PROMPT-AHF and PACT-HF tried different approaches—EHR nudges and smarter transitional care. Success proved elusive.
PRAGUE, Czechia—Everyone working in heart failure (HF) care agrees: there’s no shortage of powerful therapies, with more looming on the horizon, but the trouble is getting them to patients. And compared with the explosion of studies testing new and established HF drugs and devices, along with a growing body of evidence supporting their use swiftly after a hospitalization, there remains a dearth of science establishing the best ways to make that happen.
Now, two such studies, both presented at the recent European Society of Cardiology Heart Failure 2023 Congress, have come up empty-handed.
PROMPT-AHF tested whether targeted electronic health record (EHR) alerts would improve the use of guideline-directed medical therapy (GDMT) among patients hospitalized for acute HF, compared with usual care. As reported by TCTMD, the earlier PROMPT-HF study showed that similar alerts improved GDMT use in the chronic heart failure setting. But as presenter Tariq Ahmad, MD, MPH (Yale University School of Medicine, New Haven, CT), showed, the EHR nudges yielded no such success in the acute setting.
“The results were completely negative,” Ahmad told TCTMD. The percent increase in the number of patients discharged on guideline-recommended meds was no different in patients whose care teams had received the reminders than in those whose teams had not. “We also found that only one in nine patients were discharged on all four pillars of guideline-directed medical therapy,” Ahmad lamented.
These are high-, high-risk patients and we ought to use this period when they’re hospitalized and when they’re returning home to really improve healthcare services . . . and to also help them live their best lives. Harriette Van Spall
Taking a very different approach, the PACT-HF trial, presented by Harriette Van Spall, MD, MPH (McMaster University, Hamilton, Canada), tested the ability of a multifaceted, patient-centered, transitional-care model combining guideline-recommended care with evidence-informed services to improve outcomes in patients being discharged after an HF hospitalization. Primary results of the trial, published in 2019, showed no difference between the approaches. Out to 3 years, as Van Spall showed here, patients randomized to the transitional-care intervention, “including long-term HF clinic visits, did not reduce the risk of composite all-cause death, readmission, or ED [emergency department] visit relative to usual care,” she said.
Investigators for both trials, said Adam DeVore, MD, MHS (Duke University, Durham, NC), who commented on the findings for TCTMD, need to be congratulated for trying to fill the “implementation” void. “Somebody even trying different things and trying to study them in a rigorous fashion really needs to be applauded because it’s the only way we’re going to figure out how to do this better,” he said.
Trials of new drugs and new devices typically hog the headlines by showcasing how such therapies might work in a controlled setting, he said, “but when we’re talking about trying to implement those in practice, we’re talking about all the challenges of trying to deliver care in the real world . . . . A lot of times that is not as exciting, not as sexy, but it has a huge public health impact in terms of trying to make people feel better and live better lives.”
In PROMPT-AHF, 1,012 patients hospitalized for acute HF with reduced ejection fraction (HFrEF) not already taking the four classes of GDMT were randomized within the first 24 hours to usual care, or to a pop-up alert in the EHR reminding physicians of the evidence-based order set for each of the recommended medications. Physicians receiving the alert had the option to open the order set or not.
Not only was the proportion of patients taking GDMT at discharge similar between groups, but also rates of 30-day rehospitalization, percent increase in each class of GDMT, number of prescriptions filed, and mortality at 1 year were all unchanged.
Speaking with TCTMD, Ahmad pointed to what he called “a behavioral economic question” in heart failure whereby no single person is tasked with getting patients on GDMT and it’s not a performance standard in the same way it is, for example, in the post-ACS setting. “So, no one is held accountable,” he said. “There’s also the question of clinician burnout because all these alerts that are happening on the inpatient side, mostly they don’t happen as much on the outpatient side, and it just overwhelms you.”
Somebody even trying different things and trying to study them in a rigorous fashion really needs to be applauded because it’s the only way we’re going to figure out how to do this better. Adam DeVore
One solution would be to tie reimbursement more closely to adherence to GDMT initiation, then compensating that appropriately, Ahmad said. “It’s really, to me, a tragic situation where we’re basically doing low-value interventions far before we’re doing high-value interventions, which I think is wrong. We really should be focusing much more on getting the best possible quality. Acute heart failure hospitalization is a good time to get people who qualify on the right therapies because those are the high-risk patients.”
PROMPT-AHF, he said, helps to point investigators and industry players in some new directions, however. For one, he believes healthcare systems are not using electronic health records to their full potential, with alerts and nudges rolled out without rigorous testing. “They have launched all these nudges in the EHR that have overwhelmed clinicians, and this has contributed to burnout without really using the EHR in a way that could really help clinicians and patients.”
Seeking clinician input on how to improve these alerts and nudges is long overdue, Ahmad added.
Looking more broadly, he continued, “the incentives for heart failure care need to change as well because these are relatively inexpensive high-value therapies that we’re bypassing in patients that could lead to pretty dramatic reductions in mortality, far more than you see for ACS therapy, for example. I think there’s really a need to rethink our approach to heart failure from quantity to quality.”
Taking a very different approach, the intervention tested in PACT-HF was a transitional-care model involving a patient-centered discharge summary with HF care plan, a follow-up appointment with a primary care physician within 1 week, HF nurse home visits beginning in the first week and repeated for a minimum of 4 to 6 weeks, as well as a first HF clinic visit within the first month, with additional visits continuing indefinitely.
Just as was seen at 1 year, patients treated according to the transitional-care model had similar rates of all-cause death, hospitalization, or ED visits at 3 years compared with those followed according to usual care. Uptake of GDMT at 3 years was nearly identical in each group.
Of note, said Spall, more than half of the trial patients had died within that 3-year window. Accounting for death as a competing risk, investigators did see a decreased risk of all-cause ED visits in women and a decrease in rates of HF readmissions overall, but no change in the risk of all-cause readmissions.
We’re basically doing low-value interventions far before we’re doing high-value interventions, which I think is wrong. Tariq Ahmad
Going after an endpoint that included death and all-cause hospitalizations was “ambitious,” Spall acknowledged, adding: “those are notoriously difficult to move the needle on.”
That said, she continued, outcomes important to HF patients often differ from those valued by clinicians, regulators, and public health authorities, and complex health services interventions are typically geared towards what matters to patients. “So if we are to implement them, we need to be careful that we understand what outcomes they do affect, and we have shown in the primary analysis that they improve the patient’s experience of care,” she said. “As long as we know what these services do, we can make informed decisions from a healthcare system perspective.”
She pointed to the fact that more than 90% of patients in PACT-HF had either a hospital readmission or emergency department visit, or were dead by 3 years. “That’s another point that we need to really wrap our heads around. These are high-, high-risk patients and we ought to use this period when they’re hospitalized and when they’re returning home to really improve healthcare services, improve the implementation of therapies that we know improve outcomes, and to also help them live their best lives.”
Negative Trials Still Informative
Speaking with TCTMD, Spall suggested that despite the disappointing results of both trials, it might be worth exploring some combination of the two.
“I think to make the results more reproducible and generalizable, and to show a treatment effect, one of the solutions might be to combine a patient-centered, transitional-care service with an algorithmic-driven approach that may be facilitated by digital health technologies such as electronic health record alerts,” she said. “I think that the inpatient setting might be too brief to really effectuate change, and you need a longer duration of time in which the alerts could impact behavior.”
To TCTMD, DeVore noted that the problem of getting proven therapies to patients, at the optimal time and the most effective dose, is not limited to heart failure. These two trials advance the field, he said, by illuminating problems—such as the inefficacy of EHR nudges, in their current form—which also opens up new avenues for improvement.