.3%28.7%38%37.6%----Creatinine2.62.62.52.22.31.41.52.12.2IVF pre-R----2254222620832515259119502000IVF 0-6 hrs498134992805328522791964171320001784IVF pre-R – 6h----5059551143624479430439503784UC=Usual CareR=randomization

67. Clinical Characteristics Across StudiesRivers, et al.ProcESSARISEProMISEVariableEGDTN=130UCN=133EGDTN=439Stand.N=446UCN=456EGDTN=796UCN=804EGDTN=625UCN=626APACHE II21.420.420.820.620.715.415.818.718.0% Shock54.7%51.3%55.6%53.8%53.3%70%69.8%54.1%55.6%Lactate7.76.94.85.04.94.44.27.06.8% PNA38.539.531.934.133.136.532.836.533.1% Urinary25.627.722.820.220.618.720.117.318.7Bacteremia34.2%36.1%31.7%28.3%28.7%38%37.6%----Creatinine2.62.62.52.22.31.41.52.12.2IVF pre-R----2254222620832515259119502000IVF 0-6 hrs498134992805328522791964171320001784IVF pre-R – 6h----5059551143624479430439503784UC=Usual CareR=randomizationARISE cohort not as sick

68. Clinical Characteristics Across StudiesRivers, et al.ProcESSARISEProMISEVariableEGDTN=130UCN=133EGDTN=439Stand.N=446UCN=456EGDTN=796UCN=804EGDTN=625UCN=626APACHE II21.420.420.820.620.715.415.818.718.0% Shock54.7%51.3%55.6%53.8%53.3%70%69.8%54.1%55.6%Lactate7.76.94.85.04.94.44.27.06.8% PNA38.539.531.934.133.136.532.836.533.1% Urinary25.627.722.820.220.618.720.117.318.7Bacteremia34.2%36.1%31.7%28.3%28.7%38%37.6%----Creatinine2.62.62.52.22.31.41.52.12.2IVF pre-R----2254222620832515259119502000IVF 0-6 hrs498134992805328522791964171320001784IVF pre-R – 6h----5059551143624479430439503784UC=Usual CareR=randomizationMost patients got 4-5 L of IVF

69. CriticismsMortalities in ProCESS and ARISE were lower than expected (~ 18%)Raises the possibility of a lack of treatment effect in less sick patientsProMISe mortality ~ 29% (Rivers - 46%)? Bias in usual care groups’ treatment due to knowledge of Rivers study: 4-8% transfused, 1-8% got DobutamineAll patients given ~ 2L IVF pre-randomization

70. CriticismsMortalities in ProCESS and ARISE were lower than expected (~ 18%)Raises the possibility of a lack of treatment effect in less sick patientsProMISe mortality ~ 29% (Rivers - 46%)? Bias in usual care groups’ treatment due to knowledge of Rivers study: 4-8% transfused, 1-8% got DobutamineAll patients given ~ 2L IVF pre-randomizationThis is not trivial – it will dilute out a treatment effect because it will make absolute differences between the 2 treatments smaller. Imagine studying ASA for STEMI and patients in both control (placebo) and intervention (ASA 325 mg) groups were given 81 mg before randomization…

71. Use of Lactate to Guide ResuscitationSo the routine use of ScvO2 is not necessary…Experts have also called for the use of lactate “clearance” as a non-invasive guide for IVF resuscitation(Of course “clearance” is a misnomer because can’t measure production and elimination, only random points in time…)Lactate levels may be clinically relevant though: Mortality rate for septic pts with lactate > 4 mmol/L ~ 30% and is up to 46% if also hypotensiveSurviving Sepsis Campaign guidelines:Suggestion: Targeting resuscitation to normalize lactate in patients with elevated lactate levels as a marker of tissue hypoperfusion (grade 2C)Levy et al. Crit Care Med 2010

72. Jansen, et al. looked at using lactate to guide IVF therapy in a RCT in 2010Used EGDT/physiologic endpoints to target ~ 10% lactate reduction per hourCompared with a non-lactate based approachStudy outcomes included organ severity (SOFA), mortality, length of stay, etc.Use of Lactate to Guide ResuscitationJansen et al. Am J Resp Crit Care Med, 2010

73. (A) Treatment algorithm, control group and lactate group. The goal for central venous pressure (CVP) was 12 to 15 mm Hg in mechanically ventilated patients. Besides the static CVP goals, CVP was used as a dynamic safety limit during fluid challenges. Both crystalloids and colloids could be used at the discretion of the clinician. Albumin was not a standard resuscitation fluid in the participating centers. The goal for hemoglobin was 10 g/dl in patients with cardiac ischemia. (B) Additional treatment algorithm, lactate group. If the lactate level became less than or equal to 2.0 mEq/L, a further decrease was no longer required. Fluid responsiveness was assessed by a fluid challenge of 200 ml crystalloids or colloids. The goal was an increase in blood pressure, ScvO2, or stroke volume, or a decrease in heart rate. CVP was used as a dynamic safety limit: if CVP increased less than or equal to 2 mm Hg, fluid administration was continued; if CVP increased more than 2 and less than or equal to 5 mm Hg, the fluid challenge was repeated after waiting for 10 minutes; if CVP increased greater than or equal to 5 mm Hg, fluid administration was stopped. Before administration of vasodilators, fluid responsiveness was assessed and fluids were infused if necessary. The recommended dose for nitroglycerin was 2 mg in 1/2 hour followed by 2 mg per hour. MAP = mean arterial pressure; NTG = nitroglycerin; RBCs = red blood cell transfusions; SaO2 = arterial oxygen saturation; ScvO2 = central venous oxygen saturation; T = time; UP = urine production.But Jansen study algorithm quite complexLactate checked q2h in Lactate GroupIf lactate remained > 3 mmol/LAll of these are designed to improve O2 delivery

74. Lower adjusted mortality by 10 percentage points (adj. HR 0.61; P=0.006)Lower SOFA scores (i.e. organ failure) at 72 hours (6.4 vs. 7.0, p=0.009)Shorter time to weaning from mechanical ventilation and inotropes (but not from vasopressors or dialysis)Shorter ICU length of stayJansen Study Clinical OutcomesCompared to Control group, the Lactate group had:

75. Use of Lactate to Guide ResuscitationLactate Level (mEq/L)P ValueHours after Start of TherapyControl GroupLactate GroupBaseline (0 h)4.7 (3.9–5.5)4.6 (3.9–5.4)0.7582.7 (2.3–3.2)2.6 (2.2–3.1)0.590–83.3 (2.8–3.9)3.2 (2.7–3.8)0.809–721.7 (1.4–2.0)1.6 (1.3–1.9)0.17Jansen et al. Am J Resp Crit Care Med, 2010;182(6):752-61.Why was this?

76. Use of Lactate to Guide ResuscitationLactate Level (mEq/L)P ValueHours after Start of TherapyControl GroupLactate GroupBaseline (0 h)4.7 (3.9–5.5)4.6 (3.9–5.4)0.7582.7 (2.3–3.2)2.6 (2.2–3.1)0.590–83.3 (2.8–3.9)3.2 (2.7–3.8)0.809–721.7 (1.4–2.0)1.6 (1.3–1.9)0.17Jansen et al. Am J Resp Crit Care Med, 2010;182(6):752-61.Interestingly, there was no difference in lactate levels between control and “lactate-driven protocol”Why was this?

77. Use of Lactate to Guide ResuscitationLactate Level (mEq/L)P ValueHours after Start of TherapyControl GroupLactate GroupBaseline (0 h)4.7 (3.9–5.5)4.6 (3.9–5.4)0.7582.7 (2.3–3.2)2.6 (2.2–3.1)0.590–83.3 (2.8–3.9)3.2 (2.7–3.8)0.809–721.7 (1.4–2.0)1.6 (1.3–1.9)0.17TreatmentControl GroupLactate GroupP ValueFluids, ml* 0–8 h†2,194 ± 1,6692,697 ± 1,9650.011 9–72 h‡10,043 ± 6,1418,515 ± 4,9870.055Red blood cell transfusion, ml 0–8 h†196 ± 495322 ± 10370.15 9–72 h‡345 ± 667423 ± 13000.59Jansen et al. Am J Resp Crit Care Med, 2010;182(6):752-61.Interestingly, there was no difference in lactate levels between control and “lactate-driven protocol”Why was this?

78. Use of Lactate to Guide ResuscitationLactate Level (mEq/L)P ValueHours after Start of TherapyControl GroupLactate GroupBaseline (0 h)4.7 (3.9–5.5)4.6 (3.9–5.4)0.7582.7 (2.3–3.2)2.6 (2.2–3.1)0.590–83.3 (2.8–3.9)3.2 (2.7–3.8)0.809–721.7 (1.4–2.0)1.6 (1.3–1.9)0.17TreatmentControl GroupLactate GroupP ValueFluids, ml* 0–8 h†2,194 ± 1,6692,697 ± 1,9650.011 9–72 h‡10,043 ± 6,1418,515 ± 4,9870.055Red blood cell transfusion, ml 0–8 h†196 ± 495322 ± 10370.15 9–72 h‡345 ± 667423 ± 13000.59Jansen et al. Am J Resp Crit Care Med, 2010;182(6):752-61.Interestingly, there was no difference in lactate levels between control and “lactate-driven protocol”Again we see more upfront IVF in the Lactate group and late catch up in the control groupWhy was this?

79. Lactate monitoring may be beneficial in guiding sepsis resuscitationBut does not accelerate lactate “clearance” compared to usual careLactate is likely a biomarker for, rather than a “cause” of, organ failureReinforces the principle that the first six hours is a crucial time for aggressive IVF, vasopressors, achieving physiologic goalsAnd that playing catch up after 6 hours is insufficient to reverse organ failureWhat can we learn from Jansen study?

80. Additional Reading on Lactate and SepsisSuetrong B and Walley KR. Lactic Acidosis in Sepsis: It’s Not All Anaerobic: Implications for Diagnosis and Management. Chest. 2016;149(1):252-61.Nice review on lactate production and metabolism in health and sepsis.Jones AE, et al. Lactate clearance vs central venous oxygen saturation as goals of early sepsis therapy: a randomized clinical trial. JAMA. 2010;303(8):739-46.Jones et al. found that targeting lactate clearance of at least 10% was non-inferior to targeting SvO2 > 70%. There were no differences in mortality between the two groups.

81. SummaryThe 3 most important components of sepsis care are:Rapid administration of the right antibiotic(s)Source control within first hospital dayAggressive resuscitation with IVF, pressors, etc. during first 6 hours of careHow much fluid is enough? Most patients in Rivers, ProCESS, ARISE, and ProMISe received 4-5 L and this is likely a good estimate, acknowledging every patient is different.How do you judge fluid responsiveness? By all of these: Physical exam, straight leg raise, IVC ultrasound, CVP, MAP, UOP, lactate, ScvO2, trial and error, etc. etc.

82. SummaryThe 3 most important components of sepsis care are:Rapid administration of the right antibiotic(s)Source control within first hospital dayAggressive resuscitation with IVF, pressors, etc. during first 6 hours of careHow much fluid is enough? Most patients in Rivers, ProCESS, ARISE, and ProMISe received 4-5 L and this is likely a good estimate, acknowledging every patient is different.How do you judge fluid responsiveness? By all of these: Physical exam, straight leg raise, IVC ultrasound, CVP, MAP, UOP, lactate, ScvO2, trial and error, etc. etc. No single test will tell you what to do!

83. SummaryFluids should be administered as long as the patient is fluid responsive, to a point.Avoid fluid overload (edema worsens O2 delivery due to increasing distance needed for O2 diffusion)Conversely, using high dose vasopresors in a hypovolemic patient will cause digital necrosisSo the take home message is give enough fluids up front to fill the tank (~4-5 L may be sufficient) but not too much to cause significant fluid overload…

84. SummaryToo little fluidToo much fluidMore organ injuryHigher mortalityVasopressor-induced digital necrosisWorse edemaImpaired oxygen diffusionWorse ARDSFinding the middle for each patient requires combining many variables: physical exam, straight leg raise, IVC ultrasound, lactate, ScvO2, urine output, trial and error, etc.

85. SummaryThe ProCESS, ARISE, and ProMISe studies are not an assault on protocolized sepsis managementHowever, routine use of ScvO2 > 70% for resuscitation goal cannot be recommendedBut ScvO2 monitoring may still be highly indicated in select patients that are not responding to non-invasive approach alone, i.e.Rising lactate despite adequate fluid resuscitationActive cardiac ischemia (heart regulates VO2 based on DO2 rather than O2 extraction)

86. SummarySo back to where we started – should we still try to optimize oxygen delivery as a treatment for sepsis?Yes, to a point. We know that raising DO2 to supra-physiologic levels is bad, but returning DO2 to physiologic range is still valid based on the best available data and first principles. But not everyone needs ScvO2 monitoringAnd each organ has different metabolic needs, so using the ScvO2 to make assumptions about local tissue oxygen delivery comes with all of the caveats discussed

87. Additional Reading on SepsisATS Fellows Reading ListNEJM 2014; 371(15):1381-91. Low vs. high Hb goalsNEJM. 2014; 24;370(17):1583-93. High vs. low MAP targetsAnesth Analg. 2010 Aug;111(2):444-50. Practice malalignment in RCTs, specifically as it relates to ICU transfusion goals

88. Feel free to email questions:bryan.kraft@duke.edu