Introduction

Heart failure (HF) results from abnormalities or damage to the heart’s structure or function, reducing cardiac output and raising intra-cardiac pressures at rest or during exertion [1]. About 64 million people worldwide have HF, with 50% having severe HF (symptomatic HF at rest) [2]. Malaysia had one of the highest HF prevalence rates in Southeast Asia, with 721 cases per 100,000 persons in 2017, an increase of 7.7% from 669 cases per 100,000 persons in 1990 [3].

HF requires frequent hospitalisation and reduces life expectancy and quality of life. HF’s 30-day readmission rate ranged from 6.3% to 56% [4, 5]. Six-month all-cause readmission was 23% to 50% [6–8]. In Malaysia, 13% to 18.1% of HF patients were readmitted within 30 days [9, 10]. Furthermore, the readmission rate of HF increased steadily from 16.6% in 2007 to 19.6% in 2015 [10].

Management of HF is associated with increased healthcare resource utilisation and cost, leading to an increase in economic and clinical burden during the acute decompensated state and for a prolonged period after that [11]. HF consumes about 1–2% of developed countries’ annual healthcare budgets [8]. Cook et al. estimated the cost burden of HF across 197 countries using published estimates of HF economic burden [12]. The direct cost of HF consumed 1.1% and 0.11% of total health expenditure in high and middle-low-income countries, respectively. However, recently published data from Korea [13] and the Philippines [14] showed that the direct cost of HF was higher than the value estimated by Cook et al. These changes could be due to the increase in the countries’ gross domestic product (GDP) [15], and the direct cost of HF is positively correlated with the GDP [12, 16]. In addition, the risk of developing HF increases with age, so HF prevalence and costs are expected to rise with an ageing population [13, 17, 18]. Furthermore, different HF phenotypes categorised using left ventricular ejection fraction (LVEF) accrued cost differently [5, 19–21]. In Malaysia, 66.7% of HF patients were classified as LVEF <40% [9]–which was higher than in other European (59.0%) [22] and East Asia countries (21.0–35.2%) [23, 24]. Considering the increasing prevalence of HF, growing GDP, and different HF phenotypes distribution in Malaysia, the annual economic burden of HF is estimated to be higher than the value reported by Cook et al. for Malaysia (i.e., USD 12 million).

The previous cost analysis study focused on estimating the mean annual cost of HF by cost components for different HF phenotypes (categorised by LVEF), disease severities (categorised by New York Heart Association functional classification [NYHA]), and underlying comorbidities [25]. Subsequently, the main focus of the current study was to estimate the population-level economic burden of HF in Malaysia. Furthermore, the present study is a continuous work from the previous study by comparing the costs estimated using the unweighted average and inverse probability weighting (IPW) methods to provide insight into the impact of censored data on the average costs.

One approach to justify allocating healthcare resources to a specific condition would be to provide data on the extent of our healthcare system’s burdens for that condition. Thus, the present study aimed to estimate the overall economic burden of HF from the perspective of the public healthcare system. Identifying the HF phenotypes with the highest cost burden will help prioritise the scarce healthcare resources.

Methodology

Results

Direct cost

The mean (SD) annual cost per patient estimated using the IPW method was RM 8,295 (RM 5,274) [USD 5,217 (USD 3,317)], with inpatient costs accounting for 74.8% of the total costs (Fig 1). HFmrEF encountered the highest costs [RM 10,410 (RM 4,162), USD 6,547 (USD 2,618)], followed by HFpEF [RM 10,128 (RM 5,117), USD 6,368 (USD 3,218)] and HFrEF [RM 8,033 (RM 5,430), USD 5,052 (USD 3,415)]. The differences in the mean annual cost across HF phenotypes were not statistically significant (p = 0.778).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Annual cost per patient (in RM).

HFmrEF: heart failure with mildly reduced ejection fraction; HFpEF: heart failure with preserved ejection fraction; HFrEF: heart failure with reduced ejection fraction; IP: inpatient; OP: outpatient; RM: ringgit Malaysia. Inverse probability weighted costs were estimated using Bang & Tsiatic estimator [27]. Unweighted average costs were estimated by dividing the total cost by the total months of follow-up and transformed to annual cost by multiplication of 12 months.

https://doi.org/10.1371/journal.pone.0288035.g001

The mean annual cost per patient calculated using the unweighted average was RM 8,146 (RM 5,187) [USD 5,123 (USD 3,262)]. S2 Table (see ESM) summarises the mean difference of annual HF cost per patient estimated using the unweighted average and IPW method. The mean total annual costs estimated by the IPW method were higher than the unweighted average but not statistically different [mean difference (MD) = RM 148, 95% CI: -RM 870–731 (MD = USD 93, 95% CI: -USD 547–460), p = 0.865]. The annual costs estimated using both methods were not statistically different across HF phenotypes and cost components, p>0.05.

Economic burden of heart failure

Cardiologists from tertiary hospitals estimated that the overall diagnosis rate of HF in Malaysia was 80%. Based on the proportion of medicine use in public and private sectors in 2016, 64.5% of the diagnosed HF patients were followed up in public healthcare institutions [33]. Therefore, 87,448 HF patients (95% CI 75,798–99,523) received HF treatment in public healthcare institutions. According to MyHF, the prevalence of HFpEF, HFmrEF, and HFrEF were 21.8%, 11.5%, and 66.7%, respectively [9]. Thus, the total number of HFpEF, HFmrEF, and HFrEF patients who sought medical treatment in the public healthcare system was 19,064 (95% CI 16,524–21,696), 10,057 (95% CI 8,717–11,445) and 58,328 (95% CI 50,558–66,382), respectively.

The base case estimated annual direct cost of HF management from the public healthcare system’s perspective in Malaysia was RM 766.3 million (USD 481.9 million) (Table 1). Managing HFrEF patients accounted for 61% of the total financial burden of HF in Malaysia despite having the lowest mean annual cost. The high disease burden of HFrEF drove the main economic burden of HF in Malaysia.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Base-case estimated number of patients categorised in different phenotypes and age groups and their base-case cost of management.

https://doi.org/10.1371/journal.pone.0288035.t001

The annual cost of HF per patient increased with age, from RM 4,086 (USD 2,570) for the age group 20–29 to RM 9,271 (USD 5,831) for the age group 80+ years. Although the mean annual costs of HF per patient were not significantly different between patients with ages <60 years and ≥ 60 years [MD = RM 997, 95% CI: -RM 1,053–1,275, (MD = USD 627, 95% CI: -USD 666–802, p = 0.851)], the overall cost of HF increased by 68.7% to RM 226.0 million (USD 142.1 million) for age group 60–69 years as compared to 50–59 years due to growing number of patients living with HF. Conversely, the financial burden of managing HF decreased to RM 80.4 million (USD 50.6 million) for the age group of 80+ years. The cost of managing HF in patients from 50–79 years accounted for 74.1% of the total financial burden of HF in Malaysia.

Varying the direct cost of managing HF (first one-way sensitivity analysis) has a more significant impact on the overall financial burden of HF than changing the number of patients (Table 2). In the first sensitivity analysis, the direct cost of managing HF was RM 58.2 million– 1,695 million (USD 36.6 million– 1,066 million) compared to RM 664.2 million– 872.1 million (USD 417.7 million– 548.4 million) in the second one-way sensitivity analysis. Sensitivity analysis confirmed that HFrEF has the highest economic burden. The two-way sensitivity analysis considered the best- and worst-case scenarios showed that the financial burden of HF cost between RM 50.5 million– 1,929 .0 million (USD 31.7 million– 1,213.2 million). Thus, the HF financial burden was estimated to account for 1.05% (range: 0.07%–2.66%) of Malaysia’s total health expenditure in 2021 (S3 Table, see ESM) [34].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. One- and two-way sensitivity analyses for the estimated annual direct cost of HF in each phenotype and age groups.

https://doi.org/10.1371/journal.pone.0288035.t002

Discussions

The estimated mean annual healthcare cost per HF patient in Malaysia was RM 8,295 (RM 5,274) [USD 5,217 (USD 3,317)], with inpatient costs contributing about three-quarters (74.8%) of the total costs. This translated to a significant financial burden on the public healthcare system, especially for patients with HFrEF and aged between 50 and 79 years, which accounted for 61% and 74.1% of the total cost burden of HF in Malaysia, respectively.

This study employed broad inclusion criteria to achieve a sufficient sample size. However, the main drawback of such an approach was that not all the patients had complete cost data for one year of follow-up. Considering all observed or uncensored cases only (ignoring censored costs) will provide an inaccurate estimate of the actual mean total cost [35, 36]. An IPW estimator was applied to accurately estimate the total healthcare cost of HF [27]. The IPW estimator was chosen over other techniques that could handle the time-restricted mean cost because the former used single-record-per-subjects survival data. Furthermore, the IPW estimator with known cost histories provides robust mean and standard error estimates across different censoring mechanisms [37] and with mild to moderate degrees of data censoring (<18%) [38]. The periods with data censored observed in this study were 1.44% (S4 Table, see ESM). Thus IPW estimator is a reasonable and appropriate approach to handling censored cost data [39].

There was no significant difference in the mean annual cost of HF estimated using the unweighted average and IPW method in the current study. This might be due to the current study duration not being long enough to observe a significant difference in the cost. Another study that employed a similar method and duration of follow-up also found that the mean total annual cost of HF estimated by the unweighted average and IPW method showed no significant difference [40]. However, another study which estimates the cost of oral cancer patients with a longer duration of follow-up (4–5 years) observed a significant difference in the mean cost estimated using the unweighted average and IPW estimator [41]. The probability of having censored costs being low with a short duration of follow-up could cause the low impact of censored costs on estimating the actual mean total cost in the current study [42].

A systematic review that included cost studies from North America, South America, Europe, Asia and Africa reported that the annual healthcare cost related to HF per patient was between USD 1,096 and USD 101,886 in 2021 [43]. Another systematic review by Lesyuk et al. that included the USA, Europe, Africa and Asian countries found that inpatient admission costs accounted for about 44% to 96% of the overall direct cost of HF [44]. The wide variation in the costs reported across different countries may be mainly attributed to the relative cost variation of healthcare resources utilised to treat HF, treatment protocol, and, most importantly, the methodology used in calculating the costs. Additionally, the differences could also be explained by the vast discrepancy in resource allocation by each country for their healthcare system in managing HF [12].

From the public healthcare perspective, the direct cost of managing HF in Malaysia was RM 766.3 million (range: RM 50.5 million– 1,929 .0 million) (USD 481.9 million, range: USD 32.0 million– 1,213.2 million), which account for 1.05% (range: 0.07%–2.66%) of Malaysia’s total health expenditure in 2021. The global economic burden of HF was USD 108 billion per year in 2012 [12]. Developed countries devoted 1–2% of their total health expenditure to managing HF [8]. On average, the direct cost of HF consumed 1.1% and 0.11% of total health expenditure in high and middle-low-income countries in 2012, respectively [12]. The annual economic burden of HF in developed countries, such as the USA (USD 42.1 billion) [45], Korea (USD 625.6 million) [13] and the UK (USD 3 billion) [46], was much higher than in Malaysia. Healthcare expenditure is positively correlated with GDP [12, 16]. Thus, developed countries can allocate more resources to the healthcare sector. Conversely, developed countries like Ireland reported a lower economic burden (USD 243.9 million) [47] than Malaysia because the Irish population is significantly lower than Malaysia [48] and hence lesser people living with HF. Although Brazil has a higher GDP than Malaysia [49], the HF economic burden of Brazil (USD 420.3 million) [50] was lower than Malaysia because the study was conducted two decades ago. Advancement in the treatment of HF, including device therapy and newer pharmacology therapy, which are more expensive, has increased the healthcare cost of managing HF. Even though the GDP of Malaysia and the Philippines is similar, the economic burden of the Philippines was lower (USD 52.6 million) because only hospitalisation cost was considered when estimating the economic burden [14].

The direct cost burden of Malaysia reported by Cook et al. was USD 12 million in 2012 or 0.11% of total health expenditure, much lower than the cost burden estimated in this study. The discrepancy was caused by the economic burden of HF in countries without published data was extrapolated by Cook et al. using data from countries with published estimates, such as the USA and Brazil. In addition, the increase in the economic burden of HF in Malaysia was mainly attributed to the steady rise in the prevalence of HF [3].

The current study estimated the economic burden of HF in Malaysia using cost data calculated from the bottom-up approach [25]. Unlike other studies in which the costs were derived from insurance claims [13, 45] and hospital casemix database [50], the bottom-up approach can segregate the costs according to disease severities and other categories because detailed data was collected from each patient [51]. The current study reports the economic burden by HF phenotypes (categories using LVEF) and age groups, which enables the decision-makers to identify the main cost components and hence design policies to curb the huge economic burden of HF in Malaysia. The costs estimated from the casemix database were attributed to the primary diagnosis, and segregation of costs would be difficult [51].

We observed that the cost burden of managing HFrEF patients was the highest among the three HF phenotypes because the prevalence of HFrEF was the highest among the HF population in Malaysia [9] and Asia [52]. HFrEF was more commonly found among younger patients [17, 53] and having coronary artery disease (CAD) as the aetiology of HF [53, 54]. The mean age of this study population was 54.6 years, consistent with the population of HF-related studies previously conducted in Malaysia [9, 40, 55]. The Japanese HF cohort found that the mean age of HFpEF was significantly older than the other two phenotypes and was the most prevalent HF phenotype (61.9%) among the study population [24]. Besides, the prevalence of the traditional risk factors for CAD, such as diabetes, hypertension, and dyslipidaemia, has increased steadily among Malaysia’s population since 2011 [56], translated into a higher risk of developing myocardial infarction and eventually leading to HFrEF [17, 53].

In our study, we observed that the financial burden of HF increased with age until age 69 years because the prevalence of HF increases with age [13, 17, 18]. In addition, advancing age further increases the existing risk of traditional risk factors for developing HF [17]. In addition, the cost burden of HF peaked at the age group 60–69 years because HF patients in Malaysia were diagnosed at a much younger age (approximately 60 years) [9, 40, 55]compared to patients from other Asia countries (range: 66–81 years) [5, 57, 58] and western countries (range: 71–77 years) [21, 59, 60]. Furthermore, the cost of HF was the highest during the initial HF diagnosis [21]. Besides, the prevalence of HF decreased after 69 years in Malaysia, leading to a decrease in the cost burden of HF. This observation could be due to the life expectancy of patients with HF being 5–10 years after diagnosis [61, 62]. Secondly, the mean annual cost of HF in geriatric patients (80+ years) was lower than those of younger patients because older patients were treated with more conservative strategies [58] or were unable to seek medical treatment in the hospital due to their frail condition [40].

The growing economic burden of HF can substantially affect the clinical decisions and management strategies of healthcare providers and physicians. The rising economic burden of HF may reduce the availability of specialised clinics, multidisciplinary care teams, or equipment critical to managing HF patients. N-terminal pro-brain natriuretic peptide (NT-proBNP) is a biomarker for diagnosing HF and monitoring disease progression. However, this testing service is not widely available in Malaysia due to its impact on the financial budget. In addition, the high economic burden of HF can influence the selection of treatment options. Even though more expensive therapies may be more effective, clinicians may prioritise less expensive ones due to cost considerations [63]. Besides, the economic factor also affects the treatment compliance of HF patients. Patients may have trouble accessing their medications if the cost of their prescriptions is high or their insurance does not cover them [64]. This may result in non-adherence, underdosing, or medication discontinuation, all of which can negatively impact clinical outcomes [65].

This study has the potential risk of underestimating HF’s economic burden in Malaysia. First, the economic burden was calculated from the perspective of the public healthcare system only, and the indirect cost was not considered. The direct cost associated with private healthcare institution follow-up was not considered as funding mainly sourced from patients’ out of pockets expenses or private insurance companies. However, MoH Malaysia remains the most prominent financing source for healthcare in Malaysia (i.e., 52.5% of the total health expenditure) [66]. Future studies on the economic burden of Malaysia could include the healthcare cost from the private setting because one-third of the Malaysian population received treatment in the private setting [33]. In addition, future studies can be conducted from the societal perspective to include the indirect cost associated with absenteeism and productivity loss. Secondly, the prevalence of HF in Malaysia used in this study was the estimated value for 2017, but the prevalence of HF in 2021 could be higher due to the high prevalence of risk factors for HF and longer life expectancy in patients with comorbidities [67]. Nevertheless, sensitivity analyses have been performed to address these uncertainties around HF’s economic burden. Thirdly, this study had inherent limitations associated with retrospective data collection from patients’ medical records. However, data collectors were trained to find all HF-relevant details from patients’ medical records to increase the estimation accuracy. The strength of this study was that the direct cost of HF was calculated from the resources utilised in managing HF patients in three tertiary centres using the standard cost-of-illness method and local epidemiological data to estimate the HF’s economic burden. To the best of our knowledge, the present study is the first to estimate the financial burden of HF in Malaysia nationally. The current study’s findings could help stakeholders allocate resources for funding prevention programmes by identifying risk factors for HF and early diagnosis and treatment of HF.

Conclusion

The present study showed that HF posed a substantial financial burden to the public healthcare system. The cost of HF estimated using weighted and unweighted approaches did not differ significantly in the current study. The higher prevalence of HFrEF in Malaysia and the inpatient costs are the main financial burden of HF. Demographic shifts in Malaysian society will inevitably increase the prevalence of HF, which would further increase the financial burden of HF. Thus, interventions that reduce hHF and target HF risk factors to prevent new HF may have economic benefits.

Supporting information

Acknowledgments

We would like to thank the Director General of Health Malaysia for his permission to publish this article. We acknowledge the contribution of data collectors from all the participating hospitals. Boehringer Ingelheim (Malaysia) Sdn Bhd was given the opportunity to review the manuscript for medical and scientific accuracy as well as intellectual property considerations.