Portal de Periódicos da CAPES

Acute Heart Failure in the Emergency Department: Just a One Night Stand?

2016; Wiley; Volume: 24; Issue: 3 Linguagem: Inglês

10.1111/acem.13151

1553-2712

Peter S. Pang, Sean P. Collins,

Cardiac, Anesthesia and Surgical Outcomes

It begins with a chief complaint. It almost always ends with an admission. Somewhere in the middle, after the diagnosis is made, we treat patients like it's 1974.1 Standard therapy hasn't changed. Followed by a ritual washing of hands, we move on to the next patient. This is the state of acute heart failure (AHF) managed in the emergency department (ED) today. Will we be doing the same thing for the next 40 years? The vast majority of AHF patients feel better after ED management; yet, for 85% of them, admission is the final common pathway.2 Intravenous (IV) loop diuretics are often the only treatment given, despite large amounts of discussion regarding vasodilators.3 In fact, IV loop diuretics are occasionally withheld completely in the ED—for fear of causing some yet unknown and unproven poor outcome—only to see them used aggressively once admitted. Why do we keep doing the same thing over and over again? We have limited prospective data to guide us and are thus left walking the road of clinical inertia. No question, diagnosis has improved due to both natriuretic peptides and lung ultrasound:4, 5 But therapy and risk stratification lags behind. The major reason: patients feel better and, thus, we feel better.6 Patients no longer suffer in front of us and we believe our job complete. Yet the standard traditional therapy we deliver is neither standardized nor well defined. There have been no meaningful changes in morbidity or mortality as a result of AHF therapy. Outcomes are strikingly poor, with over 25% of patients dead or rehospitalized within 30 days, figures worse than acute coronary syndrome.7 Patients indeed feel better. Is this all we need to do? Is this all we can do? Over a million hospitalizations for AHF occur every year.8 The ED is the entry point for over 80% of these. While arguing over IV loop diuretics highlights the need for better evidence, as well as the unmet need for novel therapies, neither, however, will solve the most pressing ED AHF question: risk stratification. In-hospital risk-stratification instruments don't address the issue of ED decision making: they are confounded by the act of hospitalization itself.9, 10 Given the 85% admission rate and multiple comorbidities encountered in this complex cohort, developing methods to better risk stratify patients remains challenging. The prognostic importance of one of these overarching AHF comorbidities, frailty in the elderly, was examined by Martin-Sanchez and colleagues.11 Frailty has been defined as a “clinically recognizable state of increased vulnerability resulting from aging-associated decline in reserve and function across multiple physiologic systems such that the ability to cope with an everyday or acute stressor is compromised.”12 Objectively, the definition put forth by Fried et al. was the one used in this study—meeting three of the five following criteria—low grip strength, low energy, slowed walking speed, low physical activity, and/or unintentional weight loss. Frailty is associated with worse outcomes in AHF.13 However, the authors had a more focused goal: the impact of a frailty assessment on ED disposition decision making, evaluating its association with 30-day all-cause mortality. This retrospective analysis of a convenience sample conducted in three Spanish EDs enrolled patients ≥ 65 years old with a clinical diagnosis of AHF. Only STEMI patients or those refusing to consent were excluded. The principal investigator (PI) at each site reviewed each case independent of the ED treating team and decided on final inclusion into the study. Those with severe functional dependence, defined by a Barthel index (BI) ≤ 40 points were excluded. Frailty was assessed by a trained physician at each site. Ultimately, after multivariable adjustment, frailty was independently associated with 30-day mortality (HR = 2.5; 95% CI = 1.0–6.0; p = 0.047) in older AHF patients. At first glance, this finding seems intuitive. Frail, older patients are at greater risk for worse outcomes. However, how confident should we be of the results? The primary endpoint was all-cause mortality, not cardiovascular specific. Thus, it is unclear if frailty is merely a marker of risk in AHF; mediating frailty may not lead to improved AHF outcomes. We also don't know whether death occurred during hospitalization or postdischarge, an important consideration when planning strategies to improve outcomes. Knowing length of stay (LOS) would also provide context, as a previously published Spanish AHF registry reported a median LOS of 7 days, nearly double that of AHF patients in the United States.14 Of note, in a separate single-center study from Madrid by a different group of authors, frailty was not associated with 30-day mortality or rehospitalizations.13 Several other methodologic limitations are important to highlight. Patient selection was performed by the PI at each site. There was no adjudication. The proportion of patients initially diagnosed with AHF by the treating physician, but later found to be misdiagnosed by PI chart review and excluded from analysis, is unknown. While the European Society of Cardiology Guidelines are listed as the criterion standard, the guidelines essentially emphasize AHF as a clinical diagnosis supported by ancillary studies. This could have resulted in misclassification bias, excluding patients whom the treating MD felt had AHF, the cohort of interest in an ED-based prognostic study. Also, frailty was not directly measured but self-reported, a reasonable approach in a busy ED setting. Controlling for other known risk factors better highlights the incremental discrimination afforded by the frailty assessment. The authors themselves acknowledge frailty did not improve discrimination versus a known risk instrument. Further, to avoid confounding by severity of baseline dependence, BI > 40 was used to define “nonseverely dependent.” For reference, in stroke, a BI score of 75 had 95.7% sensitivity and 88.5% specificity for a modified Rankin Score (mRS) of 3, where stroke patients are able to walk unassisted but need help.15 A mRS of 4 is defined as moderately severe disability, unable to walk unassisted. Intriguingly, several of the authors previously published a large registry stating a BI score of 21–60 as “severe dependence,”14 although in a different review, 0–40 was defined as severe. Where the severe dependence threshold is defined, 60 or 40, may significantly impact the outcome, given a borderline statistical significance of p = 0.047. The CONSORT diagram in Martin-Sanchez et al.11 is also worth examining in more detail. Ultimately, only 63% of eligible patients were included, of whom 25 were missing vital status. Given only 34 total deaths, a few more deaths in either direction would profoundly affect the results. While not directly mentioned, ACS patients were likely mixed into the groups. This is based on the following: 1) only STEMI patients were excluded, not NSTEMI, and 2) troponin values were not reported. A previously published Spanish registry showed 45% of AHF patients with a positive troponin value at the 99th percentile cut-point.14 While troponin release occurs frequently in AHF, this detail would aid the clinician to place the study in context. Finally, nearly one-third of patients did not have a frailty assessment performed. This introduces significant bias. Were these patients not able to be assessed because they were too sick? Did they go home sooner? Or were they just unavailable at the time the study investigator was present? Unfortunately, whether or not frailty assessed in the ED for AHF should impact decision making remains to be determined. Ultimately, recognition and treatment of life and limb threats will always be the priority. But emergency medicine has evolved from the “emergency” room, to an environment where we must act as stewards for those who are less acutely ill. Our increasing role in the management of acute on chronic disease has become our daily bread. This is most important for patients with the greatest needs—the frail, the underserved, and those who commonly experience health disparities. The impact of emergency care extends beyond a single transaction, a brief episode. By exploring issues such as frailty, we can better care for our patients.