Professional Documents
Culture Documents
ACUTE Heart Failure Risk Stratification: Circulation
ACUTE Heart Failure Risk Stratification: Circulation
EDITORIAL
M
ore than 80% of all emergency department (ED) visits result in discharge.
Conversely, >80% of patients with acute heart failure (AHF) in the ED are
admitted to the hospital. It is disappointing that this practice persists de-
spite 20 years of effort, with little reason to believe it will change.1 Although some
patients clearly benefit from hospitalization, up to 50% of patients with AHF in
the ED may be discharged or placed in observation.2,3 It is important to note that
nearly half of all patients hospitalized with AHF present with lower-risk features,
such as a blood pressure >140 mm Hg and a B-type natriuretic peptide <1000 pg/
mL, supporting the idea of a lower-risk cohort embedded within the overall AHF
population.4 However, this cohort has many associated comorbidities. It is more
important that a significant proportion is likely to experience an adverse event at
Downloaded from http://ahajournals.org by on February 26, 2019
Stiell et al7 2017 1100 Canada History of stroke or TIA 30-day serious Yes Maybe: clinicians
Ottawa Heart History of intubation for respiratory adverse events told not to
Failure Risk distress solely base their
Scale decision on
Heart rate on ED arrival ≥110
OHFRS
Room air Sao2<90% on EMS or ED
arrival
ECG has acute ischemic changes
Urea ≥12
Serum CO2 ≥35
Troponin I or T elevated to MI level
NTproBNP ≥5000
During walk test, Sao2<90% on room
air or usual O2, or HR ≥100 during
3-min walk test, or too ill to walk
Miro et al8 2018 4711 Spain Barthel index at admission 30-day mortality Yes No
MEESSI Systolic blood pressure
Age
NTproBNP
Potassium
Troponin
NYHA at admission
Respiratory rate
Low-output symptoms?
Oxygen saturation
Episode associated with ACS?
Hypertrophy on ECG?
Creatinine
Downloaded from http://ahajournals.org by on February 26, 2019
Lee et al9
2018 1983 Canada Age 7- and 30-day Yes No
EHMRG Arrival by ambulance mortality
Systolic blood pressure (triage)
Heart rate (triage)
Oxygen saturation (triage)
Potassium
Creatinine
Troponin
Active cancer
Metolazone use before ED arrival
ST depression on 12-lead (30-day
model)
(Continued )
Table. Continued
EDITORIAL
Auble et al11 2008 8384 USA Sex Death or Yes No
Acute Heart History of (h/o) MI serious medical
Failure Index h/o angina complication
h/o PTCA before discharge.
h/o diabetes mellitus Secondary:
h/o lung disease inpatient death
Heart rate alone and 30-day
Respiratory rate mortality alone
Systolic blood pressure
Temperature
Sodium
Potassium
BUN
Creatinine
Glucose
WBC count
Arterial pH
ECG findings: MI
ECG findings: ischemia
CXR: pulmonary congestion
CXR: pleural effusion
ACEI indicates angiotensin-converting enzyme inhibitor; ACS, acute coronary syndrome; BMI, body mass index; BNP, B-type natriuretic peptide; BUN, blood urea
nitrogen; CXR, chest x-ray; ED, emergency department; EHMRG, Emergency Heart Failure Mortality Risk Grade; EMS, emergency medical services; h/o, history of;
HR, heart rate; MEESSI, Multiple Estimation of risk based on the Emergency department Spanish Score in patients with AHF; MI, myocardial infarction; NTproBNP,
N-terminal pro b-type natriuretic peptide; NYHA, New York Heart Association; OHFRS, Ottawa Heart Failure Risk Scale; PTCA, percutaneous transluminal coronary
angioplasty; STRATIFY, Improving Heart Failure Risk Stratification in the ED; TIA, transient ischemic attack; and WBC, white blood cell.
than those we admit,12 admit patients who have an un- low-risk (518 patients) categories had 7-day and 30-day
eventful and brief hospital stay, and remain surprised mortality rates of 0%. The discrimination for physician-
by the proportion (4%) who experience death within estimated risk (area under the curve=0.71) was improved
30 days.13 This also makes one consider the competing (area under the curve=0.82) with use of the EHMRG7
Downloaded from http://ahajournals.org by on February 26, 2019
risk of death and nonfatal rehospitalization. Which risk model (area under the curve=0.81). The EHMRG30 had
is more concerning? Should we consider the event with slightly lower (0.77) discrimination in comparison with
the higher financial penalty? Both are important, espe- EHMRG7. Another important key finding is the overesti-
cially to patients. However, stratifying the risk for death mation of 7-day and 30-day mortality at the low end of
is paramount. the risk spectrum by providers, and an underestimation
In this issue of Circulation, Lee and colleagues14 exter- of mortality at the higher end.
nally tested their 7-day (EHMRG7: Emergency Heart Fail- From the ED standpoint, 0% mortality at 7 and 30
ure Mortality Risk Grade) and 30-day (EHMRG30) risk days in the low-risk group is very reassuring. Emergency
score in nearly 2000 patients at 9 hospitals in Ontario, physicians’ overestimation of risk in these same patients
Canada. In comparison with most previous risk-stratifi- highlights the need for an objective score. Still, it would
cation studies, the authors externally tested their orig- be good to know what element of the risk score drove
inal rule in a prospective manner, in a separate cohort categorization. Dichotomizing certain variables, such
of patients, simultaneously determined physician-esti- as emergency medical services transport (at times in-
mated risk, and performed comprehensive follow-up. appropriately used in the United States) and troponin
This step is critical before an implementation study. It is (how positive?), may sway the risk rule. Using an online
important to note that a waiver of informed consent fa- EHMRG calculator, it is possible to categorize patients
cilitated enrollment along the entire spectrum of disease with either a systolic blood pressure of 80 mm Hg, a
severity. Their patients were older (median 81 years), very high troponin value, or significantly worse acute
with 71% having a previous diagnosis of heart failure, kidney injury into the low-risk group. This brings the
and a fair proportion of cardiovascular and noncardio- challenge of real-world applicability into the crosshairs;
vascular comorbidities. Of these patients, 21% were dis- for the decision rule to be used, it must account for
charged from the ED. Those patients discharged home patients who clearly need admission, but are catego-
had <1.5% 7-day and 3.3% 30-day mortality. Within rized by the risk score as low risk. This is arguably unfair
7 days, 39 patients died (2%), and by day 30 this rose to the decision rule and discounts the rigor by which
to 138 patients (7%). Of the 138 deaths, only 17 oc- this rule was developed. Furthermore, it renders clinical
curred outside the hospital. They assigned patients into judgment obsolete. Nevertheless, it highlights the need
5 prespecified risk categories: very low, low, intermedi- for an implementation study. The absence of high-risk
ate, high, and very high. Patients in the very-low-risk or features in EMHRG suggests a lower-risk patient; how-
ever, they still may not be eligible for discharge because will have an established patient–provider relationship in
of other complicating and competing conditions. which outpatient follow-up is easily facilitated. How-
This study advances our understanding of the ever, rapid outpatient access for all patients regardless
EDITORIAL
EHMRG rule, and risk stratification in general, but of the time of ED discharge and previous provider rela-
there are several limitations to consider. Nearly 30% tionship is crucial for success.
of patients have no history of heart failure. Although Over the next decade there are great opportunities
management of de novo heart failure varies country to increase the proportion of patients with AHF in the
by country, in the United States such patients gener- ED who can be safely discharged home. Although such
ally warrant admission. The need for decongestion, needed progress is unlikely to match the state of dis-
identification, and management of precipitants, and position decision-making in other cardiovascular pro-
investigating underlying cardiac structure and func- cesses in the ED, such as chest pain, studies such as
tion, as well, is challenging to accomplish outside of ACUTE (Acute Congestive Heart Failure Urgent Care
the hospital setting. Similarly, how well does the model Evaluation) are a necessary step in the right direction.
discriminate when high-risk patients are excluded? A Other AHF rules require similar external testing and im-
risk rule for discharge has less utility in patients with plementation studies to determine their optimal role in
hypotension, who require noninvasive or invasive ven- the ED. Such continued advances will help drive further
tilation, have very high troponin values, or have severe improvements in early treatment and local support for
acute kidney injury. Once all the appropriate reasons to rapid outpatient follow-up: necessary items to safely
admit are taken out, how then do we decide what to discharge a larger proportion of patients with AHF.
do? Data collection was not standardized, and this can
introduce inconsistency and inaccuracies. The authors
discuss the use of a net reclassification index to sug- ARTICLE INFORMATION
gest how the EHMRG rule could impact physician de- Correspondence
cision making. However, this may not be a completely Sean P. Collins, MD, MS, 1313 21st Ave South, Nashville, TN 37232. Email
accurate picture of the rule’s impact. Although there Sean.collins@vumc.org
is a clear need to identify lower-risk patients safe for
ED discharge, provider decision-making accounts for Affiliations
the possible success of outpatient management given Department of Emergency Medicine, Vanderbilt University Medical Center,
the severity of both AHF and non-AHF symptoms. Fi- Nashville, TN (S.P.C.). Department of Emergency Medicine, Indiana University
Downloaded from http://ahajournals.org by on February 26, 2019
type natriuretic peptide levels and in-hospital mortality in acute decom- gent care: a cohort study. Ann Intern Med. 2012;156:767–75, W. doi:
pensated heart failure. J Am Coll Cardiol. 2007;49:1943–1950. doi: 10.7326/0003-4819-156-11-201206050-00003
10.1016/j.jacc.2007.02.037 10. Collins SP, Jenkins CA, Harrell FE Jr, Liu D, Miller KF, Lindsell CJ, Naftilan AJ,
5. McCausland JB, Machi MS, Yealy DM. Emergency physicians’ risk atti- McPherson JA, Maron DJ, Sawyer DB, Weintraub NL, Fermann GJ, Roll SK,
EDITORIAL
tudes in acute decompensated heart failure patients. Acad Emerg Med. Sperling M, Storrow AB. Identification of emergency department patients with
2010;17:108–110. doi: 10.1111/j.1553-2712.2009.00623.x acute heart failure at low risk for 30-day adverse events: the STRATIFY deci-
6. Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr, Colvin MM, sion tool. JACC Heart Fail. 2015;3:737–747. doi: 10.1016/j.jchf.2015.05.007
Drazner MH, Filippatos GS, Fonarow GC, Givertz MM, Hollenberg SM, 11. Hsieh M, Auble TE, Yealy DM. Validation of the Acute Heart Failure Index. Ann
Lindenfeld J, Masoudi FA, McBride PE, Peterson PN, Stevenson LW, Emerg Med. 2008;51:37–44. doi: 10.1016/j.annemergmed.2007.07.026
Westlake C. 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/ 12. Lee DS, Schull MJ, Alter DA, Austin PC, Laupacis A, Chong A, Tu JV,
AHA Guideline for the Management of Heart Failure: a report of the A- Stukel TA. Early deaths in patients with heart failure discharged from
merican College of Cardiology/American Heart Association Task Force the emergency department: a population-based analysis. Circ Heart Fail.
on Clinical Practice Guidelines and the Heart Failure Society of America. 2010;3:228–235. doi: 10.1161/CIRCHEARTFAILURE.109.885285
J Am Coll Cardiol. 2017;70:776–803. doi: 10.1016/j.jacc.2017.04.025 13. Sax DR, Mark DG, Hsia RY, Tan TC, Tabada GH, Go AS. Short-term out-
7. Stiell IG, Perry JJ, Clement CM, Brison RJ, Rowe BH, Aaron SD, McRae AD, comes and factors associated with adverse events among adults discharged
Borgundvaag B, Calder LA, Forster AJ, Wells GA. Prospective and explicit from the emergency department after treatment for acute heart failure. Circ
clinical validation of the Ottawa heart failure risk scale, with and without Heart Fail. 2017;10:1–14. doi: 10.1161/CIRCHEARTFAILURE.117.004144
use of quantitative NT-proBNP. Acad Emerg Med. 2017;24:316–327. doi: 14. Lee DS, Lee JS, Schull MJ, Borgundvaag B, Edmonds ML, Ivankovic M,
10.1111/acem.13141 McLeod SL, Dreyer JF, Sabbah S, Levy PD, O’Neill T, Chong A, Stukel TA,
8. Miró Ò, Rossello X, Gil V, Martín-Sánchez FJ, Llorens P, Herrero-Puente Austin PC, Tu JV. Propsective validation of the emergency heart failure
P, Jacob J, Bueno H, Pocock SJ; ICA-SEMES Research Group. Predicting mortality risk grade for acute heart failure: the ACUTE Study. Circulation.
30-day mortality for patients with acute heart failure in the emergency 2019;139:1146–1156. doi: 10.1161/CIRCULATIONAHA.118.035509
department: a cohort study. Ann Intern Med. 2017;167:698–705. doi: 15. Collins SP, Storrow AB. Moving toward comprehensive acute heart failure
10.7326/M16-2726 risk assessment in the emergency department: the importance of self-
9. Lee DS, Stitt A, Austin PC, Stukel TA, Schull MJ, Chong A, New- care and shared decision making. JACC Heart Fail. 2013;1:273–280. doi:
ton GE, Lee JS, Tu JV. Prediction of heart failure mortality in emer- 10.1016/j.jchf.2013.05.002
Downloaded from http://ahajournals.org by on February 26, 2019