Medical Therapy for Heart Failure With Reduced Ejection Fraction: The CHAMP-HF Registry

Introduction

For patients with chronic heart failure with reduced ejection fraction (HFrEF), contemporary therapy includes multiple medications proven to decrease mortality and hospitalization rates in large randomized controlled trials (1,2). The robust survival benefits of these medical therapies in clinical trials have been generally shown to apply in routine clinical practice (3). Accordingly, these medications form the cornerstone of contemporary evidence-based HFrEF care and are supported by class I indications in clinical treatment guidelines (1,2). Guidelines also recommend target doses for each medication, as tolerated, based on doses tested in landmark studies (1,2).

Despite proven benefits and strong guideline recommendations, medication use and dosing in routine clinical practice have traditionally fallen short of levels achieved in clinical trials (4,5). As rates of mortality and morbidity for the general HFrEF population remain high, improving real-world use of proven medications remains a critically important quality improvement initiative (6–8). However, the utilization patterns of chronic HFrEF medical therapies and the barriers to their uptake and dosing in contemporary U.S. outpatient practice are poorly understood. Such information is even more relevant now in light of recent regulatory approval of sacubitril/valsartan and well-described concerns over traditionally slow adoption of novel medical therapies (9,10). Better understanding of current practice patterns, gaps in medication delivery, and barriers to receiving guideline-directed medical therapy (GDMT) are critical to development of targeted initiatives aimed at improving patient outcomes and quality of care. In this context, the CHAMP-HF (Change the Management of Patients with Heart Failure) registry offers the opportunity to study a contemporary U.S. outpatient HFrEF cohort and explore patterns of GDMT use and dosing, the clinical and social patient profiles associated with GDMT use, and the patient-level factors associated with medication use and target dosing.

Methods

Results

Baseline medication use and dose

The present analysis included 3,518 HFrEF patients with available baseline medication data from 150 U.S. primary care and cardiology practices. Overall, mean age of the study cohort was 66.4 ± 12.6 years, 29.2% were females, 74.9% were white, and mean EF was 29 ± 8%. For each class of medical therapy, proportions of patients: 1) with a contraindication; 2) treated; and 3) without a contraindication but not treated are displayed in the Central Illustration. For all therapies, the percentage of patients with a documented absolute contraindication was low, with the highest rate of 1.8% for ACEI/ARB/ARNI. Among eligible patients, 2,536 patients (73.4%), 2,351 patients (67.0%), and 1,163 patients (33.4%) were treated with ACEI/ARB/ARNI, beta-blocker, and MRA, respectively. ACEI/ARB was prescribed to 2,107 eligible patients (60.5%) and ARNI to 452 patients (13.0%). Among patients receiving each medication, <30% of patients were prescribed target doses of ACEI/ARB/ARNI or beta-blocker therapy (Central Illustration). In contrast, >75% of patients receiving an MRA were prescribed the target dose. Among patients eligible for all classes of medication, 755 patients (22.1%) were simultaneously prescribed some dose of ACEI/ARB/ARNI, beta-blocker, and MRA therapy, and 37 patients (1.1%) were simultaneously prescribed target doses of all 3 therapies. Medication use and doses among patients enrolled in the first half (December 15, 2015 to August 22, 2016) were similar to those among patients enrolled in the second half (August 23, 2016 to March 6, 2017) (Online Tables 2 and 3). In contrast, with the exception of ACEI/ARB, use of medication was substantially lower among patients enrolled from family medicine or internal medicine practices (n = 485) than among patients enrolled from cardiology outpatient practices (n = 2821) (Online Table 4). Among patients treated with each medication, doses were relatively similar by practice type (Online Table 5).

Discussion

In this large contemporary U.S. registry of stable outpatients with HFrEF, significant gaps in guideline-directed use and dosing of evidence-based medications remain. Despite <2% of patients having a documented absolute contraindication to any specific therapy, use of each guideline medication fell below 75%. MRA and ARNI therapies were particularly underused, with only 33% and 13% of eligible patients treated, respectively. When GDMT was used, it tended to be used at lower doses with the majority of patients prescribed subtarget doses of ACEI/ARB/ARNI and beta-blocker, and high proportions prescribed <50% of target dose. Only 1% of eligible patients were simultaneously treated with target does of ACEI/ARB/ARNI, beta-blocker, and MRA therapy, and <25% of patients simultaneously received any dose of all 3 medications. For each medication class, patient profiles varied with dose of medications prescribed. Notably, with exception of MRA therapy, rates of prior HF hospitalization and NYHA functional class III/VI status were generally inversely related to medication dose, with highest rates among patients receiving the lowest doses. In multivariate analysis, for each medication class, several patient characteristics were independently associated with patients receiving each therapy and the specific dose.

To our knowledge, we present the most comprehensive and contemporary analysis of outpatient HFrEF medical therapy dose in U.S. clinical practice. Although prior HF registry data informed previous quality improvement efforts, recent programs have focused on patients hospitalized with HF, a notably different population from the stable outpatients recruited in landmark registration trials for GDMT (13,14). In this respect, the current CHAMP-HF data from stable ambulatory HFrEF patients evaluates use of GDMT in a U.S. cohort more representative of the populations from which clinical trial evidence was derived. In addition, the present analysis offers several strengths that provide more granular insight into the pattern of current GDMT use in the United States. First, the registry included pre-specified and detailed dosing information to allow description of target and subtarget doses, as well as associated patient profiles. Importantly, data capture included ARNI therapy, to be consistent with class I recommendations from the recent treatment guideline update (2). Second, recognizing the potential contributions of social factors to the use of HFrEF therapy, CHAMP-HF offers a comprehensive socioeconomic characterization of the patient population. Third, to mitigate potential influences of medical record accuracy and completeness of registry medication data, CHAMP-HF included an electronic case report form with pre-specified capture of medication data, presence of contraindications, and multiple questions regarding tolerability. Fourth, rigorous multivariate modeling inclusive of both clinical and social factors was performed to assess independent patient-level associations with medication use and dose, and hierarchical models were used to account for patient clustering by enrolling study site.

Prior data examining dose of GDMT in U.S. outpatient clinical practice comes largely from the IMPROVE HF (Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting) study (15). Specifically, this large-scale quality improvement initiative found rates of baseline target dosing of ACEI/ARB, beta-blocker, and MRA among eligible patients to be 36.1%, 20.5%, and 74.4%, respectively (16). Despite IMPROVE HF data reflecting HFrEF care approximately a decade ago, contemporary target doses in CHAMP-HF are remarkably similar for beta-blockers (27.5%) and MRA (76.6%) and are considerably lower for ACEI/ARB/ARNI (16.7%). Overall use (i.e., any dose) of GDMT within the 2 cohorts was also similarly low (e.g., baseline use of MRA among eligible patients was 34.5% in IMPROVE HF and 33.4% in CHAMP-HF) (15).

Given that significant gaps in provision of HFrEF medical therapy persist, next steps must focus on why underuse and underdosing continues. Recent prospective investigations strongly suggest that inadequate rigor of outpatient follow-up may not be the causative factor. For example, the prospective BIOSTAT-CHF (Biology Study to Tailored Treatment in Chronic Heart Failure) program recruited 2,100 HFrEF patients in Europe, who were either not receiving ACEI/ARB and beta-blocker at baseline or were receiving <50% of target doses (4). Despite the pre-specified intent of the program to encourage medication up-titration, at median follow-up of 21 months, only 22% and 12% of patients achieved target doses of ACEI/ARB and beta-blocker, respectively. Likewise, findings from the GUIDE-IT (Guiding Evidence Based Therapy Using Biomarker Intensified Treatment in Heart Failure) trial showed that, despite an intensive outpatient follow-up regimen that included a median 12 clinic visits over 12 months and care guided by natriuretic peptide levels, only 15% of patients reached target doses of beta-blockers, and 31% reached target doses of ACEI/ARB (17).

As a complement to these recent prospective experiences, the current study from CHAMP-HF directed attention to the role of patient-level factors in the likelihood of receiving target doses. Although results for specific therapies were varied, lower systolic blood pressure and other unfavorable prognostic factors such as more severe NYHA functional class, older age, chronic renal insufficiency, and recent hospitalization for HF generally favored lower medication use or dose. These findings are compatible with the previously described “risk-treatment paradox,” where HF patients with greatest need are less likely to receive appropriate therapy (18). Although these observations from CHAMP-HF may be partly intrinsic to greater severity of disease and potential overlapping side effects of many HFrEF medications (e.g., hypotension, worsening renal function), it is conceivable that GDMT underuse and underdosing in higher risk patients may be driven by disproportionate emphasis among providers on potential patient destabilization with medication changes rather than clinical benefits of therapy. Although clinical judgment remains paramount, randomized evidence supports the relative safety of higher doses of GDMT as compared with lower doses. For example, a meta-analysis of ACEI/ARB trials found that, compared to low-dose ACEI/ARB therapy, patients randomized to higher doses experienced improved survival without excess rates of drug discontinuation (19). Similarly, despite enrolling a severely symptomatic HFrEF population, the COPERNICUS (Carvedilol Prospective Randomized Cumulative Survival) trial demonstrated the relative safety and efficacy of carvedilol initiation among >2,200 patients (20).

Numerous studies have explored interventions to improve provision of medical therapy by providers and/or to improve medication adherence among patients (21–23). Although the broad application of most data remain limited by small numbers of study centers, limited numbers of patients, and heterogeneity of interventions, evidence does support the potential impact of some strategies such as quality metric checklists, pharmacist-driven medication programs, and patient education (21–23). Likewise, the recent American College of Cardiology Expert Consensus Document provides recommendations for addressing challenges to appropriate use of HFrEF medical therapy, including clinical, economic, and social barriers (12). Nonetheless, CHAMP-HF highlights the persistently unmet need for generalizable strategies to improve widespread use and dosing of GDMT. Specifically, the current “risk-treatment paradox” findings suggest that patients with the highest clinical risk may particularly benefit from future quality improvement strategies. Moreover, although most patients received care at cardiology practices, larger gaps in GDMT use among patients enrolled at family medicine and internal medicine clinics suggest such practices may have the most to gain from future efforts to improve quality of care. Aside from more conventional approaches, novel interventions grounded in behavioral economics or technological innovation (e.g., mobile applications) may be particularly promising and are being tested (24).

Conclusions

In this contemporary registry of outpatients with chronic HFrEF, there remain significant gaps in guideline-directed use and dosing of HFrEF medications. Despite guidelines, educational efforts, and quality improvement initiatives, comparison with prior registry data approximately a decade ago show that outpatient use and dosing of GDMT has generally not improved. Several clinical factors show strong independent associations with baseline use and dose of GDMT, and these findings may inform targeted efforts toward optimal implementation of medical therapy for HFrEF. In the setting of continued high rates of morbidity and mortality in the general HFrEF population, effective strategies to improve use and target dosing of outpatient medical therapy remain urgently needed.

Footnotes