A review of appropriate indicators for need-based financial resource allocation in health systems

Background Optimal, need-based, and equitable allocation of financial resources is one of the most important concerns of health systems worldwide. Fulfilling this goal requires considering various criteria when allocating resources. The present study was conducted to identify the need indicators used to allocate health resources in different countries worldwide. Methods A systematic review conducted on all published articles and reports on the need-based allocation of health financial resources in the English language from 1990 to 2020 in databases, including PubMed, Cochrane, and Scopus as well as those in Persian language databases, including magiran, SID, and Google and Google scholar search engines. After performing different stages of screening, appropriate studies were identified and their information were extracted independently by two people, which were then controlled by a third person. The extracted data were finally analyzed by content analysis method using MAXQDA 10 software. Result This search yielded 823 studies, of which 29 were included for the final review. The findings indicated that many need-based resource allocation formulas attempt to deal with health care needs using some weighting methods for individuals. In this regard, the most commonly used indicators were found as follows: age, gender, socio-economic status or deprivation, ethnicity, standardized mortality ratio (SMR), the modified health indicators (disease consequences, self-assessed health, and disability), geographical area / place of residence (geographical) (rural versus urban), cross-boundary flows, cost of services, and donations. Conclusion The indicators used in allocating the health systems’ financial resources in each country should be designed in order to be simple and transparent and in accordance with the moral norms of that society. Moreover, these should be a good representative of the health needs of people in different geographical areas of that country. In addition, their related data should be available to an acceptable extent.


Introduction
Having access to health services with the aim of promoting, maintaining and ensuring the individuals' health is one of the important pillars in the development of any society (1). However, health systems are faced with the shortage of resources and are not able to provide all the services required for all society members on the one hand (2) and seek to increase justice in people's access to health care and services on the other hand. Are (3). Concerning limited nancial resources, world health systems face various challenges in terms of service quality, e ciency, effectiveness and justice; equitable resource allocation is one of the major concerns of any health system and is also a key part of the decision-making process in this regard (4,5). As a moral issue, it plays a critical role in creating justice in health care services and its health consequences (6). Resource allocation is fair when health care resources are distributed among competing consumers (e.g. regions) based on the need for health care (7). Need-based resource allocation is one of the methods which has been considered for equitable allocation of resources in recent decades in most publicly nanced health care systems (8) (9,10). Need-based resource allocation is a process in which nancial resources are distributed among individuals or populations based on the need for health care services (11,12). This method is considered the most ethical and fair mechanism for allocating nancial resources (13). In fact, need-based resource allocation tries to eliminate budget inequalities among the areas of a region or a country (14). It can ensure that government's public resources are distributed to different regions in accordance with the goals of the health system (15).
Since it is di cult to measure health care needs directly in a society, indirect methods such as indicators are used to measure health needs (16). So far, no standard gold indicator has revealed the need for health care services in communities; this has caused challenges in selecting the indicators required to compile need-based resource allocation formulas (17). Therefore, the models and indicators used in this regard are signi cantly diverse (18).
According to the World Health Organization, health and treatment conditions have improved in most countries in recent decades; this improvement has been re ected in indicators such as "life expectancy at birth, life expectancy with balanced disability and decreased mortality resulted from health measures".
However, these improvements have been accompanied by some concerns, the most important of which are inequalities in access to health care services and health consequences among different economic groups and individuals inside and outside countries (19,20). Therefore, concerning numerous challenges that exist due to the shortage of resources to provide equitable health care services, the equitable allocation of available resources is of great importance; This requires changing the current resource allocation approaches to the need-based allocation method. Since creating a suitable indigenous model requires studying the models of other countries in order to identify the principles governing them and the type of indicators used by them, the present study aimed to identify the indicators used for need-based resource allocation in world's health systems through a comprehensive review of studies.

Search strategy
Authoritative English language databases such as "PubMed, Cochrane and Scopus", Persian language databases of "Magiran and SID" and Google scholar search engines were reviewed to nd the studies published between 1990 and 2020. Studying the articles found in these databases and reviewing their references, other related articles were also identi ed and added to the previous ones.
Selection of articles: A total of 823 articles were extracted from the databases. Of them, 641 articles were excluded based on their titles and also due to duplication. The abstracts of the remaining articles (182) were reviewed, and 95 were excluded. The full text of the remaining 87 articles was studied; another 58 articles were excluded due to their non-compliance with the inclusion criteria. nally, 29 articles were entered into the nal phase of the study (Figure 1). Inclusion and exclusion criteria: English or Persian language articles which examined speci c criteria and indicators in the need-based nancial resource allocation were included in the study. Studies whose abstracts were not in English or Persian or whose full texts were not available were excluded. Data Extraction and Analysis: the information in the selected articles was extracted using MAXQDA software (version 10) and a pre-designed form. Data related to each study was extracted independently by two people; some minor disagreements were resolved by the third person, and the extraction was nalized if all three people agreed (Table 1).

Findings
Examining the results of studies conducted on the need-based allocation of nancial resources in the health system of different countries has revealed that different indicators and methods have been used to allocate nancial resources according to the speci c conditions of each country, and that the main purpose of all methods is to promote justice in having access to health services, make use of these services and thus increase the e ciency of services. Table 1 shows the common indicators used in the distribution of nancial resources in the health system of countries.
As the data in the above table shows, different countries use one or a combination of two general "individual-level data and area-level data" methods to allocate resources. The most commonly used indicators are age, gender, socio-economic status or deprivation, ethnicity, standardized mortality ratio (SMR), modi ed health indicators (disease consequences, self-assessed health and disability), geographical area / place of residence (geography) (rural versus urban), cross-boundary ows, cost of services and donations.
Reviewing the existing formulas has also shown that all models based on per capita initially take advantage of the indexes "population size and the age-sex distribution of the population of the mentioned areas". In fact, this index forms the basis for all further calculations.

Discussion
In this study, common indicators used in compiling the formula of allocating nancial resources were examined. Need-based formulas, as a more equitable allocation of health budgets to geographical areas of a health system, are increasingly being used as an alternative to the historical methods (8,23,42,43). To develop a need-based formula, it is important to have a practical de nition of justice (13). In other words, de ning a need-based approach is the rst essential step in selecting the indicators "necessity and need to compile a formula for allocating need-based resources" (44). According to Star eld, justice is de ned as "No difference in access to health services for equal health needs or greater access for the population de ned in terms of social, demographic or geographical status with greater health needs" (45). The theoretical basis of the need-based allocation formula is that the need for health care in populations of equal size is not necessarily equal, and that the population characteristics are the basis for inferring the population's relative needs .
As mentioned in the ndings section, the most common need indicators used to measure the relative need for health care services are: age, gender, socio-economic status or deprivation, ethnicity, standardized mortality ratio (SMR), modi ed health indicators (disease consequences, self-assessed health and disability), geographical area / place of residence (geography) (rural versus urban) and crossboundary ows. However, although indicators such as the cost of services and donations are not considered a need indicator, they are used in the resource allocation formula in some countries.

Population size
Personal characteristics of individuals determine their needs for health care services. Owing to the wide variation of the population size in provinces, the population size in a geographical area is the rst important indicator of the need for health services in the resource allocation formula (16,34,43,49).

Demographic indicators
Population composition in a region or country (especially age and gender) is a key demographic factor in estimating the relative need for health care services in a geographical area; this is because there is a close relationship between age / gender and the need for health care services (16,23,43,(50)(51)(52).
Therefore, demographic composition can have more weights in the resource allocation model compared to other factors (53). Three main age / gender groups of children, women of reproductive age (childbearing) and the elderly people are considered the most vulnerable population groups to diseases and thus need more health care services (34,(54)(55)(56). Thus, the population size in these groups is an important factor affecting the need for health care services and, consequently, health resources in different regions and areas. For example, the Resource Allocation Working Party (RAWP) claims that demographic features affect the need for health care services and weigh the population of each region according to the national use of health services by age and gender groups (57). In the British formula, 18 age groups within health trusts were adjusted through the national use of health services in the trusts (58). In South Africa (8), the age / sex ratio adjusted based on the national use of health services in each group is used as a need factor in the health resource allocation formula. Children under 5 are selected as a demand criterion for child care services; women aged 15-49 are chosen as an indicator of the increased need for health care services experienced by women who are mainly in childbearing age; and people aged 65 and more are recognized as the criterion of the need for the elderly care. According to the data written in table 1, it can be seen that the age/ gender factor has been used in all the studied countries (except for Israel and France which have only used the age factor).

Ethnicity
Ethnicity is often used in terms of race, citizenship and country of birth in both matrix and ecological models (11). In some countries, some ethnic groups do not use health services (e.g. the Maori people in New Zealand and non-Nordic immigrants in Sweden). In New Zealand, it is estimated how much Maori does not use health care services and thus weighting is adjusted accordingly, while ethnicity is not considered in the Swedish model. It is clear that this index can be used in a country where there are ethnic differences between different regions.

Socio-economic status
Socioeconomic status or deprivation is often used as an indirect indicator of the relative need for health care services (13,23,42,46). Since the relationship between socioeconomic status and the need for health care services is not simple and straightforward (59), weights less than one are assigned to socioeconomic factors in a need-based formula. In different countries, various socio-economic indicators are used as indicators of the need to adjust the models for allocating health care resources (for example: income / assets (Netherlands, South Africa, Malawi), homelessness and education (New South Wales), unemployment (Belgium and Stockholm), welfare status (Alberta, Netherlands, New Zealand, Northern Ireland and USA), marital status (Norway, Stockholm), family structure (Norway), quality of housing (Belgium), housing ownership and social class (Stockholm) and cohabitation (Stockholm, Northern Ireland). In South Africa, unemployment, people living in poor housing conditions, lack of access to tap water, poor toilet facilities, lack of access to clean energy sources, illiteracy of the household head and female-headed households have been used in the resource allocation formula in order to compile socioeconomic indicators (42) Namibia has used a deprivation index using household assets including electricity, radio, television, refrigerator and motorcycle, as well as drinking water and toilet type for equitable health resource allocation among the provinces (13).

Population mortality rate
Owing to some features such as having a familiar concept, reliability and ease of data collection, mortality indicator is considered one of the most common indicators selected to indicate the need for health (49,60). Standardized mortality ratio (SMR) and age / sex speci c mortality indicator have been used as indicators of need to know the health resource allocation (17). For example, raw and standardized mortality ratios have also been used as indicators of need in Per capita schemes in Belgium, Italy, Namibia, Northern Ireland, Norway, Scotland Wales and Zimbabwe. Making use of mortality indicators may have some disadvantages because the health system provides health services for the living not for the dead, so resource allocation indicators should be directed to the living as much as possible. Moreover, using mortality statistics such as SMR to allocate resources may not be appropriate in some cases because a signi cant portion of health care services is provided for people whose treatment does not lead to death. Finally, the geographical distribution of health needs may not be in line with the geographical distribution of death in different parts of a country. For example, some people who die in one area may die from diseases they contracted in another area some years ago (61).
Since the relationship between mortality and need is not straightforward, the mortality rate cannot be mechanically considered in the resource allocation formula, as it may lead to unreasonable and unrealistic allocation patterns (62). Hence, weights less than one are applied to this index in the resource allocation formulas (53).

Disease complications in the population
Disease complications directly indicate poor health (ill-health) conditions in the population (12). The prevalence of some chronic diseases such as diabetes, cardiovascular problems and osteoarthritis as well as the occurrence of acute complications such as the gastrointestinal and respiratory injury or infection are examples of appropriate indicators of disease complications to assess the need for health care services (17). Disease complication data have been used, in combination with socioeconomic factors, to allocate health nancial resources in Stockholm, Sweden (38) and the NHS (60). Self-reported health which is people's perception of their health compared to the peers' health is considered an appropriate indicator of disease complications because it is closely related to many other health indicators and is independent of the indicator "health services use" (15,63). However, making use of disease complications is not popular as a need indicator due to its technical problems. For example, data on disease complications may be biased owing to differences in records of institutions and regions (64).
In addition, disease complication indicators may not cover all the health conditions people need to enjoy health care services (16). This may underestimate the need for health care resources in areas which need these resources more. Moreover, there are always limitations in the frequency, timing or availability of disease complication data for the entire populations and regions; in turn, they impose some restrictions on assessing the need for health through disease complications (65).

Geographical factors
Geographical area is usually an indicator used to decide on the allocation of resources in most health systems. Reasons have been given to justify the allocation of resources based on geographical areas. For example, making use of the geographical area-based resource allocation approach, differences in the cost of providing health services in different regions can be offset by appropriate reimbursements (64).
The United Kingdom, Scotland or Ontario Canada have considered these differences in their formulas. The resource allocation based on the geographical area has the potential to include both the justice goals and the goals of the e ciency of health systems (66). Allocating a larger share of the health budget to geographical areas that need more health care services can increase the e ciency in the use of health services (11). Fair distribution of health credits among geographical areas can also improve the previous inappropriate distribution of health facilities and trusts in the regions (67,68).

Place of residence
Place of residence (province, cities, towns, urban-rural areas and slums) can affect health and the chance of improving living conditions (69,70). Geographical differences are regarded as an ethical concern in terms of having access to health care services and health outcomes (10). Therefore, geographical classi cation is considered an important tool for health promotion and proper distribution of health resources among regions (71). Living in urban or rural areas affects the people's health in different ways (72). Urban life provides citizens with many opportunities and excellent and better living standards. However, urban environments can increase health risks and reshape population health issues from infectious diseases and malnutrition to non-communicable diseases, violence and injuries and deaths from accidents and the effects of environmental disasters (73).

Cross-boundary ow
Cross-boundary ow is where patients may cross health care boundaries to access neighboring health services because the required services are not available at their place of residence or there is an unreasonable delay in obtaining care services (32 , 74, 75). Cross-border use of health services is often enjoyed by temporary guests who include people who use the facilities provided in the border regions, people who seek treatment in other cities or abroad, and people who are sent to other cities or abroad by their health sponsors. (74,76). Cross-boundary ows are considered an element in some resource allocation formulas (including Alberta, Canada, New South Wales and Spain). However, in many cases there is a lack of information about cross-boundary ows, especially in developing countries. This places limitations on the inclusion of "cross-border use of health services" in need-based resource allocation formulas (77).

Costs of providing health services
Costs of providing similar services can vary greatly from region to region (77). For example, costs can be much higher in remote rural areas due to higher transportation costs and perks given to employees in order to encourage them to travel to these regions. Moreover, owing to a tiny number of people in a region or country with very low population density, the cost may be wasted (78). In addition, due to different input costs, service costs may vary among buyers. (79). These factors implicitly indicate the need to adjust a need-based resource allocation formula based on differences in service provision costs resulted from the impact of geographical factors. However, appropriate data must be provided to include various costs in the formula. Additionally, decisions made about adjusting different service costs are often a political issue (11). Alberta, the United Kingdom, Ontario, Scotland, the United States and Wales are examples of countries which have used this indicator in their formulas.

Donations by donors
Alternative nancial resources provided by donors and NGOs especially for low-income countries is another indicator used in compiling a need-based allocation formula. The challenge this indicator poses for health policymakers and planners is whether the government should allocate fewer resources in areas with higher donations (41) or not. Uganda is an example of countries that uses this index to allocate health care budgets according to the following weighting: 60% based on the population size index in different age groups, 20% based on human development index (per capita income, life expectancy and school enrollment rates) and 20% based on donations and NGO expenditures in each region (11).
As mentioned above, the geographical area is usually the most common decision-making factor for resource allocation in most health systems; thus, it is the basis of the need-based allocation formula because geographical conditions can affect health and therefore the use of health care services (10,63). In most of these formulas, the weighted capitation is used to estimate the relative need for health care services in each geographical area. Concerning the main indicators of population composition (age, gender) and especially age (because the gender distribution is usually very similar in different regions) (11), socio-economic factors (education or occupation, income, wealth, marital status and employment status,…) and geographical factors, this approach ensures more equitable distribution of resources among geographical areas in accordance with the principle "equal access to health care services for all people with the equal need" (10, 11, 63 , 80) (8,11,43). Age, with higher weight for newborns and population over 75, is the most prominent factor used to pay per capita in high-income countries, while socioeconomic factors as well as factors associated with disease complications are considered a less important criterion except for psychiatric and society-based care services. In low-and middle-income countries, however, the population under 5, poverty indicators and rural population have the highest frequency and are of great importance in the development of need-based nancing formulas (11). The essential point about the possibility of using these indicators is that they must meet the requirements so that they can be used as indicators of need in developing resource allocation models. Seven main criteria which the "need indicators" must meet are: universally recorded, veri able, consistent, no incentive for gaming, no vulnerability to manipulation, con dentiality respected and plausible (10,16,64).

Conclusion
There are various methods for need-based resource allocation; they can vary from simple indicators such as population size and composition mainly used in developing countries to complex models used in developed countries. Each of these methods is designed according to the conditions of those countries.
Access to data and the possibility of calculating the index for the given region, values and ethical criteria and cultural, economic and social conditions are among the most important factors which are considered in the allocation of resources. According to the ndings of the present study, an appropriate combination of demographic indicators, mortality indicators, socio-economic indicators and geographical location seems to be effective in developing a need-based allocation formula and thus improving justice in the distribution of nancial resources. Although the allocation of nancial resources in health systems seems economic in nature, ethical standards of society must be taken into consideration for fair allocation of resources. Therefore, the most appropriate method of need-based resource allocation in the health system in each country is to design and choose a method which both is simple and transparent and uses indicators that meet the moral norms of that community and be a good representative for people's health needs in different geographical areas of that country. Moreover, the information about the characteristics of that model must also be available to a great extent.

Declarations
Ethics approval and consent to participate: Not applicable Consent for publication: Not applicable Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: Not applicable Authors' contributions: MR, FE and RM conceived the study and designed its method. MR performed the computations and applied the model, with help from AA for revision of the analytical method. All authors discussed the results and contributed to the nal manuscript. MR BN and AA carried out the analytical experiment. MR and BN wrote the manuscript All authors contributed to the development and approved the nal manuscript. MR is the guarantor.