Optimizing Management of Sepsis in the ED

October 2010

Optimizing Management of Sepsis in the ED

Search Strategy: You conduct a PUBMED Clinical Query using a narrow search strategy for clinical prediction guides and the search term “emergency department sepsis score” to obtain 21 citations. To reproduce this search strategy go to http://tinyurl.com/26d2bn8.

You are working a TCC shift with Big Brian Bausano when a 66 year old male presents with 3 days of cough, congestion, subjective fevers, and “just feeling poorly”. He has a history of hypertension and coronary artery disease and only takes “a few medications” at home. Recently, he has been at his baseline state of health, with no recent antibiotic exposure or other healthcare exposures. His vital signs are: respiratory rate 21, heart rate 100, temperature 38.6°C, blood pressure 120/80, and O2 saturation 92% on 3L NC. On exam, you note an alteredmental status and has some crackles at the right base. Serendipitously his CXR is shot immediately and demonstrates a right lower lobe pneumonia. As you await your lab workup and figure out your treatment strategy, you begin to ponder….”Is this guy septic? I think he is sick, but I’m not sure how sick. Is there anything I can do to risk-stratify him to tell me how sick he might be, help me with disposition, and give me an idea about mortality?” You seek the sage advice of the wise Dr. Bausano, who unfortunately is upset because Notre Dame lost another football game. He says, “I’ll Facebook it and try to find the answer.” Realizing this probably isn’t the best search strategy, you devise PICO question #1 and conduct a PUBMED query using the terms “sepsis emergency department risk stratification” and find an Evidence Based Medicine review on the topic from which you select two trials to analyze [Search Strategy #1]. An hour has passed, and your patient’s vital signs are unchanged. His labs have returned, showing a lactate level of 6.1. The rest of his labs are normal. You now think out loud to anyone who will listen, “Is this guy septic or severely septic? I think the latter. Is there anything I can do in this patient with clear evidence of tissue hypoperfusion to improve his trajectory and improve outcome?” Dr. Bausano is still on Facebook, but allows you to use the computer long enough to do a PUBMED search [Search Strategy #2] using the terms “severe sepsis bundled care” and PICO Question #2 to identify a recent meta-analysis. Noting a clear mortality benefit, you institute early and appropriate antibiotic therapy, and begin a titrated hemodynamic resuscitation hoping to improve oxygen balance in this very sick gentleman. Unfortunately after 3 liters of crystalloid, his blood pressure is 70/40. Now you wax philosophically…. “This guy is in shock. He needs a vasopressor, but I have no idea which one to choose.” Dr. Bausano is trying to get onto YouTube, but the very vigilent ED computers have blocked all access. He again allows your PUBMED search [PICO Question #3, Search Strategy #3]. You now have guidance.


 

PICO Question #1

Population: Adult ED patients with suspected sepsis

Intervention: Addition of a risk-stratification instrument to clinical gestalt

Comparison: Clinical gestalt alone

Outcome: Short-term mortality, ICU admission rates, ED & ICU length-of-stay

Search Strategy #2: You conduct a PUBMED Clinical Query using a narrow search strategy for clinical prediction guides and the search term “sepsis bundled care” to obtain 3 citations. To reproduce this search strategy go to http://tinyurl.com/2bpr88g.


PICO Question #2

Population: Adult ED patients with suspected sepsis

Intervention: Sepsis bundled care

Comparison: Routine clinical care

Outcome: Short-term mortality, ICU admission rates, ED & ICU length-of-stay

Search Strategy #3:You conduct a PUBMED Clinical Query using a narrow search strategy for therapy combining the search terms “shock” yielding 2278 original research publications. Next, you search “vasopressor” (227156 citations) and then combine the “shock” findings with the “vasopressor” findings to yield a manageable 152 citations to evaluate. Although one of the citations is a Cochrane review from 2004, you choose to instead review a recent randomized controlled trial of 1679 patients instead.


PICO Question #3

Population: Adult ED patients with suspected sepsis and fluid-refractory hypotension

Intervention: Norepinephrine vasopressor therapy

Comparison: Dopamine vasopressor therapy

Outcome: Blood pressure, short-term mortality, ICU admission rates, ED & ICU length-of-stay, neurological outcomes


Years

First years: Mortality in Emergency Department Sepsis (MEDS) score: a prospectively derived and validated clinical prediction rule, Crit Care Med 2003; 31: 670-675. (http://pmid.us/12626967)

Second years: Validation of the Mortality in Emergency Department Sepsis (MEDS) score in patients with the systemic inflammatory response syndrome (SIRS), Crit Care Med 2008; 36: 421-426. (http://pmid.us/18091538)

Third years: Comparison of dopamine and Norepinephrine in the treatment of shock, JAMA 2010; 362: 779-789. (http://pmid.us/20200382)

Fourth years: Bundled care for septic shock: an analysis of clinical trials, Crit Care Med 2010; 38: 668-678. (http://pmid.us/20029343)


Articles

Article 1: Mortality in Emergency Department Sepsis (MEDS) score: a prospectively derived and validated clinical prediction rule, Crit Care Med 2003; 31:670-675
ANSWER KEY

Article 2: Validation of the Mortality in Emergency Department Sepsis (MEDS) score in patients with the systemic inflammatory response syndrome (SIRS), Crit Care Med 2008; 36:421-426
ANSWER KEY

Article 3: Comparison of Dopamine and Norepinephrine in the Treatment of Shock N Engl J Med 2010; 362:779-789
ANSWER KEY

Article 4: Bundled care for septic shock: an analysis of clinical trials Crit Care Med 2000: 38:668-678
ANSWER KEY


Bottom Line

MEDS score: A reliable and accurate predictor of 28-day mortality in ED patients admitted to the hospital with SIRS (http://pmid.us/20162763) (infectious and non-infectious etiologies). Future impact analyses (http://pmid.us/10092723) will need to assess whether EP’s and admitting physicians accept the MEDS score as a SIRS prognostic instrument AND incorporate it into management protocols. In addition, the impact of the MEDS score on overall resource utilizations, costs (http://pmid.us/18497416) and patient outcomes (http://pmid.us/18588253) will need to be assessed.


MEDS Score

Risk Factor Points
Rapidly terminal co-morbid illness 6
Age > 65 3
Bands > 5% 3
Tachypnea or hypoxia 3
Shock 3
Platelet < 150,000 mm3 3
Altered mental status 2
Nursing home resident 2
Lower respiratory infection 2

Interpretation of the MEDS score from the accumulated validation trials through late 2007 (Carpenter 2008 http://pmid.us/19427752):

Score Label Range in 28-day Mortality
0 – 4 Very low risk 0.4% – 11.0%
5 – 7 Low risk 3.3% – 5.0%
8 – 12 Moderate risk 6.6% – 19.0%
12 – 15 High risk 16.1% – 32.0%
> 15 Very high risk 39.1% – 40.0%

Dopamine versus Epinephrine:

In adult sepsis patients in the ICU dopamine offers no survival advantage over norepinephrine and may increase mortality in cardiogenic shock. In undifferentiated ED patients with shock, NE is the preferred first-line vasopressor for now. Future trials should verity these findings in ED populations.

Sepsis Bundles:

The jury is still out. Lesser quality (non-randomized controlled trial) designs and the absence of trials reporting compliance with early goal directed therapy at baseline or therapy-specific responses of hemodynamic benchmarks (Scv02, CVP, etc) all severely restrict astute clinician’s ability to interpret clinical trials of sepsis bundled care. Future trials are needed to individually assess the effectiveness of components like rhAPC, PRBC, vasopressors, and lung protection mechanical ventilation before these interventions become part of sepsis bundled care pathways. Before IHI or CMS or JC incorporate bundled care “all or none” reimbursement pathways, ongoing trials’ evidence should be obtained. Future trials will also need to assess bundled care effectiveness and cost-effectiveness in heterogeneous settings in the era of electronic medical records and computer physician order entry.

Print Friendly, PDF & Email