Down with the Surviving Sepsis Campaign
Anand Swaminathan MD, Rory Spiegel MD and Josh Farkas MD
Take Home Points
- Many of the recommendations of the new Surviving Sepsis campaign bundle are not evidence based.
- There is no real evidence of benefit and these treatments are cumbersome and logistically difficult.
- There is evidence that portions of these guidelines may be harmful.
- In May of 2018, we saw the release of the Surviving Sepsis campaign bundle for 2018. Everyone has a sepsis protocol and many of the recommendations are derived from recommendations of this group. This is problematic because sometimes these recommendations are not evidence based. Weingart discussed this in May.
- What was in the 2016 Surviving Sepsis campaign bundle? It was an extensive document that essentially involved a 3 and 6 hour bundle.
- The 3 hour bundle included 4 requirements; measuring a lactate level, obtaining blood cultures prior to administration of antibiotics, initiating broad spectrum antibiotics and administering 30cc/kg crystalloid bolus. This was for patients who were either hypotensive or had lactate greater than 4.0
- The 6 hour bundle was less discussed but involved giving vasopressors for patients that were hypotensive with a target of a MAP of 65 mmHg or greater. If patients remained hypotensive, they were assessed for fluid responsiveness in some measure if they remained hypotensive. The lactate level was re-measured if it was elevated. One of the most difficult aspects of compliance was the repeated and documented focused exam which included two of the following elements; measuring the CVP, measuring the SCVO2, having a bedside cardiac ultrasound or some dynamic assessment of fluid responsiveness.
- This is probably discussed less in emergency medicine as we hope the patient will be out of the ED by six hours. This may be why we focus on the 3 hour bundle.
- It can be difficult to complete all of these tasks within 3 hours, especially if the patient had cryptic septic shock that wasn’t easily identified.
- In the most recent update, they combined some elements of the previous 3 and 6 hour bundles into a single 1 hour bundle. Within 1 hour, we are supposed to somehow magically achieve the following; we need to check a lactate level, get blood cultures, start antibiotics and give 30cc/kg of fluid for patients who are hypotensive or have a lactate greater than 4.0 mmol/L. We need to start pressors if the patient is hypotensive despite starting fluid.
- Is there anything good about these guidelines? Logistically, it may make it easier to be compliant with the guidelines. If you look at the data on how emergency departments fail in compliance; they fail to recognize the sepsis until later and don’t complete everything in the first 3 hour bundle or they fail to document the focused re-examination in the 6 hour bundle. These are the patients who are boarding in the emergency department way beyond the three hour time limit. We have already moved on and are seeing other patients. We forget to document the bundle. By combining it into a one hour bundle, it may be easier to turn into a protocol via EHR.
- What is the downside?Just because it is good for the hospital’s bottom line doesn’t mean it is good for patients. If you look at the data, we are already terrible at initiating these treatments and completing them within the 3 hour time window. Typically, the way we make improvements is with blanket treatment protocols that ensure that everyone with even the slightest chance of infection gets placed in these sepsis pathways.
- By condensing the time allowed until completion, most emergency departments will turn to an even more rapid and non-discriminatory approach.
- We have been through this before. Remember the guidelines that we need to give antibiotics to every pneumonia patient within 4 hours? That was a disaster. It led to incomplete diagnosis, increased pressure and inappropriate antibiotics. There are some reports of hospitals that had C. difficile outbreaks linked to these pneumonia pathway packages. These recommendations were harmful and were eventually withdrawn.
- The current 1 hour timeline for sepsis is even worse.
- We rarely discuss trade offs when examining these guidelines. We only look at the subset of patients that may potentially benefit from this type of aggressive care. But by being forced to treat these patients in the time allotted, we are forced to treat a larger cohort of patients who likely don’t have sepsis and almost certainly don’t benefit from this care. We never consider the logistical burden these guidelines place on the emergency department as a whole and how it distracts from the care of other patients.
- There is little to no evidence demonstrating that this will have any benefit on our patient outcomes.
- The studies cited as the basis for these protocols have numerous flaws. They are often before and after studies on the implementation of a bundle or studies that look at harms when the bundle wasn’t completed on time. None is a randomized controlled trial. None has high quality evidence. All have so many inherent biases that it reduces the meaningfulness of their findings.
- This data is suboptimal. Retrospective studies correlating time to intervention with outcome are junk. These studies mostly measure confounding factors. For example, patients getting better care overall are going to receive faster treatment. That doesn’t prove causality. The fact that the guidelines are based on this sort of evidence is bizarre.
- Even worse than the fact that there is no real evidence of benefit and these treatments are cumbersome and logistically difficult, there is mounting evidence that portions of these guidelines may be harmful. There have been multiple studies demonstrating the detrimental effects of aggressive fluid resuscitation. Given these data, the continued demand for an empiric 30 cc/kg fluid bolus that is mandatory even if the treating clinician thinks it is harmful, is intolerable.
- These guidelines are out there. They going to be thrust upon us. We need to know about them so we can push back. We need to understand why these are not just unattainable goals but also not good for the majority of our patients. Protocols are good when we don’t know the best care or for people who don’t know what they are doing. Emergency physicians who practice in emergency departments are experts. We need to move outside of protocols and say that our patient doesn’t fit them.
Recent Related Material
EMRAP 2018 May - SNACK - Surviving Sepsis
Ian L. - October 2, 2018 7:17 PM
The utilisation of Healthcare prior to Sepsis diagnosis Hospitalisation by LIu VX et al in Critical Care Medicine 2018 .
They concluded that over 45% of sepsis patients had clinician based encounters in the week prior to hospitalisation.
Therefore it ought be concluded that there is a "PreSepsis Incubation Time " .
LiuVX et al stated in their conclusion that these PreSepsis encounters offered potential opportunities to improve recognition risk stratify and treat prior to hospitalisation that one might conclude could well reduce the incidence of current Sepsis Morbidity and Mortality and deserves
investigation .
Crit Care MED 2018 April ; 46(4) 513-516 Liu VX et al .
Patrick S. - October 4, 2018 1:47 PM
Surviving Sepsis Campaign is currently recommending hospitals do NOT implement the 1 hour bundle due to vigorous opposition of clinicians:
http://www.survivingsepsis.org/News/Pages/SCCM-and-ACEP-Release-Joint-Statement-About-the-Surviving-Sepsis-Campaign-Hour-1-Bundle.aspx
Hopefully this is the first crack in the dam...
Ian L. - October 4, 2018 2:26 PM
To diagnose blood stream infection in Adults Low Supine Systolic Blood Pressure less than 100mmhg in the context of an infection or a drop from normal systolic Blood Pressure of 20 mmHg is indicating SBI .
Drop between supine and Standing systolic Blood Pressure don't seem to have time to measure in the chaos of triage .
In Pneumonia Severity Index Scores a Respiratory Rate greater than25- 30 per minute scores danger but in qSOFA it's Respiratory Rate above 22
Primary Care Doctors ought test the vital signs in sicker looking adults or adults at risk by age Comorbidities and Reduced Immune Competence to diagnose blood borne infections before sepsis
In a Study by Yaazoe M et al of Nursing Home patients with pneumonia comparing positive .blood culture patients with negative blood culture patients the Odds Ratio for Systolic blood pressure <= 90mmhg was 6.03 95%CI 1.06- 34.25 .
Respiratory Rates Pulse Rate and Po2 oximetry or Altered Mental Status was not mentioned .
Altered Mental Status could be graded in a discriminatory way : Months of the Year Backwards Serial 7s Does a stone float on water? like Questions.
Reference :Yamazoe M et al Journal of Infection and Chemotherapy Dec 29 2017.
The British Royal College of Physicians developed theNational Early Warning Score (NEWS) based on six physiological parameters for primary care providers prehospital ambulance services and hospitalised patients .
Regular Review of Signs and Symptoms is also important like serial ECGs and Serial Troponins in ACS .
The recommended Physiological Parametres :
Respiratory Rate
Oxygen Saturation
Temperature
Systolic blood Pressure
Pulse Rate
level of consciousness
But Not some Circulatory Parameters Skin Color Skin Turgor Tears Eyes Mucous Membranes Capillary Refill.
Rory S. - October 5, 2018 11:20 AM
Thanks for the comments Ian. While Liu et al present interesting data, I think that is a very dangerous proposal. It only looks at the few patients who end up having sepsis. What about the thousands of patients that present with simple infections that will never go on to develop into sepsis? Imagine how many patients would receive needless antibiotics in the hopes on intervening on the few that actually end up having sepsis. Not only would such a screening screening cause an overwhelming burden on the healthcare system, we have no idea whether this type of early intervention has any effect on the downstream consequences of sepsis.
Ian L. - October 6, 2018 7:52 AM
Thanks for your careful attention to the dangers of slippimg into knee jerk over treatment .
But I believe it will not be a policy of treating simple infections with antibiotics to prevent sepsis .
There is the concept of the Early Warning System of vital sign abnormalities that portend danger .
In the UK a National Early Warning System Study has developed Red Flag Signs and Symptoms that alert health care .p rofessions to initiate careful review .
In adults in the context of suspected infection feeling "Off" with also hypotension Systolic BP below 100mmHg Tachycardia Pulse rate at rest above 100 Tachypnoe Respiratory Rate above 21 Altered Mental acuity and Pulse oximetry Po2 below 94 alert the physician to the need to regular review of the signs and Symtoms indicating the descent into serious illness that will need escalated attention and intervention .
Rory S. - October 13, 2018 6:16 PM
Thanks for all who have listened and commented. Just a brief correction which was pointed out by one of our listeners (thanks Shannon). In the episode we state that the previous 3 and 6-hour bundles originated from the 2016 surviving sepsis update (Rhodes et al., Intensive Care Med, 2017). This is in fact incorrect. The 3 and 6-hour bundles first appeared in the 2012 guideline update (Dellinger et al., Crit Care Med, 2013), and the 2016 guideline did not publish any revisions to these bundles. The 1-hour bundle discussed in this segment is intended as the revision to the 3 and 6-hour bundles that best reflects the 2016 updates to sepsis care.
Anand S. - October 15, 2018 6:32 AM
Via Rory:
Thanks for all who have listened and commented. Just a brief correction which was pointed out by one of our listeners (thanks Shannon). In the episode we state that the previous 3 and 6-hour bundles originated from the 2016 surviving sepsis update (Rhodes et al., Intensive Care Med, 2017). This is in fact incorrect. The 3 and 6-hour bundles first appeared in the 2012 guideline update (Dellinger et al., Crit Care Med, 2013), and the 2016 guideline did not publish any revisions to these bundles. The 1-hour bundle discussed in this segment is intended as the revision to the 3 and 6-hour bundles that best reflects the 2016 updates to sepsis care.
Michael M., MD - November 22, 2018 3:39 AM
In cases where the lactate has cleared with the initial fluid bolus, I'm struggling with the concept of chasing hypotension with pressors, CVL's, and enormous volumes of fluid beyond the initial 30 cc/kg. If lactate is the final common pathway, and the best measure of end-organ hypoperfusion, and the lactate clears, are pressors still needed anyway? My understanding of this literature (I'm probably grossly oversimplifying here) is that the only consistent predictors of improved survival in sepsis are 1) time to antibiotic administration; and 2) whether the patient had fever in the ED (Cochrane review). A case I saw recently may help clarify my question:
72 yo man (vasculopath with known AAA, and recent outpatient antibiotics for UTI) whose only symptom was a rigor and a feeling of impending doom at home two hours PTA, rectal temp of 102.6 in ED, HR of 100, initial BPS of 90. Ordered 30 cc/ kg bolus, Ceftriaxone for presumed UTI, first lactate was 3.6. After bolus and about one hour later, lactate was down to 2.3 and BP had come up to 100. Awake, alert, and comfortable sitting semi-recumbent throughout (i.e. he looked much better than his numbers did). About that time he went into narrow SVT at about 150s, pressure dropped into 80's, but without very little change in how he felt. I then followed the unstable SVT algorithm and sedated/ tried to cardiovert him at 50j without success. He exclaimed "don't ever do that to me again". Gave adenosine total 18 mg without success, then I cajoled him into another attempt at cardioversion, this time at 150j. He converted to Afib/ RVR in the 140's. Through all this his BP never got above 90, but he didn't look bothered much by it. I didn't give any more fluid at that point--my concern was that he was hypotensive due the SVT, and was hoping if I converted him to SR the BP would come up. But no such luck, he didn't convert to SR, and pressure stayed in the 80's. So I put in a CVL, started pheynylephrine, sent him to ICU, and pressure gradually came up over the next several hours. But was it the phenylephrine, or just "tincture of time"? My gut told me to be permissive about the blood pressure, but the protocol, and the nurses, said otherwise. I've had a fair number of these cases, so any evidence that would shed light on this issue would be appreciated.