Journal Club Synopsis: ED treatment of CHF. Early Discharge of Low Risk Chest Pain
Thanks to Andrea and Keith Carlson for hosting, especially for the wicked ping-pong!
Outstanding critiques of the articles by Nathan West, John Paik, Mike Kennedy, Aris Alexander, Katie Burns and Jessica Ede.
1. Collins, SP, et al: Early Management of Patients With Acute Heart Failure: State of the Art and Future Directions—A Consensus Document from the SAEM/HFSA Acute Heart Failure Working Group. Acad Emerg Med 2015; 22:94–112.
Heart failure results in 1 million ED visits annually. Based on guidelines from national and international cardiology societies, this consensus document from SAEM and the Heart Failure Society of America discusses best practices for the diagnosis, treatment, and disposition of ED patients with acute heart failure (AHF).
Unfortunately, literature suggests that there is no historical or exam finding with both high sensitivity and high specificity for the diagnosis of AHF. Prior HF is the most useful historical feature, but the differential diagnosis of dyspnea remains broad even in these patients. The elusive S3 gallop has the highest LR+, with modest LR+ for the historical features of PND, orthopnea, and peripheral edema, and for the exam findings of HJR and JVD. Up to 20% of patients with AHF will have no congestion on their ED CXR, especially in late-stage heart failure. ECG may suggest underlying precipitants for AHF, and A fib has a high LR+ for AHF. BNP testing is strongly recommended, although levels can be affected by age, gender, weight and renal function, and can be elevated in other conditions causing myocardial stretch such as ACS and PE. Patients with HF and preserved LVEF have proportionally lower BNP levels. Ultrasound can be extremely useful, both to identify AHF (better test characteristics than CXR!), and to use IVC collapse index to evaluate volume status.
Currently, treatment recommendations vary between societies, reflecting in part the need for high quality ED clinical trials with clinically meaningful outcomes. Using an algorithm to classify patients, both by initial blood pressure and an assessment of volume status, can help to more effectively tailor treatment to the individual patient. IV loop diuretics are used frequently, but should be started at relatively low doses, e.g. the IV equivalent of the patient’s usual oral daily dose (DOSE trial). Although still included in some guidelines, nesiritide is not useful, and low dose dopamine has also not been shown to be helpful (ROSE trial). Especially in patients presenting with elevated BP, first line treatment should include high dose IV nitrates, as these patients usually have a volume distribution more than volume overload problem. Early non-invasive ventilation in severe AHF is commonly used and associated with improved symptoms and lower rates of intubation.
Morphine is mentioned as an anxiolytic, however benzodiazepines are more generally ordered for this indication. Starting inotropes and management decisions about patients in cardiogenic shock will usually be made after early consultation with cardiology. AHF requiring hospitalization is recognized as a fundamental change in the clinical trajectory of a patient with heart failure. Our hospital treatment does not lead to clear improvement in survival or re-hospitalization rates.
Currently 80% of patients with AHF are admitted. Features including renal dysfunction, low BP, hyponatremia, and elevated troponin or BNP have been associated with increased morbidity and mortality. However, the absence of high risk features does not equate to the patient being low risk and safe for discharge. While several decision instruments for early discharge of patients with AHF have been derived, there is no accepted, well-validated published tool. Our ACMC CHF Observation Protocol is an option in selected patients. Early follow-up and a stable social situation are key factors to the safe implementation of observation protocols.
2. Mahler S, et al: Performance of the 2-hour Accelerated Diagnostic Protocol Within the American College of Radiology Imaging Network PA 4005 Cohort. Acad Emerg Med 2015;22:452-60.
Bottom line: This trial of 1,140 ED patients being evaluated for ACS failed to validate ADAPT as a safe early discharge strategy, however questions were raised about the use of revascularization as a valid endpoint. Differences in patient populations and health care systems may explain the lower sensitivity of ADAPT in this trial, but further studies are needed.
This was a secondary analysis of 1,369 patients already enrolled in a multi-center CTCA trial in 5 tertiary care US EDs. Data elements were collected prospectively. Patients were at least 30 yo, with symptoms suggestive of ACS, and either classified as “low risk” or “at risk” by ADAPT, the 2 hour accelerated diagnostic protocol which includes 0 and 2 hour troponins, ECGs, and a TIMI score. Eligible patients had initial TIMI scores of 0 to 2, and a non-ischemic ECG on presentation.
Patients designated low risk and appropriate for early discharge by ADAPT have a TIMI score of 0, ECG with no new ischemic changes, and negative troponin x 2.
Primary outcome was the rate of MACE (cardiac death, MI, or coronary revascularization) within 30 days of presentation. Secondary outcomes included the rate of patients identified as safe for early discharge, and abnormal stress or CCTA testing.
Of the 1,140 patients with complete biomarker data, 2.7% had MACE. ADAPT identified 551 of 1,140 (48%) as low risk and appropriate for early discharge; 5 of these 551 (0.9%, 95% CI=0.3% to 2.1%) had MACE at 30 days. Of these 5 patients, there were no deaths, one patient with MI, and 5 with revascularizations. Sensitivity of ADAPT for MACE within 30 days was 84%. Serial troponins alone were 32% sensitive for MACE.
These findings conflict with prior very positive studies of the ADAPT strategy. In part this may be due to the inclusion criteria, as the umbrella CCTA trial excluded patients with elevated troponin, whereas prior ADAPT validation studies included patients with elevated troponins who would be identified as “at risk” by ADAPT, thereby boosting the sensitivity of those studies. In addition, 4/5 of the low risk patients with MACE had revascularization procedures without MI. The relevance of revascularization as an outcome measure is uncertain. As discussed at Journal Club, the revascularization rate for patients in the USA is high, and validations of ADAPT in other parts of the world demonstrated much lower percentages of revascularizations in their MACE composite endpoints. It was also discussed that only 1 of the 5 “missed MACE” patients really appeared to be a true miss.
There were a number of low risk patients with abnormal CCTA or stress testing in this study. Authors emphasize the importance of early therapy in these patients, reinforcing the need for close follow-up for patients discharged who do not receive higher level diagnostic testing in the ED.
3. Mahler S, et al: The HEART Pathway Randomized Trial: Identifying emergency department patients with acute chest pain for early discharge. Circulation Cardiovascular Quality and Outcome 2015:8;195-203.
Bottom line: Use of the HEART Pathway (HEART score + 2 negative troponins) allowed for a higher rate of early discharge from the ED without stress testing compared with usual care in this study of low risk ED chest pain patients. No patients identified for early discharge had missed MACE at 30 days.
In this study of 282 ED patients with symptoms concerning for ACS, patients were randomized to either the HEART Pathway or Usual Care. The HEART score includes history, ECG, age, risk factors, and troponin, and a HEART score of 0 to 3 is consistent with low risk. The HEART Pathway consists of the HEART score + troponins at 0 and 3 hours. Patients with low risk HEART scores and negative troponins were recommended for discharge without further cardiac testing. Usual care was based on ACC/AHA guidelines. The primary outcome was rates of objective cardiac testing (stress or angio) and secondary outcomes included length of stay, early discharge, cardiac related ED revisit, admission, and MACE, all assessed at 30 days. Eligible patients were age at least 21 yo, being evaluated for ACS with ECG and troponin. Exclusion criteria included STEMI, instability, or additional serious acute medical problems.
Compared with usual care, use of the HEART Pathway significantly decreased objective cardiac testing at 30 days by 12% (69% vs 57%), length of stay by 12 hours, and increased early discharges by 21%. MACE occurred in 6% of patients overall.
No patients identified for early discharge in either HEART Pathway group or usual care group had MACE at 30 days. It’s important to note that the primary outcome was objective cardiac testing, rather than a comparison of rates of MACE in the 2 groups, and the study was not powered to detect differences in MACE. However, the safety of the HEART Pathway has been consistently validated in other trials.
Decision instruments such as the HEART score are useful to support medical decision making, both when having conversations with patients about risk and their disposition preference, and when documenting in the medical record. Although “gestalt” can be reliable for more senior clinicians, decision instruments in general are particularly useful as an explicit reminder of high-risk features, and help build more reliable heuristics.