EMRA*Cast & EMRA Critical Care Committee Journal Club

Fluid resuscitation in sepsis is a hotly debated and much-studied topic - and for good reason. Listen in as three emergency physicians discuss some of the landmark papers directing emergency care of these critically ill patients. Join host Dr. Ranjita Raghavan of Mount Sinai and guests Dr. Chad Meyers of Elmhurst Hospital Center and Dr. Dustin Slagle of ChristianaCare.

OVERVIEW
This two-part episode is the inaugural quarterly Journal Club episode, done in conjunction with the EMRA Critical Care Committee. In these episodes we will look at landmark trials in emergency medicine.

Kicking off this new feature, host Dr. Ranjita Raghavan speaks with Dr. Chad Meyers, Emergency Medicine and Critical Care Boarded Physician working at Elmhurst Hospital Center, and ChristianaCare EM/IM resident Dr. Dustin Slagle, Vice Chair of the EMRA Critical Care Committee, about the origins of early goal-directed therapy and where we are today with resuscitation.

• Rivers Trial: Single center RCT 1997-2000 that looked at adult patients with severe sepsis or septic shock and compared using early goal-directed therapy (EGDT) to standard therapy given to patients at the time. The study found that outcomes for the patients in the EGDT group who generally received more fluids, antibiotics, and more aggressive care earlier, were significantly better than the patients in the standard therapy group.
• ProMISe, ARISE, and ProCESS: All were multicenter RCT that followed Rivers, comparing EGD therapy to usual care in patients meeting sepsis criteria as well. ProMISe also had a third group (novel therapy) that tended to be more aggressive than the EGDT group and give the most fluid of the groups. In all 3 trials, there was no statistical difference between the standard care groups and EGDT.

KEY POINTS
The Rivers Trial made a huge impact on the kind of care emergency physicians could and should deliver. It caused us to start treating patients aggressively within the ED, not wait until they got to the ICU.

The following trials all did not have significant results between the EGDT group and standard care groups. One might think, then, that Rivers was possibly wrong about his idea of EGDT, but actually what it showed was the impact of the Rivers Trial on what we now considered to be standard care. Standard care had become much more aggressive with large amounts of fluids, early antibiotics, and use of pressors and blood to support patients meeting sepsis criteria.

Now, we still have more to learn in terms of resuscitation but generally, early detection of sepsis, early antibiotics, and clinician-directed resuscitation and pressor selection to target a MAP > 65 are the current best-evidenced interventions.

When deciding how much fluid to administer, you should be ready to provide aggressive critical care once a patient is determined to be septic and you have considered all the possible diagnoses and determined fluids would help. Also, consider giving pressors early while still fluid resuscitating to support blood pressure.  However, take a holistic approach to the patient; do not just follow a strict bundle. Decide if the patient is fluid tolerant by using your physical exam and your ultrasound to do a RUSH exam and look for signs (is the IVC is collapsible? is the heart hyperdynamic?) to guide you. Then, remember to reassess the patient often for euvolemic state.

RIVERS TRIAL

• Single-center RCT (Henry Ford, Detroit) 1997-2000
• N=263
• Adult patients with severe sepsis or septic shock (SIRS plus sys < 90 after 20-30 mL/kg or LA > 4)

Excluding: pregnancy, ACS, arrhythmia mediated shock, GI bleed, seizure, OD, trauma/burn, CVC contraindicated, cancer on chemo, immediate surgery, DNR, iatrogenic immunosuppression, status asthmaticus

Groups

• Early-goal directed therapy (n=130)

• Early ABx but left up to physician
• A-line, CVC
• CVP 8-12mmHg, achieved with fluid boluses, 500ml q30mins
• MAP > 65mmHg vasopressors or vasodilators if MAP > 90mmHg
• ScvO2 >70%, achieved with packed RBC transfusions to maintain Hct > 30%, then dobutamine started at 2.5 mcg/kg/min and increased until max 20 mcg/kg/min until ScvO2 >70% if necessary but stopped if patient MAP dropped or HR > 120
• UOP > 0.5 mL/kg/hr
  • Standard therapy (n=133) 

Primary outcome: In-hospital mortality

• 5 vs 46.5 P < 0.009

Secondary outcomes

• Sudden CV collapse 10.3 vs 21% p < 0.02
• MOD 16.2 vs 21.8 (p=0.27)
• 28d mortality: 33.3 v 49.2 p=0.01, 60d mortality: 44.3 v 56.9% (p=0.03)
• EGDT did have more patients above MAP 65 for initial 6 hours, more got SVO2 above 70 (95 v 60), and have more patients meet appropriate UOP/MAP/SVO2 all by 6 hours (86 v 99)
• Standard care had lower CVPs and persistent base deficit at 6 hours but LA was similar
  APACHE II, SAPS II, and MODS higher in standard care 7-72 hrs

Strengths

• Incredible NNT (6)
• Physiologically reasonable
• Replicated in other studies

Criticisms

• ~half patients enrolled on elevated lactate and not resistant hypotension
• Bundle of interventions—impossible to determine which caused result
• CVC and aline is arguably not standard of care
• Blood transfusions shown to increase mortality in 1999 TRICC trial
• EMShockNet did not support SVO2 monitoring
• CVP correlates poorly to volume status
• Control group with above average mortality
• 3 large multicenter RCTs that followed with conflicting findings

ProCESS

• Multicenter RCT of academic hospitals with EDs > 40k (31 sites), USA, 2008-2013
• N=1,343
• Adult patients with severe sepsis or septic shock (SIRS plus sys < 90 after 20 ml/kg over 30 min (after 4/10 changed to 1L/30min) or LA > 4) Enrollment w/in 12 hrs of arrival and within 2 hours of shock/LA

Excluding: stroke, ACS, acute pulmonary edema, status asthmaticus, "major" cardiac arrhythmia, active GIB, seizure, OD, burn/trauma, immediate surgery, CD4 < 50, DNR, refuse blood, pregnant, transferred

Groups: 3 (whereas ARISE had 2)

• EGDT (n=439)
• CVC (not a-line)
• CVP 8-12mmHg, achieved with fluid boluses, 500 mL q30min
• MAP > 65mmHg vasopressors (not specified) or vasodilators if MAP > 90mmHg
• ScvO2 > 70%, achieved with packed RBC transfusions to maintain Hct > 30%, then dobutamine started at 2.5 mcg/kg/min and increased until max 20 mcg/kg/min until ScvO2 >70% if necessary but stopped if patient MAP dropped or HR >120
• UOP > 0.5 mL/kg/hr
• Novel Protocol (n=446)
• "Adequate IV access" 2 18g IVs or CVC if not possible
• O2 +/- intubation
• 500-1000mL bolus with min 2L if Sys<100 or SI>
• mIVF 250-500mL/hr
• Repeat protocol if hypoperfused
  • Usual care (n=458)—aside from no SVO2 monitoring and couldn't have a site protocol

Primary outcome: All-cause mortality

• 21 v 18.2 v 18.9 (p=0.31-.089)

Secondary Outcomes

• No difference in all-cause mortality at 90d and 1 year
• EGD and novel protocol with trend to worse CV outcomes—not significant—days of support the same
• Novel protocol with more renal failure, others equal—no difference with RRT days
• No difference across three with respiratory failure, no difference with vent days
• Discharge location percentages (ie home, SNF, LTAC) same
• Care metrics: getting abx/vent/steroids pre-randomization same across 3, novel strategy got more fluids 0-6 hrs (2.8 v 3.2 v 2.2, EGT and Novel got more pressors 0-6 (55 v 52 v 44 p=0.003), EGT got more dobutamine (8 v 1 v 1 p < 0.0001), EGT got more blood 14.4 v 8.3 v 7.5 p=0.001

Strengths

• Large sample
• Strong methodology, protocol adherence high
• Many hospitals

Criticisms

• Rivers likely influenced "standard care" and if not 10 years of progress made—lung protective strategies, glycemic control
• EGDT arm received on average 2.2L pre-randomization
• EGDT group had lower LA, higher SVO2, were younger, and had fewer major comorbidities
• Patient population not reflective of River’s population
• Again, lower than average mortality

ARISE

• Multicenter RCT (51 centers, mostly in Australia but also New Zealand and then Finland, Hong Kong, Ireland) 2008-2014
• N=1,591
• Adult patients with severe sepsis or septic shock (SIRS plus sys<90 after 1L over 60 minutes (although on average got 2.5L) or LA>4) randomized within 2 hours of meeting criteria
• Excluding: pregnancy, contraindication to blood transfusion, active bleed, transferred, cannot complete protocol or start within one hour of randomization, <90-day life expectancy prior to sepsis, CVC contraindicated, DNR, iatrogenic immunosuppression

Groups: 2 (whereas ProCESS had 3)

• EGDT (n=793)
• CVC, aline
• CVP 8-12mmHg, achieved with fluid boluses, 500 mL q30min
• MAP >65mmHg vasopressors or vasodilators if MAP > 90mmHg
• ScvO2 >70%, achieved with packed RBC transfusions to maintain Hct > 30%, then dobutamine started at 2.5 mcg/kg/min and increased until max 20 mcg/kg/min until ScvO2 >70% if necessary but stopped if patient MAP dropped or HR > 120
• UOP > 0.5 mL/kg/hr
  • Usual care (n=798)—aside from no SVO2 monitoring

Primary outcome: All-cause mortality

• 6 v 18.8 (p=0.90)

Secondary Outcomes

• no difference in all-cause mortality at 28d, in ICU, and at 60d
• EGD with shorter ED stay (1.4 v. 2 hours p < 0.001) but same in ICU and hospital overall
• Same amount and duration on vent as well as RRT
• Fluids EGDT 1964+/- 1415 vs 1713 +/- 1401 P < 0.001
• RBCs 13.6 v. 7, dobutamine 15.4 v 2.6, pressors 66.6 v 57.8

Strengths

• Large sample
• Strong methodology
• ~70% met refractory hypotension
• Many hospitals (mix community and academic) and multinational

Criticisms

• Lower than average mortality and APACHE scores leading to underpowering
• Not perfect protocol adherence—90% EGDT received CVC
• Average SVO2 in EGDT 72+/- 10
• Trickle in of EGDT to control group given Rivers

ProMISe

• Multicenter RCT (56 sites), double-blind, England, 2011-2014
• N=1,260
• Adult patients with severe sepsis or septic shock (2 SIRS plus sys<90 after 1L) LA>4

Excluding: pregnancy, CVA, ACS, pulmonary edema, asthma, arrhythmia, seizure, OD, burn/trauma, GI bleed, immediate surgery, AIDS, DNR, not initiated within 1 hour or didn’t complete 6 hours, transferred

• EGDT (n=630)
• CVP 8-12mmHg, achieved with fluid boluses, 500ml q30mins
• MAP >65mmHg vasopressors or vasodilators if MAP>90mmHg
• ScvO2 >70%, achieved with packed RBC transfusions to maintain Hct >30%, then dobutamine started at 2.5 mcg/kg/min and increased until max 20 mcg/kg/min until ScvO2 >70% if necessary but stopped if patient MAP dropped or HR>120
  • Usual care (n=630)

Primary outcome: All-cause mortality at 90 days

• 5 v 29.2 (p=0.90)

Secondary Outcome

• EGDT with higher SOFA at 6hr 6.4 v 5.6 P<0.001, 4 v 3.7 at 72h p=0.056
• Advance CV support 37 v 30.9 p=0.026
• ICU stay 2.6 v 2.2 p=0.005
• QOL similar at 90d and cost not significantly different (17.6k vs 16.2k)
• Fluids 2000 v 1784, dobutamine 18 v 3.8, RBCs 8.8 v 3.8

Strengths

• Large sample
• Strong methodology, more quickly done
• Many hospitals mix community and academic

Criticisms

• Trickle in of EGDT to control group given Rivers
• Again, lower than average mortality but higher than ProCESS and ARISE
• SVO2 70+/- 12 in EGDT group