Abstract
Introduction
Acute heart failure (AHF) syndrome is defined as the new-onset, gradual or rapid worsening of chronic heart failure (HF) symptoms requiring urgent therapy. 1 There has been an increased prevalence and incidence of AHF syndrome. Although data on local outcomes remain scarce, globally 25–50% of patients face readmission within six months, and 25% die within a year of initial admission.2,3
Malaysian cardiovascular services adopt a ‘spoke-and-hub’ model, where the majority of public hospitals do not offer cardiology services but instead are required to refer patients externally to designated cardiac centres. In total, there are only 18 cardiac centres out of a total of 144 public hospitals nationwide. 4 The large mismatch means that the majority of AHF patients are cared for by non-cardiac centres. Furthermore, evidence has point towards great heterogeneity between AHF admissions globally, including varying demography, co-morbidities and provision of care which restricts extrapolation of shared data in the current literature. Granular data remain key in understanding local needs and developing care systems that can provide means. Such data unfortunately remain scarce within most South-East Asian countries, including Malaysia.
The aim of this study was to illustrate the baseline characteristics and clinical profile, emergent and essential AHF care provided, and readmission and mortality outcomes of patients admitted within a single non-cardiac centre in Malaysia. Attempts were made to compare our data set to those from existing registries, both regional and international.
Methods
A retrospective cross–sectional study was conducted focusing on patients aged ⩾18 years admitted to Hospital Sungai Buloh, Selangor, for AHF between the 1 January 2012 and 31 December 2016. Eligible patients were identified through key terms used on diagnosis, including ‘acute heart failure’, ‘heart failure’, ‘decompensated congestive cardiac failure’, congestive cardiac failure’, ‘fluid overload’, ‘acute pulmonary oedema’, ‘pulmonary oedema’ and ‘cardiac failure’. Selection was through universal sampling. Patients with multiple admissions beyond their first during the study period were only considered for inclusion once, although subsequent readmissions were reviewed to supplement data collection.
We identified 1307 patients in total. Information obtained included (a) demographics and clinical profile, (b) acute management and medication prescription and (c) number of patients with at least one readmission and/or who died at various time intervals (in-hospital, 30 days, 3 months, 6 months and 12 months). Readmissions were divided into all-cause and HF-related readmissions. Mortality was described as all-cause mortality.
Descriptive statistics were used to summarise baseline characteristics, clinical management and outcomes. Continuous variables are expressed as means with standard deviations, or as medians with an interquartile ranges. Categorical data are summarised as frequencies and proportions.
Results
Demographics and clinical profile
Demographics and clinical profile are shown in Table 1. The mean age of patients admitted was 63.4 years (±16.2 years). A total of 60.7% were <65 years of age – an arbitrary definition of ‘young’ HF based on several previous studies. 5 Men made up less than half (46.4%) of our sample population. There was a high proportion of Malays (63.5%) as opposed to the other major ethnic groups. The majority of patients (57.2%) had established congestive cardiac disease prior to admission. The majority of cases (61.8%) were presumably precipitated by an ischaemic event.
Baseline characteristics, symptoms and signs, NYHA class, baseline vital signs, electrocardiography findings, baseline ejection fraction on echocardiography and blood test results on admission of acute HF patients.
HF: heart failure; NYHA: New York Heart Association;
There was a high proportion of patients with ischaemic risk factors, including hypertension, diabetes and known ischaemic heart disease (IHD; 70.9%, 62.2% and 57.8%, respectively). There was a high rate of under-diagnosed dyslipidaemia when comparing data on known dyslipidaemia to lipid profile tests on admission (Table 1). Only 23.0% patients self-reported the condition, but the majority had deranged lipid profile levels on admission (Table 1). However, less than a third of patients (18.1%) had an actual previous history of major adverse vascular events, including cerebrovascular accidents and myocardial infarctions (21.7%). Active smokers and ex-smokers accounted for 28.0% and 27.5%, respectively.
The majority of our patients were of New York Heart Association (NYHA) Class II (31.9%) and Class III (25.6%) prior to admission. A large proportion of patients reported symptoms associated with lung congestion: dyspnoea (82.9%), orthopnoea (65.9%) and paroxysmal nocturnal dyspnoea (58.0%). The commonest clinical signs on examination included lung crepitations (76.8%) and peripheral oedema (63.7%).
The average heart rate on presentation was 89.7±21.9 bpm, and the average systolic blood pressure (SBP) was 144.6±31.9 mmHg. Only a minority of our patients had SBP readings ⩽120 mmHg (17.7%) and <100 mmHg (2.5%). This parallels our data showing low rates of inotropic and vasopressor use (14.5%) on initial presentation. Sinus rhythm was the most common electrocardiogram (ECG) finding (84.5%), with low rates of arrhythmias all around. Although rates of known atrial fibrillation (AF) was low (13.1%) in our population, our database indicates high rates of newly diagnosed AF (15.2% of ECG on admission), which highlights great under-diagnosis of AF amongst our population. A total of 40.8% of our patients had left ventricular ejection fraction (LVEF) <40%, 10.6% had LVEF 40–49% and 23.5% an LVEF ⩾50% on echocardiogram done within six months of initial admission. On average, patients were mildly hyponatraemic (
Acute management and medication prescription
Acute management and medication prescription are shown in Table 2. Nearly all (96%) patients had some form of diuretic during their acute admission, the majority being furosemide-based administered intravenously as boluses (85.9%) or infusions (7.7%). Up to 14.5% of patients required inotropes and vasopressors, 12.9% required intravenous nitrates and 8.6% required dialysis or ultrafiltration. Only 4.9% of patients required intubation and mechanical ventilation, whereas a higher proportion were started on some form of non-invasive ventilation (25.9%). A quarter (26.9%) of patients required either coronary care or intensive care unit admission. The average length of inpatient stay was 5.3±3.4 days.
Emergency management and medication, and medication prior to admission and on discharge in acute HF patients.
ACE: angiotensin-converting enzyme; ARB: angiotensin receptor blockers; NOAC: novel oral anticoagulants; ARNI: angiotensin receptor-neprilysin inhibitors.
A high proportion of patients (54.7%) were already on loop diuretics prior to admission. The proportion of patients on medication for IHD prevention (primary and secondary) was also high. The majority were on aspirin (58.0%), clopidogrel (26.6%) and statin therapy (60.8%) prior to admission, with further increased prescription rates of all three upon discharge (65.9%, 50.3% and 71.78%, respectively). There was also a slight increase in prescription of beta-blockers (from 42.9% pre-admission to 50.2% on discharge) and minerocorticoid receptor antagonists (from 13.3% pre-admission to 19.7% on discharge), but surprisingly there was a reduction in the prescription of angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blocker (ARB; from combined 50.4% pre-admission to combined 48.2% on discharge). There were no prescriptions for angiotensin receptor-neprilysin inhibitors and novel oral anticoagulants, likely due to their costs and ‘private purchase only’ status.
Readmissions and mortality
Readmissions and mortality are shown in Tables 3 and 4. The cumulative proportion of patients with at least one readmission within 30 days, 3 months, 6 months and 12 months from initial presentation were 53 (4.1%), 464 (35.5%), 516 (39.5%) and 994 (76.1%) patients, respectively. Of this, the number of patients with HF readmissions were 22 (1.7%), 229 (17.5%), 281 (21.5%) and 528 (40.4%), respectively. The cumulative proportion of all-cause mortalities in hospital and at 30 days, 3 months, 6 months and 12 months were 22 (1.7%), 205 (15.7%), 293 (22.4%), 493 (37.7%) and 650 (49.7%), respectively.
Time intervals and cumulative proportion of patients experiencing their first all-cause readmission, HF-related readmission and non-HF-related readmission.
Cumulative proportion of patients suffering mortality at various time intervals from initial presentation.
Discussion
The younger and Malay-predominant population reflects the distribution IHD within the population. The high proportion of patients with ischaemic risk factors and the high rates of pre-admission antiplatelets and statin prescriptions further support this theory. The large proportion of ‘young’ HF could be accounted for by the younger mean age of Malaysian patients developing IHD, supported by our national cardiovascular disease database. 6 With regards to ethnic distribution, there is also evidence supporting a high proportion of ischaemia amongst Malays and Indians, as seen in both local7,8 as well as regional 9 studies. When compared to other global AHF registries, our presumed ‘ischaemic burden’ was relatively higher (KorAHF, OPTIMIZE-HF, EHFS-II and REPORT-HF registries quoting a 38.0%, 46.0%, 54.0% and 37.0% rate of ischaemia-related HF, respectively), highlighting an urgent need to tackle metabolic risk factors that contribute to such rates.10–14 However, when compared to regional data from the ASIAN-HF registry South-East Asian subgroup, our rates were comparable to theirs (62.0%). 15 There was also fewer patients with AF versus other registries (KorAHF, OPTIMIZE-HF, ATTEND, ADHERE-International, ASIAN-HF and EHFS-II reporting rates of 27.0%, 31.0%, 40.0%, 24.0%, 19.8% and 38.0%, respectively).10–13,15,16 However, when compared to the REPORT-HF study, the current study was comparable to the Asian cohort reported (8.0%). 14
Creatinine and eGFR values were on average deranged in our population, consistent with substantial proportions of chronic kidney disease (CKD), and deteriorated closer to discharge (Table 1). As noted by several other studies, there is great hesitancy in prescribing ACE inhibitors or ARB in our population, likely due to the worsening renal function (Table 2).16–19 A large meta-analysis has shown that in HF patients, concomitant CKD doubles the risk of mortality. However, the role of decongestion and neurohormonal blockade remains essential, and guideline-directed medical therapy (GDMT) targeting this should not be prematurely ceased – a common occurrence when managing AHF patients. 20
The majority of our patients were NYHA Class II or Class III (57.5%) on arrival. Data on heart rate and SBP mimic that of international data (KorAHF, OPTIMIZE-HF, ASIAN-HF and ATTEND reporting average heart rates of 91.0±26.0, 87.0±22.0, 78.4±15.6 and 99.0±30.0 bpm, respectively, and average SBP of 136±31, 143±33, 123.4±21.9 and 147±38 mmHg, respectively).10,11,15,21 There was also much lower use of inotropic and vasopressor support in our cohort compared to other registries (KorAHF, EHFS-II and ADHERE-AP reporting rates of 32.0%, 28.0% and 15.0%, respectively).11–13
In our cohort, patients of NYHA Class I and Class II were routinely being admitted for inpatient management due to various reasons, including immediate albeit brief management, which is supported by the short average length of stay (median 5.3 days (interquartile range 4.1–6.2)) and heavy loop diuretic use in our study – an unfortunate common practice locally. Patients receive ‘quick fixes’ during these periods often before being prematurely discharged, driven mainly by overcrowding in most public hospitals in Malaysia. 22 We also note that the majority of patients were deemed to have a possible ischaemic precipitant to their HF, which meant possible admissions for further management, which often include subcutaneous heparin or anti-thrombin (i.e. fondaparinux) injections between three and five days. This is further supported by the increase in antiplatelet and statin prescriptions post discharge in our study. 6
Readmission and mortality rates were extraordinarily high in our population. Although data for direct comparison are scarce, statistics from the ASCEND-HF registry revealed rates of 5.1% and 6.6% for all-cause mortality and all-cause readmission, respectively, at 30 days, and a rate of 7.4% rate for all-cause mortality at 180 days. 23 Data closer to home via the REPORT-HF cohort study revealed rates of 21.0% and 19.0% for all-cause mortality at 12 months in its Asian cohort and middle-income cohort, respectively. 14 Similarly, data from the ASIAN-HF registry revealed rates of 9.6% and 13.0% for all-cause mortality at 12 months in its overall study population and South-East Asian cohort, respectively. Several reasons come to mind, including the short inpatient stay and long gaps in follow-up following discharge, which restrict optimisation of essential HF medication (especially within the ‘vulnerable’ three-month window) and gaps in provision cardiac services, which have similarly been reported in other studies (fivefold difference in provision of coronary intervention and stress testing between regions).23–25 There are compelling data to suggest that (a) AHF patients benefit from transitional clinics where follow-up occur almost immediately from discharge to allow better up-titration of medication, and (b) earlier access to specialised HF clinics to risk stratify and select patient who would benefit from more aggressive therapy, including devices. There are also data to suggest (c) better use of essential ‘triple therapy’ to target dose in HF amongst cardiologist versus health-care professionals from a primary care or general medicine background.23,26,27
Due to the health-care structure that existed in our centre during the study period (where patients were seen in standardised general medicine outpatient clinics between an average of four and six months), this limits the ability to monitor patients during the ‘transition phase’, implement up-titration of medications and expedite referrals for IHD evaluation that were crucial in preventing such outcomes. 1 Since 2016, our centre has established a dedicated HF clinic to tackle this issue through regular outpatient reviews within shorter durations and more frequent intervals, data of which are currently being analysed for future publication.
When compared to other registries such as the ADHERE-AP and ASIAN-HF, we note variation in data surrounding mortality and use of GDMT. However, it should be noted that both registries incorporated data from Malaysian centres with in-house cardiology services, and we argue that these registries are not fully reflective of the true practice in the majority of hospitals within Malaysia and that clinical practice in our hospital (a non-cardiac centre) may be a better reflection of how HF management truly is.
The limitations of this study include a lack of data on brain natriuretic peptide (BNP) and N-terminal-pro BNP (due to unavailability in most public hospitals), liver function tests and cardiac enzymes, which each carry prognostic value. We also note that majority of AHF episodes precipitated by ischaemia were based on the treating physician’s assumption at the time of admission, often based on a suggestive presenting complaints, risk factors and biochemical parameters, including raised cardiac enzymes on arrival (although we understand that this could be due to a myriad of causes). Unfortunately, data on external referrals to tertiary cardiac centres were not readily available for analysis, and therefore we were unable to substantiate evidence of true ischaemia (either via functional testing of coronary angiography) in our study. Patients with significant valvular lesions and congenital heart diseases were also not excluded in our study. We appreciate that the management and efficacy of standard HF therapy in patients with valvular disease and congenital heart diseases differ greatly compared to other causes. However, the objective of the study was to describe and illustrate the true burden of AHF regardless of aetiology. We also note the lack of data on left ventricular diameter measurements and degree of mitral regurgitation, which can influence LVEF values, which were not accessible during the data-collection period.
To our knowledge, this is only the third study published on AHF admission in Malaysia, and the only one published within the last decade. Similarities were seen when compared to the study by Chong et al., including on under-utilisation of ACE inhibitors and younger affected population. 28 In contrast, our study included a much larger cohort and followed patients through 12 months from their admission dates, allowing for a more robust data set compared to both Chong et al. and Chin et al.17,28 Other larger AHF registries involving a Malaysian cohort include the ADHERE-AP and ASIAN-HF, although each have been amalgamated into a larger ‘Asian’ data set.15,16 Our data mirror that of a recent cohort study by Tromp et al. (REPORT-HF) with regards to mortality rates and under-utilisation of GDMT amongst low- to middle-income and Asian (especially South-East Asian) communities, both of which encompass Malaysia. 14
Conclusion
In conclusion, this study highlights a relatively high readmission and mortality rate within our Malaysian cohort treated in a non-cardiac single centre. We believe this provides a good insight into the disparity in AHF outcomes not only between global databases, but also within a local setting, highlighting the need to incorporate data from local non-cardiac hospitals to truly understand the burden of AHF in Malaysia. We hope for more studies and collaborations from other non-cardiac centres in Malaysia to support our findings, and establish a local database that is representative of HF care in Malaysia.
Footnotes
Authors’ contributions
Availability of data and materials
Conflict of interest
Ethics approval
Funding
Informed consent
ORCID iD
