Guideline-directed Medical Therapy and Outcomes in Heart Failure with Reduced Ejection Fraction: A Multicentre Physician‑led Clinic Experience from Sarawak, Malaysia

Globally, the prevalence of heart failure (HF) has steadily increased due to an ageing population, as well as improved survival of patients with chronic coronary syndrome and comorbidities, such as hypertension.¹ HF clinics have been shown to be instrumental in providing multidisciplinary care to patients with HF.² Despite recent advancements in HF care, HF mortality remains high at 67% within 5 years following diagnosis.³

Sarawak is Malaysia’s largest state – located on the island of Borneo, spanning approximately 800 km along its northwest coast. The state of Sarawak is divided into 11 divisions, with its capital, Kuching, located on its southwest tip. As of 2021, it is home to 2.8 million people from almost 40 different sub-ethnic groups with varied socioeconomic backgrounds. Many patients still live in remote interior jungles, with nearby towns that serve as connecting hubs to Kuching. Sarawak Heart Centre, Sarawak’s sole cardiology centre, is located in Kota Samarahan (20 km away from Sarawak’s capital, Kuching), with a limited number of cardiologists who provide care for HF patients throughout the state. Despite efforts to improve access to public transportation, patients who live in the interior jungles of Sarawak often have to rely on communal modes of transport, travelling for hours via river and land, to reach their nearest tertiary centre. Many of these HF patients eventually forego regular follow-up or continuation of treatment due to these challenges. Thus, the care for these HF patients is often shared by other hospitals without cardiology specialists.

One of the key components in reducing HF mortality and hospitalisation is the optimisation of guideline-directed medical therapy (GDMT), especially in HF with reduced ejection fraction (HFrEF). Multiple randomised controlled trials on the treatment of patients with HFrEF over recent decades have shown a reduced risk of HF hospitalisation and cardiovascular mortality. The latest European and US guidelines for HF management recommend early initiation and dose optimisation of four pillars of medications, namely, renin–angiotensin–aldosterone system inhibitors (RAASi), β-blockers (BB), mineralocorticoid receptor antagonists (MRA) and sodium-glucose cotransporter 2 inhibitors (SGLT2i) in the management of HFrEF.^4,5 In addition, the latest European Society of Cardiology (ESC) guidelines have also updated and extended their recommendation for the use of GDMT in patients with HF with mildly reduced ejection fraction.⁴

A prospective observational study by Ling et al. in 2020 on the characteristics of patients with acute HF admissions in Sarawak showed shortfalls in the implementation of evidence-based management of HF in a non-cardiology centre (Sarawak General Hospital) compared with surrounding regions.⁶ Patients in the study were slightly younger (mean age 59 years) compared with the national average (mean age of 60 years).⁷ The authors showed that HF pharmacotherapy was frequently not optimised upon discharge, largely due to the time-consuming nature of medication titration and the need to manage adverse effects.⁶ Overcrowding and limited inpatient capacity further hindered efforts to optimise HF medications before discharge.

The ESC HF management guidelines have included strategies to overcome these circumstances, and have also published a consensus statement highlighting the importance of patient profiling for tailoring GDMT, based on blood pressure, heart rate, renal function, serum potassium level and the presence of AF in HFrEF patients.⁴ Despite the available recommendations, the adverse effects of HF therapy usually result in disruption of GDMT and, as a consequence, delayed optimisation of GDMT. In addition, the conventional sequencing method in the uptitration of GDMT has been shown to be time-consuming. McMurray et al. published an article on the rapid sequencing method, with the initial commencement of a BB and SGLT2i, followed by RAASi, then finally MRA, which can facilitate faster GDMT optimisation.⁸ The STRONG-HF trial showed that, among patients admitted for acute HF, an intensive treatment strategy of rapid uptitration of guideline-directed medication and close follow-up reduced the risk of 180-day all-cause death or HF readmission compared with usual care.⁹

Different models of HF clinics have been reported, namely, the outpatient multidisciplinary team HF clinics, home management programs and home telemonitoring. These models have been introduced to overcome limitations in the optimisation of GDMT.¹⁰ However, the implementation of home management programs and home telemonitoring remains challenging in our setting due to geographical reasons and the limited accessibility to telecommunication networks in the rural areas. Jain et al. reported that staff nurse- and pharmacist-guided titration of GDMT based on a titration protocol showed improvement in HF outcomes.¹⁰

An initiative to set up general physician-led HF clinics across district hospitals in Sarawak was started in 2020 to improve the management of HF patients who could not access specialist cardiology care at Sarawak Heart Centre. The first physician-led HF clinic in Sarawak was established at Sarawak General Hospital in the city of Kuching. Over the next 3 years, physician-led HF clinics were started in neighbouring district hospitals – Miri, Sarikei, Serian, Sibu, Bintulu, Sri Aman, Kapit and Limbang. The physician-led HF clinics were run by general physicians who had not received any prior formal cardiology specialty training. Physician-led HF clinics were established independently, based on each institution’s available resources and physician initiative.

Thus, to equip these general physicians with the necessary HF knowledge, multiple HF workshops were carried out by cardiologists at the respective hospitals to educate medical officers and general physicians about the latest guidelines in HF management. In general, patients with HFrEF, as defined by ESC, were recruited. To ensure consistency, all centres adhered to ESC definitions of HF when identifying eligible patients, but the final decision to enrol remained clinician-driven. A ‘buddy system’ with cardiologists in Sarawak Heart Centre was also set up.

In this study, we describe the clinical characteristics of patients managed at physician-led HF clinics across Sarawak from 2021 to 2023.

Methods

Study Design

This was a retrospective, observational, multicentre study with the aim of assessing the demographic, baseline clinical characteristics and prescription of GDMT. Baseline left ventricular ejection fraction (LVEF) and N-terminal pro-B-type natriuretic peptide (NT-proBNP), whenever available, were acquired prior to recruitment into the HF clinics.

We also assessed the clinical outcomes of 3- and 6-month all-cause mortality, HF readmission, and HF urgent visits.

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Medical Research and Ethics Committee, Ministry of Health Malaysia. The requirement for individual informed consent was waived due to retrospective analysis of de-identified data.

Patient Population

Patients were identified from a registry of 10 centres across Sarawak. For the present analysis, we excluded data from the cardiology specialist-led HF clinic at Sarawak Heart Centre to focus exclusively on patients managed in non-cardiologist physician-led HF clinics. Thus, the study cohort comprised patients followed in nine physician-led HF clinics: Sarawak General Hospital, Sibu Hospital, Miri Hospital, Sarikei Hospital, Sri Aman Hospital, Serian Hospital, Bintulu Hospital, Limbang Hospital and Kapit Hospital.

Eligible patients were adults (aged ≥18 years) with chronic HFrEF (defined as LVEF ≤40%) who attended follow-up in these clinics between 1 January 2021 and 30 June 2023.

Follow-up

Patient data on demographics, clinical characteristics, including comorbidities, New York Heart Association (NYHA) class, GDMT, and outcomes of all-cause mortality and history of hospitalisation for decompensated HF were collected at baseline (0M), 3 months (3M) and 6 months (6M) of follow-up.

Statistical Analysis

Categorical variables are expressed as absolute numbers and percentages, whereas continuous variables are presented as mean ± SD or median (interquartile range; IQR). Analyses were performed using IBM SPSS version 28.0.1.0. In all cases, p<0.05 was considered significant. Normality of continuous variables was assessed using the Shapiro–Wilk test, and visual inspection of histograms and Q-Q-plots. Normally distributed variables were summarised as mean ± SD; non-normal variables were summarised as median (IQR).

Results

We enrolled 384 patients from nine physician-led HF clinics in Sarawak. The mean age was 56.4 ± 13.7 years and 289 (75.3%) were men (Table 1 ). Among the patients, local ethnic groups comprised the largest proportion (49.9%), followed by Malay (28.8%) and Chinese (21.4%).

Comorbidities

Of the 381 patients with available data, hypertension was present in 261 (68.1%), diabetes in 159 (41.5%), dyslipidaemia in 207 (53.9%), ischaemic heart disease in 175 (45.7%), AF/flutter in 87 (22.8%), prior cerebrovascular accident in 31 (8.1%), chronic kidney disease in 103 (27.4%) and prior MI in 77 (20.1%; Table 1). Current or ex-smoking was reported in 167 out of 381 patients (43.8%).

Functional Status

NYHA class was available for 326 patients (84.9%) at 0M, 276 patients (71.9%) at 3M and 214 patients (55.7%) at 6M (Table 2). At 0M, 25.8% were class I, 54.6% class II, 18.4% class III and 1.2% class IV. By 3M, this shifted to 51.1, 42.0, 6.5 and 0.4%, respectively; and by 6M to 60.3, 35.5, 4.2 and 0%. Thus, the proportion in NYHA I rose from 25.8 to 60.3%, while class III/IV fell from 19.6 to 4.2%.

Left Ventricular Function

At baseline, patients had a mean LVEF of 28% (SD 7.62). By 3–6M after HF clinic enrolment, the mean LVEF had increased to 39.7% (SD 13.32). Examination of distributional assumptions indicated that baseline LVEF deviated significantly from normality (Shapiro–Wilk p=0.002), while follow-up LVEF did not. Given this, a Wilcoxon signed-rank test was applied, and revealed a statistically significant improvement in systolic function between the two time points (Z=8.34; p<0.001; Supplementary Table 1 ).

N-terminal pro-B-type Natriuretic Peptide

Median NT-proBNP fell from 5,159 pg/ml (IQR 1,730–9,000; n=19) pre-clinic to 2,050 (IQR 675–5,555) at 0M (n=60) and 611 (IQR 178–2,472) at 3M (n=30; Supplementary Table 2). At 6M, a small subset (n=3) had levels of 1,789 (IQR 857–1,789).

Guideline-directed Medical Therapy

Prior to the physician-led HF clinic, use of RAASi, BBs, MRAs and SGLT2i was 60.7, 76.3, 62.5 and 17.7% (n=384; Supplementary Table 3). At enrolment (0M; n=381), use increased to 79.8, 94.0, 81.9 and 42.5%, respectively. At 3M (n=363), uptake was 85.7, 93.9, 88.4 and 55.6%; at 6M (n=312), 86.2, 95.5, 87.8 and 60.3%. Overall, GDMT – especially SGLT2i – rose markedly from pre-clinic to follow-up (Figure 1 ).

Number of Guideline-directed Medical Therapy Pillars

GDMT intensity rose markedly; pre-clinic, 52.1% were on three or more pillars (11.5% on all four; Table 3). At baseline, this increased to 75.8% (four pillars 28.3%), at 3 months to 84.3% (43.8%) and at 6 months to 84.2% (49.2%).

Clinical Outcomes

By 3 months, HF readmission occurred for 22 out of 380 patients (5.8%), and all-cause mortality in 16 out of 381 patients (4.2%; Table 4). By 6 months, HF readmission occurred for 26 out of 338 patients (7.7%) and all-cause mortality for 35 out of 338 patients (10.4%).

Discussion

In this multicentre, physician-led HF clinic programme across Sarawak, we observed substantial improvements in the number of GDMT pillars prescribed, alongside favourable trends in symptoms, cardiac function and biomarkers.

From pre-clinic to 6M, use of RAASi, BBs, MRAs and SGLT2is increased. This trajectory aligns with guideline recommendations that emphasise early initiation and uptitration of the four foundational therapies for HFrEF. The prescription of SGLT2is rose substantially, from 18% at 0M to 60% at 6M. The initially low rate reflected limited access to our clinics, primarily due to financial constraints; however, despite subsequent improvement, usage remains suboptimal in the context of our resource-constrained setting.

Clinical status improved in parallel with GDMT uptake. The proportion of patients in NYHA class I more than doubled (from 26 to 60%), while class III/IV fell from 20 to 4% over 6 months. In our cohort, 39% underwent repeat echocardiography within 3 to 6 months, while others may have had their follow-up echocardiograms after 6–12 months, which were not captured here. Based on the available data, LVEF improved by approximately 12 absolute percentage points, increasing from 28.0 to 39.7%. Unfortunately, NT-proBNP measurement was not routinely performed for follow-up in our region due to limited availability. Nevertheless, the small sample of NT-proBNP data we were able to collect demonstrated a stepwise decrease across time points.

Although causality cannot be inferred in an observational cohort, these concordant improvements are consistent with the benefits of optimised GDMT shown in randomised trials and implementation studies.

Our findings also speak to the feasibility of a physician-led clinic model in a geographically vast region with limited specialist availability. Prior to the implementation of physician-led HF clinics, HF patients were seen in high-volume outpatient medical clinics with long appointment wait times. Sarawak’s physician-led HF clinics were run by general physicians supported by cardiology-led workshops, with a stated focus on HF GDMT optimisation. The physicians had access to a ‘buddy cardiologist’ system, giving them specialist cardiologist support in complex HF management, usually via text message or email.

The literature supports multidisciplinary or protocol-driven models – including nurse- and pharmacist-led titration – to close evidence-to-practice gaps. While our clinics did not fully replicate a formal multidisciplinary team structure because of staff constraints, the observed GDMT uptake and clinical improvements suggest that a pragmatic, protocolised physician-led approach can deliver meaningful gains, even with limited access to cardiologists.

A lower-volume physician-led HF clinic with shorter appointment wait times led to improved outcomes for HF patients. Rather than the conventional stepwise titration that delays target doses, our physician-led clinics emphasised rapid sequencing and early multi-drug initiation. Reflecting this, the proportion on three or more GDMT pillars was already 52% pre-clinic, and rose to 76% at enrolment and 84% by 3–6M; use of all four pillars nearly quadrupled (11.5 to 49.2%).

Despite progress, several gaps remain. First, access to angiotensin receptor/neprilysin inhibitors and SGLT2is was constrained by limited availability, which likely tempered the extent of optimisation. Second, biomarker and echocardiographic monitoring are underutilised owing to limited availability. Third, our 6-month all-cause mortality and HF admission rates underscore the persistent risk in this relatively young cohort (mean age 56 years) with a high comorbidity burden (notably hypertension, diabetes and ischaemic heart disease). Strengthening upstream prevention, risk-factor control and post-discharge care pathways remains essential.

Study Limitations

This was a retrospective, non-randomised analysis without a concurrent control group; confounding by indication and regression to the mean are possible. Missing data were non-trivial (e.g. NYHA and echocardiography at follow-up), NT-proBNP sample sizes were small at later time points and outcomes may be under-ascertained if patients sought care outside participating hospitals. Due to the absence of an integrated cross-centre registry, dual follow-up at the physician-led clinics and a tertiary centre cannot be entirely ruled out. We did not capture medication doses, time-to-initiation, adherence and device therapy. These factors may limit generalisability and preclude causal inference.

Implications

Even without full multidisciplinary staffing, a structured physician-led clinic network can drive GDMT implementation at scale in a large, resource-limited state. Priorities for program strengthening include: standardised titration protocols with rapid sequencing; early, frequent follow-up in the first 4–6 weeks after initiation or hospitalisation; wider access to angiotensin receptor/neprilysin inhibitors /SGLT2i and laboratory/biomarker monitoring; task-sharing with pharmacists and nurses using protocolised titration; and simple digital registries to track all the data. A prospective registry or pragmatic cluster trial comparing physician-led clinics with usual care would help define effect sizes and cost-effectiveness in this context.

Conclusion

Across nine physician-led HF clinics in Sarawak, GDMT adoption improved markedly, and was accompanied by better symptoms, higher LVEF and decreasing NT-proBNP levels over 6 months, with acceptable readmission rates in a large, high-risk, relatively young population. While observational in nature and limited by missing data, these findings support the scalability of a protocolised physician-led clinic model to close evidence-to-practice gaps where specialist access is limited. Future work should prioritise rapid GDMT sequencing, enhanced follow-up intensity, broader access to advanced therapies, and prospective evaluation of clinical and economic outcomes.

Click here to view Supplementary Material.