HEALTH MICROINSURANCE

Transcription

1 HEALTH MICROINSURANCE A REVIEW OF THE LANDSCAPE & DETERMINANTS OF PERFORMANCE July 2016 Report for the social enterprise Dreamlopments LTD. Author: Dr. Carole Déglise i

2 TABLE OF CONTENT EXECUTIVE SUMMARY... iii ABBREVIATIONS... viii 1. INTRODUCTION METHOD & DEFINITIONS THE HEALTH MICROINSURANCE LANDSCAPE The Microinsurance Sector in Asia The Outreach of Health Microinsurances Migrants and Health Protection in Asia DETERMINANTS OF PERFORMANCE Understanding needs and demand Socio-economic characteristics Risk prioritization and coping strategies Understanding of the insurance concept Trust Designing a valuable product Product Access Cost Experience with the scheme Contracting quality healthcare Type of healthcare providers Quality of care Provider payment mechanisms Achieving institutional efficiency and sustainability Administration and management Risk management Institutional arrangements Environment KEY CONSIDERATIONS for the Thai-Myanmar border CONCLUSION ii

3 EXECUTIVE SUMMARY Health microinsurances (HMI) have the potential to improve access to healthcare for vulnerable populations and to reduce health-related catastrophic expenditure. The HMI sector has been expanding over the past 10 years and has evolved from simple to complex products, and towards integration of technology and multi-partnership involving more the private sector. Yet, many HMIs still face challenges to reach sufficient scale and become sustainable on the long term. This independent review was commissioned by the social enterprise Dreamlopments as part of a feasibility assessment to develop a non-profit health microinsurance targeting migrants and underserved populations along the Thailand-Myanmar border. It aims at summarizing the scope and the determinants of performance of HMI and at highlighting key considerations for the targeted context. DETERMINANTS OF PERFORMANCE Understanding needs and demand Although health coverage is recognized as a priority for low-income households, this perceived need does not always translate into demand for HMI. Many factors influence the decision to enroll in HMI. The main identified barriers include a low level of income and liquidity constraints, poor understanding in the insurance concept and lack of trust in the product or in the insurer. Some factors may facilitate uptake, such as a strong social cohesion in the community (eg. participation to community association) and the use of informal risk sharing mechanisms, when communities mutually helps each other to cope with unexpected health-related expenses. Designing the product Demand is strongly linked to the value of the HMI product as perceived by the targeted community. The PACE client value assessment tool from International Labor Office was used in this report as a framework to describe 4 key dimensions of product design: Product. HMI struggle to provide comprehensive benefit package at a cost that is affordable for the population and sustainable for the HMI. The majority of schemes focus on hospitalization products in order to simplify processes, while only a minority offer comprehensive benefits. Yet, it is nowadays recognized that HMIs should find ways to offer more comprehensive coverage by including prevention services and out-patient iii

4 consultations in their core packages, or as value-added services (VAS). Access. Factors limiting enrollment are numerous, notably complex enrollment processes and distance to the HMI points of service. The approaches used to address these limitations usually aim at improving communication and customer care, reaching out to communities, and leveraging the infrastructure and services used by the target group. Innovative distribution mechanisms are increasingly promoted, such as establishing cashless and paperless enrollment and premium payment bundled with financial or other services, or facilitated by mobile phone. Cost. Defining adequate premium is a challenge as data on disease burden and cost of health services may be lacking, driving many HMIs to adjust premium as more information on actual cost and market penetration becomes available. Balancing cost and coverage is another issue and HMI use different approaches to keep the premium to a level that their members can afford. They can involve limiting the coverage breadth (exclusion of some categories of population), depth (level and type of health services covered) or height (proportion of the cost covered). Many HMIs use subsidies to increase insurance coverage, especially for poor households or those at greater health risk. Other indirect costs have been considered by some HMI reimbursing travel and opportunity costs. Experience. As products and distribution models become more complex, good customer care strategies that educate on essential HMI features, provide support when needed, assess client satisfaction and establish grievance mechanisms are important. Settling claims is particularly critical, as it is when the HMI becomes real for its members. This process is often supported by peers, community-based volunteers or HMI agents, directly or via a hotline. Technology has helped to verify eligibility using smart cards, and to reduce the claims processing time. Finally, some HMI establish a third-party claim payment system with selected healthcare providers, so that their members do not need to pay for the health service and seek reimbursement from the HMI afterwards. Contracting quality healthcare providers Some of the most successful schemes have progressively adopted a more active purchasing role, through an active search for the best providers, the best health services to purchase, and the best contracting arrangements. HMI have applied a mix of various provider payment methods, such as fee-for-service, capitation or case-based payment. Each method gives different incentives and disincentives for providers to control the cost and quality of services. Finding the optimal payment system that incentivizes performance and quality alike remains iv

5 challenging but essential. HMI also implement quality assurance and monitoring mechanisms, through accreditation system, standard protocols, capacity building and presence of medical advisors within the contracted health facilities. Achieving institutional efficiency and sustainability Maintenance of long-term stability of the HMI scheme or organization involves effective financial management, permanent risk mitigation strategies and engagement of several partners, including the public sector. But management capacity may be weak in some schemes, in particular community-based models. Building capacity or partnering with a third party administrator has helped to improve efficiency in some HMI. Solid computerized system has also become essential to improve data collection, verify eligibility and facilitate near-real time analysis of enrollments and claims, and thus control moral hazard, fraud and adverse selection. Many HMI try to adopt a community-based approach, while developing partner-agent arrangements including public-private partnerships and alternative distribution channels as key drivers to improve efficiency and scale. Public-private partnerships can build upon strengths of each partner to serve a larger effort to achieve Universal Health Coverage (UHC), but challenges to streamline various financing options in the broader perspective of UHC remain important. KEY CONSIDERATIONS FOR THE THAILAND-MYANMAR BORDER Addressing health needs of migrants and border populations has been complex along the Thailand-Myanmar border. The health system is very different on both sides of the border, notably in term of availability and readiness of the health services. As part of the efforts engaged by Dreamlopments to design a HMI for migrants and other communities living along the Thai-Myanmar border, it will be essential to translate lessons-learnt from other settings into strategies that are demand-driven and highly context-specific. Assess demand Surveying the target population is critical to inform product design and develop effective distribution strategies that address identified barriers. The main factors to explore at a preliminary stage are the socio-economic characteristics of the target population, the priority risks they face and their consequences, the existing coping strategies and their effectiveness, the access to and utilization of health services, and the preference regarding product design and distribution mechanisms. v

6 Design the HMI product Important lessons-learnt from other settings that are particularly relevant for the Thailand- Myanmar context are summarized as follows: Developing simple products and processes will be essential to ensure that the scheme is understood and valued by the target community. This will be also particularly important for a service that aims to target hard-to-reach mobile and cross-border populations, as it needs to be easy to distribute and monitor. Establishing demand driven process to guide the product design, with strong involvement of the target group and of existing community-based organizations will be crucial to build trust in the HMI. This will also help to understand the complex cross-border coping mechanisms already in place and develop services that fills gaps and/or complements them. Reaching out to the migrant and cross-border population will require the optimization of distribution mechanisms by leveraging network and specific services already used by the target population. It will be essential to explore innovative distribution channels, such as using mobile technology to enable on-spot paperless and cashless enrollment and payment. Overcoming low capacity to pay with well-designed financing strategies should be put in place to address the liquidity constraints reported by the target groups, which often has an irregular income flow, and to set an affordable premium. Develop an efficient and integrated system Strong strategies need to be developed at the beginning to ensure efficiency, mitigate the risks of adverse selection, moral hazards and fraud, and adapt the scheme to the broader and evolving health protection mechanisms in both countries: Integrating technology across insurance functions, using mobile phone and computerized information management system should be established to strengthen distribution and communication with members, implement VAS, and enable near-real time monitoring and adjustments. Promoting group enrollment may contribute to improve efficiency by reducing administrative cost and by enhancing social cohesion for support and control. vi

7 Contracting healthcare providers shall include solid quality assurance and monitoring mechanisms through accreditation, implementation of standards of care and presence of medical advisors in the contracted health facilities. Provider-payment mechanisms need to be negotiated considering efficiency and quality alike. Establishing a flexible and iterative process based on a solid monitoring and evaluation system is necessary since the pilot implementation to early identify possible deficiencies, and refine processes before scale-up. Exploring public-private partnership will be important to ensure that the scheme is well integrated and adapted to the UHC approach and evolution in both countries. Other forms of partnership may be explored to improve efficiency. Analyzing legal and regulatory requirements in both countries will be particularly important to develop a transnational mechanism. vii

8 ABBREVIATIONS ARY CBO CHAT CHI CMHIS EHO GK HMI IEC ILO IP M&E MoH MW NGOs OOP OP PPP RSBY SKY SMS SSS TPA TPP UHC VAS WHO WTP Arogya Raksha Yojana Community-based organization Choosing health plan together Community health insurances Compulsory migrant health insurance scheme Ethnic health organization Grameen Kalyan Health microinsurance Information, education and communication International labor organization Inpatient Monitoring and evaluation Ministry of health Migrant workers Non-governmental organization Out-of-pocket Outpatient Public-private partnership Rashtriya Swasthya Bima Yojana Sokapheap Krousat Yeugn Short message service Social security scheme Third-party administrator Third-party payment Universal health coverage Value-added service World health organization Willingness-to-pay viii

9 1. INTRODUCTION Impoverishment as a result of catastrophic health expenditures, especially out of pocket (OOP) payment, has been a concern globally and more particularly in developing countries.(1) Health financing reforms are at the core of strategies to move towards universal health coverage (UHC), which the World Health Organization (WHO) defines as ensuring that everyone, everywhere, can access quality health services without facing financial hardship as a result.(2) Many governments or policy makers specifically pursue health insurance as a primary path towards UHC, but face significant challenges in making progress. A particular challenge exists with providing coverage for people outside of the formal sector. Thailand achieved UHC with the introduction of the Universal Health Scheme (UHS) in However, migrant workers (MW) and other people with citizenship problem 1 remained uncovered.(3) A large number of MW 2 entered into Thailand illegally and consequently were not officially recognized although they had been living and working in Thailand for a long time. Policies for administration and management of these groups were without concrete standard practices until 1999.(4) Since then, the Ministry of Health (MoH) implemented guidelines to provide health insurance for MW, which were enforced on a year-by-year basis, requiring annual Cabinet resolutions for renewal. Currently, healthcare for MWs is addressed under two insurance schemes: (i) the Social Security Scheme (SSS), managed by the Ministry of Labor and applicable for registered migrants from the formal sector; and (ii) the Compulsory Migrant Health Insurance Scheme (CMHIS) managed by the MoH targeting registered MWs who are not covered by SSS, and unregistered MWs entering the registration process, and their accompanying dependents.(4) 1 There is an estimated 450,000 people with citizenship problem (PCHIS). They are defined as citizens without a birth registration document stating Thai nationality: most of them were born in Thailand but lack documentation due to geographical, economic or educational barriers; they include ethnic minorities. Since 2010, they have access to the Health Insurance Scheme for people with citizenship problem (HIS-PCP), providing a similar coverage as the UHS. The Ministry of Public Health (MoH) undertakes both roles of purchaser and provider.(3) 2 The majority of MWs are from Myanmar (~80%) 1

10 The UHC movement has been picking up in Myanmar as well, although the country still has one of the highest OOP payment in the world, reaching close to 70% of the total health financing.(5) The government has endorsed the goal to achieve UHC by Concrete steps are proposed in a drafted roadmap towards UHC in Myanmar ( ).(5) But as of today, both the health service delivery system and the health financing function need to be urgently strengthened. With the objective to address unmet needs and gaps, the social enterprise Dreamlopments is planning to develop a private non-profit health insurance scheme for migrants living in Thailand and other vulnerable people living along the Thailand-Myanmar border. Previous research conducted by Dreamlopments among migrants living in the targeted areas identified important gaps under the CMHIS, and interest for a concept of low-cost, not-for-profit and needs-tailored health insurance.(6) As part of the feasibility assessment to develop this private health microinsurance, this review aims at: 1. Understanding the landscape of health microinsurances in other contexts, with a focus on Asia. 2. Identifying opportunities, challenges and lessons-learnt for the delivery of health microinsurances, specifically regarding determinants of demand and supply. 3. Discussing key considerations to designing a health microinsurance scheme in the intended context 2

11 2. METHOD & DEFINITIONS A rapid review of the literature was conducted to identify peer-reviewed publications and reports of research studies and project evaluations, in which HMI were delivered to vulnerable population within low and middle income countries. The detailed methodology is described in the separated ANNEX I. The findings are summarized in this report and detailed in the enclosed data extraction files ANNEX II. Key points on the case studies cited in this paper are summarized in the separated ANNEX III. In this review, microinsurance is defined according to Churchill et al, as the protection of lowincome people against specific perils in exchange for regular premium payments proportionate to the likelihood and cost of the risk involved.(7) The definition of HMI is less clear as they exhibit a broad range of objectives and institutional arrangements. They can be categorized according to their level of formality and integration within the national health financing system, as summarized hereafter and illustrated in Figure 1 (8): Informal schemes are managed by communities and are not underwritten by any formal or semi-formal institution. Formal schemes are underwritten either by a formal institution or organization (e.g. nongovernmental organizations-ngos), or by an insurer, while the premium is entirely paid by individuals (contributory schemes) or co-contributed by the managing institution or government (co-contributed schemes). Social schemes are underwritten by insurance companies or by government, which pays the full or large part of the premium. It also includes programmes where government fully subsidizes the operation of the scheme. Besides contributory schemes, there are non-contributory mechanisms relying on direct or indirect taxes or external source of funding. Special taxes such as sin taxes on alcohol or tobacco, as implemented in Thailand are an area of growing interest for financing health systems.(9) 3

12 Figure 1: Characterization of microinsurances and social security. Source: (8) In this report, challenges, opportunities and lessons-learnt were extracted from a broad range of HMI schemes. However, a focus was placed on formal and informal schemes that offer voluntary affiliation and are driven by non-profit motivations. The report first provides a summary of the HMI landscape in Asia, including schemes targeting migrant populations. It then examines the determinants of performance for HMI delivery and summarizes lessons-learnt. The findings are presented according to the four main operational aspects of HMI: (i) understanding needs and demand, (ii) designing the HMI product, (iii) contracting health providers, and (iv) achieving efficiency and sustainability. The report ends with a discussion on key considerations to design the pilot HMI being planned for the specific context of the Thailand- Myanmar border. 4

13 Because of the broad variety of HMI models, it should be highlighted that it was difficult to generalize determinants of performance, as they are highly context and scheme specific. In addition, the availability and the strength of the peer-reviewed literature was limited for private HMI (but more abundant for social health insurances), and for more innovative implementation strategies. It was thus important to also consider case studies or anecdotal evidence to inform this report on the latest evolution in the sector. Finally, the time limitation for the review process may have resulted in language bias as only literature in English was selected, or selection bias and gaps in data extraction, as priority was put on studying reviews on the topics of interest and key papers, rather than conducting an extensive literature searching. 5

14 3. THE HEALTH MICROINSURANCE LANDSCAPE 3.1. The Microinsurance Sector in Asia Microinsurances (MI) are developing at a breathtaking pace covering nowadays an estimated 280 million lives worldwide.(10) Asia, notably India, has the highest number of people covered (172 millions in 2013), but population coverage remains relatively low (<5%) as compared to other regions.(8) Since 2010, the annual growth rates have been compellingly at 30%, driven by some dynamic countries in the sector such as India, The Philippines, Thailand, Bangladesh, China, Indonesia, Cambodia and Malaysia. The growth has been fostered by government support, public-private partnerships and the development of innovative effective payment systems using new technologies or other mechanisms. Expansion of regulations also played an important role in the development of the MI sector in Asia. Hence, many of the informal community-based initiatives moved into partner-agent arrangements (as defined in section 4.4.3) to avoid regulatory problems. Currently, institutions or regulated commercial insurers take a lead role in the sector in Asia. Microinsurance in Thailand is defined through a framework developed by the Office of the Insurance Commission. Although the Thai government has been supportive of microinsurance, it is recognized that substantial work still needs to be done to achieve good coverage of quality microinsurance products. Rural populations in particular have little access to formal microinsurance mechanisms. Process, diversification and regulation of distribution channels are key areas that need to be addressed to increase expansion of microinsurances to low income populations in Thailand.(11) Myanmar has authorized 12 private insurances to enter the market since mid-2013, but their operation still seems severely restricted.(12) Microinsurances have not yet emerged, mainly because of regulatory restrictions. 6

15 3.2. The Outreach of Health Microinsurances HMI schemes emerged in the late 1980s in Africa and in the late 1990s in Asia to offer financial protection against catastrophic health expenditure.(13) Since 2000, the International Labor Organization (ILO) has endorsed HMI in its programme Strategies and Tools against social exclusion and poverty.(14) However, the HMI sector remains marginal within the microinsurance environment, because HMI products tends to be more complex. As illustrated in Figure 2, India is home to half of the formal HMI provision in Asia, while Pakistan, the Philippines, Thailand and Bangladesh constitute most of the other half. Countries like Cambodia, Indonesia and Sri Lanka already have some outreach.(15) Figure 2: Number of population covered by formal HMI in Asia, Source: (8) The HMI picture in Asia is very heterogeneous in terms of institutional and organizational models. Formal HMI are mostly implemented in partnership with regulated commercial insurers 7

16 (particularly because this approach is common in India). NGO are the second type of HMI providers (mainly in Bangladesh, Pakistan and Cambodia), whereas cooperatives and community-based organizations (CBO) only represent a minority of the supply in Asia.(11) Governments are also taking a leading role in the HMI space in Asia through social schemes, i.e. co-contributory schemes with insurance companies. Those schemes are characterized by a large or full contribution of the government to pay the premium. The New Rural Medical Scheme in China is the leading social microinsurance scheme in Asia followed by Rashtriya Swasthya Bima Yojana (RSBY) in India. The Philippines (PhilHealth) and Thailand (CMHIS) are other countries that run subsidized health insurance programmes Migrants and Health Protection in Asia The nature and level of inclusion of international migrants in national health care systems greatly varies in South-East Asia. On the one hand, sending countries like the Philippines and Indonesia implement policies to protect their overseas migrant workforce. On the other hand, receiving countries such as Thailand, Malaysia and Singapore still implement restrictive policies to discourage permanent settlement, despite the high demand for MW. Among all health protection mechanisms developed, Thailand seems to stand out with the SSS and the CHMIS, because it provides a coverage similar to UHS for MW who can show evidence of registration.(16) Nevertheless, both schemes, in particular the CHMIS face several implementation issues resulting in important gaps in population coverage, which will be discussed in Section 5. Overall, three types of models are emerging to address the insurance needs of migrants and their families, as described in a publication by Powers et al, supported by ILO (17): Host country model: the insurer is located in the country in which the migrant has immigrated. It usually covers only the migrant and its dependents in the host country, but family members in the home country are not covered. eg. CHMIS in Thailand Home country model: the insurer is located in the migrant s country of origin. The insured is the migrant (with products such as a repatriation insurance) or its family members at home, with the migrants paying for the policy from abroad or during home visits. The policy is usually sold prior to migration. 8

17 eg. In the Philippines, the government social health insurance PhilHealth has a specific Overseas Filipinos Program to protect workers abroad; registration is made prior to departure or online at a later stage; compensation is via reimbursement.(16) Seguro Popular, Mexico s public health insurance, began a pilot program to facilitate affiliation by having migrants sign-up their family members back home through a number of consulates in the United States.(18,19) Hybrid model: the insurer has a presence in both the home and host countries. Although it offers significant potential to address needs of transnational families, it is much more complex to manage. There are only a few such models, mostly at experimental stage. eg. Sekure Healthcare, based in the USA, charges low premium but only covers partial costs of health through discounts of 20-35% in medical and dental care in the USA and Mexico.(19) Most microinsurances targeting migrants are at pilot stage and few models are yet viable because of the regulatory challenges and the cost of the additional steps needed to reach this target group.(19) Legal and regulatory frameworks pose the biggest challenge. These constraints are often under-researched or underestimated, yet are one of the main causes of failure. It is therefore essential that the legal and regulatory issues are fully analyzed before launching a microinsurance targeting migrants, in particular defining: (i) which countries law(s) apply, (ii) what are the licensing conditions, and (iii) what are the regulation regarding insurance intermediaries.(17) Developing the adequate product and defining distribution mechanisms is also complex. Enrollment may be limited by the need of identification documents that migrants, particularly undocumented ones, may not be able or willing to share. Early experiences point to lessons and innovative approaches, such as partnering with remittance sending or receiving agencies, money-transfer agents, diaspora organizations, repatriation agencies (IOM/UNHCR), and migrant placement agencies or broker.(17) Marketing and distribution challenges could be mitigated with internet or mobile phone strategies, but face-to face approach remain important, as community leaders or peers can help locate the migrants, built trust and provide education on the microinsurance concept and product. 9

18 4. DETERMINANTS OF PERFORMANCE HMI have the potential to expand access to healthcare for the vulnerable population and to reduce catastrophic health expenditures. Strong evidence supports that HMI contribute to financial protection by reducing OOP, catastrophic and total health expenditure, as well as household borrowings and poverty (20 22). However, the evidence today also shows that HMI is often not an affordable option for the poorest and the most vulnerable.(14,20,23,24) Some observations suggest that HMI improve utilization of health services, but it is less clear in what way it affect behavior and health outcome. Impact assessments are rare, as they need longterm sustained intervention and are more difficult to implement.(15) In term of efficiency, many HMI face challenges to reach scale and become sustainable on the long term.(25) Several reviews report that only few schemes cover large populations and/or high proportions of the eligible population.(26 28) Small scale HMI lead to insufficient risk pooling, adverse selection 3 and high administrative cost.(28) Nonetheless, the case for HMI is still considered as part of the multiple strategies to achieve UHC given the resource constraints to scale up national health programmes. Also, promising avenues for efficiency and scale improvement are foreseen from innovative approaches that have been recently implemented, mainly in term of partnership for distribution, evolution in the benefit packages and technology integration. The determinants of performance identified in this review will be discussed hereafter as four main elements of an HMI operation: understanding needs and demand (section 4.1), designing a valuable HMI product (4.2.), contracting quality health care providers (4.3), and achieving 3 Adverse selection occurs when the risk profile of the group insured is worse than what would be expected in the general population. Consequently, the cost of insuring these people will be higher than expected and often the claims are higher than the total premium collected, resulting in the scheme making a substantial loss. There are 2 main causes: - The insured group is not a true pre-existing group and sick persons have come together specifically to gain insurance benefits - Pre-existing groups with higher-than-expected numbers of sick members join an insurance programme when other healthier groups do not. 10

19 institutional efficiency and sustainability. Although described as stand-alone blocks, these 4 main determinants of performance are closely inter-related and should be considered comprehensively Understanding needs and demand Health coverage is recognized as a priority for low-income households in most countries. Yet, this perceived need does not always translate into demand for HMI. Many factors influence a household s decision to enroll into such schemes. The determinants of demand described hereafter include socio-economic characteristics, use of other risk coping strategies, understanding of the insurance concept and trust in the HMI scheme offered Socio-economic characteristics Several socio-economic factors seem to influence the demand for HMI, often assessed in terms of willingness-to-pay (WTP). In a review from Adebayo and colleagues summarizing finding from both quantitative and qualitative studies, it is suggested that low levels of income and lack of financial resources (especially liquidity constraints) are major factors affecting enrollment.(29) WTP is also positively associated with increased age and higher education level. Mixed results are observed for other factors such as gender and household size depending on the setting where the HMI is implemented.(30,31) In Myanmar, a qualitative survey conducted in 2015 identified that higher education level, larger household size and urban residence were significant determinants of demand.(32) The household composition, like having child-bearing age women, pregnant women, elders, or children under 5 also positively seem to influence the demand in some settings.(33) Poor health status and previous experience of catastrophic health expenditures in the household may also encourage uptake, which highlights the risk of adverse selection in HMI. This was highlighted in Thailand, where the presence of illness was strongly related to the purchase of the voluntary health insurance before the introduction of the Universal Health Scheme in 2000.(34) Risk prioritization and coping strategies Low-income households are used to managing risk and commonly resort to informal risk sharing agreements and to the use of credit, savings, or remittances. Through community risk-sharing 11

20 mechanisms, households participate in an important social function. Social cohesion (i.e. participation in a community association) has been associated with HMI uptake in several contexts.(27,35,36) Insurance being new in most settings, it is likely that it may be perceived as most valuable when it complements, rather than substitutes, existing arrangements.(37) Understanding of the insurance concept Several studies report that target populations of HMI often lack understanding of the concept of insurance, and that this is linked to poor enrollment.(7,29,33) In Myanmar, where the communities have not been exposed to health insurance, Oo et al reported that awareness of health insurance was low (<10%) and that awareness was a significant determinant of HMI acceptance.(32) However, following thorough explanation, the acceptance towards the health insurance scheme increased markedly in that context (78%) Trust Trust in the insurance or in the contracted health facilities is often reported as an essential determinant of HMI uptake.(29,38 40) Trust and WTP are enhanced when there is a participative process in designing and implementing the scheme, and when people see that their preferences matter.(26,41) Engaging with community leaders, or with pre-existing groups in which members have interpersonal relationships (such as cooperatives) may facilitate the establishment of HMI. Indeed, experience-sharing and peer support are key leverage mechanisms to foster understanding and trust. LESSONS - LEARNT Understanding needs and demand Ø Conducting research on the target population to identify and understand: - major health risks faced and their consequences - social cohesion mechanisms, coping strategies and their effectiveness - knowledge and exposure to insurance or risk pooling concept - health seeking behavior and preferred healthcare providers - potential geographical, cultural, and socio-economic barriers to HMI uptake Ø Ensuring participation of the target group at all stage of the HMI design and involving 12

21 gatekeepers or pre-existing social groups to enhance trust. Ø Assessing the added-value of the proposed HMI product in comparison with other risk mitigation strategies developed and used by the target population. Ø Developing a plan to mitigate identified barriers to HMI access and uptake, and exploring the need for specific approaches to reach vulnerable groups Ø Carefully pilot how simple changes in marketing, consumer education, and product design resonate with the target group Further reading: Guideline for Market Research on Demand for Microinsurance. Sebstad J., Cohen M. and McGuinness E., 2006 Why people do not buy microinsurance and what can we do about it. Microinsurance paper No.20. Matul M.et al, Designing a valuable product The demand for and perception of the value of HMI are interlinked. It is reported that a demanddriven process is essential to ensure enrollment and retention in HMI scheme.(41) The PACE client value assessment tool was developed by the ILO team to support product design at the initial stage and provide a framework for analysis that can be complemented by market research and impact studies.(42) In this framework, it is assumed a MI product should be: - appropriate: matching the most important risk management needs - accessible: being explained simply and delivered in the vicinity of the target groups - affordable: providing good value for money at a price that the target group can afford - responsive: providing a timely response to shocks through prompt claims settlement and good customer care - simple: being simple to understand and use The PACE framework further links these characteristics to four key components of product design, i.e.: Product, Access, Cost and Experience (illustrated in Figure 3), which are used to report the findings in this section. 13

22 Figure 3: PACE Client value analysis framework. Source: (43) Product Eligibility, exclusion and waiting period A unique aspect of HMI is the willingness to be broadly inclusive and open to vulnerable groups. But, as many private HMI face considerable issues with adverse selection, (28) some schemes decide to have criteria to exclude high risk populations or individuals with pre-existing conditions (26,44). In these schemes, preliminary health examination or declaration of good health could be part of the underwriting process. Other HMI decide to avoid exclusion out of both equity and economic rationales: the cost of monitoring and enforcing complex claim screening and validation process has to be weighed against the claims avoided.(44,45) To avoid exclusion, some strategies are identified to stabilize the scheme. Group enrollment is frequently used as it helps to reach more people at lower cost (the group is contracted under the responsibility of one main policyholder). HMI works also well 14

23 with pre-existing social groups, because it leverages social capital for support and peer pressure to control for adverse selection, fraud and moral hazards 4.(46) Different approaches are used for group enrollments: - Household or village enrollment. To minimize the risk for adverse selection, a minimum percentage of the group is required to subscribe to the HMI in some cases (eg. ARY- India, where 30% of the village is required, or, Kasapi-iGroup-The Philippines, where 70% of the group is required) - Enrollment of other pre-existing groups, such as groups of microfinance clients (eg. Uplift in India, Nirapotta and Grameen Kalyan in Bangladesh), of cooperative members (eg. Yeshasvini in India), or of factory workers through partnerships with employers (eg. HIP-Cambodia) Individual enrollment is also offered as the main product in some schemes, or as an option in addition to group enrollment (eg. Nirapotta-Bangladesh, or PhilHealth-Philippines). These schemes often require a higher participation rate and/or a higher premium to be sustainable, as expenses are increased for sales, underwriting, administration and claims costs. Another approach used by HMI to mitigate adverse selection is implementing waiting periods before all or specific benefits can be received.(44) Waiting periods are used to identify preexisting conditions, to build up premium resources to cover claims, and to capitalize reserve funds. Benefit package A benefit package should cover the appropriate risk faced by the community, based on analysis of disease burden, frequency of event and level of risk, availability of health services and assessment of the perceived needs and demand from the target population. Because of huge variations in those factors, there is no single optimal benefit package that has received universal acceptance.(46) Conducting household surveys and ensuring strong participation of the target group seems crucial to ensure the product s appropriateness. Experience shows that it is feasible to involve 4 Moral hazard occurs when people with insurance use more services then they would if they did not have coverage only because they know that they are protected. 15

24 the community without compromising the judiciousness of rationing choices, even with populations with low education levels.(26,47) While using the tool Choosing health plan together (CHAT) 5 in India, people could consensually select benefit packages within their willingness to pay and perceived priorities.(48) Yet, many HMI face challenges to develop a benefit package that comprehensively covers the needs of the target population at an affordable cost that is sustainable for the scheme. Many HMI had to revise the benefit package following difficulties to balance expenditures and revenue collection.(26,28) As illustrated in Figure 4, the majority of formal HMI schemes (65%) in 2012 still focused on in-patient (IP) care and hospital cash 6 products in order to simplify design and processes. Only a minority of HMI offered comprehensive benefits.(8) Yet, it is recognized that a narrow approach limiting the benefit package Figure 4: Type of benefits. Source: (8)AM and the financial protection may result in a lack of perceived value (49). Low-income households who usually suffer routine OP expenditures may perceive OP coverage as more valuable than IP coverage. In addition, incorporating OP coverage could encourage early diagnosis and treatment, thus possibly reducing the need for expensive IP treatment.(26,50) For example, VimoSEWA in India included access to OP services after observing that one-third of 5 CHAT is an interactive game that was developed by the Micro Insurance Academy (MIA) to help communities assess their most common health needs and risks, and define a benefit package that best addresses these. 6 Hospital cash is a simple HMI product that pay the policyholder a fixed amount of cash per day of hospitalization. 16

25 its hospitalization s cost was for common and preventable diseases.(50) It is nowadays recognized that HMI should evolve towards offering more comprehensive coverage. More schemes start to pilot different mechanisms to expand IP coverage, and increasingly include preventive care and OP services in their core packages, or as value-added services (VAS). Value added service VAS are defined as services that supplement a HMI core product. They are offered as benefits outside of the insurance contract and are not linked to a claim. A publication from ILO describes different types of VAS (50): Consultations: they are offered as in-person consultations (at fixed points-of-care, or through mobile clinics), Dial-a-Doctor services and remote assisted consultations systems (telemedicine). Although they may be costly for the insurer, they have a significant potential for IP claims savings. Other mhealth approaches reinforcing capacity of frontline health workers (such as remote diagnostic support systems), or promoting patient adherence to treatment (such as consultation or medication reminders by SMS) may also offer significant potential. eg. Naya Jeevan - Pakistan, Care India Access to low-cost supplies or services in contracted health facilities or pharmacies: some HMI offer reduction for health services or medicine outside of the contracted network of healthcare providers. eg. Nirapotta, Grameen Kalyan - Bangladesh; Uplift, VimoSEWA, ARY India Preventive services: they are offered as health check-up, health talks, health camps, or awareness campaigns. These VAS show some prospects of being a cost-effective approach to reducing overall claims costs, but may be less popular with members because of their lack of immediate and tangible impact on OOP expenses. eg. SKY - Cambodia; Naya Jeevan - Pakistan; VimoSEWA and Uplift - India An increasing number of HMI offer these VAS, most of them as a complement to hospitalization coverage. Although there are limited data to enable robust quantification of their effectiveness, it is anticipated that VAS can have a positive effect on how a target population values a HMI, and hence on renewal rates.(50,51) 17

26 LESSONS - LEARNT Appropriate Product Eligibility Ø Implement strategies to avoid adverse selection but pros and cons need to be thoroughly assessed Ø Use simple eligibility criteria since waiting period and exclusion may reduce health protection and increase complexity for the members, distributor and administrator Ø Explore the possibility of group enrollment to promote social capital, increase scale and minimize administrative cost Benefit package Ø Involve target communities in defining the benefit package and its cost (possibly using existing tools, such as CHAT) Ø Start with a simple package limiting choices and options Ø Find strategies to include prevention and OP services for improved health protection and client value, as well as reduced risk of high expenditures in case of complications. It could be done in the core package or as VAS VAS Ø Introduce VAS to expand benefits and increase perceived value Ø Promote use of technology-enabled VAS, using the mhealth approach Ø Start with progressive phase-in after launch of the HMI, and promote it to raise understanding and awareness on its value Further reading: Value-added service in health microinsurance. Pott J et al, 2013 Improving client value: microinsurance insight from India, The Philippines and Kenya. Matul M. et al, 2011 Innovation and barriers in health microinsurance. Leatherman S, 2010 Lessons learnt and good practices in health microinsurance, a guide for practitioner. Chandani T. and Garand D,

27 Access Information and understanding A major challenge in providing HMI is overcoming the lack of awareness and understanding in the insurance concept, as discussed above. Information, education and communication (IEC) strategies are important to overcome this issue, and involve a systematic effort to teach risk management, the concept of insurance and risk pooling and the specificities of the HMI offered, in relation to existing health seeking behavior. Evidence of the positive effects of IEC efforts is particularly strong after enrollment, as it helps the members to recognize the value in health insurance and builds the know-how and confidence in utilizing the scheme. (52) Characteristics of IEC strategies that have been identified as effective include (15,53): Using simple messages: Uplift in India used simplified description of benefit categories according to illnesses to make it easier for members to understand the product. CARE Foundation in India did not use the term insurance, since most members did not understand what it meant. They simply emphasized that the card coverage entitled access to a defined set of services.(44) Developing a step-by-step approach, from the broader risk management and insurance concepts, to the specific HMI product benefits and limitations, as well as its logistics and practicalities. Using a mix of participatory methods, such as group-based training or interactive games. Delivering on-going education integrated in product delivery. Measuring the effect of consumer education on knowledge, uptake, renewal and claims. Enrollment choice and processes With voluntary affiliation, private HMI rarely enroll more than 30% of their target population.(37) Retention in voluntary schemes seems challenging as well, especially for members who did not use the scheme.(7,54,55) In the SKY scheme in Cambodia, the members who made claims were far more likely to renew the contract than the ones who did not. In addition to demand factors affecting uptake, it is recognized that the enrollment process may be too long or complex, and that the beneficiaries may lack the correct identification papers required for enrollment.(51) 19

28 Several strategies are used to promote uptake or retention (51) : Establishing automatic enrollment (with opt-out option) embedded in services contracts, such as micro-finance services, mobile phone subscriptions, or employment contracts. eg. In the HIP scheme in Cambodia, the employers from garment factory could choose to make enrollment compulsory, in which case they fully pay the premium. Improving user-friendliness with quick and efficient enrollment process supported by local motivators, community health workers (CHW) and/or HMI agents (with or without incentives). eg. VimoSEWA in India use existing self-help groups or champions clients to enhance trust in the product, while Nirapotta in Bangladesh supports CHW who also provide basic frontline health services. Using technology with biometric identification to remove the need for identity documentation at enrollment. eg. Naya Jeevan in Pakistan provides HMI to low-income workers via a system of sponsorship by the employer. They use an online enrollment system where employees provide minimal information through their employer (notably their mobile phone number) and then complete information with the support from a call center. However, although technology has simplified procedures (and reduced enrollment costs), human interaction continues to play an important role in helping to integrate it in use of smooth enrollment processes. Several forms of incentives are also offered in some HMI to encourage enrollment until the target population become more familiar with the scheme (46): Try before you buy initiatives: PharmAccess in Nigeria offered heavily discounted premium for the first 2 years of operation thanks to subsidies from the government and donors. Although the enrollment rate was high in the initial phase, it was difficult to increase the members contributions afterwards, because the subsidies were not explicit and the members did not understand why the cost increased after 2 years. Time-sensitive vouchers: offering an explicit reduction of premium during a clearly defined time has shown promising effects on enrollment rates in some settings.(56,57) Premium discounts and rewards: they may be offered when members committed for long-time enrollment at renewals, or when they enroll the whole family. In India, 20

29 VimoSewa offered an option for children to join at a lower incremental fee. Proximity HMI products that sells far away and demand travel to the HMI point of service often result in low uptake. In the GK scheme in Bangladesh, the level of membership among the two lowest socio-economic groups appeared to be related to distance: subscription was as high as 90% of the population in villages near the service point, and of only 35% in distant villages.(26) Modest enrollment results is also associated with only passively selling the HMI product at the point of care.(58) Experiences clearly show that the target population should have the possibility to enroll on-thespot, or, to at least have an easy access to the HMI point of service. Outreach has been promoted through community organization, HMI agents or financial services used by the target population, as illustrated in Figure 5. Figure 5: Distribution channel by microinsurance type Source: (8) 21

30 More innovative distribution channels through employers, retailers or mobile network operators remained marginal in 2012 in Asia, but are becoming more frequent nowadays. Indeed, optimizing distribution mechanisms is recognized as essential to reduce the distance gap for the target population as well as the cost for the distributor. The proliferation of internet and mobile networks in particular offer interesting opportunities, as they may not only minimize access barriers and administrative costs, but could also facilitate premium payment and enhance monitoring capacity.(47,54) For example, the mobile network provider Tigo partnered with Bima and MicroEnsure in Senegal and Tanzania to has launched a life insurance in 2012 that was agent-free and linked to mobile phone airtime, enabling customers to enroll at any time and from anywhere.(55) Mobile microinsurances continue to expand. In 2015, 120 mobile microinsurances were available in 33 emerging markets, predominantly in Africa (58%), in South Asia (19%) and in East Asia-Pacific (18%).(59). Although, the majority still covers other risks (such as life insurance), some starts to include health coverage as well. For example, an hospital-cash HMI model launched in 2015 in Pakistan in partnership with Teleonor and Micro-Ensure gained more than 100,000 subscribers in just 1 year.(60) In Cambodia, Bima provides accident, life and health microinsurances in partnership with Smart, and has operated already in 10 provinces after 2 years of service only. While mobile microinsurances using a strategy with low human interaction reduce their distribution cost, the target group may face challenges to understand the HMI processes. Hybrid approaches using sales agents and call centers are valued for the high-quality education and service offered, but may also show limited scalability because they are more expensive. Premium payment mechanisms Microinsruances face unique challenges to collect premiums, as they target low income populations in the informal sector that often have irregular incomes and liquidity constraints. This is particularly problematic for HMI having a rigid enrollment window, or if there is a lack of access to financial services.(54) In many community-based HMI, the premium are collected by going door-to-door, during community meetings, or by asking the policyholders to come to the HMI point-of-service. These collection methods represent high transaction cost for the insurer and may be more prone to fraud because of the large number of people involved in the premium collection. 22

31 Several strategies are developed to address challenges regarding premium payment: Increasing flexibility of the premium payment to match the income flow of the target population: some schemes ensure that the premium collection match the post-harvest season in rural areas, while others ensured that payment modalities are flexible and fragmented with monthly, quarterly or semi-annual payments (PhilHealth, SKY in Cambodia). It has been highlighted that payment for longer periods may increase retention, lower transaction costs and reduce risks of fluctuation of revenues into the scheme.(51) Introducing cashless payment mechanisms: premium are embedded in service contract such as savings, loan or remittance contracts, or are part of the employer s benefit plan (with automatic salary deduction or sponsorship from the employer). Mobile phone technology has been a key evolution to facilitate cashless premium collection from unbanked populations, through loyalty-based premium, normal premium or freemium. 7 In Kenya for example, Safaricom and Changamka aimed to bundle the HMI product with a maternity saving card using mobile phone technology to facilitate transaction and premium payment. The feasibility and efficiency of such solutions still need to be further assessed.(44,61) LESSONS - LEARNT Accessible product Understanding & Information Ø Develop step-by-step IEC strategy starting with financial literacy and adding on the insurance concept and the HMI product benefits and practicalities Ø Simple messages, with adequate language adapted to the target group should be used 7 Loyalty-based models encourage customers to spend a certain amount of airtime or keep a certain balance in their mobile money account to qualify for insurance (usually calculated on a monthly basis). Premium models are more like a traditional premium payment for a specific insurance coverage. However, with mobile insurance, this may be monthly, weekly or daily payments, which differs from traditional yearly premium payments. Freemium models are a combination of the two: customers can subscribe to loyalty-based insurance, and can increase their cover by paying a fee. 23

32 Ø Provide ongoing education after enrollment Ø Monitor the effectiveness of IEC strategies Enrollment Ø Develop a simple, quick and efficient enrollment process Ø Explore strategies to make automatic enrollment embedded in other services contracts (microfinance loan, mobile phone contract) or employment benefits Ø Consider short-term explicit incentives for enrollment (voucher, cashback), with caution as they may have long-term implications Proximity Ø Ensure proximity of the HMI point-of-service and promote on-spot enrollment if possible. Ø Develop outreach strategies via existing community groups or associations that already work with the target group. Ø Explore innovative distribution mechanisms leveraging: - existing infrastructures serving the community (retailers, financial organizations, or outlets of mobile network operator) - mobile HMI, using paperless and flexible enrollment systems, but still ensuring face-to-face contact Premium payment mechanisms Ø Establish flexible payment mechanisms in line with income flows, addressing liquidity constraints of the target group Ø Explore cashless approaches with premium payment made through a service or employment contract, supported by mobile technology or bundled with financial services (health saving accounts, or loan) Further reading: Removing obstacles to access microinsurance. Cimon E. et al, 2013 Beyond slogans, good practices in promoting microinsurance. Lee NR., 2013 Microinsurance distribution channel, insight for insurer. Merry A et al, 2014 Emerging practices in mobile microinsurances. GSMA, 2015 The emerging global landscape of mobile microinsurance. CGAP,

33 Cost Premium to benefit and to members income Defining adequate premium in a new HMI market is a challenge because of the lack of data on health conditions and costs of services. Inadequate actuarial data forces many insurers to price the premium using a trial-and-error approach and to adjust it when more information on actual cost of services and market penetration becomes available.(44) In many HMI reviewed, the premium was improperly priced, most often being too low, as it did not account for what was really required to cover costs and generate minimum margins for expansion and sustainability.(46) Balancing cost and coverage is another important challenge. Some private HMI offer premium to the target population based on assessment of the individual risk profile for profitability or financial sustainability. However, it is found that an actuarially sound premium, if unsubsidized, may be unaffordable, especially the very poor population.(26,62) Other schemes, mainly community-based or non-profit schemes, offer premiums based on the willingness and/or capacity to pay. They use a flat-rate premium across individuals or groups, or sliding scale discounts for group enrollment, giving deeper discounts for each additional member and an incentive for larger families to enroll.(7,46) The contribution from members has been reported to be between 1 and 2 % of the annual household income in some settings.(45,63) Efforts to keep the premiums to a level that members are willing to pay can also involve: - Limiting coverage (to a specific list of health services) - Limiting the networks to pre-approved healthcare providers offering discounted rates - Including copayments (to reduce moral hazard) - Subsidizing premiums Some studies suggests that premium subsidies in HMI contribute to decrease OOP expenditures and to increase demand and use of services.(64) Different forms of premium subsidies are used (65): - explicit premium subsidies from government or donors - sponsorship mechanisms: eg. Naya Jeevan scheme in Pakistan, in which employers subsidized the premium for their workers - cross-subsidization of premium between classes of insured beneficiaries (effective only in case of large risk pooling), or with the revenue from a commercial venture, which profit is invested in the HMI: eg. GK - Bangladesh 25

34 Many HMI use premium subsidies to increase insurance coverage, especially for poor households or those at risk (e.g. free insurance for children).(64) To improve equity, it is reported that targeted premium subsidies have the potential to be more effective than universal subsidy (for all policyholders).(64) But, it has been challenging to establish effective targeting mechanisms, as highlighted in the RSBY scheme in India which faced important challenges in identifying populations below the poverty line. In addition to ensuring equity, subsidies have also been used to address market challenges, such as lack of data to price the premium, poor understanding of insurance concepts and low uptake. In the case of Hygeia in Nigeria, the short-term subsidized premium resulted in high enrollment rates and enabled the HMI management team to reduce the risk of adverse selection and gather information on health-care behavior, claim cost, and customer preferences, that facilitated subsequent product improvement. But, as described in Section 4.2.2, it also had an important negative impact on members renewal rates. The above highlights that using direct subsidies on premium may have lasting effects on the demand and on the sustainability of the scheme, and therefore need to be carefully designed. Testing of subsidy mechanisms are still underway, with little evidence on what really works best and how. Some practitioners seem to recommend that subsidies be used to build the institutional capacity of the scheme (as discussed in Section 4.4.1), rather than to directly support the premium.(52) Other fees and cost Many health schemes have co-payments to limit moral hazards. But, these OOP co-payments may have negative impact on the most vulnerable groups preventing them to seek care, thereby rationing care instead of rationalizing it. Rather than putting in place co-payments, some HMI prefer to limit the benefit package, for example by rationing the number of OP consultations. Indirect costs in accessing healthcare, such as travel costs and opportunity costs in the form of loss of income are other barriers to access the HMI services. Some schemes, like RSBY in India, provide a travel allowance to remove geographical barriers and improve the value of insurance in rural areas. 26

35 LESSONS - LEARNT Affordable cost Premium pricing Ø Hire insurance experts to define premium costs based on actuarial data. Premium may need to be adjusted when more data is available Ø Assess capacity and willingness to pay of the target group Ø Consider different options to keep premium low. Identify gaps between actuarial cost and affordability, and assess pros and cons of subsidy, if required. Smart subsidies may be designed with a clear purpose, monitoring plan, long-term financing and exit strategy. To address equity concern, they should have an efficient targeting strategies. Other costs Ø Use co-payment with caution as they increase OOP expenditure and decrease value. Ø Assess indirect cost for accessing healthcare (transport cost, loss of income) and explore the possibility to cover them Further readings Making health microinsurance work, ten recommendations for practitioners. Holtz J. et al, Using subsidies for inclusive insurance: lessons-learnt for agriculture and health. Vargas Hill R et al., Experience with the scheme Claim procedures Claim settlement in HMI is more complicated than in any other MI, because the products are more complex and involve another party, the healthcare provider. Yet, settling claims is essential, as it is when the HMI becomes real for its members. HMI have used various strategies to improve or facilitate the claim process. First, support from community-based volunteers, HMI agents or hotlines are established in some settings. Then, several HMI work to accelerate the claim processing time. Technology and decentralization of the claim validation have helped to improve the claim turn-around time, as experienced in the RSBY scheme in India.(66) In the Nirapotta scheme in Bangladesh, the eligibility, and simple claims are verified on the spot by a liaison officer supported by a computerized claim decision 27

36 tool. A medical advisor intervenes only for a more comprehensive clinical review of problematic cases. In this case, the turnaround time of claim processing has been decreased from 25 to 10 days. The conditions of access to healthcare are also essential for the HMI members. Some HMI schemes require that their members pay cash at the time of receiving healthcare services, and then seek reimbursement from the insurer. Schemes, such as Uplift in India, argue that reimbursement creates a stronger sense of ownership amongst members, decrease moral hazards and avoids inflated costs from healthcare providers. But, for low-income households, reimbursement may represent a critical financial barrier to accessing care. HMI providers are increasingly offering cashless claims system, when a thirdparty payment (TPP) mechanism with selected healthcare providers is established: in this system, the insured patients are not required to pay the cost of health services at the point of service, and a third party (the HMI scheme or another entity on behalf of the HMI scheme) pays the healthcare provider for the covered services it provided to the patient.(67) Both reimbursement and TPP mechanisms are illustrated in Figure 5. Reimbursement mechanism 28

37 Third-party mechanism Figure 6: Third party payment mechanism, Source: (67) An online survey conducted by ILO in 2009 among 65 HMI reported that 52% of the schemes surveyed used TPP, 22% used reimbursement, 18% used a mixed approach and 8% provided integrated care.(67) In South-East Asia, TPP mechanisms are a common feature amongst HMI schemes since national health financing systems are oriented towards TPP.(67) Although TPP as cashless mechanisms may reduce barriers to access care, and reduce the administrative burden for the HMI, their reported challenges include moral hazard, the need to require permission before admission for IP care, the challenges of contracting partner health facilities and the additional burden for the healthcare provider to verify eligibility. Policy administration and tangibility In some HMI, it s been reported that members wait a long time (up to 3-4 weeks) to get their insurance cards, or may even not receive a tangible proof of being insured.(43) Rapidity in benefit enforcement and tangibility have been improved thanks to the provision of an insurance card with a photograph and biometric information. In the RSBY scheme in India, utilization rates increased when photo identification cards were issued immediately.(52) In the ILO online survey, it was found that 67% of the schemes use an identification card with a photograph, and 29

38 14% use a more sophisticated smart card that allows electronic verification of eligibility. (43) Tangibility has been further reinforced in some schemes by sending follow-up SMS with information related to the policy as done in Vimo SEWA in India. Customer care As products and distribution models become complex, the role of customer care has become an important factor of satisfaction and renewal. Good strategies enable members to understand the essential features of the HMI product, provide support when needed, assess client satisfaction and establish a feedback and grievance mechanism. Placing a representative of the insurer at hospitals to assist members has been a way to improve their experience when accessing care and making claims, and to collect their feedback.(44) With advancements in mobile telecommunications, providing 24/7 toll-free hotlines are being tested (e.g. RSBY, Uplift).(43) LESSONS - LEARNT Responsive product Claim processing Ø Establish simple claim mechanisms with support (face-to-face or via 24/7 call center) Ø Improve rapidity of the claim process by using technology and/or decentralization of the eligibility and claim validation process Ø Encourage cashless approach with TPP mechanisms to reduce barrier to seek care, while using mechanisms to control moral hazard and verify eligibility at the point-of care Policy & tangibility Ø Make the HMI concrete by providing an insurance card (with identity and biometric information) and clear policy documents Customer care Ø Commit to provide user-friendly servicing along the HMI processes to improve satisfaction and retention Ø Monitor the customer care strategy through direct contact, satisfaction survey, and easy complaint mechanisms Ø Promote forum for experience-sharing among members 30

39 Further reading: Third party mechanisms in health microinsurances. Le Roy P. et Holtz J., 2011 The moment of truth; claim management in microinsurances. Rendek K. et al, Contracting quality healthcare providers The perception of the value of a HMI scheme is also linked to the ability to access quality healthcare. Some of the most successful schemes have progressively adopted a more active purchasing role.(28) Strategic purchasing is present when there is an active search for the best providers, the best health services to purchase, and the best contracting arrangements Type of healthcare providers HMI cover access to health services through a variety of providers depending on the proximity, availability quality and readiness of these health services. The choice of contracted providers is also done based on the existence of pre-existing free public services to avoid duplication and keep costs down. Some insurances have provided different benefit packages depending on the type of healthcare facilities contracted. A review in 2002 reported that out of 132 community-based insurances assessed, health services were provided in 61% of the cases by a public provider; in 17% by the HMI own facility; in 4% of the cases by a private provider; and in 18% of the cases by a mix of selfowned, public and private providers.(27) This pattern seems to evolve with a trend for increasing participation of private providers in formal HMI schemes. Inclusion of private providers could contribute to expand service coverage and client satisfaction. In RSBY in India, inclusion of private facilities has led to high satisfaction levels among clients and was critical to the success of the OP pilot, as many public providers were not available in remote areas. HMI also increasingly engage less formal providers such as community health workers, or private pharmacies because of the lack of formal providers in remote areas.(52) Quality of care The lack of quality of health centers is often identified as one of the strongest impediments to take-up,(26,29,38,68) or renewal of HMI.(51,54) Managing a network of healthcare providers is 31

40 crucial to ensure the provision of high quality services (and to control costs), although there are yet weak evidences on the effect of HMI on the quality of healthcare. Some HMI require that all contracted facilities or providers are accredited and audited by an internal review process or by an external authority. Some have agreements with contracted health facilities on the use of standard protocols for the most common diseases, and monitor compliance.(44) Others developed a capacity-building program designed to enhance standardization of care (e.g. SafeCare program by Hygeia in Nigeria). Effective quality assurance mechanisms also entail that information and data are collected and analyzed. Liaison officers and medical advisors are regularly present within the contracted health facilities to assess that insured patients receive appropriate care, according to standard guidelines.(67) It has been reported that involvement of the clients in the process is important, through assessment of their satisfaction with the provider and provision of actionable information on the providers.(69) Provider payment mechanisms Various methods of provider payments are described as retrospective (fee for service, casebased, or per day of hospitalization), or as prospective (capitation, or direct payment of salary of the contracted providers).(67) According to Robyn et al,(70) many HMI have applied a mix of provider payment methods, most of them using some form of fee for service, followed by salarybased payment, while fewer have used capitation or case-based methods. Each method gives different incentives and disincentives for providers to control the cost and quality of services, as described in Table 1 hereafter. Payment methods such as case-based, per-day payment and capitation, which transfer some financial risk to healthcare providers, may have greater potential to contain costs than fee-for-service payment, but need additional measures to ensure quality of care. They are also usually more difficult to negotiate with the healthcare providers. Some qualitative evidence supports that dissatisfaction with the choice of payment methods has led to reduced provider participation in some HMI schemes.(70) Little evidence exists on the effect of the different types of provider payment on patient demand for services. However, provider payment methods may affect population enrollment or renewal both directly (e.g. through the level of patient copayments for healthcare) and indirectly (e.g. through their impact on the quantity and quality of healthcare services provided).(70) 32

41 Table 1: Pros and Cons of the different provider payment mechanisms (26,44,52,67,70) Methods of payment Fee-forservice Pros Easy to understand and implement Well accepted by providers as it may be their usual method of payment, and their risk is reduced May be used when utilization and charges are acceptable Cons Drive-up costs by including a risk to over-utilize services Present a financial risk for the HMI, which may tend to decrease benefits, introduce co-payment or reject claims Per case or Per day Simplify claims administration Transfer of part of the financial risk to the providers Encourage efficient care management (shorter length of stay) Difficulty to define a fair cost/ case Present incentive to - diagnose more complex cases - make unnecessary admission - reduce services, even when they are necessary Capitation Simplify claim administration Transfer of the financial risk to providers Reduce provider-induced demand Ensure a steady revenue stream Encourage provider to provide preventive care and improve early diagnosis and treatment Present incentive to limit care or exclude high-risk patients Encourage inappropriate referral to other health facilities Are difficult to negotiate with providers Are challenging to implement if the pool of patients is small, as the variation of cost of care within that pool can be important. 33

42 LESSONS - LEARNT Contracting quality healthcare providers Type of provider Ø Map and assess different providers according to proximity, availability quality and readiness of their services. Ø Private providers are sometimes included in the coverage and may improve client satisfaction, although they are more expensive and more difficult to negotiate with Quality Ø Define quality standard and improvement strategies with the contracted providers, with implementation and monitoring of: - accreditation system and compliance monitoring - standards protocols for disease of high burden - capacity-building program if needed - strong data and information management Ø Ensure ongoing quality assurance and by medical officer in the contracted facilities Ø Involve client in monitoring quality Providerpayment mechanisms Ø Identify the optimal payment system that incentivizes performance and quality alike. Payment mechanisms that transfer part of the risk to the provider have greater potential to control cost, but need strong quality monitoring system. Ø Involve providers in the decision process regarding the provider-payment mechanism Further reading: Assessing health provider payment system: a practical guide. Joint Learning Network, 2015 Improving health system quality in low- and middle-income countries that are expanding health coverage: a framework for insurance. Mate K. International Journal for Quality in Health Care 2013; Volume 25, Number 5: pp Provider payment in community-based health insurance scheme in developing countries: a systematic review. Robyn PJ. Health Policy and Planning 2013;28:

43 4.4. Achieving institutional efficiency and sustainability In addition to product design, various supply determinants also influence the efficiency and sustainability of HMI schemes. These determinants are linked to the scheme administration and management, the control of cost and risk, the delivery model and partnership arrangement, and the regulatory environment Administration and management Maintenance of long-term stability involves effective financial management and permanent risk monitoring. Some HMI show weak management capacity, especially community-based schemes, whose managers are often volunteers who may lack the skills and the time to improve performance.(26) Management gaps can include a failure to adjust the product to perceived needs, to set up adequate premium pricing, and to efficiently monitor the scheme and manage insurance risks. One approach to strengthening the management of the scheme has been to subcontract the task to a third-party administrator (TPA), although the cost may be prohibitive for small-scale HMI. This approach combines the advantages of supporting a sense of strong group identity, and bringing in professional management. Ownership remains with the small-scale associations, which carry out the tasks of community sensitization, premium collection, and control of possible misuse, while the TPA takes care of technical aspects such as the financial management, and strategic purchasing. The ability of the HMI scheme to manage moral hazard, fraud, claim and administrative costs heavily depends on the quality and efficiency of its information system and processes.(67) Paper-based information management systems have shown important efficiency limitations. More HMI now heavily invest (sometimes with external subsidies) in developing computerized information management systems that help capture data and facilitate near-real time analysis of the sales and claims payment. In the large scale RSBY in India, all data on enrollment and claims are stored daily in a central server and automatically checked in near real-time to control fraud and moral hazards. Besides regular monitoring, it is also recognized that more solid evaluations on performance and impact should be done. Some microinsurances working groups, such as the 35

44 Microinsurance Network s Impact Working Group, have developed standard guidelines and tools.(71) Risk management While it has been reported that high distribution and administrative costs were the most important challenges to HMI sustainability, controlling claims ratios is also a key issue, as highlighted in a recent study of 5 South Asian schemes.(65) Factors such as adverse selection, fraud and moral hazard may result in high claim incidence, which is further exacerbated by poor performance monitoring that prevent organizations to early identify claim trends.(45,72) Lessons from case studies show that the risk of fraud can be mitigated with several management and monitoring arrangements, implemented in addition to a strong technologyenabled information management system(44): Use of smartcards to improve accuracy in verifying members identity and eligibility, thereby reducing fraud and increasing administrative efficiency. Involvement of community committees to assess claim and promote transparency and solidarity, eg. Uplift in India Presence of liaison officers to verify eligibility and coordinate with the healthcare providers. Use of updated lists of eligible members, which is critical for schemes with capitation payment. This strategy is simpler for HMI schemes that limit enrollment to defined period, but more demanding for HMI schemes that maintain open enrollment. Moral hazard has been controlled by a set of measures in various HMI, including: Introduction of co-payments or benefit limits, but these measures increase the financial risk for the beneficiary. Negotiation with providers (capitation or case-based payment mechanisms), but strong mechanisms for quality control must then be put in place. Clear referral system and rapid pre-authorization mechanisms for high-cost services at the secondary level of care. A gate-keeping function is especially important when providers receive fee-for-service payments. (eg. Yeshavisni - India) Another particular risk faced by HMI is this of covariant risks such as related to natural disasters or epidemics, notably in schemes of small size and of limited geographical scope.(7) A measure to protect HMI against unexpected high-level expenditure is reinsurance.(26) Governments may 36

45 provide implicit reinsurance to HMI for losses due to substantial covariant risk.(7) While the theoretical benefits of social reinsurance are widely understood, practical experience in this area in developing countries seems limited. Commercial reinsurance services do exist, but these are more hardly accessible to small HMI: only 21% of the formal HMI in Asia are reinsured.(8) Yet, it has been reported that HMI may benefit more from paying the reinsurance premium than keeping potential equivalent surplus as a safety margin in a reserve account.(15) Institutional arrangements The insurance value chain is composed of three essential institutional entities: - The distributor, in charge of marketing and selling of policies, as well as collecting and paying funds from the target groups - The administrator, in charge of the product and process design, administration of the claims, and monitoring and evaluation - The risk carrier, in charge to make actuarial decisions, to assume the financial risk, and ensure the long-term stability of the scheme. A reinsurer may assume part of the risk when outsourced. Each of these functions can be entailed by various actors, thus resulting in a broad range of institutional arrangements. Among the classifications proposed, Radermacher and Dror propose a four-model typology in the microinsurance compendium, as illustrated in Figure 7 hereafter. (45) The community-based model: Community Health Insurances (CHI) have involved a wide range of insurance arrangements, with variation in terms of ownership, management, membership, and service. In theory, CHIs share common characteristics (14): voluntary affiliation, non-profit motivation; resource pooling based on mutual assistance within a group usually presenting similar characteristics (eg. location, job); risk sharing based on solidarity, intended broad inclusion with premiums independent of individual risk; and risk on the insured population which owns or participates in the scheme management. The social cohesion of the group represents in large the strength that facilitates the functioning of this kind of model. However, the voluntary and limited nature of the insurance target often results in small risk pools, which are more vulnerable to adverse selection..the scheme management may also lack sufficient managerial knowledge to ensure effectiveness. 37

46 The non-profit or social model: This model involves a wide range of institutional arrangements, which are all non-profit and aim at not putting the risk on the insured persons. The degree of risk on the insured and their involvement in the HMI management distinguish it from the CHI, while the solidarity approach differentiates it from the partneragent model. The provider may be an NGO or a social enterprise. The model can be applied to social HMI subsidized by governments. Maintenance of long-term stability is arguably the weakest point of this model, although sustainability without subsidies may not be the primary objective because of the social benefit motivation. The partner-agent model: In this model, the relationship between the beneficiaries and a licensed insurance company ( the partner ) is facilitated by an intermediary ( the agent ), which can be any other organization with close contacts to the target group (often a microfinance institution, or retailers). The partner carries the risk, and thus often keeps control over strategic decisions affecting maintenance of long-term sustainability. The strength of this model is that the agent has an existing effective interface with the target population. For the agent, it may be easier and less risky to offer insurance in partnership with a formal insurer, which has knowledge of market regulation, capital, reserve and reinsurance. The agent could approach insurers with a product prototype with a price range that the target group would be willing to pay, and organize a bidding process to select the insurer. Challenges reported with this model include: ensuring that client needs are represented; building capacity of the agent; defining roles and responsibilities; and setting the incomes and commissions. The provider-driven model: The unique feature of this model is the involvement of the healthcare provider in the design and management of the HMI. Healthcare providers may launch an insurance scheme to facilitate access to health services for specific segments of population, and/or generate larger volumes of activity in a network of facilities. The unification of the roles of provider and purchaser of services may create conflicts of interest and the provider may have incentives to compromise on quality. Difficulties can also lie in the lack of technical and managerial capacity of the provider in management of such schemes, to ensure effectiveness and long-term stability. 38

47 Figure 7 : Different delivery model, Source:(42) While CHI arrangements still account for a sizable portion of the HMI outreach, they are being surpassed by other partner-agent arrangements, including public-private partnerships.(15) For HMI, it is increasingly recognized that an integrated strategy of social protection should be conceived in collaboration with several partners. A study of microinsurances schemes that achieved large scale identified partnerships as one of the key driver of success.(72) Yet, the many potential players involved in delivering HMI (i.e. commercial insurers, CBOs, NGOs, delivery channels, TPAs, and healthcare facilities) have very different characteristics and objectives, which may lead to challenges in sustaining the partnership. Problems arise when there is no business case for each partner, or when there is a serious power imbalance between partners or when interests are not aligned. Most commonly, partnership challenges occur between insurers and distributors.(72) 39