HF: Can observation units reduce readmissions?

There is consensus that reducing hospital admissions and readmissions for heart failure (HF) is critical, but not on the best path to follow to reach those goals. Within 30 days after hospitalizations for HF, nearly 10 percent of patients die and about 25 percent of patients are readmitted to the hospital. Reducing readmissions for HF has become a national quality improvement initiative and a priority of the National Heart, Lung, and Blood Institute. The Journal of the American College of Cardiology presented two viewpoints on improving outcomes and reducing readmissions in its Jan. 15 issue.

Sean P. Collins, MD, of Vanderbilt University in Nashville, Tenn., and colleagues argued that increased use of observation units (OUs) for eligible HF patients who cannot be discharged home from the emergency department (ED) may be one answer. They noted that the majority of patients who are admitted with an HF diagnosis require only symptomatic treatment, most often intravenous diuretics, and rarely receive invasive procedures or diagnostic tests that require admission. For a large subset of HF patients, those who do not respond to treatment quickly enough to be discharged home from the ED but who respond partially and do not worsen with ED treatment, the required monitoring and therapy can be provided safely and more economically in an observation setting.

Using the ED to triage HF patients for disposition home, to the OU or hospital room would require a two-level assessment in the ED. The first assessment would occur at presentation, as ED treatment begins and a diagnostic workup gets under way. Factors that put the patient at high risk would be identified at this time and many high-risk patients would be slated for hospital admission. Patients who are not severely ill and who respond partially to ED treatment are candidates for treatment in the OU for up to 24 hours, where their responses can be monitored and post-discharge planning can take place. Patients who are not high risk and who respond completely to ED treatment would be eligible for discharge home from the ED.

Collins et al estimated that up to 50 percent of the ED patients diagnosed with HF could safely be discharged home after a period of observation. Noting that 400,000 non-high-risk patients present at the ED with HF annually, “changing the disposition status from admission to an OU stay in merely 5 percent (of these cases) would result in savings of 80,000 hospital days and more than $80 million in savings annually in the United States.”

They proposed a randomized clinical trial to assess outcomes using the approach. During ED assessment, responsive patients could be discharged home after treatment and high-risk or nonresponsive patients would be admitted to the hospital. The subset of patients who are not high risk but who do not respond quickly and completely to ED treatment would be randomized to an OU with early discharge and follow-up or to hospital admission. Both groups would receive care according to standardized care plans. Endpoints would be mortality after six months, 30-day bed days/HF rehospitalization and all-cause rehospitalization. The researchers also would study time to symptom relief, adequacy of optimal medical therapy, documented patient education and compliance with early discharge follow-up.

Akshay Desai, MD, and Lynne Stevenson, MD, both of Brigham and Women’s Hospital in Boston, questioned whether the time is right for a randomized clinical trial such as Collins et al propose. One concern is that “[t]he term ‘observation’ itself is appropriate for chest pain of unknown etiology but seriously misleading when applied to heart failure decompensation, which may be mild but is never entirely benign.” They claimed that many steps are necessary to stabilize a decompensating patient, and therefore this transition hub between ED and discharge is not properly called an observation unit.

They questioned the triage structure that Collins et al set forth, asserting that the clinical measures the ED personnel would use to stratify risk were not validated for use in early discharge decisions. Geographical differences in personnel, practice methods and standards also concerned Desai and Stevenson, who emphasized the difficulty of implementing standardized protocols across different practice environments. And they stressed that there was little, if any, evidential data suggesting when a hypothetical HF patient in an OU would be ready to be discharged home, noting that available discharge risk scores applied only to patients who had been admitted to the hospital.

In addition, Desai and Stevenson emphasized that factors such as cognitive impairment, at-home support systems and ability to adhere to post-discharge instructions are all significant factors in outcomes for HF patients, and the Collins proposal does not address such issues.

They concluded, “The emergence of a single, uniformly effective strategy is not likely. An environment of such limited standardization of care and widespread practice variation may not be hospitable to implementation of a randomized trial, because the outcome of a fixed protocol may be outdated before it is analyzed.” Instead, they advocated continuing research into optimal care strategies and careful selection of clinical processes that will function best in each distinct clinical environment.