Trial DesignsDesign and pilot implementation for the BETTER CARE-HF trial: A pragmatic cluster-randomized controlled trial comparing two targeted approaches to ambulatory clinical decision support for cardiologists
Section snippets
Background
Heart failure is a major public health problem in the United States that affects over 6 million people, costs over $30 billion annually,1 and is a leading cause of hospitalization and re-admission.2 Despite proven benefits in randomized trials, class I guideline recommendations,3 and published clinical performance measures,4 patients with heart failure and reduced ejection fraction (HFrEF) are often not prescribed guideline-directed medical therapy (GDMT).5., 6., 7. An estimated 68,000 deaths
Overview of BETTER CARE-HF trial design
BETTER CARE-HF is a pragmatic, cluster-randomized, three-arm intervention trial to assess the effectiveness of two targeted CDS intervention tools (alert and automated EHR message) as compared to usual care on the primary outcome of MRA prescription. Patients were randomized as a cluster at the level of the cardiologist. Figure 1 displays study flow. The study has a pragmatic design, with all interventions, data collection, and data extraction occurring via the electronic health record. The
Study population
Eligible patients included those with an EF ≤ 40% on their most recent echocardiogram, no active prescription for an MRA, age > 18, and an outpatient cardiology encounter during the study period. We excluded patients with ventricular assist device, a diagnosis of cardiac amyloid, and/or on hospice. To ensure CDS tools are only targeting patients appropriate for MRA therapy, we utilized the following additional exclusion criteria specific to MRA prescribing:9,19 most recent systolic blood
Study setting
New York University Langone Health (NYULH) is a large, urban health system with >60 cardiology practices and >250 practicing cardiologists in diverse practice settings ranging from community health centers to tertiary academic faculty group practice. All practices have the same electronic health record (Epic, Verona, WI). Echocardiogram reports at most locations include a structured field for EF that is imported as discrete data into the electronic health record, allowing for automated
Interventions
BETTER CARE-HF is a 3-arm trial with 2 interventions (alert and automated EHR message), and one control arm.
Methods guiding intervention development
The electronic algorithm upon which patients were identified for CDS interventions has been previously described by our group. This algorithm was created using an iterative process that included validation manual chart review, multidisciplinary discussion, and algorithm refinement.20
We used a 4-part process to inform the development and design of the two CDS interventions. First, we performed semistructured individual interviews with cardiologists to inform CDS design. We then performed pilot
BETTER CARE-HF trial randomization
We used a cluster-randomization design by cardiology provider in a 1:1:1 fashion (alert vs automated EHR message vs control). We further stratified randomization by cardiology provider subspecialty (ie, general cardiology, electrophysiology, advanced heart failure, interventional) and provider volume (by quartile). Randomization was conducted using an electronic random number generator. The two pilot practice sites were not excluded from randomization as they were chosen specifically for their
Outcomes of interest
Table II depicts outcomes of interest. The primary outcome is the proportion of eligible patients in each arm who are prescribed MRA at the end of the study period we will assess for MRA prescription by using data on signed orders. Secondary clinical outcomes will include assessment of BB and ACEI/ARB/ARNI prescriptions and prescribing rate by provider (Table II). We will also assess markers of CDS implementation, including time to prescription and provider engagement with CDS tools. Safety
Statistical considerations
Patient characteristics in each CDS intervention arm will be compared to usual care arm using χ2 test or Fisher's exact for categorical variables, and t test or Wilcoxon rank sum for continuous variables. To account for the clustered nature of the data, logistic mixed effect model will be used to assess MRA prescription among different groups at the patient level. For secondary outcomes, linear mixed effect model will be performed to measure the prescription rate change over time among
Discussion
Shortfalls in prescribing of proven therapies for HFrEF, particularly mineralocorticoid receptor antagonist (MRA) therapy, account for several thousand preventable deaths per year nationwide.8 Targeted electronic CDS tools are a low cost, scalable method with the potential to improve prescribing, but the optimal format and timing of CDS tools in this setting is unknown. BETTER CARE-HF is a pragmatic, cluster-randomized, 3-arm clinical trial comparing 2 different selective and individualized
Funding
This study was funded by the National Center for Advancing Translational Sciences (NIH/NCATS UL1TR001445). A. Mukhopadhyay is supported by the National Heart, Lung, and Blood Institute (NIH/NHLBI 2T32HL098129-11). We thank Allen Thorpe for funding the learning health system program and NYU Langone Health for providing in-kind contributions.
Disclosures
H.R. Reynolds discloses in-kind donations for unrelated research from Abbott Vascular, Philips, Siemens. L.M. Philips discloses consulting for Novo Nordisk. All other authors have no relevant disclosures.
Acknowledgment
We thank Felicia Mendoza and Jay Stadelman for assistance with semistructured interviews, and William King and Zaina Laarousi Tribek for assistance with randomization protocol.
References (25)
- et al.
2020 ACC/AHA clinical performance and quality measures for adults with heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures
J Am Coll Cardiol
(2020) - et al.
Medical Therapy for Heart Failure With Reduced Ejection Fraction: The CHAMP-HF Registry
J Am Coll Cardiol
(2018) - et al.
Interrupting providers with clinical decision support to improve care for heart failure
Int J Med Inform
(2019) - et al.
Electronic alerts to improve heart failure therapy in outpatient practice: a cluster randomized trial
J Am Coll Cardiol
(2022) - et al.
2021 Update to the 2017 ACC expert consensus decision pathway for optimization of heart failure treatment: answers to 10 pivotal issues about heart failure with reduced ejection fraction: a report of the American College of Cardiology Solution Set Oversight Committee
J Am Coll Cardiol
(2021) - et al.
Clinical Decision Support to Efficiently Identify Patients Eligible for Advanced Heart Failure Therapies
J Card Fail
(2017) Chapter two - Cognitive load theory, in psychology of learning and motivation
(2011)- et al.
Care gaps in adherence to heart failure guidelines: clinical inertia or physiological limitations?
JACC Heart Fail
(2020) - et al.
Heart disease and stroke statistics-2020 update: a report from the American Heart Association
Circulation
(2020) - et al.
Association of changing hospital readmission rates with mortality rates after hospital discharge
JAMA
(2017)
2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on clinical practice guidelines and the Heart Failure Society of America
Circulation
Improving evidence-based care for heart failure in outpatient cardiology practices: primary results of the registry to improve the use of evidence-based heart failure therapies in the outpatient setting (IMPROVE HF)
Circulation
Cited by (3)
Impact of Visit Volume on the Effectiveness of Electronic Tools to Improve Heart Failure Care
2024, JACC: Heart FailureCluster-Randomized Trial Comparing Ambulatory Decision Support Tools to Improve Heart Failure Care
2023, Journal of the American College of Cardiology