in a randomized trial conducted at several centers in the same health care system.
The results of the PROMPT-AHF trial, which assigned such patients to have or not have the GDMT-promoting physician nudges as part of their in-hospital management, were “not entirely surprising,” Tariq Ahmad, MD, MPH, of Yale University, New Haven, Conn., said in an interview.
“We have created an environment in the hospital that makes care quite fractured for patients with heart failure,” he said. “They are cared for by many different clinicians, which leads to well-known behaviors such as diffusion of responsibility.”
Moreover, many clinicians focus on stabilizing patients “rather than starting them on a comprehensive set of medications, which most think should be done after discharge,” Dr. Ahmad added.
“Importantly, there has been a logarithmic increase in alerts while patients are hospitalized that has caused clinician burnout and is leading to even very important alerts being ignored.”
Likely as a result, the trial saw no significant difference between the alert and no-alert groups in how often the number of GDMT prescriptions rose by at least one drug class, whether beta blockers, renin-angiotensin system inhibitors, mineralocorticoid receptor antagonists, or SGLT2 inhibitors. That happened for 34% of patients in both groups, reported Dr. Ahmad at the Heart Failure Association of the European Society of Cardiology (HFA-ESC) 2023 sessions
Nor was there a difference in the secondary endpoint of increased number of GDMT meds or escalated dosage of prescribed GDMT drugs.
GDMT ‘uncommon’ in AHF
In an earlier trial in outpatients with chronic HF, conducted by many of the same researchers, use of a targeted EHR-based alert system was associated with significantly higher rates of GDMT prescriptions 30 days after discharge, compared with usual care, Dr. Ahmad observed in his presentation.
Because GDMT is similarly “uncommon” among patients hospitalized with acute HF, the team designed the current trial, a test of the hypothesis that a similar system of nudges would lead to higher rates of prescriptions of the four core GDMT drug classes.
The study enrolled 920 adults with acute HF, an EF of 40% or lower (their median was 28%), and NT-proBNP levels higher than 500 pg/mL. The patients received IV diuretics for the first 24 in-hospital hours and were not taking medications from any of the four core HF drug classes. Their mean age was 74, 36% were women, and 25% were Black.
Physicians of patients who were randomly assigned to the intervention received the alerts as they entered information that involved ejection fraction, blood pressure, potassium levels, heart rate, glomerular filtration rate, and meds they were currently or should be taking, “along with an order set that made ordering those medications very easy,” Dr. Ahmad said.
“There was absolutely no evidence that the alert made any difference. There were zero patients on all four classes of GDMT at baseline, and at the time of discharge, only 11.2% of patients were on all four pillars – essentially, one in nine patients,” Dr. Ahmad said. Nor were there any subgroup differences in age, sex, race, ejection fraction, type of health insurance, or whether care was provided by a cardiologist or noncardiologist physician.
The study was limited by having been conducted within a single health care network using only the Epic EHR system. The alerts did not go exclusively to cardiologists, and patient preferences were not considered in the analysis. Also, the study’s alerts represented only some of the many that were received by the clinicians during the course of the trial.