- Temple EM
Ottawa Heart Failure Score
Posted by: Kelly Goodsell
The article:
Steill, IG et al. “Prospective and Explicit Clinical Validation of the Ottawa Heart Failure Risk Scale, with and without use of Quantitative NT-proBNP”. Academic Emergency Medicine. 2017 Mar;24(3):316-327. doi: 10.1111/acem.13141.
Background:
-CHF exacerbation/acute heart failure (AHF) is a common ED complaint, accounts for a lot of admissions and healthcare $
-some of these admissions may be unnecessary
-quality evidence to inform potential dispo guidelines for these pts previously lacking
-the Ottawa Heart Failure Risk Scale (OHFRS) –previously derived scoring system for ED pts w/ sx of AHF, estimates probability of a serious adverse event (SAE) in the next 14 days
SAEs defined as: death <30 days, ICU admission, or any of the following in next 14 days -re-admit, need for PPV after admission, intubation, MI, need for major procedure including PCI, CABG
OHFRS specifics-
10 items on a 15-point scale:
hx stroke/TIA
hx prior intubation for respiratory distress
HR > 110 on arrival
O2 sat <90% on arrival
ECG w/ acute ischemic changes
CTNI >0.10
Serum CO2 >35
Pro-BNP>5000 ng/L
BUN >12 mmol/L
Too sick/tachy to ambulate in ED
Score >3 = high to very high risk of SAE within 14 days (16% for a score of 3 to 90% for score of 9)
baseline risk of 2.8% with a score of 0; score <2 correlates to < 10% risk SAE within 14 days
The idea:
-to prospectively evaluate the accuracy/clinical usefulness of the OHFRS in risk stratifying patients presenting to the ED with sx of AHF and assisting w/ dispo decisions
The study:
-Prospective observational cohort study, 6 tertiary hospitals in Canada
-pts w/ sx of AHF were evaluated by ED physicians using the OHFRS criteria then followed for 30 days looking for SAEs
+/- pro-BNP measurement
-1100 patients included: age >50, p/w SOB clinically concerning for acute heart failure
-excluded
sx > 7 days, no evidence of volume overload on CXR, HR >120, O2 sat <85% (on RA after tx), SBP <85, ECG changes c/w STEMI, hx dementia, terminal illness or NH resident
-for eligible patients ED docs used the OHFRS instrument and
calculated a score (correlated with an estimated category of risk)
rated comfort w/ using the scale to aid in dispo decisions
-outcomes:
primary– incidence of SAEs vs OHFRS score
secondary– clinical sensibility of OHFRS, evaluated in 2 ways
overall accuracy – 95% CI for interpretation of original OHFRS by treating ED doc vs score generated by hospital committee (control) based on chart review w/ no knowledge of pt outcome status for SAE
physician comfort w/ using scale (descriptive format)
assessed for potential impact of scoring system by comparing actual admission rates vs predicted rates
Results:
–1100 total enrolled; mean age 77, 57% admitted to the hospital on index visit
-overall SAE rate 15%; 19% in admitted pts, 10% for those d/c from ED; total deaths (inside and out of hospital) 41 (4%)
primary outcome:
Performance of OHFRS score w/o including BNP measurements:
-using a score cutoff of >1 led to increased sensitivity for predicting SAE (91% from 71% in actual practice), but higher admission rate (77% from 57%)
-using a score >2 led to comparable sensitivity at 71% but less admits (48% from 57%)
-pro-BNP measurements available for 684 of 1100 pts. Performance of OHFRS in this group:
-using a score cutoff of >2 showed similar admission rates (63% from 60% in actual practice) w/ better sensitivity (79% from 69%)
secondary outcome:
-overall agreement in risk category classification between ED docs and control committee; 59% agreement for exact category, 95% agreement for score +/- 1
-only 12% of ED docs reported discomfort using the OHFRS
The takeaway:
– OHFRS a risk stratification tool that offers insight into likelihood of short term SAE in pts presenting with acute heart failure, can help ED docs inform dispo decisions
-using an OHFRS threshold of >2 to guide decisions to admit would offer similar sensitivity in predicting which patients are at greatest risk for short term adverse events but potentially reduce admission rates.
-incorporating BNP values may increase sensitivity
-ED docs showed general acceptance, fair accuracy in using the OHFRS