I spent the afternoon wading through the proposed CMS sepsis measure. Anyone who has any experience with national quality measures understands just how painful a way this is to waste a few hours. Here is the issue:
You and I have a pretty clean definition of Severe Sepsis and Septic Shock:
- A suspected infection
- with >=2 of the SIRS criteria
- along with a Lactate >=4 mmol/L or hypotension (SBP<90, MAP<65) after initial fluid resuscitation
This was the definition used in the EGDT Trial [PMID 11794169], the ProCESS Trial [PMID 24635773], the Arise Trial [PMID 25272316], and the ProMISE Trial [PMID 25776532]. It is not perfect, but at least we are all on the same page.
In a patient meeting the above definition, we are all pretty clear that if the goals of care are curative, we should pursue aggressive sepsis treatment. You can debate endlessly exactly what that care should be, but I think we all agree that we should be focusing our efforts on this group.
But of course, this is just a bit too clean cut for the world of quality.
The fun started in 2012 with the revision of the Surviving Sepsis Campaign Guidelines [PMID 23353941].
Somehow this table managed to find its way into the guidelines:
You can see at the bottom of the table, that these new criteria for Severe Sepsis are seemingly derived from this 2003 paper [PMID 12682500]. I still can't find where in the paper (which is actually a consensus document from a conference held in 2001), precisely where these variables emerged from–I imagine they were from some of the illness severity scoring systems, but who knows? But put that aside for a second, even if you give credence to these variables, note that in the table's heading it states …THOUGHT TO BE DUE TO THE INFECTION. This will become important in just a bit.
National Quality Forum
The NQF has been pushing for Measure 0500 (this link will take you to a version downloaded on 6/8/15) for a while now. We worked hard to get this one quashed just prior to the publication of ProCESS, as putting forth a new measure demanding CVP and ScvO2 seemed a bit silly at the time, but I digress. The NQF document has some new and fun definitions for severe sepsis:
- Severe sepsis is defined as a suspected source of clinical infection, 2 or more manifestations of systemic infection (SIRS criteria) and
the presence of sepsis-induced organ dysfunction.
- SIRS criteria include: Temperature >38.3 C or <36.0 C, Heart rate >90 beats per minute, Respiration > 20 breaths/min, White blood
cell count >12,000 or <4000/mm3, or >10% bandemia.
- Organ dysfunction variables include:
- (SBP)<90 mm Hg or mean arterial pressure <70 mm Hg or a SBP decrease >40 mm Hg or <2 SD
below normal for age or known baseline
- Creatinine > 2.0 mg/dl (176.8 mmol/L) or Urine Output < 0.5 ml/kg/hour for > 2 hours,
- Bilirubin > 2 mg/dl (34.2 mmol/L),
- Platelet count < 100,000,
- Coagulopathy (INR >1.5 or aPTT >60 secs),
- Lactate > 2 mmol/L (18.0 mg/dl).
- (SBP)<90 mm Hg or mean arterial pressure <70 mm Hg or a SBP decrease >40 mm Hg or <2 SD
- Septic shock requires the presence of severe sepsis as above AND as sepsis-induced hypoperfusion persisting despite adequate fluid resuscitation OR lactate > 4 mmol/L.
- Sepsis induced tissue hypoperfusion is present with (SBP)<90 mm Hg or mean arterial pressure <70 mm Hg or a SBP decrease >40 mm Hg or <2 SD below normal for age or known baseline.
So what can we say about this little bit of delight. Well, they have re-dubbed severe sepsis to be something very different than we are used to or have read in any of the major sepsis studies. They seem to have conveniently left out the, “thought to be due to infection” that was in the SSC table 2. I am sure they will say, “we have the line sepsis-induced in there,” but unfortunately the measure itself doesn't list your clinical feelings anywhere as an exclusion–so that would be bunk. I am still fairly desperate to know the evidence for this new definition, but I haven't found it yet. Pretty unacceptable to hold every hospital in the US accountable to an arbitrary definition that has not been tested in large-scale trials.
Then we have a reimagining of septic shock. Somehow, septic shock now includes a lactate >4. The lactate >=4 was supposed to be the severe sepsis that went into the phrase that EGDT built: Severe Sepsis and Septic Shock. I'm not sure whose idea it was to conflate these two, but it is a bad idea. And where the hell does MAP<70 factor into this, every other shred of sepsis evidence uses a MAP<65.
Now what would they have you do in these patients?
A. measure lactate level
B. obtain blood cultures prior to antibiotics
C. administer broad spectrum antibiotics
D. administer 30 ml/kg crystalloid for hypotension or lactate = 4 mmol/L
E. apply vasopressors (for hypotension that does not respond to initial fluid resuscitation to maintain a mean arterial pressure >=65)
F. in the event of persistent hypotension after initial fluid administration (MAP < 65 mm Hg) or if initial lactate was = 4 mmol/L, reassess
volume status and tissue perfusion and document findings.*
* To meet the requirements, a focused exam† by a licensed independent practitioner (LIP) or any 2 other items are required:
• Measure CVP
• Measure ScvO2
• Bedside cardiovascular ultrasound
• Dynamic assessment of fluid responsiveness with passive leg raise or fluid challenge
• Focused exam† including vital signs, cardiopulmonary, capillary refill, pulse and skin findings.
G. remeasure lactate if initial lactate is elevated
A is good–I'm with them there.
B is problematic. Yes, when possible get the blood cultures first, but the Surviving Sepsis Campaign actually addressed this issue in the wording of their recommendation:
We recommend obtaining appropriate cultures before anti-microbial therapy is initiated if such cultures do not cause significant delay (> 45 minutes) in the start of antimicrobial(s) administration (grade 1C).
- Kumar did a retrospective study over a sig. chunk of time that showed early antibiotics were associated with decreased mortality in patients with septic SHOCK. (Crit Care Med. 2006 Jun;34(6):1589-96.)
- Puskarich et al showed the same thing, again in septic SHOCK. Delay for patients not in shock did not seem to have an effect on mortality (Crit Care Med 2011;39:2066)
- Gaieski showed time from EGDT Criteria to antibioitcs was associated with mortality. This was the only one that showed Lactate>=4 also had an association and they didn't actually subset this group out. (Crit Care Med 2010; 38:1045–1053)
- de Groot showed early antibiotics didn’t matter unless the pt was in SHOCK [Crit Care 2015;19:194]
So the only real evidence (albeit not great evidence) is for patients who have septic SHOCK. And when I say shock, I mean shock the way everyone else in the world thinks of it as opposed to NQF–that is hypotension after initial fluids. Though the evidence isn't there, I think giving antibiotics within 3 hours for the Lactate>=4 group makes sense as well. It was done in the 3 recent big sepsis trials for those folks. But there is NO evidence for the NQF-redefined severe sepsis group, this portion of the measure should be rewritten to only include the EGDT criteria patients.
D of course should only list the 30 ml/kg bolus for patients who meet the EDGT definition as well.
E seems fine unless you have a 20 year-old, 50 kg female whose baseline BP may very well be a MAP of 60, but why should we let individualized medicine get in the way of quality measures
F is just plain silly. A desperate grab to keep big-brother ownership of what ProCESS, ARISE, and ProMISE demonstrated to be a purely clinical game.
and then G which seems fine, though we could quibble.
So in conclusion for NQF, go back to some real evidence based definitions of severe sepsis and septic shock, keep A, alter B, C&D get fixed with the repaired definition, E is fine, strike F, and keep G.
which brings us to… CMS
Antibiotic Delays and Feasibility of a 1-Hour-From-Triage Antibiotic Requirement: Analysis of an Emergency Department Sepsis Quality Improvement Database. Ann Emerg Med. 2019 Sep 24. pii: S0196-0644(19)30593-1. doi: 10.1016/j.annemergmed.2019.07.
And now we tie the pain to the money.
We have the proposed SEPSIS BUNDLE PROJECT (SEP) v 5.0a, which I believe will go into effect in October 2015. From this document comes yet another definition of severe sepsis:
In order to establish the presence of severe sepsis, there are three criteria, all three of which must be met within 6 hours of each other.
- Documentation of a suspected source of clinical infection. There may be reference to “possible infection from xx”, “suspect infection from xx”, or similar reference in progress notes, consult notes, or similar physician/APN/PA documentation
- Two or more manifestations of systemic infection according to the Systemic Inflammatory Response Syndrome (SIRS) criteria
- Organ dysfunction,evidenced by any one of the following:
- Systolic blood pressure (SBP) < 90, or mean arterial pressure < 65, or a systolic blood pressure decrease of more than 40 mmHg from the last previously recorded SBP considered normal for that specific patient
- Creatinine > 2.0, or urine output < 0.5 mL/kg/hour for 2 hours
- Bilirubin > 2 mg/dL (34.2 mmol/L)
- Platelet count < 100,000
- INR > 1.5 or aPTT > 60 sec
- Lactate > 2 mmol/L (18.0 mg/dL)
When determining organ dysfunction, any single blood pressure or mean arterial pressure reading in the first hour after presentation that is abnormal, as described above, will satisfy the criteria for organ dysfunction.
So if a patient at any point has a single drop in BP or a lactate of 2.1 or an INR of 1.6 they are now severe sepsis in this imaginary evidence-free world. Note, now even the sepsis-induced has disappeared. So patients with a baseline Cr of >2 = you are screwed on the measure, sucker.
Looks pretty similar to the NQF measure right? That's b/c it is the same docs shepherding both of them.
What would they have you do with the patients that meet this definition?
Received within three hours of presentation of severe sepsis:
• Initial lactate level measurement
• Broad spectrum or other antibiotics administered
• Blood cultures drawn prior to antibiotics
AND received within six hours of presentation of severe sepsis:
• Repeat lactate level measurement only if initial lactate level is elevated
AND ONLY if Septic Shock present:
Received within three hours of presentation of septic shock:
• Resuscitation with 30 ml/kg crystalloid fluids
AND ONLY IF hypotension persists after fluid administration, received within six hours of presentation of septic shock:
AND ONLY if hypotension persists after fluid administration or initial lactate >= 4 mmol/L, received within six hours of presentation of septic shock:
• Repeat volume status and tissue perfusion assessment consisting of either
A focused exam including:
- Vital signs, AND
- Cardiopulmonary exam, AND
- Capillary refill evaluation, AND
- Peripheral pulse evaluation, AND
- Skin examination
Any two of the following four:
- Central venous pressure measurement
- Central venous oxygen measurement
- Bedside Cardiovascular Ultrasound
- Passive Leg Raise or Fluid Challenge
So in order to understand this we need to know how they are defining septic shock:
a. There must be documentation of severe sepsis present.
b. Tissue hypoperfusion persists in the hour after crystalloid fluid administration, evidenced by either
• systolic blood pressure (SBP) < 90, or
• mean arterial pressure < 65 or
• a decrease in systolic blood pressure by > 40 mmHg from the last previously recorded SBP considered normal for that specific patient
• Lactate level is >= 4 mmol/L
Back to a mixed up vision of septic shock.
So how do we fix the CMS measure?
- Keep the Lactate
- Change the ABX within 3 hours to their new definition of Septic Shock (i.e. the old definition/EGDT definition of Severe Sepsis and Septic Shock), and not their new vision of severe sepsis
- Eliminate the Blood Culture rule
- Keep the 30 ml/kg of fluid for the septic shock patients (by their definition) unless clinicians document why they felt the pt should not get this volume
- Keep the vasopressors
- Eliminate the focused exam silliness
Alternatively, everything could be fixed by defining the denominator as patients with persistent hypotension after fluids or Lactate>=4
Why Should You Care?
If the CMS measure goes through as currently written, we are hosed. Since all of this data will be collected retrospectively, there will be hundreds of patients deemed severe sepsis that you never even thought were sick.
And we are not even bringing up the Time Zero ridiculousness that this is all based on.
So why are we complicit?
Because these government organizations do not come up with this stuff themselves. Behind every silly CMS reg you are subjected to, there is, somewhere, a group of doctors that made it happen. Just as medical malpractice would not exist without a host of plaintiff witnesses, these measures would not exist without us.