Health system efficiency and equity in ASEAN: an empirical investigation

Abstract

Background

Equity and efficiency are two fundamental principles for the sound development of health systems, as advocated by the World Health Organization (WHO). Despite the notable progress made by the Association of Southeast Asian Nations (ASEAN) in advancing their health systems, gaps persist in achieving global health goals. This paper examines the efficiency of health system stages and the fairness of health resource distribution in ASEAN countries, analyzes the underlying causes of the existing gaps, and suggests potential solutions to bridge them.

Methods

Data spanning 2011 to 2019, sourced from the WHO Global Health Observatory and the World Bank Database, form the foundation of this study. This study employs an enhanced two-stage data envelopment analysis (DEA) to assess the efficiency of health system stages in ASEAN countries. Equity in health resource distribution is evaluated using health resource agglomeration degree and concentration curves across demographic, geographic, and economic aspects. Furthermore, the Entropy-Weighted TOPSIS method is utilized to integrate equity across these dimensions, measuring the overall fairness in health resource allocation across different countries. Finally, rankings of health system fairness and efficiency are compared to assess the overall development level of health systems.

Results

The overall efficiency of the ASEAN health systems from 2011 to 2019 averaged 0.231, with an upward trend in the first stage efficiency at 0.559 and a downward trend in the second stage at 0.502. The health resource agglomeration degree indicated that Singapore, Brunei, and Malaysia had HRAD and HRPD values significantly greater than 1, and Cambodia, Myanmar, and Laos predominantly had indices significantly less than 1. The concentration curve for hospital beds was the closest to the line of absolute equity. During the study period, the health resource concentration curve increasingly approached absolute equity, shifting from above to below the concentration curve. Singapore, Brunei, and Malaysia consistently remained in the first quadrant of the quadrant plot, and Myanmar and Cambodia were consistently in the third quadrant.

Conclusion

ASEAN countries face two key challenges in their healthcare systems: first, while many nations such as Indonesia, Thailand, and Vietnam have improved resource allocation efficiency, this hasn’t yet translated into better health services. To address this, establishing national health sector steering committees, focusing on workforce training and retention, and implementing centralized monitoring systems are crucial. Second, there is a growing disparity in healthcare development across ASEAN. Promoting balanced resource distribution and leveraging ASEAN’s economic integration for regional collaboration will help bridge these gaps and foster more equitable healthcare systems.

Introduction

It has become increasingly evident that a solid health system is an essential component of a healthy and just society. The World Health Organization (WHO) underscores the significance of robust health systems as essential to achieving Universal Health Coverage (UHC) and health-related Sustainable Development Goals (SDGs) [1]. Moreover, equity and efficiency are two fundamental values of high-quality health systems [2]. The interplay between efficiency and equity in health systems often operates in both directions, with improvements in performance enhancing equity and increases in equity boosting efficiency [3]. The harmonious development of equity and efficiency within health systems ensures the delivery of high-quality medical services and the achievement of optimal health outcomes [4].

The efficiency of a health system, as defined by the WHO, refers to the extent to which system goals are achieved with a given level of resource input [5]. Current research on health system efficiency primarily follows the classification by Farrell, dividing efficiency into technical efficiency (TE) and allocative efficiency (AE) [6]. TE is reflected in the healthcare provider’s ability to fully utilize its available production factors or inputs to achieve maximum output [7]. On the other hand, AE focuses on maximizing output at a given level of input cost or minimizing input cost at a given output level [7]. Only when both technical and allocative efficiencies are met can overall economic efficiency be achieved.

Equity in the health system means equal access to care based on need, equal utilization for similar needs, and equal quality of care for all individuals [8]. The fair allocation of resources is a fundamental aspect of health equity, as it directly influences access to healthcare services and outcomes across different populations [9]. In this study, equity specifically refers to the fairness of healthcare resource distribution. Generally, the fairness of a health system is indirectly reflected by measuring the degree of health inequity [10]. Asada’s three-step approach to measuring health equity-defining health equity, selecting a framework or indicators for measurement, and quantifying the degree of health inequality using concentration indices or Gini coefficients-lays the groundwork for the measurement of health equity [11].

The Association of Southeast Asian Nations (ASEAN), comprising ten independent countries-Indonesia, Myanmar, Brunei, Cambodia, Laos, Malaysia, the Philippines, Singapore, and Vietnam, is the fifth-largest economy globally. Despite political, economic, and social diversity, ASEAN countries have made significant strides in healthcare, particularly in strengthening health systems and achieving UHC [12, 13]. Under ASEAN Health Cluster 3’s 5-year work program, which prioritizes strengthening health systems and access to care, UHC has become a central goal for member states [14]. Countries such as Malaysia, Singapore, and Thailand have implemented extensive social health insurance (SHI) policies, which have contributed to increasing UHC coverage for over 90% of their populations, including initiatives like Malaysia’s PeKa B40 Scheme and Singapore’s “3M” system: Medisave, MediShield, and Medifund [15, 16]. Myanmar’s National Health Plan 2017–2021 outlines a phased strategy aimed at increasing UHC by 2030, focusing on accessibility, equity, and financial protection [17]. Cambodia’s 2024–2028 universal health coverage roadmap aims to meet key UHC targets by 2035, focusing on expanding coverage and reducing healthcare costs [18]. Thailand, Singapore, and Malaysia, as Southeast Asia’s medical tourism hubs, leverage their healthcare system structures to offer high-quality services competitively, generating additional investment resources and curbing the brain drain of health professionals [19, 20]. Despite advancements, health systems in ASEAN face challenges. Budgetary constraints on health expenditures in countries like Malaysia, the Philippines, Vietnam, and Laos lead to significant out-of-pocket payments, potentially compromising the welfare of the ASEAN population [21, 22]. Furthermore, disparities in healthcare resource allocation, with wealthier provinces receiving disproportionate investments, exacerbate inequities [23, 24]. Additionally, the region faces increasing pressure from shifting epidemiological patterns and aging populations, driving up healthcare costs, particularly for non-communicable diseases [25, 26]. Therefore, the study of efficiency and equity in ASEAN’s health systems is urgently necessary, as it is crucial for the rational and efficient utilization of health resources, ultimately promoting the sustainable development of health systems across the region.

This study will address two key questions: What is the current state of efficiency and equity in the health systems of ASEAN countries? What are the underlying causes of this efficiency and equity? By analyzing these questions, the study will provide new insights into the efficiency of ASEAN health systems and the equity of health resource allocation, along with valuable recommendations for policymakers in the region. The organization of this article is as follows: “Literature review” section reviews relevant studies, discusses the limitations of existing research, and outlines the contributions of this paper. “Methods” section details the methodology. “Data collection” section discusses the selection of indicators and data sources. “Results” section analyzes the data. “Discussion” section presents the results and their implications. “Conclusions and policy implications” section summarizes the main findings and offers policy recommendations.

Literature review

Health system efficiency

The majority of studies on health system efficiency utilize quantitative methods, with Data Envelopment Analysis (DEA) the most widely adopted non-parametric approach, and Stochastic Frontier Analysis (SFA) being the leading parametric method [27, 28]. Arhin et al. [29] employed a variable returns to scale (VRS) output-oriented DEA method to evaluate the efficiency of healthcare systems in sub-Saharan Africa towards UHC, revealing a potential efficiency enhancement of 19%. Zhou et al. [30] utilized a DEA model and Malmquist index to assess the efficiency of China’s urban and rural healthcare institutions, suggesting that technological upgrades are needed to enhance rural primary healthcare efficiency. DEA evaluates relative efficiency by comparing the weighted sum of outputs to the weighted sum of inputs for production units. DEA compares the relative efficiency of production units by analyzing the weighted sum of outputs relative to inputs, but its deterministic nature makes it vulnerable to measurement errors and statistical noise [27]. In contrast, SFA uses regression analysis to estimate the production frontier and calculates efficiency based on the residuals of the estimated equation [31]. Ngami and Ventelou [32] employed SFA to analyze the impact of environmental factors on OECD countries’ healthcare system performance, finding that environmental quality significantly affects system rankings. Wu et al. [33] applied SFA to gauge the efficiency of global COVID-19 pandemic responses and their influencing factors, concluding that the average global efficiency in responding to COVID-19 is poor, with significant potential for enhancement. Although SFA accounts for the impact of measurement errors, its reliance on functional form and the distribution of random errors can also bias the results [34]. Most research tends to focus on overall health system efficiency, often treating the system as a “black box” without considering the internal dynamics between subsystems. However, studies have emphasized that health systems consist of various interconnected subsystems, and the overall efficiency is closely tied to the performance of these subsystems [35, 36]. Traditional efficiency measurements fail to break down the system into its components, neglecting the internal mechanisms and interrelationships between subsystems [28, 36]. On one hand, decomposing overall system efficiency into subsystem efficiencies provides valuable insights into the internal factors driving poor performance, offering a more granular understanding than traditional DEA models [35]. On the other hand, accounting for the interconnections between subsystems and their operational variations enhances the accuracy of overall efficiency assessments [37]. To address these gaps, this paper adopts an improved two-stage DEA approach to analyze the ASEAN health systems, dividing them into two interrelated subsystems and measuring both overall and subsystem efficiencies across different countries.

Health system equity

Numerous studies have examined health equity, with a particular focus on evaluating the fairness of health resource allocation. Commonly used methods in these studies include the Gini coefficient, concentration index, and agglomeration degree. Most research integrates these approaches to assess health system equity, which can generally be classified into two categories.

The first type involves directly using the Gini coefficient, concentration index, and Theil index to measure the fairness of health resource distribution, often at a micro level. The Gini coefficient and concentration index are used to assess the distribution of specific health resources across different income groups, with the concentration curve serving as a graphical representation. The Theil index, which can be decomposed into within-group and between-group inequality, identifies the main sources of inequity. For instance, Feng et al. [38] used the concentration index and curve to examine income-based inequities in Vietnam’s health resource allocation and applied the Theil index to measure regional disparities, revealing increased pro-poor inequity in resources like per capita health budgets and community hospital beds. Similarly, Chai et al. [39] employed these methods to analyze the fairness of Health Human Resource (HHR) allocation in China from 2012 to 2021, highlighting significant geographic inequalities but relative population-based fairness.

The second category adopts a more macro perspective, where studies combine the Gini coefficient, concentration index, and agglomeration degree to evaluate broader systemic inequities. In these studies, the concentration index or curve often measures economic disparities in resource allocation, while the agglomeration degree supplements this by assessing the degree of alignment between the distribution of health resources and factors such as population and geographic area to evaluate equity. Dai et al. [40] used the Gini coefficient to evaluate the distribution of Traditional Chinese Medicine (TCM) hospital resources in China, finding better equity in population-based distribution compared to geographical allocation, though regional disparities persisted. Some research has also adopted the entropy-weighted TOPSIS method for a more comprehensive evaluation of resource distribution equity. For instance, Liu et al. [41] analyzed economic, population, and geographic aspects of health resource distribution in the Yangtze River Economic Belt using the concentration curve and agglomeration index while incorporating the entropy-weighted TOPSIS method to calculate an overall equity score. Similarly, Chen et al. [42] applied this method to assess the fairness of basic health resource allocation across cities in Hubei Province, taking into account geographic, population, and economic factors. Building on previous research and considering data availability in ASEAN member states, this study adopts a macro-level approach, employing three methods to comprehensively assess the equity of healthcare resource allocation. First, we use the concentration index to measure equity in terms of geographic and demographic distribution. Second, concentration curves illustrate the distribution of various healthcare resources among different income groups. Finally, we apply the entropy-weighted TOPSIS method to normalize multiple indicators of healthcare resource equity across different dimensions and compute an overall equity score and ranking.

This study evaluated the effectiveness of healthcare systems and the equitable allocation of healthcare resources in the ASEAN region from 2011 to 2019, aiming to contribute in the following three areas:

• (1) The research delves into two subsystems of health system efficiency: resource allocation and service capacity. By applying an enhanced two-stage DEA model, we measure the overall and segmented efficiency across ASEAN countries’ health systems, thereby overcoming the traditional DEA’s “black box” approach that disregards the internal dynamics of subsystems.

• (2) Health resource agglomeration degree and concentration curves are used to analyze the equity of health resource distribution from geographical, population, and economic perspectives. This methodology compensates for the shortcomings of previous studies that evaluated equity solely based on single dimensions such as economic status or population size.

• (3) The study employs the Entropy-Weighted TOPSIS method to rank the fairness of health system resource allocation among ASEAN countries. By plotting the rankings for equity alongside the rankings for the overall efficiency of health systems in a scatter diagram, we perform an integrated analysis to evaluate the general level of fairness and efficiency across the health systems of ASEAN member countries.

Methods

Improved two-stage data envelopment analysis

This study employed the relational two-stage DEA developed by Kao and Hwang [35] to evaluate health system efficiency in ASEAN. The method can decompose the efficiency of the whole process into the product of the efficiencies of the two sub-processes and identify the causes of inefficiency more accurately. The assumption in this study is that a production process consists of a sequence of two sub-processes, as illustrated in Fig. 1. The whole process uses $\text{m}$ inputs to produce $\text{s}$ outputs ${{Y}_{rk}},r=1,\ldots,s$. In contrast to the traditional one-stage production process [43], here the overall process consists of two sub-processes involving $\text{q}$ intermediate products ${{Z}_{pk}},p=1,\ldots,q$. Additionally, the intermediates ${{Z}_{pk}}$ serve as both the outputs of stage 1 and the inputs of stage 2. Drawing from the two sub-processes stated above, we consider $DMU\text { k}$ as a composite of two divisions.

Fig. 1

Mechanism of the improved two-stage DEA model

Different from the conventional two-stage DEA method [44], the connection between the two sub-processes is established through a description of a relationship between the overall process and the two sub-processes. Consider $DMU\text { k}$, $u_r^*$, $v_i^*$ and $w_p^*$ are identified as the multipliers selected by $DMU\text { k}$ to determine its overall efficiency $E_{k}$ as the multipliers that has selected to calculate its overall efficiency and sub-process efficiencies $\textbf{E}_k^1$ and $\textbf{E}_k^2$. Upon converting the linear fractional program to a linear program, the resulting form is as follows:

$$\begin{aligned} \begin{aligned} E_{k}& =\max \sum _{r=1}u_rY_{rk} \\ & \quad \text {s.t.} \sum _{i=1}^m\nu _iX_{ij}=1 \\ & \quad \sum _{r=1}^su_rY_{rj}-\sum _{i=1}^m\nu _iX_{ij}\le 0,j=1,2,\ldots,n \\& \quad \sum _{p=1}^qw_pZ_{pj}-\sum _{i=1}^m\nu _iX_{ij}\le 0,j=1,2,\ldots,n \\ & \quad \sum _{r=1}^su_rY_{rj}-\sum _{p=1}^qw_pZ_{pj}\le 0,j=1,2,\ldots,n \\& \quad u_r, v_i, w_p\ge \varepsilon ,r=1,2,\ldots,s;i=1,2,\ldots, m;p=1,2, \ldots,q \end{aligned} \end{aligned}$$

(1)

The efficiencies are obtained subsequently as ${{E}_{k}}=\underset{r=1}{\overset{s}{\mathop {\sum }}}\,u_{r}^{*}{{Y}_{rk}},E_{k}^{1}=\underset{p=1}{\overset{q}{\mathop {\sum }}}\,w_{p}^{*}{{Z}_{pk}}/\underset{i=1}{\overset{m}{\mathop {\sum }}}\,v_{i}^{*}{{X}_{ik}},$ and $E_{k}^{2}=\underset{r=1}{\overset{s}{\mathop {\sum }}}\,u_{r}^{*}{{Y}_{rk}}/\underset{p=1}{\overset{q}{\mathop {\sum }}}\,w_{p}^{*}{{Z}_{pk}}$. Thus, ${{E}_{k}}=E_{k}^{1}\times E_{k}^{2}.$

Nevertheless, the optimal multipliers derived from Eq. (1) may lack uniqueness, leading to a non-deterministic decomposition of ${{E}_{k}}=E_{k}^{1}\times E_{k}^{2}$. Consequently, comparing either $E_{k}^{1}$ or $E_{k}^{2}$ among all DMUs lacks a common basis. To address this issue, a potential resolution is to identify the set of multipliers that yield the largest $E_{k}^{1}$ while ensuring that the overall efficiency score, ${{E}_{k}}$ calculated from Eq. (1) is maintained. This idea can be expressed as follows:

$$\begin{aligned} \begin{aligned}E_{k}^1&=\text { max }\sum _{p=1}^{q}w_{p}Z_{pk}\\&\quad s.t.\sum _{i=1}^{m}\nu _{i}X_{ik}=1\\&\quad \sum _{r=1}^{s}\varvec{u}_{r}Y_{rk}-E_{k}\sum _{i=1}^{m}\nu _{i}X_{ik}=0,\\&\quad\sum _{r=1}^{s}\varvec{u}_{r}Y_{rj}-\sum _{i=1}^{m}\varvec{\nu }_{i}X_{ij}\le 0,j=1,2,\ldots,n\\&\quad \sum _{p=1}^{q}w_{p}Z_{pj}-\sum _{i=1}^{m}\nu _{i}X_{ij}\le 0,j=1,2,\ldots,n\\&\quad\sum _{r=1}^{s}u_{r}Y_{rj}-\sum _{p=1}^{q}w_{p}Z_{pj}\le 0,j=1,2,\ldots,n\\&\quad u_{r},v_{i},w_{p}\ge \varepsilon ,r=1,2,\ldots, s;i=1,2,\ldots, m;p=1,2, \ldots, q\end{aligned} \end{aligned}$$

(2)

After $E_{k}^{1}$ is calculated from the above model, the efficiency of the second stage is obtained as: $E_{k}^{2}={{{E}_{k}}}/{E_{k}^{1}}\;$. Figure 2 provides a more intuitive illustration of the specific application of this method in this paper.

Fig. 2

Application of the improved two-stage DEA model in health system

Health resource agglomeration degree

The concept of agglomeration degree originated in economics, referring to the concentration of enterprises within the upstream and downstream industry chain in a specific area [45]. It has since been widely applied across various research fields. Building upon this framework, Yuan et al. [46] expanded its application within the health sector, thereby introducing the related concepts of health resource agglomeration. Unlike other equity measures for health resource distribution, such as the Gini coefficient and the Theil index, the agglomeration degree offers the advantage of comprehensively considering geographical and demographic factors while maintaining a straightforward calculation method [40, 47]. The specific calculation formulas are as follows [47, 48]:

Agglomeration degree of health resources by geographical area:

$$ \text{HRAD}_i=\frac{\text{HR}_i/\text{HR}_n\times 100\%}{A_i/A_n\times 100\%}=\frac{\text{HR}_i/A_i}{\text{HR}_n/A_n} $$

(3)

Agglomeration degree of health resources by population:

$$ HRPD_i=\frac{HRAD_i}{PAD_i}=\frac{HR_i/P_i}{HR_n/P_n}$$

(4)

where

$$ \textbf{PAD}_i=\frac{P_i/P_n\times 100\%}{A_i/A_n\times 100\%}=\frac{P_i/A_i}{P_n/A_n}$$

(5)

$HRA{{D}_{i}}$: the geographical concentration of health resources in the $i\text {-th}$ region.

$PA{{D}_{i}}$: population concentration of the $i\text {-th}$ region.

$HRP{{D}_{i}}$: the population concentration of health resources in the $i\text {-th}$ region.

$H{{R}_{i}}$: the number of certain types of health resources owned by the $i\text {-th}$ region.

${{A}_{i}}$: the land square of the $i\text {-th}$ region.

$H{{R}_{n}}$: total health resources of the upper-level region.

${{A}_{n}}$: total land area of the upper-level region.

${{P}_{i}}$: the population of the $i\text {-th}$ region.

${{P}_{n}}$: the total population of the upper-level region.

In the instance when $HRA{{D}_{i}}=1$, it signifies the health resources allocation across the region adheres to absolute fairness on a geographical scale. Similarly, when the value of $HRP{{D}_{i}}=1$, it signifies the health resources allocation across the region adheres to absolute fairness based on population size.

Concentration curve

The concentration curve is a measure endorsed by the World Bank that evaluates the fairness of distributing healthcare resources across different regions, taking into account varying economic circumstances [49]. The concentration curve plots the cumulative percentage of the health variable (y-axis) against the cumulative percentage of the population, ranked by living standards, beginning with the poorest, and ending with the richest (x-axis). The calculation formula of S is [50]:

$$ S=\frac{1}{2}\sum _{i=0}^{n-1}\left(Y_i+Y_{i+1}\right)\left(X_{i+1}-X_i\right)$$

(6)

${{X}_{i}}$: the cumulative percentage of the population in the $i\text {-th}$ region.

${{Y}_{i}}$: the cumulative percentage of health resources in the $i\text {-th}$ region.

S: the area between the concentration curve and the absolute fairness line.

The farther the concentration curve is from the line of absolute equity, the more unfair the resource allocation; conversely, the closer it is to the line, the fairer the resource allocation. The concentration curve below the absolute fairness line indicates that resources are skewed toward better economic regions and vice versa.

Entropy-weighted TOPSIS method

The entropy weight method evaluates indicator dispersion with a mathematical approach; indicators with lower entropy values provide more information and carry greater weight, in contrast to those with higher entropy values [51]. The TOPSIS method identifies the optimal and worst-case solutions within the space defined by the positive and negative ideal solutions and ranks. It evaluates the alternatives based on their relative distance to the optimal solution [52]. The Entropy-Weighted TOPSIS (EW-TOPSIS) method combines the strengths of both the entropy weight method and the TOPSIS approach. It allows for the simultaneous consideration of multiple criteria, making it particularly suitable for complex decision-making scenarios. Moreover, by employing the entropy weight method, EW-TOPSIS provides a more objective assessment of the importance of each indicator, thereby reducing subjectivity in the decision-making process. This integration enhances the robustness and reliability of evaluations, especially in contexts like health resource allocation and system efficiency [53]. The specific computational steps are as follows [54, 55]:

• (1) Standardization of data. In this study, since all indicators are positive, the following formula can be employed for data standardization:

  $$ x_{ij}^{\prime}=\frac{x_{ij}-\min _ix_{ij}}{\max _ix_{ij}-\min _ix_{ij}}$$

  (7)

  ${{X}_{ij}}$: the observed data for the $j\text {-th}$ criterion of the $i\text {-th}$ evaluated object.

• (2) Determine the weights of the indicators. The weight of the $j\text {-th}$ indicator is

  $$ E_j=-k\sum _{i=1}^nf_{ij}\;\text{ln}\;f_{ij}\left(f_{ij}=\frac{x_{ij}^{\prime}}{\sum _{i=1}^nx_{ij}^{\prime}},k=\frac{1}{\text{ln}n}\right)$$

  (8)

  $\text {If}\;{{f}_{ij}}=0,\; \text {then}\;{{f}_{ij}}\ln {{f}_{ij}}=0;$

  $$ W_j=\frac{1-E_j}{p-\sum _{i=1}^nE_j}(0{\leqslant }W_j{\leqslant }1)$$

  (9)

  ${{X}_{j}}$: the information entropy of the $j\text {-th}$ criterion. ${{W}_{j}}$: the weight of the $j\text {-th}$ indicator.

• (3) Calculation of relative closeness

  $$\begin{aligned} C_i=\frac{D_i^-}{D_i^++D_i^-}\end{aligned}$$

  (10)

  ${{C}_{i}}$: the composite score index (relative proximity). $D_{i}^{+}$: the positive ideal solution distance. $D_{i}^{-}$: the negative ideal solution distance.

Data collection

Variable selection

Input variables and intermediate variables

The health system can be seen as a macroscopic unit of health production [56]. The process of decision-making, encompassing resource allocation and the realization of objectives, can be segmented into distinct stages [57]. Common input indicators for the health system encompass health expenditure, hospital bed capacity, and the healthcare workforce, which are widely utilized for assessment [27, 58]. Cheng and Qian have suggested that health financing is the initial input within the health system framework [57]. Furthermore, Evans and Estienne [59] and Chang et al. [60] have argued that health financing is an essential precondition for constructing health infrastructure and remunerating healthcare providers. However, certain studies indicate that health outcomes, such as infant mortality and life expectancy, are not strongly [61,62,63], or directly linked to the level of expenditure on health [64, 65]. In light of these findings, this study adopts current health expenditure as the primary input indicator for the health system, with the number of hospital beds, medical doctors, and nurses as the intermediate resource inputs essential for providing health services.

Final output variables

Improving health status is one of the core goals of health systems as defined by the WHO [5]. Therefore, a wide variety of research has applied life expectancy and infant mortality as the ultimate outputs of the health system [27]. The expert panel on Understanding Cross-National Health Differences among High-Income Countries states that a health system encompasses the full continuum between public health and medical care services. Public health services focus on population-wide health through collective actions like disease prevention and health promotion, which can be reflected in indicators such as vaccination rates and coverage of basic public health services [66]. In contrast, healthcare services concentrate on individual patients’ diagnosis, treatment, and disease management [66]. Based on previous research and data availability, we categorize outcome indicators into two parts: public health and healthcare services [36, 37].

Non-communicable disease mortality rates were chosen as an output indicator for public health services, as they are the leading global cause of death and disability. WHO estimates that chronic diseases account for 41 million deaths annually, representing 71% of total global mortality, underscoring their critical public health challenge [67]. Access to safe drinking water, sanitation, and hygiene are fundamental to human health and well-being and align with the United Nations’ Sustainable Development Goal 6 (SDG 6) [68]. Consequently, the proportion of people using at least basic sanitation services is included as an output indicator for public health services. Furthermore, drawing on the previous research, this study adopts neonatal mortality rate, life expectancy at birth, and maternal mortality ratio as output indicators for healthcare services [36, 37]. The specific indicators selected can be seen in Table 1.

Table 1 Indicators of health system efficiency

Data source

We collected panel data from ten ASEAN nations for 2011–2019, a period chosen for data accessibility and to pre-empt COVID-19’s effects on health system metrics. This timeframe also signifies ASEAN’s health system evolution between the Millennium Development Goals (MDGs) and Sustainable Development Goals (SDGs). The majority of health system data for these countries were sourced from the WHO Global Health Observatory and the World Bank Database. Any missing data were initially addressed by consulting national statistical yearbooks, followed by interpolation to complete the dataset.

Results

The efficiency of the health system in ASEAN countries

Analysis of overall efficiency

Figure 3 illustrates the overall efficiency of health systems in ASEAN from 2011 to 2019, as well as the efficiencies of the two sub-stages: resource allocation (stage 1) and service capacity (stage 2). The overall efficiency averaged 0.231, indicating a relatively low level with minimal change over the observed years. There is notable variance in efficiency scores among countries, with Singapore achieving the highest average score of 0.637, while Cambodia recorded the lowest at 0.066 (Table 2). The remaining countries consistently had efficiency values below 0.4 during the study period. In terms of the sub-stages, average efficiency values for stage 1 and stage 2 are 0.559 and 0.502, respectively, indicating moderate performance. Notably, stage 1 demonstrates a general upward trend in efficiency, contrasting with stage 2, which exhibits a decline. Since 2018, the efficiency of stage 2 has consistently remained lower than that of stage 1, marking a significant shift in the dynamics of the health system’s sub-stages.

Fig. 3

Efficiency of the health system in AEEAN from 2011 to 2019

Table 2 Average efficiency of the health system in ASEAN from 2011 to 2019

Analysis of sub-process efficiencies

The initial phase of health system efficiency in ASEAN countries from 2011 to 2019 reveals four distinct trends in health resource allocation efficiency. Indonesia, Thailand, Vietnam, and Malaysia exhibited upward trajectories, with Vietnam notably increasing from 0.378 in 2011 to a full score by 2019 (Table 3). Meanwhile, the efficiency of Myanmar experienced a decline from 0.318 in 2011 to a low of 0.163 in 2019. Brunei and Laos demonstrated fluctuating trends, with Brunei witnessing a decrease to 0.722 in 2016, followed by a subsequent rise to 1 in 2019. Cambodia, the Philippines, and Singapore, however, had minimal fluctuations, with Singapore achieving the highest efficiency at 0.685 and Cambodia the lowest at 0.078.

Table 3 Resource allocation efficiency of the health system in ASEAN from 2011 to 2019

In the second stage of health system efficiency, reflecting health service capacity, most ASEAN countries showed a decline in efficiency between 2011 and 2019 (Fig. 3, Table 4). Singapore led with an average second-stage efficiency of 0.929, while the Philippines had the lowest at 0.232. Notably, despite Cambodia’s low first-stage efficiency, it demonstrated significant improvement in the second stage, peaking at 1 in 2019 and averaging 0.909-second only to Singapore.

Table 4 Service capacity efficiency of the health system in ASEAN from 2011 to 2019

Equity of the health system in ASEAN countries

Agglomeration analysis of health resource allocation by geographical area and population

The results of the calculation for health resource concentration degree by geographical area (HRAD) and population (HRPD) in ASEAN countries for 2019 are presented in Table 5. Based on this, the allocation of health resources in each country can be categorized into three groups.

Table 5 Equity analysis of health resource allocation in ASEAN in 2019

Most HRAD and HRPD values for Singapore, Brunei Darussalam, and Malaysia exceed 1, with Singapore’s HRAD for medical doctors and nursing personnel reaching the hundreds, indicating a high concentration of health resources in these countries, both geographically and demographically. In contrast, Cambodia, Myanmar, Laos, and Indonesia show agglomeration degree below 1 for most indicators, reflecting inadequate access to medical resources and imbalances in their geographic and demographic distribution. Thailand has a value of 1 to 1.5 for each of the indicators of agglomeration, indicating a relatively balanced allocation of health resources. However, some countries show disparities in the development of different resource types. The Philippines and Vietnam, for instance, have both values significantly above and below 1, indicating uneven development across various health resource categories. This underscores the need for improved health resource planning in these nations.

Analysis of the concentration curve of health resource allocation

The concentration curve indicates the equity of medical resource allocation across regions of varying economic statuses. Using the cumulative percentage of population as the X-axis and the cumulative percentages of health resources as the Y-axis, the ASEAN-10 countries are ranked by GDP in ascending order to plot concentration curves, as depicted in Figs. 4, 5, 6.

Fig. 4

Concentration curve of hospital beds

Fig. 5

Concentration curve of medical doctors

Fig. 6

Concentration curve of nursing personnel

Hospital beds display a concentration curve closest to the line of absolute equity, indicating a more equitable distribution compared to medical doctors and nursing personnel (Fig. 4). The concentration curve for physicians shifts downward from above to below the equity line (Fig. 5), suggesting that the allocation of medical professionals increasingly favors economically stronger regions in ASEAN. Similarly, the concentration curve for nursing staff also trends downward, but the gap from the equity line is narrowing (Fig. 6), indicating that while nursing resources are still concentrated in wealthier areas, the distribution disparity is gradually decreasing.

Comprehensive analysis of health system efficiency and equity

A scatter plot is constructed with the ranking of the comprehensive index reflecting equity in health resource allocation on the X-axis and the ranking of the health system efficiency value on the Y-axis (Fig. 7). This study’s composite score index reflects the equity of health system resource distribution by considering health resource agglomeration from geographical and demographic perspectives. The indicator weights, composite scores, and country rankings are detailed in Tables 6, 7. The two-dimensional scatter plot is divided into four quadrants, categorizing the 10 ASEAN countries into four types based on their development in health system efficiency and resource equity. The first quadrant represents “dual high” types with relatively high health system efficiency and health resource agglomeration; the second quadrant indicates “low-high” types with lower efficiency but higher agglomeration; the third quadrant denotes “dual low” types with low efficiency and low agglomeration; and the fourth quadrant signifies “high-low” types with higher efficiency but lower agglomeration.

Fig. 7

Distribution of health resource agglomeration and efficiency from 2013 to 2019 (from a to c). The first to the fourth quadrants indicate high agglomeration and high efficiency, low agglomeration and high efficiency, low agglomeration and low efficiency, and high agglomeration and low efficiency, respectively

Table 6 Weight of each indicator in 2013, 2016, 2019 (%)

Table 7 Composite score index and ranking of ASEAN countries in 2013, 2016, 2019

Figure 7 reveals that most nations are grouped within the first quadrant (dual high type) and the third quadrant (dual low type), indicating a substantial divergence and a bipolar pattern in the integrated development of health system performance and fairness across countries. Singapore, Brunei, and Malaysia consistently reside in the first quadrant during the study period, indicating high health system efficiency but highly concentrated resource allocation (Table 5, Fig. 7). Cambodia, Myanmar, and Indonesia consistently occupy the third quadrant, suggesting a scarcity of health resources and low resource utilization efficiency. Notably, the Philippines’ health system has shifted from the second to the third quadrant. Considering Table 5 and Fig. 7 together, it can be inferred that the Philippines has seen a reduction in the previously high concentration of health resources, suggesting an enhancement in equity. In 2019, Laos and Thailand’s indicator points lie on the axes, signifying a critical phase in improving their health system efficiency and equity (Fig. 7).

Discussion

This study provides a comprehensive analysis of the health system efficiency across ten ASEAN countries from 2011 to 2019. The overall efficiency was found to be low, with an average of 0.231 and minimal variation over time. These findings align with the efficiency analyses of ASEAN health systems by Kang et al. and Singh et al. [31, 69]. The variation in the sub-stage efficiencies of ASEAN’s health systems can elucidate the reasons behind the overall low efficiency. Regarding the efficiency of the health system’s stages, ASEAN countries exhibit an upward trend in Stage 1 efficiency, while Stage 2 efficiency continuously declines. This indicates that improvements in resource allocation have not effectively translated into enhanced service capacity, leading to overall low efficiency. This issue is corroborated by a WHO survey on the efficiency of typical national health systems [70]. Potential reasons include a critical shortage of medical facilities and well-trained healthcare professionals across most ASEAN member states, with the existing workforce being overburdened, thus compromising service efficiency [71, 72]. Additionally, the lack of cohesive strategies and multisectoral collaboration may result in fragmented efforts to tackle health issues [70, 73].

In this study, we categorized the initial phase efficiency of health systems among ASEAN member countries into four distinct trends. Indonesia, Thailand, and Malaysia exhibit an upward trajectory in their first-phase efficiency, potentially due to their stable economic growth, which has underpinned continuous investment in health resources. Sustained investments in infrastructure, medical staff, and technology have effectively enhanced the efficiency of medical resource allocation [74]. Vietnam’s remarkable improvement in first-phase efficiency is attributed to the reform and vigorous implementation of various health policies. The Vietnamese government’s increasing investment in universal health insurance, aiming for a coverage rate exceeding 95% by 2025, has been a key driver [75]. Additionally, the establishment of a Public–Private Health System has stimulated private hospitals to participate in service provision, with competition among different hospital categories improving service quality and resource allocation efficiency [76]. In contrast, Myanmar’s continuous decline in first-phase efficiency may stem from its long-standing economic and political instability, with parallel public and civil community health systems leading to a lack of integration, duplicated efforts, and resource wastage [77, 78]. Singapore’s consistent top ranking in sub-phase efficiency is fundamentally due to its stable self-financing health financing system, the “3M” framework, ensuring citizens have sufficient savings to cover increasing healthcare costs while achieving risk-sharing within the national insurance scheme [16]. Cambodia, despite poor first-phase efficiency, leads in the second phase, indicating a successful transformation of health resources into effective health service utilization. This may be attributed to the “Enhancing Quality of Healthcare Activity” project, which integrates the National Quality Enhancement Monitoring (NQEM) and Performance-Based Financing (PBF) strategies. Service providers identify gaps and areas for improvement through external quality assessments and receive financial incentives through PBF, effectively promoting the enhancement of healthcare service quality [79].

The efficiency of healthcare systems varies significantly across countries. The study finds that regions with higher economic levels, such as Singapore and Malaysia, tend to have more efficient healthcare systems at all stages. Conversely, regions with lower economic levels, such as Cambodia and Vietnam, often exhibit lower efficiency in their healthcare systems. Consistent with our research, Neugebauer [80] has identified a correlation between low-income status and reduced healthcare service utilization. Their study posits that economic barriers, including high deductibles and out-of-pocket costs, are significant factors that deter individuals in low-income areas from seeking essential healthcare services [80]. Building on these insights, our findings further demonstrate that developed countries exhibit higher HRAD and HRPD values, with the concentration curve indicating a trend of healthcare resources shifting towards economically developed regions. Consistent with our findings, Le et al. [81] have identified a severe shortage and attrition of healthcare professionals in Vietnam, Cambodia, and the Philippines when compared to other ASEAN nations. This may be due to poor working conditions, job and economic insecurity, and limited career opportunities in lower economic regions, while higher economic regions offer better pay, living conditions, and family security [82, 83].

Additionally, two main patterns of health resource development disparity emerge: (1) The health resource agglomeration degree reveals significant internal imbalances in countries such as the Philippines, Indonesia, and Vietnam, with values markedly above and below the baseline of 1, indicating that certain areas are over-concentrated while others are underserved. Ghosh’s study also highlights the uneven, inequitable, and poorly distributed health workforce in India concerning numbers and types [84]. (2) Scatter plots from a comprehensive analysis of efficiency and equity show that nations like Singapore, Malaysia, and Brunei have efficient health systems with high resource concentration, whereas Cambodia, Laos, and Myanmar exhibit both low system efficiency and resource concentration. These results partially align with the findings of Humphries et al. [83]. The disparities can be explained by the Matthew Effect [85], where nations with efficient systems and high resource concentration, such as Singapore, Malaysia, and Brunei, benefit from a positive feedback loop-efficiency attracts more resources, which further enhances their healthcare systems. In contrast, countries like Cambodia, Laos, and Myanmar, with low efficiency and resource concentration, are trapped in a negative cycle where inefficiency and lack of resources hinder the delivery of effective healthcare services, leading to poorer outcomes.

Conclusions and policy implications

The strength of this study lies in its empirical findings, which provide valuable insights for policymakers to strategically enhance the development of health systems in ASEAN member states. Specifically, the results provide guidance for policymakers in terms of both equity and efficiency:

• (1) While most ASEAN countries, such as Indonesia, Thailand, Malaysia, and Vietnam, have improved the efficiency of health resource allocation, this has not been effectively translated into enhanced health service capabilities. For these nations, ensuring that limited health resources are transformed into quality health services is crucial. Specifically: on one hand, a national health sector steering committee should be established to coordinate between the government, healthcare providers, and local authorities, effectively mitigating the waste of limited resources. Additionally, train and retain healthcare professionals, focusing on skills development in emergency care, maternal health, and non-communicable diseases, thereby fundamentally saving health resources through disease prevention. On the other hand, to actively enhance medical service quality, it is recommended to implement a centralized monitoring system to track resource allocation and service delivery outcomes. Drawing on Myanmar’s experience, establishing a medical service feedback system for NQEM, along with financial incentives like PBF, can help identify gaps in service delivery and motivate providers to improve service quality.

• (2) Significant disparities exist in the development of health systems among ASEAN countries, characterized by a polarization phenomenon. Consequently, it is imperative to promote the balanced development of health resources both within and among ASEAN nations. On one hand, nations such as the Philippines, Indonesia, and Vietnam should establish national health sector steering committees to ensure the balanced development of diverse health resources. These committees will be tasked with formulating national guidelines to standardize the development of health infrastructure, human resources, and service delivery. On the other hand, ASEAN countries should leverage economic integration to foster collaboration in the health sector, bridging the development gap in health systems among nations. Firstly, ASEAN member states should effectively implement the ASEAN Economic Community Blueprint 2025 and strengthen cooperation in the areas of the digital economy, green economy, and blue economy. Secondly, regional integration of health services need to be strengthened, refining Mutual Recognition Arrangements (MRAs) to promote knowledge-sharing and collaboration among health professionals. Lastly, ASEAN nations should devise national health workforce plans that prioritize talent retention through incentives such as housing, higher salaries, and career advancement opportunities to address brain drain.

Limitations

The selection of indicators in this study is based on prior research and group discussions, with some limitations due to data availability. We recommend that future studies focus on countries within ASEAN that exhibit low efficiency and significant disparities in resource allocation, such as Cambodia, Laos, and Myanmar. Since this paper examines the overall efficiency of ASEAN, the selected indicators are mostly macro-level, while the analysis of individual countries’ health efficiency could incorporate more micro-indicators that reflect actual conditions, such as patient satisfaction. Furthermore, this paper analyzes the equity of health resource allocation in ASEAN countries from demographic, geographic, and economic perspectives, without considering the health status and medical needs of different regional populations. Future research may consider incorporating indicators that reflect the medical needs of populations in different regions, such as hospitalization rates, outpatient volumes, and average travel distances to healthcare facilities. This integration of health resource allocation with demographic needs could ensure a more comprehensive evaluation of fairness.

Appendix

An appendix contains supplementary information that is not an essential part of the text itself but which may be helpful in providing a more comprehensive understanding of the research problem or it is information that is too cumbersome to be included in the body of the paper.