Recently CMS released a new guideline on Sepsis benchmarks, called SEP-1. The protocol calls for resuscitation with fluids and antibiotics and blood cultures and serial lactates if the patient has "severe sepsis" or "septic shock" but not "sepsis" alone (sirs plus suspected source). The deep dive into this guideline is reproduced below, for the purpose of highlighting that to be excluded the patient must have a specific order for "comfort care". A better deep dive is here and here.
Now anyone who knows anything about CMS and medical review knows that if you write the words "comfort care" on the chart, the patient does not qualify for inpatient status unless they are on hospice, and that is not a benefit that can be elected in the 3-6 hour time period by an altered septic patient who cannot make decisions, and the hospital cannot admit them unless they are admitted to a hospice program through the hospice itself. Further, unless I am mistaken, SEP-1 does not include "altered mental status" as evidence of 'end organ dysfunction" and medical delirium is common in patients with severe sepsis and septic shock. These patients are not good decision-makers. I reiterate that the words "comfort care" are not reimbursed for inpatient status as medically necessary. Medicare is a defined benefit plan that denies payments for non medically necessary patient convenience or comfort.
From the reimbursement perspective, hospitals cannot bill inpatient claims for patients who are "comfort care". Patients who do not meet expectation of surviving beyond two midnights under the Medicare two midnight rules must be placed in observation, unless there is a reasonable expectation that they would survive more than two midnights and require medically necessary care. Palliative treatment in the inpatient setting for sepsis would be such a condition, only if the patient is reasonably expected to survive on admission, and with a near 40% mortality for sepsis, this is a big question. However, if this is the case, the treatment would be palliative, and therefore it would not for example be wise to give a 30 ml/kg bolus to a patient with an EF of 15% who has severe sepsis who does not want positive pressure ventilation or intubation. What is concerning is what the medical review criteria for auditors would be for these patients now that CMS has benchmarked the standard of care for patients with severe sepsis or septic shock and whether that gives ammunition to recovery audit contractors to deny claims. Further, these patients, DNR/DNI but not hospice, will apparently be quality and value based purchasing fallouts for the hospital UNLESS they have a comfort care order on the chart. This will have a hospital payment penalty in the not too distant future.
The available data from the PROCESS , ARISE and PROMISE studies do nothing to assist in answering the plight of the septic patient who needs palliative care and is not a candidate for aggressive resuscitation. These patients were all excluded from these trials. Recently, Ashley Shreves highlighted the question of what is a good death here and here.
It is obvious that a delirious death from sepsis is not a good death, whether it is associated with EGDT or not. I would hope we would focus more on addressing the care of these patients whom we know will likely die and focus on the GOALS of end of life care for these patients as THERAPY. Unfortunately, with this new SEP-1 benchmark I am afraid there will be a push to really push patients into two different black and white groups, either full code with EGDT and SEP-1 benchmark, or comfort care. This push will come from hospitals who will be under payment pressure for meeting standards of care. There will be another push when patients who then have the "comfort care" order are denied from inpatient status and the claims will trend to observation care for patients not expected to survive two midnights and not on hospice. This will result in increased patient financial obligations and lack of effective palliative care for critically ill patients who are too sick to go home and not expected to die imminently in the Emergency Department unless the Emergency and Critical Care community is vocal in voicing our concerns on this issue to CMS. CMS must reform the way that patient needs are met at the end of life and how the payments to hospitals and providers are structured to match the care that is delivered, whether in hospital or out of hospital.
CMS SEP-1 Guideline
Sepsis from CMS point of view is defined as SIRS plus suspected or documented infection, severe sepsis indicates some element of tissue hypoperfusion or end organ dysfunction and septic shock indicates the presence of a shock state (inadequate delivery of metabolites to meet tissue demand) as defined by failure of resuscitation with crystalloids. http://epmonthly.com/article/understanding-the-new-sep-1-sepsis-rollout/
The specific definition of severe sepsis for CMS purposes goes by certain criteria, notably not including “altered mental status” as a measure of end organ dysfunction and the specific criteria are a little different than what is noted in some trials. These are the definitions in the 2012 SCCM guidelines.
3 Hour and 6 Hour Benchmarks for Severe Sepsis and Septic Shock
The new CMS guideline for sepsis, termed SEP-1, calls for basic care elements within the first 3 hours of presentation of a patient with severe sepsis or septic shock, (each category having their own 3 hour and 6 hour clocks), but the time stamps and benchmarks do NOT apply to patients with sepsis only (they are in the denominator).
- 1. draw lactate
- 2. draw blood cultures
- 3. administer antibiotics before 3 hours,
- 4. if initial lactate is elevated above the lab reference, or there is hypotension to give a 30ml/kg crystalloid bolus before 3 hours.
This clock starts from the time that there is the earliest documentation of meeting sepsis criteria while the patient is in the hospital (SIRS plus a source) and can be from the nursing notes, triage vitals and triage statement, or later in the hospital stay (even on the floor). If the patient has only sepsis criteria first (SIRS plus suspected source) then LATER becomes hypotensive or has an elevated lactate, then the clock presumably only starts at the time those occur.
The second element requires by 6 hours of care
- 1. apply pressors for persistent hypotension not responding to 30 ml/kg to maintain MAP >65
- 2. to perform volume status reassessment by 6 hours if there is persistent hypotension not responding to 30 ml/kg fluid bolus OR lactate greater than 4.0
- 3. to redraw lactate only if elevated above lab upper limit of normal (even if below 4.0)
Notably absent is the old NQF requirements for CVP or ScVo2 measurement for patients who are persistently hypotensive or with elevated lactates, however this has been replaced by the “volume status reassesment” which can be met in the following way:
• Repeat focused exam (after initial fluid resuscitation) by licensed independent practitioner including vital signs, cardiopulmonary, capillary refill, pulse and skin findings
Two of the following:
Bedside cardiovascular ultrasound
Dynamic assessment of fluid responsiveness with passive leg raise or fluid challenge
Specific guidelines about how these data are abstracted were reviewed in a CMS webinar on Oct 26
and an October 6, 2015 Webinar with Dr Townsend, Dr Tefera, and Dr Rivers was broadcast and a recording can be accessed here:
Importantly, patients are not excluded from these measure collections if they are DNR, they are only excluded if there is a specific order for comfort care in the first 3 hours for severe sepsis and the first 6 hours for septic shock.