, Sean Collins2 and Gregory J. Fermann3
(1)
Department of Emergency Medicine, Vanderbilt University Medical Center, Nashville, TN, USA
(2)
Department of Emergency Medicine, Vanderbilt University, 2201 West End Ave, Nashville, TN 37235, USA
(3)
Department of Emergency Medicine, University of Cincinnati, 231 Albert Sabinway, Cincinnati, OH 45267-0769, USA
Keywords
Hemodynamic monitoringBlood pressureHeart rateRespiratory rateOxygen saturationTroponinNatriuretic peptidesEmergency department (ED) observation units (OUs) represent an emerging cost-effective treatment option for low-risk acute heart failure (AHF) patients. After initial evaluation and treatment in the ED, AHF patients can be discharged, admitted to the hospital, or undergo further management as observation patients. Observation status is independent of the actual location of care delivery and can therefore occur in beds anywhere in a hospital or in dedicated OUs. Likewise, this short-stay population can be managed by inpatient specialists, hospitalists, or emergency providers. Observation is fundamentally a billing status defined by the Center for Medicare and Medicaid Services (CMS) as care spanning typically less than 24 h but no more than 48 h. As such, the objective ED evaluation in conjunction with provider intuition, response to treatment, and assessment of self-care barriers represents an important factor in risk stratification for AHF patients and subsequent inclusion or exclusion from an OU stay.
Background
Nearly 75 % of ED visits for AHF ultimately lead to hospitalization, and this high proportion of ED visits with resultant inpatient admission has not changed over the last decade [1, 2]. The costs and morbidity associated with these hospitalizations have generated increased pressure to manage these patients more efficiently in the acute care environment [2–7]. Such inpatient admissions could possibly be avoided in a large proportion of patients, as relatively few ultimately receive intensive acute care, mechanical ventilation, and circulatory support or undergo invasive diagnostic or therapeutic interventions [8, 9]. After appropriate risk stratification, selected AHF patients may be safely and effectively managed in an OU at a lower cost compared to an inpatient stay [10–12]. In a relatively short period of time, an OU can provide frequent reassessment in a monitored setting, IV diuretics, afterload reduction, targeted patient education, and coordination with outpatient providers regarding medication regimens and close follow-up. Despite the fact that prospective, randomized, controlled trials evaluating this strategy are lacking, preliminary evidence suggests that AHF patients managed in an OU setting have similar outcomes and improved resource utilization compared to a risk-matched group of admitted patients [11, 12].
Risk Stratification on Emergency Department Presentation
The Society for Cardiovascular Patient Care (SCPC, formerly the Society of Chest Pain Centers) has published several recommendations for patient selection and management in the observation setting. Generated from existing evidence on AHF risk stratification and later externally tested in an independent data set, these recommendations can serve as a guide to identify patients who may benefit from an OU stay rather than an inpatient admission [13, 14]. Table 14.1 outlines recommended inclusion and exclusion criteria for OU entry based on the SCPC recommendations that have been updated to reflect current evidence and emerging practice patterns.
Table 14.1
Recommended inclusion and exclusion criteria for OU entry
Recommended | Comments | |
|---|---|---|
Inclusion criteria | ||
Blood pressure | SBP > 100 mmHg | |
Respiratory rate | <32 breaths/min | |
Heart rate | Less than 110 bpm | Consider atrial fibrillation if the rate can be controlled with oral meds |
Renal function | BUN < 40 | Consideration should be given for relative changes from baseline |
Creatinine < 3.0 | ||
ECG findings | No acute ischemic changes | Consider not normal but unchanged to be eligible for OU |
Natriuretic peptides | BNP < 1,000 pg/mL | Consider in context of clinical scenario |
NT-proBNP < 5,000 pg/mL | ||
Respiratory | On O2 per NC | Consider after weaned off BiPAP/CPAP |
Troponin | Nondetectable troponin | Consider low, detectable, and non-rising elevations as OU eligible |
Social support | Ability to establish | |
Exclusion criteria | ||
Mechanical ventilation | BiPAP/CPAP | |
Vasoactive medications | No active titration | |
Initial risk stratification has typically focused on the prediction of acute inpatient mortality as the primary endpoint. Further, the majority of studies have focused on identifying high-risk, rather than low-risk, physiologic markers in ED patients with AHF and have been limited by using retrospective design in hospitalized patients. ED patients have traditionally not been enrolled; thus, those patients discharged from the ED are rarely included. Despite these and other limitations, low blood pressure, renal dysfunction, low serum sodium, and elevated cardiac biomarkers (troponin [Tn] or natriuretic peptides [NP]) have been repeatedly shown to be increased risk factors for morbidity and mortality [15].
However, a recent prospective cohort study conducted at four hospitals enrolled 1,033 ED patients with AHF, including 7.7 % that were discharged from the ED, to evaluate the incidence of severe adverse events (SAE) within 30 days of ED evaluation (ACS, coronary revascularization, emergent dialysis, intubation, mechanical cardiac support, CPR, and death). The study assessed readily identifiable ED variables to select a cohort of patients who may be eligible for ED discharge, and the resultant decision tool was highly sensitive for a 30-day mortality and SAE [16]. Similarly, a prospective observational cohort study enrolled 559 AHF patients at six Canadian EDs to assess for a 30-day death and a 14-day SAE. Their Ottawa Heart Failure Risk Scale identified prior intubation for respiratory distress, vital sign abnormalities, ECG changes, laboratory findings, and history of stroke/TIA as important variables in their final risk prediction model [17].
Overall, clinical variables for risk stratification of AHF patients are often categorized broadly into demographics, cognitive/functional status, comorbidities, hemodynamics, cardiac ischemia markers, electrolytes, and heart failure biomarkers.
Demographics
The current recommendations published by the SCPC do not specifically refer to age, sex, or race as inclusion or exclusion criteria. Lee and colleagues looked at over 12,000 ED patients with AHF to derive and validate a prediction rule for a 7-day mortality. Among several other variables further outlined below, their retrospective analysis found age to be an independent predictor of a 7-day mortality risk (adjusted odds ratio [OR], 1.40 (95 % CI, 1.16–1.69), p < 0.001) [18].
Hemodynamics
As more studies have attempted to delineate high-risk versus low-risk cohorts using simple, rapidly available data points, systolic blood pressure (SBP), heart rate, and oxygen requirement have proven to be important markers for rapid assessment and disposition. In the Emergency Heart Failure Mortality Risk Grade (EHMRG) 7-day mortality risk score, mortality risk increased with higher triage heart rate (OR, 1.15 [CI, 1.02–1.30]), lower triage SBP (OR, 1.52 [CI, 1.31–1.77] per 20 mmHg), and lower initial oxygen saturation (OR, 1.16 [CI, 1.01–1.33] per 5 %) [18]. However, this study had several limitations including a retrospective patient identification, an exclusion of early readmission for AHF as an outcome, and a practice environment not reflective of the United States. Stiell and colleagues found that both heart rate >110 beats/min and oxygen saturation less than 90 % at ED arrival were independent predictors of SAE [17]. In AHF patients who are ultimately admitted, those with SBP of less than 120 mmHg had threefold higher inpatient mortality than those with SBP greater than 140 mmHg (7.2 % vs 2.5 %, p < 0.001) [19]. In the HF patient who presents in acute distress, a lower initial SBP may reflect left ventricular contractile dysfunction while a higher initial heart rate suggests the need for increased chronotropy to maintain cardiac output and increased sympathoadrenergic response. A lower initial oxygen saturation demonstrates increased pulmonary congestion and underlying respiratory compromise and therefore places the patient at increased risk for mortality [18].
Although a majority of patients who present with AHF will require oxygen supplementation, most patients can be titrated down to a nasal cannula after initial steps targeting decongestion and symptom relief, and these individuals can be easily managed in an OU. Patients requiring acute critical care interventions such as active titration of parenteral vasoactive medications, intubation, or ongoing noninvasive positive pressure ventilation (NIPPV) meet ICU-level criteria and may need to be excluded from OU stay. However, some patients who are initially supported with NIPPV may be quickly weaned from this and could be eligible for OU management. Many of these patients are hypertensive and do well with aggressive blood pressure control and may improve rapidly and avoid intubation [20–22]. If a patient can be weaned off NIPPV in the ED after initial stabilization, transitioning their care to an OU may be considered if other criteria are met.
Renal Function and Electrolytes
Elevated creatinine (SCr > 3.0 mg/dL) and blood urea nitrogen ([BUN] > 40 mg/dL) on hospital admission are strongly correlated with increased in-hospital and post-discharge mortality, and this is reflected in the SCPC recommendations [23]. In the aforementioned prospective cohort study, an elevated BUN represented one of the primary predictors of adverse events (p = 0.01), while the use of dialysis trended toward significance [16]. Two other ED-based studies similarly found an elevated BUN and creatinine to be independent predictors of SAE and mortality, respectively [17, 18].
Hyponatremia, as defined by a serum sodium <135 mmol/L, is associated with increased in-hospital mortality, post-discharge mortality, and readmission rates [24]. Hyperkalemia that may accompany renal insufficiency, resulting from excess repletion in the setting of diuretic use or from potassium-sparing diuretic use, can complicate OU management [25]. Conversely, the large prevalence of loop diuretic use can frequently lead to hypokalemia. An abnormal potassium level (<4.0 mmol/L or >4.5 mmol/L) functions as one of the elements of the EHMRG 7-day mortality risk score, with an increased mortality risk seen in those with hyperkalemia compared to hypokalemia [18].
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
