Podcast 120 – The ProCESS Trial with Derek Angus

process-trial

The ProCESS trial was published less than a week ago (The night before my SMACC lecture on severe sepsis–dohhh!)

If you have no idea what I’m talking about, climb down from your mountain-top monastery, find a damn iphone and read this:

ProCESS Trial in the NEJM

As soon as I returned to the states, I begged and pleaded with the study author, Dr. Derek Angus, to give us his thoughts, he kindly acceded.

Derek AngusDr. Angus is chair of the Department of Critical Care Medicine and Distinguished Professor and Mitchell P. Fink Endowed Chair of Critical Care Medicine at the University of Pittsburgh Schools of the Health Sciences and UPMC Health System. His accomplishments are too numerous to list here, so check out his bio page.

I will be adding a ton of stuff to this page regarding the trial ASAP; for now I just wanted to get Dr. Angus’s interview up on the site.

Most Important Table

Updated Table S4

Updated Table S4

 

Other Stuff Mentioned

Supplementary Material for ProCESS (this is the corrected version)

High versus Low Blood-Pressure Target in Patients with Septic Shock

Excellent Posts in the FOAMcc World

Now on to the Podcast…

Play

You finished the 'cast,
Now get CME credit

Already an EMCrit CME Subscriber?
Click Here to Get CME Credit for the Episode


Not a subcriber yet? Why the heck not?
By subscribing, you can...

  • Get CME hours
  • Support the show
  • Write it off on your taxes or get reimbursed by your department

Sign Up Today!

.

Subscribe Now

If you enjoyed this post, you will almost certainly enjoy our others. Subscribe to our email list to keep informed on all of the ED Critical Care goodness.

This Post was by , MD, published 7 months ago. We never spam; we hate spammers! Spammers probably work for the Joint Commission.

Comments

  1. Dr Doug Lynch says:

    Scott!

    Great work. Amazing the speed at which the study was published. Leading to an error in this case. (Hmmmm)

    It makes you wonder if all this real time Blog Podcast and #FOAMed stuff is beginning to influence how certain articles are fast tracked by even the NEJM?

    Are the two other related trails ARISE and PROMISE close to ready?

    Or is it just that this study and the whole EGDT/Surviving Sepsis story is the medical equivalent of the next Royal baby in England or the next Mrs Tom Cruise? It’s definitely PrimeTime!

    One way or the other; even as we abandon EGDT (as it was originally conceived) let’s remember that Manny Rivers (& Co.) started a process that has changed the world of critical care. I would argue that Rivers has directly or indirectly radically improved sepsis care, especially outside the bigger centres, outside the ivory towers!

    I never followed the EGDT protocol. (I work in Australia where it was never completely embraced.) However I think the fuss around his single centre study and the debate and the fall-out and the controversy and the scandals and the hundreds/thousands of journal club presentations have possibly saved thousands of lives around the world.

    Not many people get credited with that sort of thing.

    Thank you Manny.

    Just my opinion.

    (You did ask what we thought.)

    Doug

    @TheTopEnd

    (Care of lifeinthefastlane.com)

    PS

    No mention of the great Jimmy Cliff song?

    Many Rivers to Cross, released in 1969 on Trojan records.

  2. Neil Perera says:

    Sir, is there a description of what “usual care” actually is? With all of us aware; perhaps using EGDT; it seems hard to think that it would vary much from the standard care arm. IVF and early antibiotics; pressors prn. Just a thought. Thanks for all that you do for our profession!
    Neil Perera,MD
    Emergency Medicine
    US Army (community based practice)

  3. Gavin denton CCP uk says:

    Hopefully the death of blood transfusion as a standard in sepsis, blood is valuable and increasingly scarce in the UK resource. We stop wasting our time with dobutamine, now we just need to persuade cardiologists that NORAD is not the devils spawn.

    • we moved away from transfusion until they hit 7 g/dl years ago; as to dobutamine not sure where the hatred comes from in UK and Australia–prob. b/c it was crazy expensive at what time. It is a small group that we would even argue about for the drug anyway, so not worth a hullabaloo

  4. Kyle Gunnerson, MD says:

    Scott, Thanks for the interview with Derek. Overall I think you did a nice job at identifying some practical ED resuscitation issues when interpreting this trial. A few thoughts:

    I think the fact that these centers had a control group mortality of around 20% leads me to think that they were so good at general resuscitation the sample size would have to be huge (if even calculable) to show any difference at all.

    Maybe 20% mortality is now our goal for septic shock. If I remember correctly, the Surviving Sepsis Campaign has many more centers and exponentially more patients, and even though they are showing a very large improvement in mortality reduction, they are still reporting mortality in the high 28%-30% range.

    I wish you would have grilled Derek harder on what endpoints he would use – the physical exam – really? I don’t think so. I know Derek’s not titrating fluid bolus to skin turgor or the hepato-jugular reflex. If not CVP then what? Pick something other than this.

    In the end, Derek mentions that this is a study of “means” and there will still be many patients that will require tailored resuscitation. If you place a central line for pressors and find that the ScvO2 is 45% are you going to ignore it?

    Overall nice job Scott. Enjoyed the podcast.

    • classic poop sandwich brother : )

      15,000 pts in NYC are hovering around 20%, I think that’s the standard for average urban centers. Derek said offline his practice looks very much like the standardized group (hybrid) which also matches our NYC experience. The absence of ultrasound makes me sad, but it would be hard to integrate in standardized fashion–wonder how many in usual care were guided by this.

      • The one thing that really hasn’t come up, which probably will when you speak with Dr. Rivers, is the fact that the initial trial (which started almost 15 years ago) used a physiologic endpoint of ScvO2 to represent a nice balance between an oxygen delivery approach to alleviate oxygen debt. If you don’t have a delivery dependent state of shock, then increasing delivery should have no benefit and only carry the risk of harm each intervention has.

        Unfortunately without some objective physiologic endpoint (lactate isn’t it) which will help you determine whether or not your patient is in a delivery dependent state, you are essentially guessing. A cardiac echo won’t tell you, a PA cath won’t tell you (unless you send the SvO2). One is just banking on the fact that everything is preload dependent. The initial EGDT trial was designed to rapidly identify and reverse oxygen debt, with the same strategy that the good intensivists at Pittsburgh used to teach Dr.Rivers. They were still teaching this when David Huang and I were there in early 2000 (a little ironic isn’t it).

        Again, I think there was a very strong Hawthorne effect at these sites. This may be the “little secret” that surfaces. They were good to begin with, very enthusiastic with an ER champion who wanted to do a good job for the great people at Pitt. This enthusiasm improved the entire care of everybody. Not a bad thing, just my speculation.

        I like Scott’s suggestion that I’d like to see if this strategy of “status quo” works at smaller EDs with higher baseline mortality rates.

        • Yep, fantastic point. I think Manny’s and my own population were far more likely than other centers to wait to arrive at the hospital until they are way into the supply dependence portion of their illness course. Most of the other trials, including Jones, showed there would have been a minimal impact in care regardless of ScvO2. ProCESS certainly showed a markedly lower population who required blood than EGDT trial–which may represent different patients or the fact that some patients got blood for pragmatic reasons rather than ScvO2 indices in the original trial.

  5. Scott,
    Great stuff as usual. We never did get the answer to the dobutamine question!!! I still use it if the LV isn’t contracting well. Haven’t used Milrinone in this setting.

    I particularly appreciated the closing part as I don’t think this is a repudiation of EGDT so much as a validation of “there’s more than one way to skin a cat”
    Mike

    • same with me with dobutamine. I had to cut it, but Derek stated if a fellow showed him an echo with a hypodynamic, filled heart; he would consider dobutamine.

  6. Hi Scott,

    Great interview!
    A quick question (or a bit of a hijack) regarding the Caironi paper on albumin replacement. Do you think that total protein would have an impact on outcomes different from that of albumin, or is only albumin likely going to say the exact same thing?
    Cheers,
    Kyle

  7. Sean Marshall says:

    Hey Scott,
    My biggest take-away from the podcast was the point that PROCESS doesn’t investigate the ABC of early sepsis care but the DEF. The patients in all groups were screened and identified early, got early antibiotics and an early couple liters of fluids. My feeling is that ABC part is the most important to be systematic and protocolized, and if that is done well, it’s 80% of the battle won.

    Understanding that helps me jive how PROCESS’ equivocal findings fit in with hugely successful quality improvement projects like yours in NY.

  8. Michael Toolis ED Registrar Melbourne with no affiliations. says:

    Great stuff – Thanks Scott and Derek.

  9. dave barounis says:

    Hey scott,

    Thanks for putting this together!

    You state in your podcast that the amount of fluid received in the first 6 hours (T0 = randomization) was similar. However, when you look at the appendix the usual care arm received ~ 2300cc, the PSC arm received 3300cc (a liter more on AVG), and the EGDT arm around 2800 (500cc more).

    When looking at these numbers it may seem like a liter isn’t “that much more fluid” but as Dr. Angus mentions that is a 50% increase compared to the PSC arm in MEAN VALUES. Mean values cannot tell you about individual patients (regression to the mean) and if the average patient got about a liter more of crystalloid maybe the outliers were the patients who got 3 or more liters more of fluid and those patients were the ones who went on to require cvvh for volume management in the ICU (since more patients developed “renal failure” in the PSC arm).

    Maybe this excess volume explains why these patients were more likely to require CVVH, be slightly more acidemic in the first 6 hours, have slightly higher INRs and have a trend towards developing more acute cardio-respiratory failure.

    Also I am certain the reason there were so few patients getting PRBC/dobutmaine tx in the EGDT arm compared to the original rivers trial is because in that original trial the avg starting ScVO2 was 49% versus 71% in the process trial.

    Anyways, its good school for thought and goes back to your prior podcast with paul marik about giving sufficient fluid resuscitation.

    Cheers,

    Dave

    • drfasthugs says:

      Hi Dave, If you add the pre-randomisation and randomisation to 6 hrs volumes then you can see where Scott gets the figures from 5l vs 5.5l vs 4.5l….

  10. jeremysfaust says:

    Great interview! First, I think it is fantastic that you were able to suss out for us the transposition error in the supplemental figure and I hope the print version of the paper does not reflect the error–another example of why print publication should really be banned someday…

    Second (a smaller point), you mentioned that the EGDT group got dobutamine 8% vs around 1% in the other groups (and that’s the first 6 hours, on the table you link above) and that you didn’t think this was a very big difference. This is just a matter of semantics/opinion, but to me 1:100 vs 1 in around 12.5 is reasonably different, from the perspective of residents with a lower n (i.e. not working ICU/Resusc shifts every single day). It the difference of using that medication maybe once per week or less, as opposed to a few times per year. Really changes comfort/experience with a medication for residents like me if I only use a med a few times during my training as opposed to going to it constantly. So while I don’t mind not having to use dobutamine, if this trial had come out differently and all of a sudden we all really had to do EGDT to a T, that would probably change my landscape significantly with respect to that particular drug. (And on a related note, I would have gotten more central lines, which I need, but that is besides the point and I’m certainly glad our patients do not need to be subjected to quite so many…)

    Thanks again -Jeremy

  11. Faheem Guirgis says:

    THE MAIN PIECE OF DATA MISSING FROM PROCESS IS THE PERCENT ACHIEVEMENT OF SCVO2 > 70% (OR 66% USED IN PROCESS) IN THE EGDT GROUP. IF THE BENEFIT CONFERRED BY EGDT IS THE ACHIEVEMENT OF SCVO2>70 (AS THAT WAS THE ONLY DIFFERENCE BETWEEN THE EXPERIMENTAL AND CONTROL ARMS IN RIVERS’ ORIGINAL STUDY) WE NEED TO KNOW HOW MANY OF EGDT PATIENTS IN PROCESS MET THAT GOAL. IN RIVERS STUDY ALMOST 100% OF PATIENTS IN THE EGDT GROUP MET THE SCVO2 GOAL VERSUS 60-70% IN THE CONTROL ARM.

    DEFINITELY APPRECIATE THE DISCUSSION WITH ANGUS. THANKS!

    • Faheem Guirgis says:

      To address my own comment – I found in the text that the mean (+/- SD) SCVO2 after catheterization in the EGDT group was 71 +/- 13% (page 6 above Table 1 of Process study). Therefore, Process does not debunk SCVO2 measurement particularly in the sickest of patients as the SCVO2 goal was probably not achieved in a large proportion of patients receiving EGDT.

      • Not so fast, here is Dr. Angus’s reply:

        Hi Scott,

        So, a quick peak at the data suggest about 88% of patients in the EGDT arm get an ScvO2 up to >70%, and 93% get an ScvO2 up to >65% in the first 6h. Of the remainder, some died before hour 6, some requested their data be withdrawn, some got a good ScvO2 after the first 6h, some just had no more ScvO2s recorded, and, presumably, in some, the docs just couldn’t get the ScvO2 fixed in 6h despite best efforts.

        Thus, not 100%, but darn close.

        We have no idea what happened in the other two arms because serial ScvO2 was only measured in 3.5% and 4% of cases.

        Of note, I’m not sure I fully understand what 100% means in the Rivers study. In my own experience, I’m not sure I could guarantee ‘fixing’ ScvO2 in 100% of all patients with bad shock, even in the ICU, and even with blood, inotropes, and fluid. I thought some of the patients in the Rivers trial died in the first 6h – did they die with normal ScvO2?

        Hope this helps.

        • Faheem Gurgis says:

          Just wanted to also state that I really appreciate Dr. Angus’ time to answer these questions. I have the utmost respect for him. Process was an excellent study and will definitely change the way these patients are managed across the world.

          Thank you for the reply. I think understanding the SCVO2 goal is the key to determining whether classic EGDT works. I wish that data would have been available in a table or supplement somewhere. Any chance that could happen?

          From the Rivers’ study, 94.9% in the EGDT group and 60.2% in the control group met the SCVO2 > 70 goal at 6 hours. That was the only discernible difference between the two groups. So yes, not 100% but close. From their data, it seems that there was a slow downward trend in SCVO2 in both groups but the downward trend in SCVO2 was more steep in the standard care arm (at 7 to 72 hours 70.4 +/- 10.7 for EGDT vs 65.3 +/- 11.4). So I’m guessing (again, just guessing) that they didn’t die with a normal SCVO2 but died with a lower SCVO2 after it was initially normalized in that first 6 hours.

          • Joseph Bove says:

            Hi. Thanks for the great review here. I see that up to 6 hours a little more than 50 percent received central lines in the protocol and usual care groups. Are the numbers available to see how many patients received central lines after the 6 hours. It would be nice to see if care continued in the intensive care units without the need for the line as well. Thanks in advance for comments.

  12. Hi Scott,

    Great interview. Just got around to hearing it now.

    A couple of points:

    I am quite taken aback at this “transposition” error (that’d be a copy&paste error to you and me). Not so much that it happened, because God only knows how many times I have made a mess of a table on Word or deleted the wrong row on Excel when I am rushing to put together a talk or an abstract at 3 in the morning. It’s more that I am surprised that one of the investigators was putting together a crucial table/doing a crucial bit of stats late late late the night before. This is a massive multi-centre trial, and I would have thought the process around the analysis of the data and dissemination of the results would be more rigidly performed than that. Makes me wonder if there was anything else done in a rushed manner. (Sorry if that’s being a bit critical!)

    I would have liked to hear a little about the reasons for stopping the trial early. They used the mortality figures from the original Rivers study as their expected mortality rate to guide the original sample size calculation, and they quote the Jones paper from 2010 as being in keeping with the actual mortality they found. But wasn’t there evidence since 2001 that the mortality in the Rivers study was quite high compared to average. Did these authors have a baseline data from this network of study sites on what their expected mortality should be, instead of using Rivers results for their sample size? I agree that even had they kept going to the end that it is likely there still would not have been a difference between the 2 groups, I just don’t like seeing trials stopped early unless absolutely necessary.

    Thanks again

    John Cronin, EM trainee, Dublin, Ireland

    • John,

      High-level academic research has been aptly compared to a sausage factory–trust me, you do not want to know what goes into it to make the final product.

      As to the 2nd point, when you speak the SSC people, they feel the RIvers’ mortality was actually real so I get why the ProCESS guys went with that even though it seemed silly to us.

      I also hate trials getting stopped early unless the results are overwhelming and will be accepted by all.

Trackbacks

  1. […]       Join the larger conversation. Read the paper, the editorial, and the excellent online commentaries (Like this one http://academiclifeinem.com/process-study-identify-sepsis-early-treat-aggressively/) . Discuss it at your local journal club. Get online and blog about it. Join Twitter and Google + and continue debating how we can further improve sepsis care. Listen to podcasts- Scott Weingart of emcrit.org has interviewed Derek Angus  (one of the key players in the study) -listen to it here http://emcrit.org/podcasts/process-trial/ […]

  2. […] Excellent podcast by Scott Weingart interviewing ProCESS study author Derek Angus can be found here. […]

  3. […] Weingart has interviewed Dr Angus (lead author) of ProCESS on Emcrit – you will have to wait for the SMACC podcast to hear his talk from a few weeks […]

  4. […] Podcast 120 – The ProCESS Trial with Derek Angus (EMCrit) […]

Speak Your Mind (Along with your name, job, and affiliation)