CHICAGO — People with acute heart failure had improved clinical outcomes at hospitals where emergency department (ED) clinicians were provided a point-of-care decision support tool that guides risk stratification and care.
In the stepped-wedge, cluster-randomized trial COACH, the tool determined patients at low-risk to be candidates for the pathway of early discharge and prompt outpatient follow-up, whereas high-risk individuals were suggested for hospital admission. As for people deemed to be at intermediate risk, ED clinicians were advised to admit patients who had an intermediate-to-high risk, and to consider early discharge for patients at low-to-intermediate risk.
EDs given this tool saw significant improvements over usual care in terms of combined all-cause death and cardiovascular hospitalization within 30 days after presentation (12.1% vs 14.5%, adjusted HR 0.88, 95% CI 0.78-0.99) and at 20 months (36.1% vs 51.5%, adjusted HR 0.95, 95% CI 0.92-0.99), according to Douglas Lee, MD, PhD, of University of Toronto.
“Implementation of this approach may lead to a pathway for early discharge from the hospital or ED and improved patient outcomes,” Lee said in a presentation at the American Heart Association (AHA). The findings were published simultaneously in the New England Journal of Medicine.
Reassuringly, fewer than six deaths or hospitalizations occurred in low- or intermediate-risk patients before a first outpatient visit. “Our findings support the concept that not all patients who present to the emergency department with heart failure require hospitalization,” Lee and colleagues said.
AHA session discussant Harriette Van Spall, MD, MPH, of McMaster University in Hamilton, Ontario, Canada, called COACH a “winning trial” that is “cause for celebration.”
She said that in the past, health system interventions had shown limited utility in improving outcomes for the high-risk group of people with decompensated heart failure. One example is her group’s PACT-HF transitional care intervention — comprising early home visits and heart failure clinic care titrated to patient risk — that failed to reduce events at 10 hospitals in Ontario.
Lee reported that his group’s low-risk pathway sent early-discharged heart failure patients to rapid follow-up at a clinic staffed by a nurse and supervised by a cardiologist. There, patients received outpatient care — such as diagnostic testing and medication titration — for up to 30 days after leaving the hospital.
It will be important to assess how the study intervention changed the number of heart failure patients admitted versus discharged, and what the implications might be for healthcare utilization and resource allocation, according to Van Spall.
The COACH trial was conducted at 10 centers in Ontario, Canada.
EDs were randomly assigned staggered start dates for one-way crossover from usual care to the intervention phase, when they were given access to the validated risk algorithm EHMRG30-ST to determine the risk of an acute heart failure patient within 7 or 30 days, and the appropriate discharge versus admission decision for each person.
Participants were hemodynamically stable and were not eligible if they had new heart failure, active cardiac ischemia, uncontrolled arrhythmias, worsening renal function, or barriers to follow-up.
Altogether, there were 2,972 acute heart failure patients enrolled across hospitals during the control phase, and 2,480 during the intervention period. The two groups were well-matched in their baseline characteristics, with median age being 78 and women accounting for 45% of the cohort.
Lee’s group acknowledged that they could not tell which part of the intervention had the greatest effect on patient outcomes.
“Our intervention was complex and multipronged, with several potential mechanisms of benefit. Physicians tend to underestimate risk for those with the worst prognosis, which could lead to inadvertent discharge of high-risk patients. In addition, earlier follow-up may be advantageous, because readmissions occur frequently after shorter index hospital admissions,” they wrote.
They cautioned that COACH was further limited by the lack of race or ethnic information for the majority of participants, and the fact that investigators were unable to capture follow-up outcomes for all these patients.
The study was funded by the Ontario Strategy for Patient-Oriented Research Support Unit, the Ted Rogers Centre for Heart Research, the Peter Munk Cardiac Centre, the Canadian Institutes of Health Research (CIHR), and ICES.
Lee disclosed no relationships with industry.
Van Spall disclosed support from CIHR and Stroke Foundation.