Patient, Provider, and Practice Characteristics Associated With Sacubitril/Valsartan Use in the United States

Introduction

Sacubitril/valsartan, an angiotensin receptor neprilysin inhibitor (ARNI), was compared with enalapril in 8442 patients with chronic heart failure with reduced ejection fraction (HFrEF).¹ Patients treated with sacubitril/valsartan demonstrated a 20% reduction in the primary outcome of cardiovascular death or HF hospitalization (hazard ratio, 0.80; 95% CI, 0.73–0.87). These data led to a change in the HF guidelines.² Patients with prior or current symptoms of chronic HFrEF are recommended to receive an angiotensin-converting enzyme (ACE) inhibitor, angiotensin receptor blockers (ARB), or ARNI. In addition, patients with chronic HFrEF and New York Heart Association class II or III symptoms who can tolerate an ACE inhibitor or ARB are recommended to have replacement of the ACE inhibitor or ARB by an ARNI.

A prior analysis highlighted that if ARNI therapy was comprehensively utilized in eligible patients with chronic HFrEF ≈28 484 deaths per year could be prevented in the United States alone.³ Instead, adoption of sacubitril/valsartan into clinical practice has been slow since approval by the US Food and Drug Administration in July 2015. For example, in a study of data from the Get With The Guidelines–Heart Failure (HF) Registry, 2% of eligible hospitalized patients were prescribed sacubitril/valsartan at hospital discharge.⁴ Outside of this, there are limited data describing the contemporary use of sacubitril/valsartan. Moreover, no studies have yet documented effective implementation strategies for the updated guidelines and what patient or provider factors are associated with higher rates of sacubitril/valsartan use. We used data from CHAMP-HF (Change the Management of Patients With Heart Failure), a large US outpatient registry of patients with chronic HFrEF, to evaluate patient, provider, and practice characteristics associated with adoption of sacubitril/valsartan.

METHODS

The data, analytical methods, and study materials will not be made available to other researchers for purposes of reproducing the results or replicating the procedure. CHAMP-HF is a study of outpatients with chronic HFrEF (left ventricular ejection fraction [LVEF] ≤40%), the details of which have been previously published.⁵ To be eligible, all participants must be receiving treatment with at least 1 oral pharmacotherapy for management of HFrEF. Participants are being recruited and followed as part of routine outpatient HF management at 152 sites across the United States. The registry is a prospective observational cohort study without intervention and no attempt to influence clinical practice. Patients are not required to have specific follow-up visits to the investigator. Instead, postenrollment clinical data is reported by the sites at designated data abstraction time points: 30 days and 3, 6, 12, 18, and 24 months.

The study is ongoing, and the current analysis includes data collected between the start of the registry and the most recent data extraction, December 2015 and August 2017. In the current analysis, we considered patients on sacubitril/valsartan at various time points: (1) initiated before the baseline registry visit or at the baseline visit and (2) noted to be on sacubitril/valsartan at a postenrollment data abstraction time point that occurred before August 2017. The majority of the analyses focus on the use of sacubitril/valsartan before the baseline registry visit or at the baseline visit as postenrollment data is still being collected in the registry. We utilized additional data from the CHAMP-HF registry, including baseline characteristics: demographics, insurance status, level of education, employment status, symptom burden, medical history, medication use, vital signs, and laboratory data. We also included practice and provider information collected from investigators at the start of the registry. We excluded patients with a contraindication to sacubitril/valsartan, such as history of hyperkalemia or angioedema, which was prospectively collected throughout the study. We also excluded patients with missing demographics and sites that enrolled <10 patients.

Patients enrolled in the CHAMP-HF registry signed written informed consent before collection of study data and Institutional Review Board approval was obtained. The CHAMP-HF registry is sponsored by the Novartis Pharmaceuticals Corporation (East Hanover, NJ). Data are managed by the United BioSource Corporation (Blue Bell, PA), and the Duke Clinical Research Institute (Durham, NC) is the data analytic center.

For the primary analysis, we categorized patients into 3 groups according to renin-angiotensin system inhibition treatment at the baseline visit: ACE inhibitor/ARB use, sacubitril/valsartan use, and neither. At this time point, patients could have been on sacubitril/valsartan before registry enrollment or started on sacubitril/valsartan at the baseline registry visit. We described baseline characteristics of the study population by study group using frequencies with percentages for categorical variables and means with SDs for continuous variables. We tested for differences between the groups using χ² tests for categorical variables and Kruskal-Wallis tests (or Ftests from an ANOVA analysis if the distribution of data was normally distributed) for continuous variables and repeated these analyses for provider and practice characteristics, stratified by quartile of sacubitril/valsartan use. We then described the proportion of patients prescribed sacubitril/valsartan at the baseline visit over time. Data were analyzed by quarter, though the first and last quarters were not exactly 3 months. The first quarter in 2016 included data from the beginning of the CHAMP-HF registry, December 2015. The last quarter only included patients enrolled before the data cut in August 2017. We described variation in the use of sacubitril/valsartan at enrollment in the study by site.

To identify provider and practice characteristics associated with sacubitril/valsartan use, we used a hierarchical logistic regression model to account for the clustering of patients by provider. The unit of analysis was the patient, and the binary outcome was sacubitril/valsartan use. The model was adjusted for patient characteristics, as well as provider and practice characteristics. Patient-level variables were selected clinically and included age, sex, race, Hispanic ethnicity, insurance status, level of education, HF hospitalization within 12 months, ventricular arrhythmias, atrial fibrillation, cardiac resynchronization therapy, coronary disease, hypertension, diabetes mellitus, chronic kidney disease, systolic blood pressure, pulse, obesity, chronic obstructive pulmonary disease, depression, cigarette smoking, and hyperlipidemia. Rates of missing data for most variables were <1.0% with a few exceptions (New York Heart Association 3.2%, blood pressure 4.5%, and heart rate 6.1%), and single imputation was performed based on a fully conditional specification in SAS PROC MI. The provider and practice characteristics were selected using backward elimination. Variables with a P value >0.05 were removed based on the highest P value first. The remaining assessment was completed with the remaining variables and model selection continued until all variables had a P value <0.05. Practice site was entered as a random effect to account for correlation between patients at the same site.

We also estimated the cumulative adoption rate of sacubitril/valsartan over follow-up by adding sacubitril/valsartan use at the baseline visit with uptake during follow-up. The latter was estimated by using the Kaplan-Meier methods to account for censoring. All analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC).

RESULTS

The study population included 4216 patients with HFrEF from 121 sites. Of the 4216 patients, 2506 (59%) were prescribed an ACE inhibitor/ARB at baseline, 616 (15%) were prescribed sacubitril/valsartan at baseline, and 1094 (26%) were prescribed neither. Table 1 shows the baseline characteristics of the study population stratified by study group. Patients prescribed sacubitril/valsartan were younger and less likely to be of Hispanic ethnicity compared with the other 2 groups. Patients prescribed sacubitril/valsartan were more likely to have managed care and private insurance, more likely to have a graduate/professional degree and 4-year college degree, and less likely to be unemployed compared with the other 2 groups. They were also more likely to be treated with β-blockers, aldosterone antagonists, ivabradine, cardiac resynchronization therapy, and an implantable cardioverter defibrillator, as well as to have a lower LVEF compared with the other 2 groups. Patients on neither ACE inhibitor/ARB nor ARNI were more likely to have a history of chronic kidney disease and higher serum creatinine values compared with the other 2 groups. Figure 1 displays the temporal trends in the proportion of patients prescribed sacubitril/valsartan at the baseline visit in the registry over time. The proportion increased from 9% to 24% during the study period.

• Download figure
• Download PowerPoint

Figure 2 displays the proportion of patients prescribed sacubitril/valsartan at the baseline visit by practice. At 75% of sites, ≤15% of eligible patients were prescribed sacubitril/valsartan. Table 2 shows provider characteristics stratified by use of sacubitril/valsartan at the baseline visit. We found no statistically significant association between investigator specialty, cardiology subspecialty, or years of practice and quartile of sacubitril/valsartan prescription. However, cardiologists and HF subspecialists had higher rates of sacubitril/valsartan use. Table 3 shows practice characteristics stratified by use of sacubitril/valsartan at the baseline visit. Larger practices, as assessed by number of physicians, cardiologists and advanced practice providers (ie, nurse practitioners and physician assistants), were associated with the highest quartile of sacubitril/valsartan prescription. For example, the mean number of cardiologists per practice was 14.1 (95% CI, 8.8–19.5) in the highest quartile versus 4.8 (95% CI, 2.5–7.1) in the lowest, P value=0.009. The mean number of advanced practice providers per practice was 6.6 (95% CI, 4.1–9.2) in the highest quartile versus 2.7 (95% CI, 1.0–4.4) in the lowest, P value=0.017. There was no statistically significant association between hospital affiliation or components of HF management and quartile of sacubitril/valsartan use. After adjustment for patient characteristics, only the number of advanced practice providers was independently associated with sacubitril/valsartan prescription, adjusted odds ratio 1.08 (95% CI, 1.03–1.14).

• Download figure
• Download PowerPoint

Figure 3 displays the cumulative use of sacubitril/valsartan over follow-up stratified by baseline ACE inhibitor/ARB use. There was a large difference in the initiation of sacubitril/valsartan at baseline (ie, 15% of the cohort not on an ACE inhibitor/ARB at baseline initiated on sacubitril/valsartan compared with a cohort that remained on ACE inhibitor/ARB at baseline). However, the 2 curves suggest the rate of adoption of sacubitril/valsartan over time was similar regardless of baseline ACE inhibitor/ARB use.

• Download figure
• Download PowerPoint

Discussion

We examined the contemporary use of sacubitril/valsartan for patients with HFrEF. Using data from a large US outpatient registry, we found that prescriptions for sacubitril/valsartan among eligible patients were low, 15%, but increasing over time. Compared with patients treated with ACE inhibitor/ARB, patients prescribed sacubitril/valsartan were younger and had a higher socioeconomic status as assessed by education, employment status, and insurance. Patients treated with sacubitril/valsartan also had better renal function compared with patients prescribed neither ACE inhibitor/ARB nor sacubitril/valsartan. We also evaluated several provider and practice characteristics that may influence quality of care, and after multivariable adjustment, found that larger practices, as assessed by the number of advanced practice providers, were associated with sacubitril/valsartan use. We did not see a statistically significant association between use of sacubitril/valsartan and other practice characteristics, such as presence of an HF clinic or the number of HF clinicians.

Since sacubitril/valsartan was approved by the Food and Drug Administration in 2015, there have been only a few studies examining current use in the United States. One study utilized data from hospitals participating in the Get With The Guidelines–HF quality improvement program.⁴ Among 21 078 patients with eligibility for sacubitril/valsartan hospitalized between July 2015 and June 2016, only 495 (2.3%) were prescribed sacubitril/valsartan at hospital discharge. Similar to our findings, patients in that study prescribed sacubitril/valsartan were younger, had lower LVEF, and higher use of aldosterone antagonists. In contrast, patients enrolled in our study were outpatients, and only 38% had an HF hospitalization within the past year. A study of data from the OptumLabs Data Warehouse from July 2016 to December 2016 found that among 102 247 patients with HFrEF, 2244 (2.2%) were treated with sacubitril/valsartan.⁶ Similar to our study, compared with patients treated with ACE inhibitor/ARB, those treated with sacubitril/valsartan were younger, had fewer comorbidities, and were more likely to be using other commonly prescribed HF medications at baseline. In the OptumLabs Data study, most prescriptions for sacubitril/valsartan were provided as an outpatient; >90% of new prescriptions for sacubitril/valsartan were ≤30 days of hospital discharge. Notably, the most common insurer in the OptumLabs Data Warehouse was United Healthcare. In contrast, our data include patients with various types of medical insurance, including Medicare, Medicaid, and no insurance.

There are also limited data evaluating the use of sacubitril/valsartan outside the United States. In 1 study, investigators compared characteristics of patients receiving sacubitril/valsartan at a single tertiary HF clinic in Belgium to characteristics of patients enrolled in the PARADIGM-HF trial (Prospective Comparison of ARNI with ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure).⁷ Belgium patients receiving sacubitril/valsartan were older, had worse renal function, and had a lower LVEF compared with trial patients, though it is hard to draw conclusions about prescribing patterns from this single-center experience. In contrast, our study includes data from 121 outpatient practices and a recent analysis compared patients enrolled in CHAMP-HF to those enrolled in the PARADIGM-HF trial.⁸ Patients enrolled in CHAMP-HF had similar baseline characteristics and similar projected risk of all-cause mortality compared to those enrolled in PARADIGM-HF.

To the best of our knowledge, our study was the first to describe variation in sacubitril/valsartan use by outpatient practice site and to examine provider and practice characteristics that may influence prescribing patterns. Larger practices, as assessed by the number of advanced practice providers, were more likely to use sacubitril/valsartan. The CHAMP-HF site survey did not collect information on the roles that advanced practice providers may have played in medication selection for HFrEF. The site survey, collected by investigator self-report, captured the number of advance practice providers in the practice and the number of advanced practice providers that were dedicated to HF management. A prior randomized trial of outpatients with HFrEF showed that a nurse practitioner specifically tasked to initiate and titrate β-blockers improved the use and target dose achieved of β-blockers compared with internists and cardiologists.⁹ To our surprise, we did not see a statistically significant association with other practice characteristics specific to HF care. The reasons for this are unclear. One possibility is that CHAMP-HF sites could have been more likely to focus on HF management, including the use of sacubitril/valsartan, regardless of practice characteristics (eg, number of HF cardiologists, HF advance practice providers, or HF nurses, presence of an internal HF clinic, or use of HF remote monitoring). Another possibility is that HF specialty practices simply may not be better at adoption of new HF therapies compared with internal medicine and general cardiology practices.

A notable finding of our data was that there were wide variations in the use of sacubitril/valsartan based on socioeconomic status. Clinical trial data suggest patients treated with sacubitril/valsartan compared with enalapril have improved outcomes, including improved quality of life as measured on the Kansas City Cardiomyopathy Questionnaire.¹ In addition, a prior cost-effective analysis suggested that sacubitril/valsartan therapy costs payers $50 959 per quality-adjusted life-year gained compared with enalapril therapy over a lifetime of therapy (estimated at 40 years).¹⁰ This number was in line with commonly accepted thresholds in the United States for a quality-adjusted life-year.¹¹ Despite this, our data highlight the current status of our healthcare system that a life-saving, cost-effective therapy is only available to a portion of the US HF population.

Our study did not identify specific provider and practice characteristics that can be targeted for implementation interventions, including practice size. Instead, our study findings highlight the need for ongoing rigorous evaluations of implementation strategies for HF therapeutics. An important finding in our study is that there were many opportunities to improve HF care, including older medical therapies (eg, aldosterone receptor antagonists), newer medical therapies (eg, sacubitril/valsartan), and devices (eg, cardiac resynchronization therapy).¹² There are currently limited data to inform our understanding of how practices and health systems adopt evidence-based HF care, and developing an evidence base for implementation strategies would have tremendous public health implications. This was highlighted in special report by the National Heart, Lung, and Blood Institute Implementation Science Work Group.¹³ This report specifically focused on 2 key intervention strategies that are most effective in improving both processes of care delivery and clinical outcomes: (1) audit and feedback and (2) educational outreach visits. These strategies are being prospectively evaluated in a cluster-randomized trial of quality improvement interventions for patients with HFrEF, CONNECT-HF (Care Optimization Through Patient and Hospital Engagement Clinical Trial for Heart Failure).¹⁴

Our study has limitations. CHAMP-HF is composed of voluntary participating sites and includes patients that signed informed consent and have the ability to complete multiple surveys over time. The study design may select a specific population of HFrEF patients. Second, CHAMP-HF is an ongoing study and, as noted in Figure 1, the adoption of sacubitril/valsartan is increasing over time. Third, we were not able to analyze if costs, access, and formulary issues may have influenced prescribing. Finally, we were only able to adjust for variables collected in the registry. Associations between provider and practice characteristics and sacubitril/valsartan use may be confounded by measured or unmeasured variables.

CONCLUSIONS

Despite current guideline recommendations, adoption of sacubitril/valsartan among eligible HFrEF patients remains low but is increasing over time. Patients prescribed sacubitril/valsartan were younger, less likely to have chronic kidney disease, have lower LVEF, and more likely to have a cardiac resynchronization device. The only provider and practice characteristics that were associated with sacubitril/valsartan use were related to practice size. There remains an important opportunity to improve patient outcomes through rigorous evaluation of novel quality improvement strategies in chronic HFrEF.

Footnotes