|Home | About | Journals | Submit | Contact Us | Français|
The Government of Kazakhstan is engaged in a “root and branch” modernisation of the health care sector. One aspect of the raft of modernisation programmes was to revisit the State Guaranteed Health Benefits Package, with the aim to review citizen entitlements to healthcare. This paper reviews the ongoing evolution of the planning of the health benefits package in Kazakhstan, with the main challenges encountered, and critical lessons learned, to be considered for similar attempts elsewhere. The main conclusions are that: the design process requires a blend of technical and socio‐political analysis, because it attracts public interest, and therefore political risks; the scale and burden of analysis need to be kept to manageable proportions; and the relationship between the benefits package and funding modalities needs to be carefully managed by the State, to ensure access to declared entitlements to all members, including the most vulnerable, while keeping the package financially feasible. © 2017 The Authors. International Journal of Health Planning and Management published by John Wiley & Sons, Ltd.
Many countries' Governments acknowledge that achieving better health is a fundamental human right and that it is their responsibility to ensure citizens have access to health services that are equitable, accessible and affordable (WHO, 2013). However, the notion of a State guaranteed health benefit package (SGHBP) has been hard to specify, standardise and implement.
In almost every country of the world, every legally available health service is available to all of the population (OECD, 2012). What distinguishes a SGHBP is what is guaranteed by the State to be available for all or some of the population, and how it is paid for. For example, elective Aesthetic Surgery is available to all citizens of most countries, but in no country is its provision a part of an SGHBP.
Many countries assumed that a SGBHBP could be developed on purely medical grounds by determining a common level of medical need for all people, and then extrapolating from there (Callahan, 1992). However, it has proven impossible to use “medical need” (the illness model) as the single meaningful criterion for a health care package (the wellness model). Instead, the package must reflect an array of ingredients—medical, ethical, social, economic and political. The political dimension, although not often explicit, is needed to elaborate the ideals of the package or at least articulate the values underpinning what is accepted as an adequate service coverage (volume and type of care per specified target groups) and quality in a particular setting.
A generic pattern of establishing and shaping the SGHBP is found in many countries. This consists of two levels. At the higher level, legislation passed by parliament (sometimes captured in the constitution, e.g. Poland (Sagan et al., 2011)), establishes the general framework by enumerating the areas of healthcare included in the benefit package. This approach is often similar to the OECD health care categories (OECD, 2011). At the lower level, the benefit package is further shaped into benefit catalogues by specifying certain services and procedures to be provided and paid for by the State within each area of health care (Sojo, 2006). These catalogues contain recommendations as well as explicit inclusions/exclusions of services. The way in which this shaping actually takes place varies considerably from country to country and within different areas of health care. In most cases iterventions are not arbitrarily assigned to the package; rather, they result from a prioritisation process to achieve specific technical and social objectives.
Such modeling of benefits packages has important implications on access, equity and utilization of health services. Therefore determining the most appropriate and effective methodologies for designing and implementing SGHBPs in different contexts is of critical importance.
In many countries the SGHBPs matured organically over time; others have developed benefits catalogues through different approaches. In general, low‐income countries, with modest SGHBPs, have specified it according to a list of included services, and high income countries, with comprehensive packages, through a list of exclusions. However, more recently, some countries are experimenting with innovative, so called “hybrid” approaches to formulating an SGHBP, which suggests combination of inclusion/exclusion approaches, and mixed classification of services by combining specific service lists with the broader service categories.
The rest of this article discusses the issues and implications of implementing the hybrid approach in selected settings, so that other countries and health systems can learn from our experience. It presents in depths the example of Kazakhstan, where the application of a hybrid approach encountered some challenges; and briefly describes the successful cases of Indonesia and Georgia, to show how the context implies on the validity, relevance and success of a suggested method.
The Republic of Kazakhstan has been rapidly modernising its health care system, in terms of both demand and supply‐side reform. The reforms have included: the introduction of more sensitive reimbursement mechanisms such as Diagnostic Related Groups‐based reimbursement and Payment by Results; initiating results based budgeting, enabling more agile and flexible service provision through hospital autonomy; and investing in better information systems and primary care.
Since 2010 the Ministry of Health (MoH) has pursued a Concept on the Unified National Health Care System, pooling funding for hospital services at the national level, and connecting payment to an individual patient, rather than a health care provider. As reflected in national development strategies, Kazakhstan is seeking improvements in health status to be consistent with those found in OECD countries (Government of Kazakhstan (GoK), 2012). These reforms are not fully implemented yet.
In 2010 the GoK decided to review the SGHBP, with the aim to produce a package consistent with OECD countries, and also, to use the SGHBP as a tool to achieve the health system development objectives, namely: to increasing access to high quality services, prioritizing and strengthening PHC, rationalizing and strengthening referrals, and stimulating development of day care and long term care, while simultaneously rationalizing the costs of services to control the public burden of costs for the SGBHBP. These objectives were aligned with the concept of Universal Health Coverage, which has been adopted by the WHO (2010), and later by the WHA, the World Bank and the UN (Campbell et al., 2013).
The prevailing SGHBP was captured in legislation in 2009 (Kazakhstan, Republic of. 2009, 2009) and revisited in 2014. The package was considerably comprehensive and far‐reaching, including inpatient, outpatient and long‐term care. The services were provided mainly through a State‐owned provider network and some private providers. Essentially, every citizen was eligible to access the majority of legally available services, and the package specified what services (or service categories) were fully or partly State‐financed.
In regulatory documents the SGHBP was defined in a non‐specific manner. The MoH Decree on the SGBP (2009) listed wide categories of outpatient services, e.g.: “emergency medical care,” “sanitary aviation for transporting patients from remote areas” “specialist consultations,” “laboratory and instrumental examination,” etc. This created open‐ended commitments, and complicated estimation of service costs, as well as budgeting. The pharmaceutical component of the SGHBP was characterized by inconsistency in drug costs, with some drugs to be procured at a cost significantly higher than in high‐income countries.
Over the years, the package had become disjointed, reflecting inclusions which had been introduced to respond to transient or immediate needs, and some others to sustain services/infrastructure which had become obsolete (e.g. TB dispensaries inherited from the Soviet system). These inclusions were financed through diverse “vertical” programmes, creating perverse incentives around integrated care and referral patterns, and encouraging duplications in diagnostics and treatment at various levels of service delivery.
Furthermore, ambiguity regarding the entitlements, referral and reimbursement, combined with low remuneration of medical personnel, encouraged out‐of‐pocket payments (formal and informal).
Thus, the GoK had sufficient reasons for reconsidering the package. The desired aims of the revision of the SGHBP, expressed by MoH, is shown in Table 1.
Prior to moving to the revision of the SGHBP, it was deemed necessary to understand what type of package Kazakhstan wanted to achieve. The Ministry of Health (MoH) set up a working group to work with development partners and national counterparts on this issue.
The classification by RAND Corporation (Cahill et al., 1997), which reflects wider social policy positions for a health benefit package, was used as a reference. It identifies five possible models:
Given the history of social protection and the aforementioned universal health goals set out for the SGHBP, a Comprehensive Benefits Package was considered the favored option for Kazakhstan. This decision enabled the MoH to move towards objectives 1, 2, 3 and 7, outlined above.
The Ministry recognized that the specification of the SGHBP had to be partly technical (i.e. reflecting the prevailing and anticipated burden of disease, issues of capacity etc.), but was also a product of socio‐political considerations including: the conviction that SGHBP inclusions/exclusions are the preserve of political decisions, conformity to multi‐national agreements (e.g. WHO policy (Campbell et al., 2013)), the desire to preserve and enhance social solidarity, the affordability of public health expenditure within the prevailing macro‐economic conditions and the relationship between government and non‐government health expenditure.
Once the decision to pursue a comprehensive benefits package was made, subsequent steps included specifying:
The MoH acknolwdged international practices showing that the most frequently used criteria for definig SGHBP include accessibility (40.7%), cost‐effectiveness (29.6%), efficacy and cost (22.2%) (Kabir et al., 2013). At first the emphases were given to deciding whether the package should be inclusion or exclusion based, and the methodology for categorization of services.
Initially, the GoK's preference was to define a SGHBP catalogue by a list of inclusions, as a generic extension of the existing approach. The MoH favored the World Bank inclusion algorithm (Musgrove, 2004) be used as the basis for planning because this would align with the GoK's desire to establish consistency with international standards and conventions, assuming that the process would be modified to reflect the specificities in Kazakhstan.
This, now well‐established algorithm, was presented by Musgrove in 2004 (Musgrove, 2004) on behalf of the World Bank (Figure 1). It was designed to expose potential candidate inclusions to an analytical process which confirms if services should be guaranteed by the State. The low burden of analysis required by the Musgrove algorithm and focus on high impact health gain means that it is a very useful tool for low‐income countries and international donors (e.g. IDA, and the Global Fund) attempting to determine an absolute minimum level of care. Furthermore, because the main interventions for the highest impact conditions remain relatively static in Low Income Countries, re‐analysis of the services to be included in the SGHBP is not regularly required.
However, there were some weaknesses with the Musgrove algorithm's practical application in Kazakhstan. The algorithm was originally presented to help countries develop a purely inclusion‐based policy for a small‐scale benefit packages. For the comprehensive SGHBP in Kazakhstan, exposing every possible service to this rigorous analytical process would be an unfeasibly large task, especially given the number of expected inclusions and the technical analytical resource available. It would also require frequent re‐analysis as medical treatment evolves and epidemiological transition is observed. Additionally, the necessary data were not readily available.
In recognition of these weaknesses, the “selective” use of this algorithm was adapted. Namely, given that the Kazakhstan SGHBP would be a fairly comprehensive inclusion package, the default position for most services would therefore be inclusion. Table 2 shows examples of some of the implicit inclusions in the SGHBP. This meant that the World Bank algorithm would only be used to determine inclusion/exclusion for services where inclusion was ambiguous. Ambiguity may occur where there is a lack of evidence for the service or inclusion would be contestable or contentious. Therefore, only a small subset of services would need to be exposed to the analytical regime implied by the WB algorithm (Musgrove, 2004).
To facilitate this process, detailed technical guidance on how to conduct the analysis was prepared for each of the elements in the algorithm, particularly around the calculation of Disability‐adjusted Life Years averted as a measure of effectiveness.
As the technical exercise unfolded, it became evident that selected services would need to be excluded from the existing SGHBP. An important determinant for decision‐making became whether the Ministry was prepared to inform the public that a service would no longer be financed by the State as part of the SGHBP. Thus, exclusion was relatively straight‐forward in cases where the service was not part of the prevailing SGHBP, and far more difficult if it needed to be removed from the current entitlements. Overall, the process became time‐ consuming and full of controversy, which confounded the decision‐making.
The next step was to decide on service categorization. Assembling services by the disease groups seemed a simple exercise, however, it was considered less useful, reducing possibility to associate costs per specific service. Categorization by population group was considered too broad and “data‐heavy” because it was questionable as to whether the MoH would be able to generate evidence to rigorously complete the exercise. Grouping services by International Classification of Diseases Group (WHO, 2016) coding seemed an appealing option, but was considered too complex because it potentially required the explicit inclusion or exclusion of up to 68 000 codes as well as co‐morbidities.
The concurrent introductions of new purchasing mechanisms, such as Diagnostic Related Groups (DRGs), and payment‐by‐results, supported thinking that classification by specific services could be more appropriate. This meant that most of the inpatient services could be grouped according to the DRG catalogue, and outpatient services listed by ICPC2 classification, and these specific list could have been defined by the hybrid methodology. In this case, the MoH would only need to update the outpatient component of the SGBHBP to introduce categorization by ICPC2 instead of the broad categories defined in regulations, as the DRGs were already in place. It would allow the MoH to have full clarity on service entitlements and related costs.
Although the above proposal originated from discussions among the MoH working group members, it was rejected after a lengthy consultation period, because the MoH was not prepared to spend resources on introduction of the ICPC2 and respective service cost estimations.
While the decision on categorization of services was on hold, the MoH pursued with the next step to determining the costs of the SGHBP, to match it with the available budget.
The approach to determining what services would be State‐financed or partly State–financed was developed from a concept originally used to classify the clinical commodities to be stocked in different health institutions (WHO, 2003), and further developed by Jones and Chanturidze (2013) to apply to health services more generally (Table 3). This approach focuses on determining the “criticality” of services by distinguishing between “vital;”, “essential”, “necessary” and “desirable” sub‐categories.
Implicit in this classification is that “vital” services are included in the SGHBP and they should be fully financed from State budget sources. State or SHI sources would also finance “essential” services, with only exceptional contributions (in the form of co‐payments) from voluntary health insurance (VHI) or out‐of‐pocket (OoP) sources. The default position for “necessary” services is that they would be expected to be financed through SHI or VHI, perhaps including an element of co‐payment. “Desirable” services would be financed by VHI or OoP sources. The main exceptions suggested to this classification was when economic hardship would result from co‐payment, or where individuals belong to priority population groups (15 categories specified in the Kazakhstan regulations). In this case their care would be mostly state financed. Table 3 summarizes the payment regimen by criticality of services. This classification was adopted by the Government.
The outcome of this process was a catalogue of services with clear definitions for how each service should be paid for, either from the State budget, SHI, VHI or OoP. This was based on the need for the service (Vital, Essential, Necessary or Desirable), cost‐efectiveness analyses, the key beneficiaries of the service and any beneficiaries exempt from payment, and where the service should be provided. Tables 4 and 5 illustrates the summary of the catalogue.
As mentioned above, the main reasons for the government of Kazakhstan to revise the SGHBP were to rationalize and more clearly define the SGHBP, but also to ensure that a fairly comprehensive SGHBP was accessible to all citizens yet financially viable for the government. Whilst there was a consensus around applying the suggested hybrid approach for revision, the complexity arose at implementation stage, due to lack of consensus on conceptual and procedural aspects.Subsequently, after the technical exercise was completed, the Ministry requested to switch back to a wholly inclusion‐based package, with modest improvements in the SGHBP catalogue, generated through the application of a hybrid methodology.
The blockages that led to the government stalling the full adoption of the revised SGHBP can be divided into four main categories related to the design process; political sensitivity of the revision, budgeting/financial issues; and service distribution and availability.
As described above, the design blockages included disagreement regarding the format of the package, questioning whether to introduce inclusion or exclusion‐based setup, and to construct the SGHBP catalogue either by service groups or service categories.
Quite legitimately, the Ministry felt it necessary to have an explicit list of services to link budgeting process with the costs per service. The application of a hybrid model will lead to substitution of categorical classification of outpatient services by explicit service lists, thus stimulating cost estimation.
The issues were not resolved partially because of complexity and scale of the technical exercise, and more though, inability to align different interest groups and overcome stakeholder resistance.
As a result, various components of the SGHBP remained to be designed differently, with outpatient services component being based on the broad service categories, leaving vast flexibility with service providers to decide which services to be provided to beneficiaries, and restricting estimation of costs per service. Advantageously, components II and III on hospital services and pharmaceuticals were developed by specific service /pharmaceutical lists, complying with the MoH vision to have full clarity on the entitlements and service costs. Hence, all three components were refined (to a different extend) based on the WB /Musgrove algorithm.
One of the factors reducing incentive to embark on a new model was the comfort with the existing system, and the resistance from outpatient health service providers to accept changes.
Another factor was related to political sensitivity of declaring changes. With the constitution of Kazakhstan entitling citizens to get free health care, the MoH hesitated to explicitly declare the benefit restrictions in an existing SGHBP, or to specify the services that would be left out of the SGHBP for being covered through the OPPs or VHI. This obstructed formation of a financially feasible SGHBP, leaving the Ministry with obligations in excess of available budget. It also left citizens with lack of clarity on services that were not “free” or were requiring co‐payments, thus simulating OPPs and illegal payments.
Another important barrier was recognizing the challenges which would be linked with assuring accessibility to the declared entitlements (not defining the entitlements per se). On one hand, the MoH wanted to increase access to public health, PHC, long term care, and day care. At the same time, it was clear that the SGHBP will not be able to fully address the populations' health need, while being financed from a single source ‐ the State budget, without significant co‐payments, and diversification of financing sources. In fact, scale up initiatives were modestly enacted for selected priority services only. This suggested that the Government should have prioritized broader health financing reforms, and consequently declare an extension of the SGHBP.
Furthermore, the MoH started to realise that a good dsign of the SGHBP was only half of solving service accessibility and utilization issues. Ministry feared that even for health services that were theoretically guaranteed by the SGHBP, personal cost, including informal payments, would still be a deterrent to seeking care for a considerable percentage of the population. Indeed, in one study in Kazakhstan, 44.3% of women respondents said that they could not afford care, and 34.9% could afford care only sometimes (Katsaga et al., 2012), while having these services formally included in the SGHBP. This suggests that an enforcement of the delivery of services was needed to improve health access.
It has been a general consensus on the fact, that the rural parts of Kazakhstan have worse access to health services than urban settings. Furthermore, rural citizens are generally poorer than their urban counterparts, and experience a greater prevalence of disabling conditions. Indeed, 4.9% of the rural population were considered to be below the national poverty line compared to 1.3% of the urban population (World Bank, 2016). Whilst these data are no longer current (2013), they do indicate a mismatch between needs and access to SGHBP services. Therefore a key objective for the SGHBP must be to minimise exposure to catastrophic health expenditure, mitigate against any deterrence to seeking care and maximise the ability of patients to access care. Mechanisms to promote these aims include:
Even if, in theory, service facilities are available to provide access to SGHBP services, it seems that the availability of patient services is sometimes limited. Of those institutions which were exposed to an accreditation process, only a very small proportion achieved the standards typical of OECD countries. The reasons cited included: the availability of skilled personnel, pharmaceutical stocks and dysfunctional equipment. As the planned share of GDP devoted to SGHBP increases, a proportion of that growth should be applied to creating sustainable stocks of pharmaceuticals, rehabilitating infrastructure and modernising technical equipment, assuming that the process is correctly planned and oversighted by the Government.
These factors appeared strong barriers to hamper realization of a package revised through a time and resource consuming exercise, with the application of a hybrid model.
The issues with the SGHBP revision in Kazakhstan were unexpected, as the exercise was driven by the evidence that a hybrid classification method had proved successful in other contexts, such as Indonesia (AIPHSS, 2014) and Georgia (Oxford Policy Management, 2009), (Zoidze et al., 2013), both of which had introduced packages which were funded through a mix of Government and SHI sources, with exclusions financed through VHI or OoP.
In Indonesia, the revision of the benefits package was allied to the extension of the existing benefits package (previously provided only to specific groups) towards Universal Health Coverage by 2018. Therefore the SGHBP had to be redefined so that it specified an intermediately comprehensive package for all citizens which was considered to be Universal Health Coverage. Indonesia is a complex country with around 8000 populated islands, a range of ethnicities, faiths, languages, an increasing degree of economic inequity (GINI coefficient currently 38.1 and increasing (World Bank, 2015)) and a unitary model of government but with complex representative sub‐national political institutions. This meant that although the benefits package had to be universal, it also had to be responsive to locally specific needs. Within these contexts, the hybrid approach was strongly indicated, allowing to address local specificities and also moderating the scale of technical exercise and related resources. Table 6 shows the key service inclusions and exclusions developed through the hybrid approach in Indonesia (AIPHSS, 2014).
Similarly, in Georgia the application of a hybrid approach to developing a State Benefit package for Primary Health care proved of be successful, mainly because of three reasons: First, the process was made relatively simple and logical because the country started by assuming a relatively comprehensive benefits package, but then excluded selected categories of services from the state guaranteed list. This meant that incontestable service inclusions did not require analysis, and exclusion decisions were made manageable because they were done at categorical, not individual service levels. Second, the package was widely accepted by a critical mass of health care providers because they were engaged with the process of determining the package content, and the Government demonstrated high political will to lead and finalize the technical exercise. Third, the package was made politically favorable by balancing the cost to the government with variable co‐payment mechanisms and ensuring that the package reflected societal perspectives.
Design of the SGHBPs in middle‐income countries and emerging economies present particular chalenge arising from the tension between the wish to devise an inclusion‐based package and the practical aspects of implementing a massive analytical process which will need regular revisiting. The hybrid approach presented in this article represents a good compromise in selected situations, where the SGHBP is not comprehensive and is structured based on wider service categories, and where the population is universaly eligible for services rendered under the SGHBP.
However, where more detailed service‐level inclusions need to be considered, thehybrid approach may be less acceptable. Such circumstances may arise from a need to integrate with other more tightly defined mechanisms, an organisational culture uncomfortable with less technocratic justifications, or the need to more closely justify eligibilities for a few target populations. In such circumstances the acceptability of a hybrid model needs to be negotiated carefully, referring to the fact that the approach is feasible and flexible enough to be able to meet the majority of key policy goals, with much fewer resources and time being required to design it, but comes with it's trade‐offs.
Stakeholder engagement in the negotiation of the poilituical, social, ethical and financial consirderations is important for the acceptability of the final SGHBP. Without such acceptance, a technically well‐designed SGHBP may never be implemented. Similarly, prudent estimation of funding for specified entitlements is vital to assure feasibility of declared benefits.
National governments should demonstrate strong political will to adhere to the agreed design and implementation pathways, averting blockages and allowing finalization of the process, and dedicating substantial resources to design, planning and re‐analysis of the SGHBPs. Defining the package is a time‐consuming, data hungry and a technically, politically and legally demanding process, which should not be underestimated.
The authors have no competing interests.
The work was supported by the Ministry of Health and Social Development (MHSD) of the Republic of Kazakhstan, under the project Development and Institutional Sustainability of Health Policy Analysis, Provider Payment System and Strengthening Single Payer Capacity; Contract No. KHSTTIRP‐A1/CS‐02.
Jones M., Chanturidze T., Franzen S., Manu A., and Naylor M. (2017) Specifying a State Guaranteed Health Benefits package for Kazakhstan: lessons for emerging economies and middle‐income countries. Int J Health Plann Mgmt, 32: 540–553. doi: 10.1002/hpm.2359.