Introduction

Heart failure (HF) is a progressive disease characterised by recurrent hospitalisations, and significant mortality [1,2,3]. Its incidence increase with age thus in the context of a rapidly ageing global population, the forecasted global economic burden of HF is expected to rise exponentially hence escalating HF as a major public health problem [4, 5]. Effective pharmacotherapy for patients affected by heart failure with reduced ejection fraction (HFrEF), defined as left ventricular ejection fraction (LVEF) ≤40%, has been identified by a series of landmark randomised controlled trials thus forming the basis for guideline-directed medical therapy (GDMT) [6,7,8,9,10,11,12,13,14,15,16].

The nurse-led HF clinic is an established model of care in Europe with studies describing its benefits in reducing hospitalisations and mortality [17,18,19]. Since the role emerged, the responsibilities of HF clinical nurse specialists have expanded to include the assessment and initiation of HF pharmacotherapy in HF patients post-hospitalisation [20, 21]. These extended capabilities make the services provided by the nurse-led HF clinic increasingly indispensable especially by accelerating the implementation of GDMT [22, 23].

The European and American guidelines were recently updated to include SGLT2i as a new pillar in HFrEF pharmacotherapy [24, 25]. Accompanying this is a paradigm shift in HFrEF management, away from the conventional stepwise approach of initiating and up-titrating one drug class at a time, towards the simultaneous initiation of multiple drug classes to achieve concurrent treatment with four cornerstone therapies known as quadruple-therapy [26,27,28]. Indeed, the independent and incremental prognostic and symptomatic benefits of each additional cornerstone therapy prescribed justify the early and widespread use of quadruple-therapy in eligible patients [29]. Despite compelling evidence and guideline updates, therapeutic inertia still exists with multiple registries demonstrating ongoing poor compliance with outpatient GDMT [23, 30,31,32].

Additionally, GDMT has grown in complexity from the increased number of cornerstone therapies, drug interactions, side-effects, and requirements for HF patient profiling [33]. While physician clinicians are more likely to have the training required to navigate these complexities, HF clinical nurse specialists may lack the necessary experience and educational supports required to confidently deliver quadruple-therapy.

From the gaps in experiential and educational supports, we proposed that the complexities of the current GDMT significantly impact the abilities of HF nurse prescribers in implementing the latest GDMT. Our study describes and compares the abilities of physician-led and nurse-led HF clinics in providing HFrEF patients with GDMT within the outpatient setting.

Methods

Study Design

We perform a retrospective multi-centre cohort study on the pharmacotherapy patterns of HFrEF patients attending either a physician-led or nurse-led HF clinic in 2021. We identified and compared the prescribing rates of HFrEF cornerstone therapies between both groups to assess if there is a significant difference in compliance with GDMT. Drug classes forming the cornerstone therapies of HFrEF were defined as angiotensin-converting-enzyme inhibitor (ACEi)/ angiotensin II receptor blocker (ARB), angiotensin receptor-neprilysin inhibitor (ARNi), beta-blocker (BB), mineralocorticoid receptor antagonist (MRA), and sodium-glucose cotransporter-2 inhibitor (SGLT2i). This observational study was designed and presented according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Patients and Clinic Setting

All adult patients (age ≥18 years) were included in this study if they met both inclusion criteria: (I) reviewed in a physician-led or nurse-led outpatient HF clinic within the University of Limerick Hospitals Group (ULHG) between January and December 2021, (II) transthoracic echocardiographic (TTE) diagnosis of HFrEF (LVEF ≤40%) at baseline. Patients were excluded from this study if they met any of the exclusion criteria: (I) not reviewed within ULHG HF services between January and December 2021, (II) no TTE confirmation of LVEF ≤40%. Outpatient HF services within the ULHG consist of one physician-led HF clinic and two nurse-led HF clinics spread across different geographical regions. At the time the study was undertaken, all cardiologists involved in HF service provision were trained in general adult and interventional cardiology.

The mode of regional heart failure service provision (physician-led vs nurse-led) within ULHG is based on the size of the acute hospital within that geographic catchment area. The denser urban region is supported by a large acute hospital (termed model 4 hospital in Ireland) and staffed with a full cardiology department. This centre had the staffing capacity to provide a physician-led HF clinic where a clinical doctor is responsible for the supervision of HF service delivery, reviewing HFrEF patients, and prescribing all HF pharmacotherapy. Two sparsely populated rural regions within the ULHG catchment area were supported by smaller acute hospitals (termed model 2 hospital in Ireland), where staffing a full cardiology department is challenging and impractical. A more feasible approach taken was to staff these hospitals with dedicated HF clinical nurse specialists who autonomously run two nurse-led HF clinics with weekly physician input on challenging cases. HF clinical nurse specialists not only reviewed HFrEF patients but were also able to autonomously prescribe HF pharmacotherapy according to a collaborative practice agreement (restricted range of drugs which nurses are permitted to prescribe) and held the responsibility of clinical decision-making for initiation and titrating HF pharmacotherapy.

Variables and Data Sources

Key variables consisted of baseline patient characteristics and pharmacotherapy patterns. Baseline patient characteristics included patient demographic (age, sex), physiologic parameters (blood pressure, heart rate, LVEF according to TTE), laboratory investigations (haemoglobin, creatinine, albumin, NT-proBNP), heart failure aetiology, comorbidities, and New York Heart Association (NYHA) functional classification. Data on the pharmacotherapy patterns of HFrEF cornerstone therapies were collected: prescription rate of each individual drug class, total number of HFrEF cornerstone therapies prescribed (no therapy, monotherapy, dual-therapy, triple-therapy, quadruple-therapy), and total number of cardiovascular drugs prescribed. All information were gathered by retrospective chart review while observing General Data Protection Regulations (GDPR).

Statistical Analyses

All calculations and statistical analysis were performed using SPSS software, version 27 (IBM Corporation, Armonk, NY, USA). P <0.05 was considered statistically significant. Descriptive statistics were expressed as mean ± standard deviation and median ± interquartile range (IQR). Statistical comparison of the baseline categorical and qualitative variables between the physician-led and nurse-led groups was performed using chi-squared test. Differences in the prescription rate of each individual drug class and the total number of HFrEF cornerstone therapies between the physician-led and nurse-led groups were statistically analysed using chi-squared test while differences in the median number of total cardiovascular drugs prescribed were statistically analysed using Mood’s median test.

Results

Patients and Baseline Patient Characteristics

A total of 395 HFrEF patients were reviewed in the ULHG HF services in 2021 and included in this study. 164 HFrEF patients were reviewed in a physician-led HF clinic (physician-led group) while 231 HFrEF patients were reviewed in a nurse-led HF clinic (nurse-led group). The baseline patient characteristics of both groups are described in Table 1. Patients in the physician-led group were generally younger when compared to those reviewed in the nurse-led group (mean age 65.1 (± 13.6) vs 70.6 (± 12.4) years, P = 0.455); however, this was not statistically significant. When comparing baseline physiologic parameters, the mean baseline LVEF was significantly lower in the physician-led group when compared to that of the nurse-led group (LVEF 26.5% (± 8.3) vs 28.7% (± 9.2), P = 0.017). Serum creatinine and NT-proBNP levels were similarly elevated in both the physician-led and nurse-led groups. Patients in both groups had similar aetiologies for HFrEF with ischaemic heart disease followed by dilated cardiomyopathy being the commonest cause. Apart from significantly more active smokers within the nurse-led group (51 (22.1%) vs 21 (12.8%), P = 0.019), both groups had similar patient comorbidities with hyperlipidaemia and hypertension being the two most commonly associated conditions. The majority of patients in both groups had a functional class of NYHA 2; at the time, they were reviewed in the clinic.

Table 1 Comparison of baseline patient characteristics. Hb, haemoglobin; NT-proBNP, N-terminal pro-B-type natriuretic peptide
Full size table

Pharmacotherapy Patterns

The pharmacotherapy patterns of HFrEF cornerstone therapies of both groups are described in Table 2. While there was a slightly lower proportion of patients within the nurse-led group on ACEi/ARB (133 (57.6%) vs 96 (58.5%), P = 0.849), ARNi (72 (31.2%) vs 63 (38.4%), P = 0.135), and BB (215 (93.1%) vs 157 (95.7%), P = 0.266) when compared to the physician-led group, this was not statistically significant (Fig. 1). Compared to the physician-led group, there was a significantly lower proportion of patients within the nurse-led group on MRA (97 (42.0%) vs 103 (62.8%), P < 0.001) and SGLT2i (18 (7.8%) vs 40 (24.4%), P < 0.001) (Fig. 1). The majority of patients within the nurse-led group were treated with dual-therapy (116 (50.2%)) while the majority of patients within the physician-led group were treated with triple-therapy (75 (45.7%)). There was no statistically significant difference in patients treated with no therapy or monotherapy between the physician-led and nurse-led groups (Fig. 2). Compared to the physician-led group, there was a significantly higher proportion of patients within the nurse-led group treated with dual-therapy (116 (50.2%) vs 51 (31.1%), P < 0.001) while concurrently having a significantly lower proportion of patients treated with triple-therapy (81 (35.1%) vs 75 (45.7%), P = 0.033) and quadruple-therapy (10 (4.3%) vs 31 (18.9%), P < 0.001) (Fig. 2).

Table 2 Comparison of pharmacotherapy patterns. ACEi, angiotensin-converting-enzyme inhibitor; ARB, angiotensin II receptor blocker; ARNi, angiotensin receptor-neprilysin inhibitor; BB, beta-blocker; MRA, mineralocorticoid receptor antagonist; SGTL2i, sodium-glucose cotransporter-2 inhibitor
Full size table
Fig. 1
figure 1

Comparison of prescribed HFrEF drug classes. ACEi, angiotensin-converting-enzyme inhibitor; ARB, angiotensin II receptor blocker; ARNi, angiotensin receptor-neprilysin inhibitor; BB, beta-blocker; MRA, mineralocorticoid receptor antagonist; SGTL2i, sodium-glucose cotransporter-2 inhibitor

Full size image
Fig. 2
figure 2

Comparison of the number of HFrEF drug classes prescribed

Full size image

While the median number of total cardiovascular drugs prescribed between the physician-led and nurse-led groups were the same (4 vs 4), a significantly greater proportion of patients were prescribed a number > median (79 (48.2%) vs 57 (24.7%), P < 0.001) with concurrently a significantly lower proportion prescribed a number ≤ median (85 (51.8%) vs 174 (75.3%), P < 0.001) within the physician-led group when compared to the nurse-led group. This demonstrates a significant difference in distribution around the median number of total cardiovascular drugs prescribed with patients in the physician-led group being more likely to receive a higher number of prescribed cardiovascular drugs than patients in the nurse-led group.

Discussion

The independent and incremental benefits of each additional HFrEF cornerstone therapy make quadruple-therapy (ACEi/ARB/ARNi, BB, MRA, SGLT2i) the current standard of care for HFrEF GDMT [24, 25, 28, 29]. Within our study cohort, rates of achieving the current standard of care remained poor across both groups with only 18.9% and 4.3% of patients receiving quadruple-therapy within the physician-led and nurse-led groups respectively.

The physician-led group were comparatively more successful in achieving the current standard of care with significantly higher proportions of patients treated with quadruple-therapy (18.9% vs 4.3%, P < 0.001) and triple-therapy (45.7% vs 35.1%, P = 0.033), significantly lower proportions of patients treated with dual-therapy (31.1% vs 50.2%, P < 0.001), and numerically lower number of patients on no therapy or treated with monotherapy. This group was more successful despite treating a cohort with more complex pharmacotherapy, suggested by the higher likelihood of patients within this cohort being prescribed a higher number of cardiovascular drugs. Patients reviewed within the nurse-led group were more likely to be prescribed less cardiovascular drugs and less HFrEF cornerstone therapies overall (majority of patients (50.2%) receiving only dual-therapy), highlighting the importance of prescriber competency and confidence when treating a disease likely to require multiple drug classes to optimise benefits. The significantly lower prescription rates of MRA (42.0% vs 62.8%, P < 0.001) and SGLT2i (7.8% vs 24.4%, P < 0.001) within the nurse-led group suggests underutilisation of quadruple-therapy is largely driven by barriers to MRA and SGLT2i prescribing. Considering the comparative effectiveness of ARNi over ACEi had been demonstrated in the pivotal PARADIGM-HF trial, the comparatively low uptake of ARNi to ACEi/ARB within both the physician-led (38.4%) and nurse-led (31.2%) groups was an interesting finding [14].

The issue of poor GDMT compliance in HFrEF is longstanding. The CHAMP-HF registry study previously brought to light its significance in 2018, demonstrating <25% compliance with the previous triple-therapy (ACEi/ARB/ARNi, BB, MRA) standard of care and only 1% of eligible HFrEF patients treated at target doses within a United States based population [30]. Despite mounting evidence of the benefits of contemporary HFrEF cornerstone therapies since, the prescribing patterns from our study highlight the underutilisation of quadruple-therapy in HFrEF as a major ongoing issue, perpetuated by HF service providers in contemporary real-world practice.

Our study demonstrates the gaps in GDMT implementation are driven by suboptimal MRA and SGLT2i prescribing, and that this issue is more pronounced within nurse-led services, suggesting autonomous nurse prescribers face greater difficulty with MRA and SGLT2i prescribing. Poor MRA uptake is a widespread and established phenomenon (despite its long-established evidence base and earlier inclusion into GDMT) with previous studies revealing the proportion of eligible HFrEF patients not receiving appropriate MRA therapy to be approximately 44–67% [30,31,32]. Although poor SGLT2i uptake could partially be explained by its novel inclusion to GDMT, the mounting evidence of its benefits extending to a broader range of heart failure patients including those with mildly reduced ejection fraction (LVEF 41–49%) and preserved ejection fraction (LVEF ≥50%) make continued underutilisation from unfamiliarity indefensible [34,35,36]. In order to improve MRA and SGLT2i uptake, we interviewed HF clinical nurse specialists and physicians of both groups to identify potential barriers to MRA and SGLT2i use.

Suboptimal MRA use appeared to be rooted in the ongoing adherence to the conventional stepwise approach to HFrEF pharmacotherapy. HFrEF pharmacotherapy is grounded on the concept of the HFrEF patient having a reserve of blood pressure, heart rate, and renal function in order to intensify HFrEF drug therapy. Adherence to conventional methodology in HFrEF management typically depleted a patient’s reserve of blood pressure through maximally titrated ACEi/ARB/ARNi and BB, thereby having no leeway to safely initiate MRA therapy. The potential of precipitating acute renal failure, hyperkalaemia, and symptomatic hypotension from the addition of MRA on top of other nephrotoxic and blood pressure-lowering drugs was the main barrier which unnerved nurse prescribers from initiating MRA. The prospect of autonomously managing potentially life-threatening complications in the outpatient setting was not an undertaking HF clinical nurse specialists were trained or prepared to do, hence their reservation in initiating MRA.

Suboptimal SGLT2i use predominantly arises from the novelty of its inclusion to GDMT. While physicians are more inclined to use this cornerstone therapy for HFrEF, HF clinical nurse specialists have more reservations from unfamiliarity with side-effects, drug interactions, and interplay with other commonly comorbid chronic conditions like diabetes. Indeed, when considering their separate vocational training, physicians were trained in the full spectrum of medical specialties (which include endocrinology), while HF clinical nurse specialists received training on only a narrow subspecialty of cardiology. The addition of a drug class with origins in diabetic therapy to GDMT would certainly prove a challenge to clinicians with no prior experience with therapeutics in diabetes. The identified barrier here is a knowledge gap in HF therapeutics. This therefore can potentially be overcome by bolstering nurse prescriber education with key information on safe SGLT2i prescribing, which would boost confidence in initiating SGLT2i.

Apart from suboptimal MRA and SGLT2i use, the disproportionately lower ARNi uptake to ACEi/ARB continues to hinder the delivery of the best HFrEF therapies to patients. Unlike in other European countries where ARNi has obtained full approval for use for HFrEF, the use of ARNi within the Republic of Ireland is approved under a managed access protocol which requires additional clinical criteria to be met before drug reimbursement is granted [37]. In order for ARNi to be reimbursed, HFrEF patients must fulfil a more restricted clinical criteria including having LVEF ≤35% and continuing to be symptomatic after already being initiated on optimal GDMT including ACEi/ARB. Prescribers initiating ARNi must also complete an additional administrative step of completing on online application form for reimbursement. This process further restricts access where if one criterion/information was omitted from the application form, the patient will permanently receive a negative reimbursement recommendation until an appeal with additional information is submitted. These administrative barriers were erected due to the exponentially higher cost of ARNi when compared to ACEi (annual cost €1677.00 vs €43.68) thus requiring rationing of public health resources for HFrEF patients who require it the most.

Implementing GDMT remains an important step in ensuring HFrEF patients reap the benefits of the scientific endeavours. Apart from the obvious abolishment of administrative barriers, multiple other creative strategies have demonstrated success in improving prescription rates of HFrEF cornerstone therapies. Mukhopadhyay et al. had recently demonstrated the effectiveness of using electronic clinical decision support tools in improving the rates of MRA prescribing (for HFrEF) by general practitioners (GP) [38]. In the BETTER CARE-HF trial, she described how GP practices employing an electronic health record–embedded alert tool more than doubled their rate of MRA prescribing in community HFrEF patients, when compared to usual care. Considering the same electronic tools may also be used by practice nurses, these findings highlight the potential role electronic health record–embedded alert tools may play in improving GDMT implementation. Specifically addressing the knowledge gap barrier in HF therapeutics, we recently described the potential for using nurse-prescriber guidelines in improving the rates of SGLT2i prescribing by HF clinical nurse specialists [39]. In a clinical audit, we described how providing key information and guidance on safe SGLT2i prescribing within a written document can enhance nurse prescriber education and have an impact on improving the rate of SGLT2i prescribing. These findings highlight the importance of updating and improving local educational resources for HFrEF prescribers as GDMT rapidly evolves with ever-changing advancements in therapeutics.

As the need for effective outpatient HF services grows, so does our reliance on HF clinical nurse specialists to deliver nurse-led HF clinics, especially in rural regions where staffing a full cardiology department remains a challenge. In order to maintain high standards of outpatient HF care, it is essential to ensure high levels of HFrEF GDMT compliance among all HF service providers and increase collaboration between physician and nurse HF service providers. By identifying areas of weakness and potential barriers to quadruple-therapy, MRA, and SGLT2i uptake within our prescribing trends, we provide the basis for further work to be performed in improving education, training, and support for HF nurse prescribers and other HF service providers.

Limitations

As SGLT2i was only added to GDMT in September 2021, the low proportion of SGLT2i may be attributed to the short duration for new GDMT implementation. The study was conducted at the tail end of the SARS-CoV-2 coronavirus pandemic causing situational disruptions of normal HF service provision. Since face-to-face clinics with blood draws are essential for the safe initiation and titration of several drug classes, the low proportion of quadruple-therapy use may have been a result of normal service disruption and not reflect real-world practice outside pandemic restrictions. As the cardiologists involved in HF service provision within ULHG were trained in general adult and interventional cardiology, the results from our study may not reflect the practices of HF services receiving further input from cardiologists with subspecialty training in heart failure.

Conclusion

Our study highlights the ongoing gaps in the implementation of current GDMT in both physician-led and nurse-led HF clinics. These gaps in GDMT implementation are driven by the suboptimal use of MRA and SGLT2i. Compared to the physician-led HF clinic, patients reviewed in nurse-led HF clinics were less likely to be treated with MRA and SGLT2i, highlighting the greater difficulties autonomous nurse prescribers face in prescribing these two drug classes. Ongoing barriers to MRA and SGLT2i prescribing have their origins in adherence to old methodology and the structure of training for HF clinical nurse specialists. Improving education and support for HF nurse prescribers may improve confidence in prescribing, thereby potentially increasing MRA and SGLT2i uptake.