Paper Trail

Expert Commentary: ARISE and ED Sepsis Resuscitation

As we discussed on our last Paper Trail post, the results of the ARISE and ProCESS trials seem to indicate that the most meaningful interventions for a patient in severe sepsis/septic shock are early antibiotics & source control and aggressive resuscitation in an appropriately-monitored environment. 

At our shop, we're currently working on analyzing our own sepsis response data and designing new protocols. In these discussions, some have expressed concern over the potential for giving too much fluid to these patients. The concepts of "fluid responsiveness" and "fluid tolerance" are the subject of much debate in the FOAMed critical care world these days. 
(See the discussions between Drs. Marik and Weingart over at EMCrit for reference.)

Our own ED Critical Care guru, Dr. Christopher Holthaus, stepped up to the plate and offered his views on fluid resuscitation of septic patients in the ED:




"The fluid issue can be challenging. Although some studies suggest that an early restrictive fluid strategy potentially may bear a signal [toward improved outcomes], the preponderance of current ED-based RCTs involving patients with severe sepsis and septic shock will have up-front ED volumes of 4-6L and are showing the best mortalities to date (approx. 18%). [See comparative table below delineating up front/early fluid volumes PROCESS vs ARISE -- Ed.] 

"Although any patient is at risk for volume overload during a resuscitation, patients in these studies appear to have tolerated these volumes relatively well with low adverse events as far as acute pulmonary edema (< 0.02%). We know patients do worse if they are dry and get placed on pressors. Somewhere in between probably lies the truth. Perhaps better exact timing of fluid administration when the heart is truly volume-responsive (i.e., on the ascending portion of the Starling curve) will be the way to go. It remains to be seen, but this approach hypothetically may also show the same overall total volume in the end...     

"Nevertheless, perhaps in the meantime a more judicious rule of thumb is to discriminate "early vs. late" in the resuscitation, and to acknowledge that patients with distributive shock from sepsis generally have trouble maintaining intravascular volume and/or blood pressure during SIRS-related events. Unfortunately we don’t have any great immediate fixes for that yet besides giving volume (or pressors), treating the source, and supportive care. 

"Providers dealing with these patients early in their presentation (i.e., first 6-8 hours) are supported in administering total fluid volumes of 4-6 L (Rivers’ EGDT, PROCESS, ARISE). The RCT using an ED early restrictive fluid strategy for these patients does not exist yet. In current practice, most providers will bolus the first 1-2 L on presentation and perhaps another 500-1000 mL "PRN" based on various factors -- fluid tolerance, tachycardia, hypotension, urine output. The argument for a maintenance rate in these patients is supported by Figure S2 from the PROCESS trial [shown below - Ed.] which used maintenance rates of 250-500 mL/hr once a SBP > 100 or Shock Index < 0.8 were achieved with boluses. This point alone was perhaps one of the best contributions from PROCESS in regards to demonstrating a proven objective alternate fluid administration strategy which went beyond the original use of CVP for quantitative resuscitations. 

"Because the ED currently does not have a go-to gold standard outcome study for gauging volume responsiveness via dynamic changes in stroke volume in these patients, we are left with the fluid volume quantity and rates which stands as the current best available evidence (i.e., 4-6 L in first 6-8 hrs parsed between boluses and maintenance rates).

"In the end, of course one size does not fit all and clinical judgment individualized to the patient is always warranted."

Adopted from ARISE & ProCESS data by Holthaus CV.




Protocol for Standard Therapy arm of ProCESS trial (Supplmental Figure S2)


What do you think? Is fluid overload from ED resuscitation in the "early" phase of sepsis a legitimate concern? When do you pull the trigger and start pressors? Let us know what you think!


Submitted by Christopher V. Holthaus MD.
Copy editing C. Sam Smith, PGY-3.



FOAMed Digest #6 Update: The Paper Trail

Today, we'd like to introduce a new section to the Everyday EBM FOAMed Digest, which will be published as a standalone post. It may seem a bit counterintuitive, but each week we're going to draw your attention to some of the latest-and-greatest papers from the primary EM literature. Most of these will not be open-access, but we hope you can still access them via the medical library at your institution. A lot of the discussion in the FOAMed world centers around recently-published papers from EM journals. The EM trainee should view FOAMed as a tool to assist with understanding and analysis of this literature, as a forum for discussion and further discovery -- not as a replacement for reading the papers for his/her self. We'll provide you a very short summary here and link you to relevant FOAMed resources if applicable.




1) This week, it's impossible to hold any discussion of EM literature this week and not start with the ARISE trial. Hot on the heels of ProCESS, the results of this trial appear to confirm what resuscitationists have long suspected: in terms of sepsis care, you don't have to do a lot of s***, you just have to give a s*** (H/t Dr. Weingart).

Briefly, this multi-center RCT enrolled over 1600 patients from 51 hospitals (mostly in Australia & New Zealand) who presented to an ED with suspected infection, 2+ SIRS criteria, and refractory hypotension (low SBP after 1L IVF) and/or hypoperfusion (lactate >4). These patients were randomized to receive EGDT (as ensured by a dedicated sepsis response team) or "usual care," which allowed for treatment at the physicians' discretion except that ScvO2 measurement was not permitted within the first 6 hours of therapy. There was no difference in the primary outcome, mortality at 90 days.

Needless to say, this publication has got the FOAMed world all atwitter (pun intended).

- The best high-yield summary with all the relevant info can be found on The Bottom Line. Also includes links to other FOAM resources, and excellent tables comparing the Rivers trial to the "holy trinity" of modern sepsis studies -- ProCESS, ARISE, and ProMISe.
- EM Lit of Note also published a short piece.
- Scott Weingart of course has an excellent summary on his latest EMCrit podcast, which has the added bonus of including references to two other recently published papers. The TRISS trial from NEJM found no differences in clinical outcomes when using transfusion goals of 9 g/dL vs. 7 g/dL. A retrospective study by Ferrer et al published in Critical Care this month again demonstrates that delayed time to antibiotics for patients with severe sepsis/septic shock increases mortality in a near-linear fashion. (The Bottom Line summary of the Ferrer study here).
- The European Society of Intensive Care Medicine posted an interview and a presentation by the lead author of ARISE, Dr. Sandra Peake.


The ARISE trial is probably more than enough to keep you busy, but a few more items for your perusal:

2) The Dalai Lama of PE diagnosis in the ED, Dr. Jeffrey Kline, is back at it again, this time with a systematic review and meta-analysis of pregnant patients undergoing investigation for PE in the ED. Seventeen studies including over 25,000 patients were analyzed, and Dr. Kline's team found a lower rate of VTE among pregnant patients compared with nonpregnant patients, with a pooled risk ratio of 0.60 (95% CI 0.41-0.87). Dr. Kline cautions, "We do not interpret these data to indicate that pregnant patients have a lower risk of PE when compared to healthy nonpregnant patients. Instead, we believe that our data illustrate that clinicians order testing at a low test threshold among pregnant patients." Well worth a read.
- Our own Captain Cranium, Dr. Chris Carpenter provided a commentary to this piece, astutely summarizing the controversy of PE overdiagnosis.
- Sarah Sanders, a 4th year medical student (!) posted an excellent summary of this paper and other current research to the EM Curious blog.

3) While this next paper is perhaps not of the most relevance to the EM trainee, its publication should be a source of pride for all of us here at Wash U. The results from the Contraception CHOICE Project, conducted entirely here in St. Louis by the Washington University OB/Gyn Department, were published in the NEJM this week. Through private funding, the investigators were able to provide young women enrolling in the study their choice of birth control method, with special emphasis on Long-Acting Reversible Contraception (LARC) such as implants or IUDs. Without cost as a barrier, 72% of the women enrolled chose a LARC method. Women utilizing LARC methods reported unintended pregnancy at a far lower rate than the national average, especially for those age 15-19 years. Really great work.

Never stop learning,
Sam Smith, PGY-3