Quality in health care is an idealized yet elusive goal. This can, in large part, be attributed to the inherent difficulty associated with establishing a precise definition of quality—a circumstance derived from the existence of multiple stakeholders (e.g., health-care providers, local administrators, patients, community, insurers, government) each with differing perspectives on what constitutes the deliverables of “good” health care. At its core, however, quality is generally regarded as an attribute of provider care, specifically technical performance (or lack thereof) as viewed through the lens of “best-practice” medicine [1]. The latter represents the summation of those actions (or inactions) that have either proven effective or are, by virtue of consensus expert opinion, considered de facto to contribute to better outcomes (e.g., smoking cessation).

Quality is thus a comparative construct which measures variance from a benchmark set by what is considered to be best care as identified by a consensus standard. But what exactly is being measured and how can one be sure that the metric is relevant at the individual patient level and attributable to the provider (or system) in question? Moreover, to what standard is the assessment held: maximal, which does not consider health benefits within the context of related cost or optimal, which places a cost-effective value on care? Understanding these issues in an era of performance measures [2] and increased accountability for health outcomes is critical.

More than two decades ago, Avedis Donabedian championed the notion that quality can and should be assessed as a function of the relationship between three essential elements termed “structure,” “process,” and “outcome” [1]. As shown in Fig. 4.1, each can exist as both a precondition for (e.g., identification of a disparate outcome at an institution leading to a change in culture or practice) and a consequence of (e.g., inability to meet time-dependent goals for therapeutic intervention because of resource limitations) the others. These relationships, however, are far from linear and can be strongly influenced by confounding variables, especially case mix.

All of this has particular relevance to acute heart failure (HF), where, despite significant advances in medicine, postdischarge outcomes remain poor [3–5]. In the following pages, we discuss the specifics of quality as they relate to HF and highlight, using the Donabedian framework, those measures being used to differentiate performance.

These are multidisciplinary, patient-focused programs that cover matters such as education about the disease and its treatment, dietary counseling, efforts to improve patients’ compliance with medical regimens, and interventions to help patients achieve and maintain control of their volume status. These programs have been shown to reduce readmissions and improve functional status but not necessarily affect mortality rates [7, 8]. Further study is needed to define their overall cost-effectiveness and the optimal strategy [9, 10], as not all approaches (e.g., postdischarge telemonitoring in those recently hospitalized with acute HF) appear to provide clinical benefit [11].

As originally written by the ACC/AHA HF Working Group, this element involved the specific decision to refer medically refractory HF patients to specialty and transplant centers. It called for health-care facilities to establish a relationship with a specialty center and coordinate a plan for transfer that is predetermined and not in response to patient crisis. In such coordinated systems, patients would be referred based on their overall prognosis and response to medical care. Indeed, the literature has shown that patients with symptoms for >3 months and a more severe initial presentation are less likely to respond to therapy and may benefit from referral to specialty centers, including transplant centers [12]. Moreover, in medically refractory patients, referral to specialty centers has been reported to result in a 98 % 1-year survival rate [13, 14] and reduce readmissions by 50 %.

Though initially centered on referral, the concept of coordinated systems has morphed into one increasingly focused on greater linkage throughout the entire continuum of HF care [15, 16]. Such systems, termed accountable care organizations (ACOs), would provide continuity for patients across different institutional settings (including ambulatory and inpatient hospital visits) and, if possible, during episodes of acute decompensation. While prospective experience with structured, shared accountability, and related outcome data in HF is lacking, there is relatively strong evidence from the OPTIMIZE-HF (Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients With Heart Failure) registry which suggests a relationship between readmission rate and early outpatient follow-up after an index HF hospitalization [17]. Among patients with acute HF who were discharged from the emergency department (ED) in the Canadian National Ambulatory Care Reporting System, an association between early collaborative HF care and increased use of drug therapies, cardiovascular diagnostic testing, and better outcomes has also been reported [18].

Positive outcomes are the ultimate goal of any health-care system and the true essence of quality. Ideal outcome measures should be measurable, sensitive to modifications in the structure and process of care, practical to use and should take into account patients’ underlying risk for good or bad outcomes. The main challenge in using outcomes as a marker of quality is that they do not depend solely on the health care provided. Age, severity of cardiac dysfunction, presenting hemodynamic profile, degree of comorbidity, and socioeconomic status have all been shown to affect outcomes for acute HF patients [21].

An additional challenge specific to the ED setting is the relative absence of data linking ED or OU acute HF processes of care with postdischarge outcomes. Consequently, it is unknown which of the commonly used outcome measures (Table 4.2) constitute a meaningful representation of what can reasonably be attributable to ED and OU management of HF patients. Thus, while a recent review of more than 50,000 acute HF patients in Ontario, Canada, found a slightly higher 90-day mortality rate (11.9 % versus 9.5 %; log-rank P = 0.016) among those who were discharged from the ED versus admitted to the hospital, its interpretation within the context of health-care quality is difficult [22]. Moreover, while 90 days is a relatively short follow-up period, it is probably long enough to introduce substantial confounding. With the advent of CMS Value Based Purchasing, shorter (30-day) postdischarge event rates are now favored and may be more reflective of an ED, OU, or even inpatient treatment period. Perhaps of greater importance, 30-day mortality and readmission data for acute HF are publicly reported by the CMS as a measure of comparative hospital quality. Regardless of the sampling period, there may be added value through use of more time-sensitive metrics such as days out of hospital and alive [15], which provide a clearer signal of causality than measurement of dichotomous (and equally weighted) outcomes that occur at any point within a prespecified time frame.