88523 SUPPLY-SIDE READINESS FOR UNIVERSAL HEALTH COVERAGE: Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia THE WORLD BANK THE WORLD BANK OFFICE JAKARTA Jakarta Stock Exchange Building Tower II/12-13th Fl. Jl. Jend. Sudirman Kav. 52-53 Jakarta 12910 Tel: (6221) 5299-3000 Fax: (6221) 5299-3111 www.worldbank.org/id/health THE WORLD BANK 1818 H Street N.W. Washington, D.C. 20433 USA Tel: (202) 458-1876 Fax: (202) 522-1557/1560 Email: feedback@worldbank.org Website: www.worldbank.org Printed in June 2014 This work is a product of the staff of The World Bank with external contributions, The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Exective Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. For any questions regarding this report, please contact: atandon@worldbank.org Report No. 88523-ID SUPPLY-SIDE READINESS FOR UNIVERSAL HEALTH COVERAGE: Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia iv Contributors This policy paper has been prepared by a team consisting of members from the Indonesia Ministry of Health’s National Institute of Health Research and Development (Soewarta Kosen, Ingan Tarigan, Yuslely Usman, Tati Suryati, Endang Indriasih, Harimat, Idawati Muas, Retno Widyastuti, Merry Luciana, Tita Rosita and Dwi Hapsari), and the World Bank’s Indonesia Health, Nutrition, and Population Team (Ajay Tandon, Wei Aun Yap, and Eko Pambudi); Additional inputs were provided by Pandu Harimurti, Puti Marzoeki, Stephanus Indrajaya, Xialu Bi, Rong Li, and Darren Dorkin. Comments from Toomas Palu, Owen Smith, Karima Saleh, David Evans, Robert Yates, Broto Wasisto, and Gayle Martin are gratefully acknowledged. This paper was edited by Diana van Walsum. All photographs used in this paper are courtesy of John Estey and Eko Pambudi. Supply-Side Readiness for Universal Health Coverage v Contents Contributors ..................................................................................................................................... iv List of Abbreviations and Acronyms .................................................................................................. viii Executive Summary ........................................................................................................................... 2 1. Introduction ............................................................................................................................... 6 2. Background ................................................................................................................................ 10 2.1 Implementation of Universal Health Coverage Reforms in Indonesia ................................... 10 2.2 Rising Non-Communicable Disease Burden in Indonesia .................................................... 11 2.3 Supply-Side Implications of Universal Health Coverage ....................................................... 14 2.3.1 Depth of Coverage in Indonesia .................................................................................. 15 2.3.2 Provider Payment Mechanisms .................................................................................... 16 3. Supply-side Readiness Assessment ............................................................................................... 22 3.1 Framework for Analysis ........................................................................................................ 22 3.2 Data Sources ........................................................................................................................ 24 4. Supply-Side Readiness Assessment for Non-Communicable Diseases.......................................... 26 4.1 General Service Readiness .................................................................................................... 26 4.2 Human Resources for Health ............................................................................................... 28 4.3 Diabetes Mellitus ................................................................................................................. 28 4.4 Cardiovascular Conditions ................................................................................................... 32 4.5 Chronic Respiratory Conditions ........................................................................................ 36 5. Conclusions ................................................................................................................................ 40 6. Policy Implications ..................................................................................................................... 44 ANNEX A ........................................................................................................................................ 48 ANNEX B ........................................................................................................................................ 49 Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia vi List of Tables Table 1: Top 10 causes of disease burden in Indonesia, 1990-2010................................................ 13 Table 2: Top 10 risk factors in Indonesia, 2010 ............................................................................. 13 Table 3: Benefit package entitlement and provider payment .......................................................... 17 Table 4: Supply-side implications for provision of DM-related care at the primary level ................ 30 Table 5: Availability of equipment for DM-related care at Puskesmas ............................................ 31 Table 6: Percent of Puskesmas reporting availability of blood glucose and urine tests for diabetes ...................................................................................................................... 31 Table 7: Supply-side implications for provision of cardiovascular conditions at the primary level .......................................................................................................... 35 Table 8: Provinces with <75% of all Puskesmas reporting captopril availability ............................. 36 Table 9: Share of DALYs lost from chronic respiratory diseases in Indonesia, 1990-2010 .............. 36 Table 10: Supply-side implications for provision of chronic respiratory disease-related care at the primary level .............................................................................................................. 37 Supply-Side Readiness for Universal Health Coverage vii List of Figures Figure 1: Burden of disease by cause in Indonesia, 1990-2010 ........................................................ 11 Figure 2: Share of population aged 65+ in selected EAP countries, 1950-2070 ............................... 12 Figure 3: Three dimensions of UHC ............................................................................................... 15 Figure 4: Supply-side assessment framework ................................................................................... 22 Figure 5: Service Delivery Indicators Model.................................................................................... 23 Figure 6: Basic amenities in puskemas in Indonesia ....................................................................... 27 Figure 7: Availability of basic amenities at Puskesmas by province ................................................. 27 Figure 8: Burden of diabetes mellitus in Indonesia, 1990-2010 ...................................................... 29 Figure 9: Availability of blood glucose and urine tests in Puskesmas by province ............................. 32 Figure 10: Hypertension prevalence among adults aged 18+ years by province in Indonesia, 2007 ........................................................................................................... 34 Figure 11: NCD Service Readiness Indicator Index, by Province, 2011 ............................................ 41 Figure 12: Prevalence of Diabetes, by Province (Riskasdes, 2007) ..................................................... 41 Figure 13: NCD Service Readiness Indicator Index vs. Selected Economic Indicators (Provincial-level) ............................................................................................................. 42 Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia viii List of Abbreviations and Acronyms Askes : Asuransi Kesehatan AusAID : Australian Agency for International Development BPJS : Badan Penyelenggara Jaminan Sosial CAD : coronary artery diseases COPD : chronic obstructive pulmonary disease CRDs : chronic respiratory diseases DALYs : Disability-adjustted life years DM : diabetes mellitus DRGs : diagnosis-related groups GIZ : German Society for International Cooperation HFCS : Health Facility Costing Study HRH : human resources for health INA-CBGs : Indonesia Case-based Groups Jamsostek : Jaminan Sosial Tenaga Kerja JKN : Jaminan Kesehatan Nasional MOH : Ministry of Health NCD : non-communicable diseases OOP : out-of-pocket PETS : Public Expenditure Tracking Survey Puskesmas : Pusat Kesehatan Masyarakat (public primary care facilities) Rifaskes : Riset Fasilitas Kesehatan Riskesdas : Riset Kesehatan Dasar SARA : service availability and readiness assessment SJSN : Sistem Jaminan Sosial Nasional UHC : universal health coverage WHO : World Health Organization Supply-Side Readiness for Universal Health Coverage EXECUTIVE SUMMARY 2 Executive Summary Indonesia is currently in the midst of major reforms of its health system, which have resulted in the institutionalization of one of the largest single-payer universal health coverage (UHC) programs in the world. At the same time, the country is undergoing a rapid epidemiological transition and non- communicable diseases (NCDs) are now the dominant share of the overall disease burden in the country. Indonesia faces several challenges to effective implementation of UHC in terms of expanding breadth, height, and depth of coverage, especially in addressing NCD conditions, which are generally chronic in nature, require careful patient case management over time, and are most cost-effectively addressed at the primary care level. Given this backdrop, this policy paper focuses on the issue of supply-side readiness from the perspective of assessing the depth of UHC in Indonesia, especially in rural and remote areas of the country where a large proportion of the poor and near-poor populations reside, and with a focus on key tracer NCD conditions – diabetes mellitus (DM), chronic cardiovascular conditions, and chronic respiratory conditions – at the public primary care (Puskesmas) level in Indonesia. The paper compares supply-side implications of UHC benefit entitlements as derived from national health facility and program-specific guidelines with the World Health Organization’s (WHO) recommendations for service availability and readiness assessment (SARA) for these same tracer NCD conditions. Using a variety of information sources – including analysis of the 2011 Rifaskes facility census – the paper assesses the ability of Indonesia’s public health system to provide effective coverage for NCDs across the country. Results from the analysis show that, even though DM is now the eighth (and rising) cause of the overall disease burden in Indonesia, service readiness for basic diagnosis and basic treatment of DM is limited. For example, only 70 percent of urban Puskesmas could conduct a blood glucose test for the diagnosis and monitoring of DM. In rural Puskesmas, this readiness indicator declined to 51 percent nationwide and in some provinces, such as Gorontalo, Papua, Southeast Sulawesi, Maluku, North Sulawesi, and West Papua, less than 20 percent of Puskesmas fulfilled this indicator. Basic pharmaceutical treatment for DM is more encouraging with 90 percent of Puskesmas stocking glibenclamide; however, only 48 percent stocked metformin. As over 95 percent of Puskesmas reported the availability of functioning blood pressure apparatus and over 99 percent reported the same of stethoscopes, most Puskesmas have the necessary equipment to diagnose hypertension. The availability of medical treatment for hypertension using captopril, a commonly used antihypertensive medication in Indonesia, is fair. Nationwide, 84 percent of Puskesmas stocked this medication, although deficiencies were notable in rural Puskesmas in provinces such as West Sulawesi, Maluku, and Papua, where captopril was available in less than 70 percent of Puskesmas. Furthermore, in the case of hypercholesterolemia, only a third of Puskesmas nationally reported the ability to conduct cholesterol tests and only 36 percent reported the availability of simvastatin, a common treatment for hypercholesterolemia. In the case of chronic respiratory diseases such as asthma, over 75 percent of all Puskesmas reported the availability of basic treatments such as salbutamol, prednisolone, and oxygen. There are geographic variations, with less than 75 percent of Puskesmas reporting availability of salbutamol and prednisolone in Central Sulawesi, West Sulawesi, and Papua. Supply-Side Readiness for Universal Health Coverage 3 Putting all the available indicators for the three NCD tracer conditions together into an index reveals that the deficiencies are especially prominent at the primary care level in 12 of 33 provinces in Indonesia, especially in the eastern part of the country where 18 percent of the country’s total population and 22 percent of Indonesia’s rural population resides: North Sumatera, Bengkulu, East Nusa Tenggara, North Sulawesi, Central Sulawesi, Southeast Sulawesi, Gorontalo, West Sulawesi, Maluku, North Maluku, West Papua, and Papua. By way of contrast, supply-side readiness for NCDs appears not to be a problem in public hospitals across all provinces in Indonesia. The analysis reported in this policy note underscores that – in order to attain UHC – there needs to be a focus not just on increasing the breadth and height but also on ensuring that effective depth of coverage exists, especially in the rural and remote areas of the country and for primary care. It is not enough to specify a comprehensive benefit package on paper if facilities do not have the basic supply-side readiness conditions to be able to provide key services. In ensuring and expanding effective depth of coverage, various policy implications should be considered. Firstly, although this paper sheds some light on the existence of deficiencies on the supply-side, further efforts and systematic analysis are required in order to understand why supply-side deficiencies exist so that the appropriate policy measures can be instituted. Furthermore, as financing gradually shifts from the supply-side to the demand-side, the mechanics of this demand-side financing should be leveraged to improve service readiness at facilities in an optimal manner. Greater clarity is required on the implications of the benefit package, both for providers (in order to understand what inputs are required to provide the benefit package) and for the insured, through improved socialization of benefit entitlements so that the insured are better able to understand what services they should be able to receive through Badan Penyelenggara Jaminan Sosial (BPJS) Kesehatan. Although the survey and census analyzed in this paper provide a snapshot of some aspects of service readiness, regular independent monitoring and evaluation of supply readiness is required: both as part of the routine due diligence functions of the insurance administrator and also as evaluation of the health provision system in order to inform Ministry of Health (MoH) policy. Regardless of whether investments in service readiness are predominantly driven by supply-side investments or by demand-side financing mechanisms, it is clear that overall investments in service readiness for NCDs are required, in order to maximize efficiency gains from the prevention and prudent management of these chronic conditions. Finally, although the scope of the paper is focused on facilities, risk factors for NCDs should be addressed holistically, through public health interventions beyond the health facility and indeed, beyond the health sector alone. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 4 Supply-Side Readiness for Universal Health Coverage INTRODUCTION 6 1. Introduction Non-communicable diseases (NCDs) now account for the largest share of the overall burden of disease in Indonesia. Whereas in 1990 only about 37 percent of morbidity and mortality in the country was due to NCDs, by 2010 this number had risen to 58 percent.1 This increase is expected to continue as the share of the population aged 65 years and older will begin to rise rapidly beginning in 2015, and ageing will bring an even higher susceptibility to NCDs in Indonesia. With this rapid epidemiological transition in the backdrop, Indonesia is currently implementing a series of health system reforms aimed at attaining universal health coverage (UHC) for its population by 2019. These reforms include the merger of several existing contributory and non-contributory social health insurance schemes with streamlined uniform benefits under a single-payer umbrella beginning in 2014, followed by a gradual expansion in breadth of coverage to those currently uncovered by 2019. There are several challenges facing effective implementation of UHC in Indonesia, both from a health financing and a service delivery perspective. Some key issues from a health financing perspective are the collection of contributions from the non-poor in the informal sector, ensuring adequate fiscal space for financing coverage for the poor and near-poor, managing an effective transition from supply-side to demand-side financing of the health system, reducing out-of-pocket (OOP) payments, and ensuring cost containment and financial sustainability of UHC. From a service delivery perspective, important challenges remain related to ensuring supply-side readiness in terms of addressing the rise in NCDs, provision of the UHC benefit package in a coordinated manner, meeting current and future demand for health care, improving quality of care, leveraging and regulating private provision, and ensuring availability of human resources for health (HRH) across the country, among others. This policy paper focuses on the issue of supply-side readiness from the perspective of assessing the effective depth of UHC in Indonesia, especially in rural and remote areas of the country where a large proportion of the poor and near-poor populations reside.2 Insurance coverage under UHC implies benefit entitlements and – in some cases – restrictions as to what services will be eligible for reimbursement. As Indonesia moves towards an increase in demand-side insurance-oriented financing of its health system, assessing the extent to which benefit entitlements are actually available is a key consideration in ensuring that UHC is effective. In particular, given the rapid rise in NCDs in the country. This policy paper assesses the ability of Indonesia’s public primary care health facilities to provide preventive, diagnostic, and curative care for selected tracer NCD conditions: diabetes mellitus (DM), cardiovascular conditions, and chronic respiratory diseases. The paper compares supply-side implications of UHC benefit entitlements as derived from national health facility and program-specific guidelines with the WHO’s recommendations 1 Institute of Health Metrics and Evaluation (2013). 2 “Effective” is a reference to the realization of the benefit package, not to the impact on health outcomes. Supply-Side Readiness for Universal Health Coverage 7 for service availability and readiness assessment (SARA) for these same tracer NCD conditions.3 Next, using a variety of information sources – including analysis of the Rifaskes facility census triangulated with other sources of information – the paper assesses the ability of Indonesia’s public primary care health system to provide coverage for the selected tracer NCD conditions across the country and discusses policy implications for effective implementation of UHC in Indonesia. The focus in the paper is on highlighting gaps in benefit entitlements and service readiness to stimulate policy dialogue on ways to enhance UHC implementation in the country. It is important to note at the outset that this policy paper is not an exhaustive study of the key determinants of tracer NCDs in Indonesia. Needless to say, even if service delivery inputs are available, the ability and efforts of providers – as well as the actions and behaviors of households and key interventions in other sectors – remain areas of consideration in ensuring that any intended reforms lead to improvements in NCDs and other health outcomes. As discussed later in the paper, there are many risk factors associated with NCDs. The availability of preventive, diagnostic, and curative medical care services is one – but only one – critical factor in addressing the burden of disease related to NCDs in Indonesia and to ensure effective UHC implementation. Addressing the rising burden of NCDs in Indonesia will require a coordinated effort from both within and outside the health sector. The remainder of the policy note is structured as follows: Section II provides some background and context related both to NCDs and to UHC in Indonesia. Section III analyzes the benefit package for key tracer NCD conditions, derives the supply-side implications of the benefit package, and compares these implications to the WHO’s SARA-recommended indicators. Section IV presents results from analysis of data related to the effective depth and distribution of coverage for key NCD conditions in Indonesia. Section V concludes with a summary and some policy implications. 3 The WHO’s SARA is a health facility assessment toolkit that is designed to help collect and analyze information on key aspects of service delivery in a health system, such as the availability of key human resources and infrastructure resources as well as basic equipment, basic amenities, essential medicines, diagnostic tools, and the readiness of facilities to provide health care interven- tions for tracer conditions. For more details see: WHO. 2012. Measuring Service Availability and Readiness: A Health Facility Assessment Methodology for Monitoring Health System Strengthening. Geneva: World Health Organization. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 8 Supply-Side Readiness for Universal Health Coverage BACKGROUND 10 2. Background 2.1 Implementation of Universal Health Coverage Reforms in Indonesia Indonesia has mixed public-private provision of health care services, with the public sector taking a generally dominant role in rural areas and for secondary levels of care. For primary care, Indonesia has more than 9,000 public Puskesmas, each serving catchment areas of 25,000-30,000 individuals (approximately a third of Puskesmas also provide inpatient services). Private provision has been increasing rapidly in recent years, including for primary care. Of the 163,000 hospital beds in the country, about 52,000 are managed by the private sector. In 2011, household survey data estimates indicated that less than half of all Indonesians reported having some form of health insurance coverage.4 An estimated 21 percent of all households had coverage under the non-contributory Jamkesmas scheme which targeted the poor and near-poor; 8.9 percent of households were covered by Askes, Indonesia’s social insurance program for the public sector; 7.9 percent were covered under Jamsostek, the private sector insurance program for the formal sector; and 7.1 percent were covered by other forms of insurance. 58.5 percent of all Indonesian households reported having no health insurance coverage in 2011. Coverage of benefits under social programs such as Jamkesmas and Askes was skewed towards provision by the public sector, although private provider enrollment in these programs has been increasing rapidly in recent years. There is evidence indicating that expansion of insurance coverage has increased utilization rates, especially inpatient utilization, among the poor and near-poor. Nevertheless, OOP spending remains the dominant source of financing, and is high even among those with coverage.5 Indonesia is currently in the midst of implementing a series of health system reforms aimed at attaining UHC for its population by 2019. The universal right to health care was included as an amendment to Indonesia’s constitution in 1999. However, the impetus for UHC came a few years later, in a 2004 landmark legislation -- the Sistem Jaminan Sosial Nasional or the SJSN Law – which formed the legal basis for attaining several social protection objectives in the country. In 2011, the government passed a ground- breaking follow-up law that defined the administrative and implementation arrangements – the Badan Penyelenggara Jaminan Sosial or BPJS Law – which stipulated that several existing contributory and non- contributory social health insurance schemes be merged to provide streamlined uniform benefits under a single-payer umbrella beginning in 2014. Following institutionalization of the single-payer insurance administrator (BPJS Kesehatan) in 2014, the government plans to incrementally extend coverage to the entire population by 2019. The unified insurance program is referred to as Jaminan Kesehatan Nasional (JKN). BPJS Kesehatan is expected to contract with both public and private providers for delivery of the JKN benefit package beginning in 2014. 4 Official MoH data indicates 63.5 percent individual coverage rate in 2011. 5 UNICO Indonesia study. Supply-Side Readiness for Universal Health Coverage 11 With these reforms in mind, service delivery issues are increasingly relevant in Indonesia. This report builds on an earlier report focused on the health workforce, which found that although the quantity of health workers had increased, the deployment and distribution of health workers was still skewed and thin in rural areas of Indonesia. Although training requirements and opportunities had expanded, the quality of public and private health services, as measured by ability to diagnose and treat, was “marginal” and overall quality remained low.6 2.2 Rising Non-Communicable Disease Burden in Indonesia Like several other countries in the region, Indonesia is undergoing a rapid demographic and epidemiological transition. NCDs now account for the largest share of the burden of disease in Indonesia. Whereas in 1990 only about 37 percent of morbidity and mortality in Indonesia was due to NCDs, by 2010 this number had risen to 58 percent (Figure 1).7 This trend is expected to continue in the coming years, with the share of NCDs in the disease burden continuing to rise as the population ages. Beginning in 2015, the share of the population aged 65 years and older is projected to increase rapidly in Indonesia (Figure 2). FIGURE 1: BURDEN OF DISEASE BY CAUSE IN INDONESIA, 1990-2010 Injuries Communicable Injuries Injuries Communicable 7% 9% 9% 43% 56% 56% 37% 49% 58% Communicable Non-Communicable Non-Communicable Non-Communicable 1990 2000 2010 Source: IHME 6 World Bank. 2010. New Insights into the Provision of Health Services in Indonesia: A Health Workforce Study. Washington, DC: World Bank. 7 Institute of Health Metrics and Evaluation (2013). Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 12 FIGURE 2: SHARE OF POPULATION AGED 65+ IN SELECTED EAP COUNTRIES, 1950-2070 Population aged 65+, 1950-2070 30 China 25 Share of total population (%) Vietnam 20 Thailand Indonesia 15 10 Philippines 5 0 1950 1970 1990 2010 2030 2050 2070 Year Source: UN Stroke, an NCD, was responsible for the largest share of the overall disease burden in Indonesia, causing 8 percent of all disability-adjusted life years (DALYs) lost due to morbidity and premature mortality in 2010 (Table 1).8 Stroke was also the leading cause of premature mortality in Indonesia. Ischemic heart disease, unipolar depressive disorders, and diabetes are other prominent NCDs in the top ten causes of the disease burden, with most of these conditions having doubled their share of the disease burden in Indonesia over the period 1990-2010 (Table 1). 8 DALYs refer to aggregated healthy years of time lost at the population level as a result of disease-related morbidity and prema- ture mortality. Supply-Side Readiness for Universal Health Coverage 13 TABLE 1: TOP 10 CAUSES OF DISEASE BURDEN IN INDONESIA, 1990-20109 Rank in DALYs lost share Top ten diseases/conditions in 2010 2010 1990 2000 2010 1 Stroke 4.3% 6.3% 8.0% 2 Tuberculosis 7.5% 7.6% 7.6% 3 Road injury 3.3% 4.6% 4.7% 4 Low back and neck pain 2.8% 3.8% 4.5% 5 Diarrheal diseases 6.9% 5.6% 4.0% 6 Ischemic heart disease 1.9% 2.8% 3.8% 7 Unipolar depressive disorders 2.6% 3.4% 3.8% 8 Diabetes 1.7% 2.6% 3.5% 9 Lower respiratory infections 13.7% 5.9% 3.0% 10 Neonatal encephalopathy 2.5% 3.3% 2.9% DALYs per 100,000 44,144 35,074 32,053 Life expectancy in years 62 66 69 Source: Institute of +Health Metrics and Evaluation (2013) The rise in NCDs in Indonesia is a result of several socio-demographic and lifestyle factors. Ageing is one contributory factor, although the prevalence of NCDs among younger age groups in Indonesia is also increasing. Physical inactivity, unhealthy diets, tobacco use, and harmful alcohol consumption are key risk factors for NCDs. Several of these risk factors – including dietary risks, hypertension, smoking, high fasting plasma glucose, and physical inactivity – are prominent among the top ten risk factors contributing to the overall disease burden in the country (Table 2). The share of dietary risks and high blood pressure as contributors to DALYs lost has more than doubled over the period 1990-2010 (Table 2). TABLE 2: TOP 10 RISK FACTORS IN INDONESIA, 2010 DALYs lost share Rank in 2010 Top ten risk factors in 2010 1990 2000 2010 1 Dietary risks 5.2% 10.2% 10.7% 2 High blood pressure 4.8% 7.4% 10.0% 3 Smoking 6.3% 6.1% 8.2% 4 Household air pollution 9.6% 6.5% 5.9% 5 High fasting plasma glucose 2.8% 3.9% 4.7% 6 Physical inactivity -- -- 3.1% 7 Occupational risks 1.9% 2.6% 2.8% 8 High body-mass index 0.4% 1.2% 2.8% 9 Iron deficiency 2.9% 2.8% 2.4% 10 Ambient particulate matter pollution 1.8% 1.8% 2.1% Source: Institute of Health Metrics and Evaluation (2013) 9 NCDs were 51.4 percent of the top ten DALYs in Indonesia, with the top ten conditions accounting for 45.7 percent of the total disease burden in the country. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 14 From a health financing and service delivery perspective, NCDs tend to be of a chronic nature, typified by long durations and slow progression rates, but they can also result in rapid premature death (for example, with stroke and cardiovascular conditions) or an expensive period of disability and morbidity. In addition, NCDs are often preceded typically by a sub-clinical phase in which preventative strategies may be effective in delaying or reversing disease progression. For these reasons, NCDs can be relatively costly to treat and require sustained case management, often requiring multiple contacts with the health system over one’s lifetime. In addition to the crucial role played by public health efforts aimed at reducing risk factors for NCDs on a population level, the management of NCDs also requires primary care services to play an important and effective role in screening and delivering preventive and promotive interventions which, for most NCDs, are far more cost-effective than treatment at advanced stages of progression. There is also evidence to suggest that NCDs are more likely to result in catastrophic health spending, placing households at risk of impoverishment.10 Given the lifestyle nature of risk factors, an additional challenge is that several promotive and preventive interventions for addressing NCDs tend to lie outside the immediate purview of health systems – in the tax, agricultural, education, urban design, and transport sectors – and some of these interventions are generally not amenable to inclusion as part of traditional insurance benefit packages. 2.3 Supply-Side Implications of Universal Health Coverage UHC – ensuring that everyone has access to quality health services when needed, without experiencing financial hardship as a result – can be conceptualized as having three key dimensions: “breadth”, “height”, and “depth” (Figure 3).11 Breadth refers to the proportion of the population that is covered; height refers to the proportion of health costs that are paid by pooled funds as opposed to via direct OOP payments; and depth refers to the benefit package of services that are covered by pooled financing (and the definition of services can be broadened to include provision of public health interventions).12 For example: by expanding coverage to include the non-poor informal sector (as intended by BJPS Kesehatan), breadth is expanded; by excluding additional services such as cosmetic surgery, depth is reduced; and by covering the costs of medications and removing co-payments and deductibles (as intended by BJPS Kesehatan) so that patients do not need to pay out-of-pocket for these, height is increased. These dimensions are often interrelated, as poor service coverage due to supply-side readiness problems can drive the insured to alternative non-empanelled private providers, in effect rendering financial coverage or even population coverage meaningless. A key challenge in implementing UHC is to balance the expansion of these three dimensions, by making the appropriate trade-offs (given resource constraints) between breadth, depth, and height. Although all dimensions are important, this paper focuses on ensuring depth of coverage – not just on paper, but also in effect – especially in rural, remote regions of the country and in light of the rapid transition to an NCD-dominated disease profile in Indonesia. 10 World Bank. 2011. The Growing Danger of Non-Communicable Diseases: Acting Now to Reverse Course. Washington, DC: World Bank. 11 WHO. 2010. World Health Report: Health Systems Financing – The Path to Universal Coverage. Geneva: World Health Organi- zation. 12 Ibid. Supply-Side Readiness for Universal Health Coverage 15 FIGURE 3: THREE DIMENSIONS OF UHC Total Health Expenditure Direct costs: proportion Reduce cost of the costs Include sharing and fees covered other services Extend to non-covered Current pooled funds Services: which services are covered? Population: who is covered? 2.3.1 Depth of Coverage in Indonesia From 2014 onwards, various social insurance schemes including Jamkesmas (for the poor and non-poor) and Askes (for civil servants) have been merged into JKN, which has a single-payer administrator (BPJS Kesehatan). The structure of the benefit packages imply that the effective depth of coverage for NCDs is sensitive to the readiness of empanelled health facilities. All social health insurance schemes prior to 2014 define their benefit packages generally, not by the coverage of specific diseases but by the types of services that are eligible for coverage (e.g., health promotion, medical screening, consultations, examination, diagnostics, and treatment); in the case of Askes, specific diagnostic tests and medical interventions were also mentioned. The benefit packages allude to treatment guidelines, which were level-specific in Indonesia and which for many diseases specify the names of drugs that should be used. For public health facilities, a national essential drug list exists and is used to define which drugs can be provided to public health facilities (these are typically procured by district health offices).13 These social health insurance schemes also defined a “negative list” of specific services and benefits that are not eligible under these schemes. The JKN benefit package is comprehensive and similar to that of Jamkesmas. The benefit package has been defined in terms of covering everything except for: • Health services that do not follow stipulated procedures, including referrals • Health services in facilities that are not contracted by BPJS Kesehatan, except for emergencies • Health services that are covered by occupational accidental insurance • Health services abroad, cosmetic procedures, health services for infertility, and orthodontic services • Health disorders/diseases caused by drug addiction and/or alcohol • Health problems caused by self-harm activities • Complementary treatment using alternative/traditional medicine unless declared effective by health technology assessments • Experimental procedures, health equipment for households, contraceptives, baby food, and milk • Health services for disaster situations. 13 Daftar Obat Esensial Nasional 2011. Ministry of Health, Indonesia. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 16 Despite the elaboration of the JKN benefit package, the effective depth of coverage of NCDs will depend on an interplay between the service readiness of these facilities (the focus of this policy note), government clinical guidelines, the national essential drug list, the service types defined in the schemes, and the exclusionary “negative list” defined in the schemes. Details of the social health insurance schemes and their benefit packages are summarized in Table 3 below. Given the importance of referrals under JKN, and for cost-effectiveness considerations more generally, it is especially important that the primary care sector play its role in the health system in terms of being ready to provide the stipulated services. All public health providers were automatically included in these health financing schemes, whereas private providers are subject to an accreditation process, with a grace period of 5 years for primary care facilities and 3 years for hospitals. Although including all public providers helps to expand the theoretical breadth (population) coverage to include the population in the catchment area of all public providers, this is likely at the expense of the depth (services) of coverage as these facilities many not be ready to provide these services. The degree of independence and robustness of the accreditation process for private providers would be an area ripe for further research. 2.3.2 Provider Payment Mechanisms Provider payment mechanisms can influence the service readiness of health facilities, and hence are summarized here in order to contextualize the findings and policy implications. Demand-side payments from social health insurance schemes are not the primary source of revenue for health facilities. For example, health facilities receive equipment and drugs (except for contraceptive, immunization, and any drugs related to vertical programs such as HIV/AIDS, TB and malaria) from (and determined by) local governments and, in some instances, through a fund operated by the central government (though with expected but often not realized contributions from the local government) called the Special Allocation Fund (Dana Alokasi Khusus). Operational budgets (including salaries) are also provided by both the central government, through the General Allocation Fund (Dana Alokasi Umum), and by local governments. Vertical programs, likewise, are funded through the central government. The supply-side readiness of Puskesmas is hence, to a large extent, influenced by a combination of local and central government supply-side financing. This caveat should be considered in the following discussion on provider payment mechanisms used by the social health insurance schemes. Indonesia’s social insurance schemes made a distinction between primary care services and secondary care (referral) services. Capitation was the predominant provider payment mechanism at the primary care level although there were important variations. Askes differentiated between the type of primary care provider: with capitation payments ranging from Rp 2,000 per month at small Puskesmas to Rp 6,500 per month for empanelled family doctors. Jamkesmas initially paid capitation payments directly to Puskesmas but, since 2011, capitation payments are made to district health offices, which then reimburse Puskesmas on a fee-for-service basis, depending on actual utilization. These arrangements are expected to be continued for Puskesmas under BPJS implementation as of 2014. Capitation payments do not currently expose primary care facilities to the costs of referrals made, leading to the risk of inappropriate referrals Supply-Side Readiness for Universal Health Coverage 17 made not for clinically justified reasons, but merely to shift the costs and burden of managing a patient to secondary care or tertiary care facilities. Furthermore, there is the risk of incentivizing undertreatment under capitation systems, which is somewhat mitigated by the fact that Puskesmas do not need to finance the purchasing of essential medicines and commodities themselves, but receive these without charge on request from district health offices. At the secondary and tertiary care (referral) levels, a form of diagnosis-related groups (DRGs) called Indonesia Case-based Groups (INA-CBGs) were used to pay providers under the Jamkesmas scheme. BJPS Kesehatan is also using this provider payment mechanism as of 2014, replacing the tariff system used by Askes currently in which a comprehensive schedule of tariffs varying by the hospital class/specialization is used to pay for specifically defined diagnostic tests and medical interventions. Additional details of this provider payment mechanism are summarized in Table 3. TABLE 3: BENEFIT PACKAGE ENTITLEMENT AND PROVIDER PAYMENT Health Provider Payment Insurance Target Group Benefit Package ‘Negative list’ Mechanism Scheme Askes1 Civil servants, Primary care outpatient At the primary care level, Selected excluded pensioners, services: the following services services: veterans, and types are included: independence pi- Rp 2,000 per member per medicines and medical • For childbirth, oneers, and their month for health centers with consumables, (ii) promo- only the first two immediate family one or no general physicians; tive and preventative care, living children are members Rp 4,000 for health center (iii) curative care, and (iv) covered. with two or more general rehabilitation. Basic inpa- physicians; Rp 5,500 – 6,500 tient services are covered for family doctor. where this is provided by the Puskesmas. Basic inpatient services: Rp 80,000 per day For formal referrals to secondary care (referral) level: A detailed schedule Secondary care (referral) of diagnostic tests and level: medical interventions are Tariff of charges for inpa- included. Interventions are tient care depending on four only permitted for specific classes of hospitals - A, B, C, medical indications. and D – and class of accom- modation (I or II). Detailed tariff of charges for specific diagnostic tests and medical interventions (varying by hospital class) Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 18 TABLE 3: BENEFIT PACKAGE ENTITLEMENT AND PROVIDER PAYMENT (CONTINUED) Health Provider Payment Insurance Target Group Benefit Package ‘Negative list’ Mechanism Scheme Jamkesmas2 The poor and Primary care outpatient At the primary care level Selected excluded non-poor, occu- services: Prior to 2011, Pusk-(Puskesmas and network services: pants of social esmas received a capitation providers), the following institutions, pris- of Rp 1,000 per member per service types are included: • Cosmetic proce- oners and victims month. From 2011, district (i) consultation and coun- dures of disasters (for health offices receive this seling; (ii) simple laborato- • General check-ups one-year) capitation of Rp 1,000 per ries (blood test, urine test, member per month instead, and feces); (iii) minor med- • False prosthetic and pay Puskesmas on a fee- ical intervention; (iv) dental teeth for-service basis depending examination and interven- • Unproven alterna- on utilization. tion, including revocation tive medicine (approx. 76.4 mil- and tooth patches by • Infertility treat- lion individuals) Secondary care (referral) dentist; (v) prenatal care, ment. facilities are paid by using postpartum visit, lactation a form of diagnosis-related consultation, neonatal and groups called Indonesia Case- under-five care, including based Groups (INA-CBGs) basic immunization; (vi) family planning consul- tation; and (vii) provision of formulary drugs. Basic inpatient services at the primary level are included as well. At the secondary care (referral) level, the fol- lowing additional service types are included: (i) medical rehabilitation and (ii) specialist consultations and examinations. Class III inpatient accommoda- tion is covered, as are the following additional service types: (i) intensive care, (ii) high-risk obstetrics, and (iii) specialist and major surgery. Supply-Side Readiness for Universal Health Coverage 19 TABLE 3: BENEFIT PACKAGE ENTITLEMENT AND PROVIDER PAYMENT (CONTINUED) Health Provider Payment Insurance Target Group Benefit Package ‘Negative list’ Mechanism Scheme JKN3 (2014 Universal cov- Primary care facilities4 (all A single benefit package Selected excluded – onwards) erage for any eligible public facilities and (for medical benefits) is services: resident (mini- empanelled private facilities) defined for all members. mum 6-months are paid by capitation based Non-medical benefits, such • Infertility treatment residency) is on the number of regis- as ambulance services • Orthodontic treat- provided through tered members. For public and accommodation, are ment defined catego- Puskesmas, the capitation dependent on the category ries: is understood to be paid to of membership, though • Complementary district health offices who upgrades of the class of medicine (not then reimburse Puskesmas accommodation are per- assessed through • Non-contribut- HTAs) ing members on a fee-for-service basis mitted if the difference is (the poor and based on utilization (as per paid separately. • Cosmetics those without the current arrangements for • Self-inflicted health means to pay), Jamkesmas). At the primary care level, problems (including whose contri- the following service types drug or alcohol butions are Secondary care (referral) are included: (i) admin- addictions) covered by the facilities are paid by using istrative costs, (ii) pro- • Work hazards and government a form of diagnosis-related motive and preventative diseases covered • Salaried em- groups called Indonesia Case- care (including medical by work insurance ployees and based Groups (INA-CBGs), check-ups, lifestyle and • Medical treatment wage-earners, using a standard tariff deter- risk factor advice, basic obtained abroad. whose contri- mined in agreement with the immunizations, family butions are par- Minister and reviewed twice planning including vasec- tially covered by yearly5. tomies and tubal ligations, the employer and medical screening), (iii) • Self-employed, Medicines and medical medical examinations, (iv) unsubsidized consumables are covered by medications and medical contributions JKN and are to be reimbursed consumables (as clinically at prices determined by the justified), (v) consultations, • Civil servants, (vi) non-specialist surgical the military, Minister. and non-surgical treat- police, indepen- ment, (vii) blood transfu- dence pioneers, Funding of medicines and sions, and (viii) first-line veterans, and commodities for vertical pro- diagnostic tests. pensioners. grams – e.g., contraceptives and immunizations - will still be covered by those vertical At the secondary care programs. (referral) level, outpatient and inpatient service types are defined. Outpatient care includes: (i) adminis- trative costs, (ii) specialist medical examinations and consultations, (iii) specialist treatment, (iv) implants, (v) second-line diagnostic tests, (vi) rehabilitation, (vii) blood transfusions, and (viii) forensic services and mortuary services. Inpa- tient service types include both intensive care and non-intensive care. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 20 Supply-Side Readiness for Universal Health Coverage SUPPLY-SIDE READINESS ASSESSMENT 22 3. Supply-side Readiness Assessment 3.1 Framework for Analysis The framework used in this paper for assessing supply-side readiness for UHC is summarized below in Figure 4. As depicted in the figure, the supply-side assessment begins with an examination of the current benefit package for Jamkesmas and Askes, two of the biggest social health insurance programs merged under BPJS in 2014, as well as the proposed package under JKN. Detailed supply-side implications for the provision of the benefit package were derived from program-specific Puskesmas guidelines that detail equipment, diagnostic tests, and medicines that are stipulated. The paper also compares the supply-side implications from program-specific Puskesmas guidelines with those recommended by the WHO SARA framework. Although the WHO SARA framework specifies various elements required for a particular category of services – for example, cardiovascular conditions – it does not generally specify the exact combination of indicators required to treat a particular (stage of ) disease, nor does it generally specify alternatives nor capture the full diversity of potential treatment options (for example, there are a wide variety of anti-hypertensive medications in reality, but only a select few “tracer” drugs are chosen). Furthermore, due to limitations of the data sources, which were not initially conceived for this form of analysis, there are important omissions in indicators captured by the census and survey. For these reasons, it may be misleading to generate disease-specific composite measures (which require the presence of a set of indicators together), although in future iterations of data collection and analysis, this would be a valuable measurement. The lack of an exhaustive list of all available drugs and commodities at the health facility limits the interpretation of some aspects of the analysis. FIGURE 4: SUPPLY-SIDE ASSESSMENT FRAMEWORK %HQHȴW SDFNDJHIRU NH\WUDFHU NCDs 3URJUDP $VVHVVPHQW VSHFLȴF 6XSSO\VLGH XVLQJIDFLOLW\ 3XVNHVPDV LPSOLFDWLRQV DQGRWKHUGDWD guidelines :+2VHUYLFH DYDLODELOLW\ DQG UHDGLQHVV guidelines Supply-Side Readiness for Universal Health Coverage 23 In assessing the depth of UHC, it is important to link the provision of covered benefits to the concept of service delivery. Service delivery can be conceptualized as consisting of three key dimensions: (i) inputs; (ii) ability; and (iii) effort.14 This paper focuses primarily on “service delivery inputs”, in the terminology of the Service Delivery Indicators model, which is analogous to “service readiness” in the WHO SARA framework. Needless to say, even if service delivery inputs are available, the ability and efforts of providers – as well as the actions and behaviors of households and key interventions in other sectors – remain areas of consideration in ensuring that any intended reforms lead to improvements in NCDs and other health outcomes. The Service Delivery Indicators model is summarized in Figure 5. By converging elements from both the Service Delivery Indicators model, and the UHC model, this framework is hence an attempt at linking health financing structures to service delivery. FIGURE 5: SERVICE DELIVERY INDICATORS MODEL15 INPUTS EXPENDITURE ȏΖQIUDVWUXFWXUH ȏ(TXLSPHQW6XSSOLHV ȏ+XPDQ5HVRXUFHV PROVIDER ABILITY $$$$$ PROVIDER EFFORT OUTCOMES ȏ:KDWSURYLGHUVNQRZ ȏ:KDWSURYLGHUVGR In addition, in assessing the depth of UHC, it is important to note that the three dimensions of UHC (depth, breadth, and height) are not independent and mutually exclusive: ensuring depth of coverage has implications for the breadth and height of UHC as well. Universal availability of the benefit package for all – not just those who are well-off and live in urban areas – is a key aspect of ensuring that UHC is not a hypothetical aspiration but a realized policy designed to enhance health and improve social protection. Also, high OOP payments – that is, low height of UHC – can be (and often are) a result of poor depth of coverage if patients have to pay OOP for drugs or seek care elsewhere in private facilities that are outside the social health insurance network. 14 World Bank. 2012. Service Delivery Indicators Concept Note. Washington, DC: World Bank. 15 http://www.sdindicators.org/ Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 24 3.2 Data Sources The assessment of the depth of coverage reported in this policy note using the framework outlined above is based primarily on analysis of the 2011 Rifaskes facility census data, triangulated with information from the 2011 Health Facility Costing Study (HFCS) as well as evidence reported in recent literature where applicable. The Rifaskes facility data was collected by the Ministry of Health’s National Institute Health Research and Development Ministry in 2011. The inclusion criterion for the Rifaskes census was that the facilities had to be established prior to 2010. The response rate of 685 hospitals and 8,981 Puskesmas was a 100 percent by all eligible public health facilities owned by central, provincial, and district governments. The data includes information on human resources, medical equipment, organization and management, health services, and outputs of most essential health services. Data was collected through interviews, observations, and also from secondary sources. Independent validation of the Rifaskes data was conducted by three public health faculties in Indonesia: University of Indonesia, Airlangga University, and Hasanudin University. Where possible, basic triangulation of Rifaskes data was done with the HFCS dataset which was a nationwide costing study commissioned by the Ministry of Health in 2011, financed by the German Society for International Cooperation (GIZ) and the Australian Agency for International Development (AusAID) with technical support from Oxford Policy Management, GIZ, and Universitas Gadjah Mada. The facility survey sampled 234 Puskesmas, 120 government general hospitals, and 80 private general hospitals across 15 provinces in 30 districts/cities. The sample was selected using a stratified random sampling method so as to ensure national representativeness of the facility data. The primary objective of the GIZ health facility study was to calculate the cost of service provision. For this purpose, the study collected very detailed information about assets, drugs, equipment, and supplies from Q4 2010 until Q3 2011. Some modules were collected monthly; others were collected quarterly or annually. To ensure data quality, an independent verification team was established consisting of staff from four universities: University of Indonesia, Universitas Gadjah Mada, University of Airlangga, and University of Hasannudin. The richness of the database and the time period of data collection make it useful to triangulate with Rifaskes data. Indicators that overlap with Rifaskes in that the same piece of equipment or drug is reported available, are noted as footnotes in the paper where and if there are large discrepancies. Although a small number (80) of private general hospitals were sampled in the HFCS, this is not nationally representative and the sampling methodology and criteria were unclear. In addition, with respect to the national insurance scheme (BPJS), private providers are subject to an accreditation process, and hence deficiencies in their supply-side readiness would be expected to result in exclusion from the scheme (and in potential impact on population coverage, or breadth, if there are no alternative providers) rather than impact the depth (services) of coverage. A Public Expenditure Tracking Survey (PETS) covering health facilities is anticipated in the coming year. It is expected that the explanatory power of this service-readiness analysis can be enhanced using advanced statistical techniques that compare facilities’ expenditure and revenue data with their readiness. Annex A summarizes key information regarding both the Rifaskes and GIZ facility datasets used in the analysis reported in this policy note. Supply-Side Readiness for Universal Health Coverage SUPPLY SIDE READINESS ASSESSMENT FOR NON-COMMUNICABLE DISEASES 26 4. Supply-Side Readiness Assessment for Non-Communicable Diseases As mentioned above, cardiovascular diseases – including stroke and ischemic heart disease -- were the leading causes of the overall disease burden and the leading causes of premature mortality in Indonesia in 2010. Although the effective availability of treatment of stroke and acute heart disease events (such as heart attacks) is a key consideration in reducing the contribution of cardiovascular diseases to the overall disease burden, it is far more cost-effective and meaningful to focus on the ability of Indonesia’s health facilities to diagnose, treat, and manage chronic cardiovascular conditions at the primary care level. In addition to cardiovascular conditions, the policy note also assesses service availability and readiness of Indonesia’s primary care facilities to provide care for two additional prominent NCDs in the country: DM and chronic respiratory diseases (CRDs). 4.1 General Service Readiness The WHO SARA framework includes two domains: general service readiness and specific service readiness for specific diseases/conditions such as DM, cardiovascular conditions, and chronic respiratory care. General service readiness encompasses the basics required to provide any medical service, such as availability of water and a private room for consultations. Almost all Puskesmas in Indonesia had electricity and a private room for consultations. Basic communications such as telephones were present in 84 percent of Puskesmas, but computers with internet access were much less common, available only in 16 percent of Puskesmas (Figure 6). There is a nearly twofold variation between provinces with the lowest and highest general service readiness. In Papua, the mean basic amenities index was 46 percent compared with 89 percent for DI Yogyakarta (Figure 7).16 As almost all Puskesmas had electricity and a private room, most of the differences between provinces appear to be driven by the availability of water, toilets, basic communications, and referral transportation. As the primary role of Puskesmas in the management of NCDs is to provide a comfortable place for outpatient medical consultations, the near universal availability of electricity and a private room was encouraging. 16 The index is the average of available indicator divide by minimum indicator’s required; for example, the basic amenities index in province A was 50 percent, meaning that on average health centers in province A only had half of the stipulated minimum indicators. Supply-Side Readiness for Universal Health Coverage 27 FIGURE 6: BASIC AMENITIES IN PUSKEMAS IN INDONESIA Transportation 82.4 Computer + internet 16.4 Communication 83.8 Toilets 74.4 Private room 99.8 Water and sanitation 71.8 Electricity 97.9 0 10 20 30 40 50 60 70 80 90 100 Puskesmas has each component (%) Source: Rifaskes (2011) FIGURE 7: AVAILABILITY OF BASIC AMENITIES AT PUSKESMAS BY PROVINCE 100% 90% Percent Health Center Has Each Component 80% 70% 60% 50% 40% 30% 20% 10% 0% Province Basic amenities index Electricity Water and sanitation Private room Toilete Communication Computer+internet Transportation Source: Rifaskes (2011) Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 28 4.2 Human Resources for Health The Rifaskes census, which was not intended to specifically explore issues related to human resources for health (HRH), nevertheless includes several questions on the numbers and qualifications of staff working at the health facility as well as in-service training received. This information has been leveraged using the SARA framework in order to inform service readiness. However, it should be noted that additional dimensions – that is, provider ability (e.g., knowledge) and provider effort (e.g., presence at the health facility) – are not included in the Rifaskes census. Complementary work on HRH focuses on human resource inputs as well as some measure of health provider ability (using health cases or vignettes, covering prenatal care, child curative care, and adult curative care) from the 1997 and 2007 Indonesian Family Life Survey.17 Regarding adult curative care, of closest relevance to NCDs, the HRH study finds that over the 10 years from 1997 to 2007, ability scores have generally increased, that public providers have marginally higher scores than private providers, and that ability scores are marginally higher in urban compared with rural areas. 4.3 Diabetes Mellitus DM, a prominent chronic NCD, occurs when the body does not produce enough insulin (Type 1 DM) or the body is not able to effectively use the insulin that is produced (Type 2 DM).18 The result is raised blood sugar levels that, over time, damage nerves and blood vessels. Globally, 6.4 percent of the population aged 20-79 is estimated to be diabetic.19 There are no known preventive interventions for Type 1 DM, treatment of which requires insulin therapy; however, Type 2 DM – which accounts for 90% of all DM cases globally – can be prevented with weight management, physical activity, and a healthy diet. Of particular relevance to Indonesia, where the prevalence of undernutrition is high, there is an established association between low birth weight and Type 2 DM.20 The management of Type 2 DM and its complications involves lifestyle adjustments, oral medication, and/or insulin injections. DM is ranked as the eighth highest cause of the overall disease burden in Indonesia, accounting for 3.5 percent of DALYs in 2010 (Table 1).21 The burden of DM in Indonesia has more than doubled over the past 20 years (Figure 8). High fasting glucose levels as a risk factor accounted for 4.7 percent of all DALYs in the same year, while related risk factors – dietary risks, physical inactivity, and high body-mass index – are also prominent in Indonesia (Table 2). Riskasdes data estimated the prevalence of diagnosed DM as 0.7 percent in Indonesia in 2007.22 The prevalence estimate was higher (1.1 percent) 17 World Bank. 2010. New Insights in to the Provision of Health Services in Indonesia: A Health Workforce Study. Washington, DC: World Bank 18 Insulin is a hormone that regulates blood sugar. 19 Shaw, JE, RA Sicree, and PZ Zimmet. 2010. “Global estimates of the prevalence of diabetes for 2010 and 2030.” Diabetes Research and Clinical Practice 87(1): 4-14. 20 Taylor, A, AD Dangour, and K Srinath Reddy. 2013. “Only collective action will end undernutrition,” Lancet S0140- 6736(13): 61084-3; De Boo, AH and JE Harding. 2006. “The developmental origins of adult disease (Barker hypothesis).” Australian and New Zealand Journal of Obstetrics and Gynaecology 46 (1): 4–14. 21 A robust comparison or analysis of the risk factors and burden of NCDs is beyond the scope of this paper. 22 Indonesia Basic Health Research (Riskesdas) is routine survey conducted by NIHRD-MoH every 3 years. The sample size is around 240.000 household and the data represented national level and provincial level Supply-Side Readiness for Universal Health Coverage 29 when including those who were untreated but reported symptoms of diabetes. Blood glucose tests among the urban population aged 15 years and older indicated a prevalence rate of 5.7 percent, with an additional 10.2 percent having impaired glucose tolerance.23 In the provinces of West Papua and North Sulawesi, more than a quarter of urban residents aged 15 years and older were either diabetic or had impaired glucose tolerance. FIGURE 8: BURDEN OF DIABETES MELLITUS IN INDONESIA, 1990-2010 3.5 3.0 Share of DALYs lost (%) 2.5 2.0 1.5 1990 1995 2000 2005 2010 Year Source: IHME (2013) As summarized in Table 4, the supply-side implications for the diagnosis and treatment of DM at the primary care level were similar in both the government’s Puskesmas basic service standard guidelines and the WHO’s SARA toolkit, except that the former also specified availability of glipizid as being essential for DM-related care at Puskesmas. As information on staff training received for NCDs was not captured in Rifaskes, the paper uses a proxy measure as to the proportion of Puskesmas that had at least one doctor on staff. 23 Impaired glucose tolerance (and the related impaired fasting glycaemia) is an early stage of DM, where the formal diagnostic criteria have not been fully met. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 30 TABLE 4: SUPPLY-SIDE IMPLICATIONS FOR PROVISION OF DM-RELATED CARE AT THE PRIMARY LEVEL Diabetes guidelines for Indicators used for Domain WHO SARA guidelines (2012) Puskesmas (2013) & assessment essential drugs list (2011) Staff & • National guidelines available for • Guidelines for diabetes diag- • At least one doctor on staff Training diabetes diagnosis and treatment nosis and treatment at Puskesmas. • At least one staff member trained • Staff trained in diabetes diag- in diagnosis and treatment of nosis and treatment. diabetes in previous two years. Equipment • Blood pressure apparatus (digital • Blood pressure apparatus • Blood pressure apparatus blood pressure machine or man- • Stethoscope • Stethoscope ual sphygmomanometer with • Adult scale • Adult scale stethoscope) • Measuring tape • Measuring tape. • Adult scale • Height board/ stadiometre. • Measuring tape • Height board/stadiometre. Diagnostics • Blood glucose • Glucometer and strips • Blood glucose test • Urine dipstick (protein) • Urine dipstick. • Urine test. • Urine dipstick (ketones). Medicines & • Metformin cap/tab • Metformin cap/tab • Metformin cap/tab Commodities • Gilbenclamide cap/tab • Glibenclamide cap/tab • Gilbenclamide cap/tab • Insulin injectable • Glipizid. • Glucose injectable solu- • Glucose injectable solution. tion. Nationally, about 95.7 percent of all Puskesmas reported having a doctor on staff (94.8 percent of rural Puskesmas and 98.4 percent of urban Puskesmas). Across provinces, only in Papua was the availability of doctors at Puskesmas below a threshold level of 75 percent (Annex B). Regarding basic equipment for DM-related care, analysis of Rifaskes data indicates that the availability of blood pressure apparatus, stethoscopes, and adult scales was excellent across the country: over 95 percent of all Puskesmas in the country reported all three key pieces of equipment being available.24 There were no significant differences in the availability of blood pressure apparatus, stethoscopes, and adult scales by type and location of Puskesmas: whether rural versus urban, with and without beds, and across provinces. Only in the province of Maluku did only 73 percent of the 103 Puskesmas without beds have blood pressure apparatus, the only subcategorical grouping of Puskesmas within any province for which the reported availability was less than the benchmark threshold of 75 percent. On the other hand, the availability of measurement tape, to estimate waist circumference, was low: only 59 percent of Puskesmas nationally reported having this.25 Only in two provinces (DI Yogyakarta and East Java) was the availability of measurement tape in Puskesmas greater than 75 percent. 24 Over 80 percent of all Puskesmas in the HFCS facility sample reported having blood pressure monitors and adult weighing scales and 68 percent reported having height measurement/stadiometre. 25 A similar magnitude was reported in the HFCS facility sample. Supply-Side Readiness for Universal Health Coverage 31 TABLE 5: AVAILABILITY OF EQUIPMENT FOR DM-RELATED CARE AT PUSKESMAS Type of Puskesmas Equipment for DM at Puskesmas Blood pressure apparatus Stethoscope Adult scale Measuring tape Rural Puskesmas 95.4% 99.0% 97.6% 58.8% Urban Puskesmas 96.8% 99.6% 97.9% 58.5% Puskesmas without beds 94.8% 98.9% 97.4% 59.0% Puskesmas with beds 97.5% 99.5% 98.2% 58.2% Indonesia 95.8% 99.1% 97.7% 58.7% Source: Rifaskes (2011) The diagnostic and monitoring capacity of Puskesmas was very low: only 54 percent of Puskesmas nationally reported being able to conduct blood glucose tests, which are crucial in the management of DM, and only 47 percent reported being able to conduct urine tests.26 Urban Puskesmas were far more likely to report having diagnostic capacity to conduct both blood glucose and urine tests for DM, although even among urban Puskesmas the share reporting availability was below the threshold value of 75 percent (Table 6). The availability (or lack thereof ) of blood glucose and urine tests was correlated across provinces: provinces with higher capacity to conduct blood glucose tests also had a generally higher capacity to conduct urine tests (Figure 9). In Gorontalo, Papua, West Papua, Southeast Sulawesi, Maluku, North Sulawesi, East Nusa Tenggara, and West Papua, less than 25 percent of all Puskesmas reported being able to conduct either blood glucose or urine tests. In Gorontalo and North Sulawesi, two provinces where the urban prevalence of diabetes was estimated to be around 8 percent for those 15 years of age or older, less than 20 percent of urban Puskesmas reported being able to conduct blood glucose and urine tests. DI Yogyakarta and East Java were the only two provinces where more than 75 percent of all Puskesmas were able to diagnose diabetes using blood glucose and urine tests.27 TABLE 6: PERCENT OF PUSKESMAS REPORTING AVAILABILITY OF BLOOD GLUCOSE AND URINE TESTS FOR DIABETES Location Percent of Puskesmas with blood glucose test with urine test Rural 51.0% 43.0% Urban 70.0% 64.0% Indonesia 55.5% 48.0% Source: Rifaskes (2011) 26 The Rifaskes questionnaire does not differentiate between the different types of urine tests (e.g., for protein, ketones, or sug- ar). Almost all hospitals, on the other hand, reported being able to conduct urine tests. 27 In a 2007 study of 252 Puskesmas and 20 private health centers in 8 provinces (DKI Jakarta, West Java, North Sumatera, West Sumatera, DI Yogyakarta, East Java, Bali, and South Sulawesi) found similar results with 82.0 percent of health centers reporting availability of blood glucose tests and 76.3 percent reporting urinary tests; in addition, a third of general physicians and 78.9 percent of other health personnel had not received any training in type 2 diabetes mellitus; see Widyahening, IS and P Soewondo. 2012. “Capacity for management of type 2 diabetes mellitus in primary health centers in Indonesia.” Journal of Indonesian Medical Association 62(11): 439-443. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 32 FIGURE 9: AVAILABILITY OF BLOOD GLUCOSE AND URINE TESTS IN PUSKESMAS BY PROVINCE 100 80 Percent of puskesmas (%) 60 40 20 0 Blood glucose test Urine test Regarding medicines and commodities related to DM, about 80 percent of all Puskesmas nationally reported having glucose injectable solution, used to treat low glucose levels occasionally caused by treatment. Rifaskes did not collect information on availability of oral diabetic medications such as metformin or glibenclamide, or injectable insulin at the Puskesmas level. In the HFCS facility sample, only about half of all Puskesmas reported having metformin, while glibenclamide was available in almost 90 percent of all Puskesmas in the sample.28 4.4 Cardiovascular Conditions Relevant cardiovascular conditions for Indonesia include hypertension, coronary heart disease, cerebrovascular diseases, and peripheral arterial disease. In addition to genetic dispositions, risk factors for cardiovascular conditions include unhealthy diets, physical inactivity, and smoking. As a result, individuals often exhibit “intermediate” risk factors such as hypertension, DM, high total cholesterol, and high body mass index. Prevention and treatment of hypertension, DM, and high total cholesterol – in addition to lifestyle change interventions – are key in preventing and managing these cardiovascular conditions. Although, as mentioned earlier, the SARA framework does not generally differentiate individual diagnoses like hypertension and hypercholesterolemia, but instead focuses on a cluster of conditions – without disaggregating service indicators by diagnoses – for the purposes of this report, three common diagnoses are highlighted: (i) hypertension (the main focus, due to relevance to Indonesia and availability of data), (ii) coronary artery disease (which can result in a “heart attack” or myocardial infarction), (iii) congestive heart failure, and (iv) hypercholesterolemia. 28 These magnitudes are similar to those reported elsewhere for Indonesia; see: Cameron, A. 2013. Understanding Access to Medi- cines in Low and Middle-Income Countries through the use of Price and Availability Indicators. Utrecht: Gildeprint Drukkerijen. Supply-Side Readiness for Universal Health Coverage 33 Hypertension (i.e., high blood pressure) is defined as systolic blood pressure equal to or above 140 mm Hg and/or diastolic blood pressure equal to or above 90 mm Hg. Normal levels of blood pressure are essential for efficient functioning of organs and body functions. Although some hypertensive people exhibit symptoms such as headaches, dizziness, palpitations, and nose bleeds, most people with hypertension do not exhibit any symptoms at all, making regular screening essential for detection and control. The WHO estimates that globally about 40 percent of all adults over the age of 25 suffer from hypertension.29 Unhealthy diets (including diets high in salt and fat, and diets low in fruits and vegetables), physical inactivity, and stress are prominent lifestyle risk factors for hypertension. Genetic factors can play a role, and hypertension can also be a result of other conditions such as during pregnancies and due to endocrine dysfunctions. Untreated hypertension increases cardiovascular risk and is a major risk factor for stroke as well as for ischemic heart disease.30 The prevalence of primary hypertension is high in Indonesia.31 The national prevalence of hypertension among adults aged 18 years or older was estimated to be 31.7 percent in 2007.32 There was a wide variation in the prevalence of hypertension across provinces in the country: West Papua and Papua had generally lower prevalence rates, while South Kalimantan and East Java had higher prevalence rates (Figure 10). An estimated 10 percent of the overall burden of disease and 370,314 deaths were attributable to high blood pressure in Indonesia in 2010. WHO SARA guidelines recommend that, in order for facilities to be able to diagnose and treat chronic cardiovascular conditions: (i) at least one staff member in the facility should have received training for diagnosis and treatment of chronic cardiovascular conditions in the previous two years, and national guidelines for the diagnosis and treatment of chronic cardiovascular conditions be available at the facility; (ii) health facilities should have, at the minimum, stethoscopes, functional digital blood pressure machines or manual sphygmomanometers with stethoscopes, and adult scales; as well as (iii) angiotensin converting enzyme (ACE) inhibitors, thiazides, beta blockers, calcium channel blockers, aspirin, metformin, and oxygen. Hypertension medications as specified in the basic treatment guidelines for Puskesmas and in the formulary include: hydrochlorothiazide, reserpine, propranolol, captopril, and nifedipine. Guidelines for Puskesmas also specify that each facility should have “blood pressure apparatus”.33 The guidelines furthermore mention that treatment of chronic cardiovascular conditions includes management of hypertension and heart failure. In addition, the National Guideline for Controlling Risk Factors of Cardiovascular Diseases includes coronary artery disease for treatment in Puskesmas. 29 WHO. 2013. A Global Brief on Hypertension: Silent Killer, Global Public Health Crisis. Geneva: World Health Organization. 30 Ibid. 31 “Primary” hypertension is defined as hypertension that is not the result of some other underlying condition; in case of the latter, the term “secondary” hypertension is used. 32 NIHRD. 2009. Report on Result of National Basic Health Research 2007. Jakarta: National Institute of Health Research and Development, Ministry of Health. 33 MOH. 2007. Guideline for Basic Medical Intervention at Health Center, Ministry of Health Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 34 FIGURE 10: HYPERTENSION PREVALENCE AMONG ADULTS AGED 18+ YEARS BY PROVINCE IN INDONESIA, 2007 40 Indonesia national prevalence 31.7% 30 Prevalence (%) 20 10 0 While diagnosis of coronary artery disease (CAD) may rely mostly on symptoms and simple equipment such as blood pressure apparatus and stethoscopes, confirmation of CAD requires referral to secondary care facilities. However, it is expected that primary providers, including Puskesmas, can provide emergency measures if a cardiovascular event is suspected. Isosorbide dinitrate, aspirin, nitroglycerine, and diltiazem are used for anginal attack in CAD before referral. With reference to congestive heart failure, digoxin and furosemide are commonly used for managing mild congestive heart failure on an ambulatory basis. These additional drugs have been added to the list of specific service- readiness guidelines summarized in Table 7. Supply-Side Readiness for Universal Health Coverage 35 TABLE 7: SUPPLY-SIDE IMPLICATIONS FOR PROVISION OF CARDIOVASCULAR CONDITIONS AT THE PRIMARY LEVEL Cardiovascular guidelines Indicators used for Domain WHO SARA guidelines for Puskesmas (2013) & assessment essential drugs list (2011) Staff & Training • National guidelines available • Guidelines for cardiovascular • At least one doctor on for diagnosis and treatment of diagnosis and treatment staff at Puskesmas. chronic cardiovascular condi- • Staff trained in cardiovascular tions diagnosis and treatment. • At least one staff member trained in diagnosis and man- agement of chronic cardiovas- cular conditions in previous two years. Equipment • Stethoscope • Stethoscope • Stethoscope • Blood pressure apparatus • Blood pressure apparatus • Blood pressure apparatus • Adult scale. • Adult scale. • Adult scale. Diagnostics -- -- -- Medicines & • ACE inhibitor (e.g., enalapril) • ACE inhibitors (captopril) • ACE inhibitor (captopril) Commodities • Thiazides • Thiazides (hydrochlorothia- • Thiazide (hydrochlorothi- • Beta blockers (e.g., atenolol) zide) azide) • Calcium channel blockers (e.g., • Beta-blocker (atenolol, pro- • Beta blocker (propranolol) amlodipine) pranolol) • Calcium channel blockers • Aspirin • Calcium-channel block- (nifedipine) ers(nifedipine) • Metformin • Metformin • Aspirin • Oxygen. • Oxygen cylinders or concentra- tors. • Diltiazem • Amlodipine • Furosemide • Isosorbid dinitrate • Nitroglycerine • Digoxin • Simvastatin. Rifaskes data indicate a generally strong readiness among most Puskesmas in Indonesia to diagnose and treat primary hypertension. Over 95 percent of all puskemas nationally reported having functional blood pressure apparatus, over 99 percent reported having a stethoscope, and 84 percent reported the availability of captopril, a common (but not the only) medication for treating hypertension.34 Availability was high even in rural Puskesmas: while 86 percent of urban Puskesmas reported availability of captopril, the corresponding number for rural Puskesmas was 83 percent (Table 8). Nevertheless, some provincial- level deficiencies were significant, especially with regard to reported availability of captopril: for example, less than 75 percent of Puskesmas in West Sulawesi, Maluku, Papua, East Nusa Tenggara, North Maluku, and Central Sulawesi reported availability of captopril (Table 8). 34 WHO reported the availability of captopril in Indonesia in 2004 as 33 percent (Originator) and 93 percent (Generics) in public sectors; For more details see: WHO. 2006. Gelders S, Ewen M, Noguchi N, Laing R. Price, availability and affordabil- ity. An international comparison of chronic disease medicines. Cairo: World Health Organization Regional Office for the Eastern Mediterranean; Captopril was available in over 95 percent of the hospitals in Indonesia. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 36 TABLE 8: PROVINCES WITH <75% OF ALL PUSKESMAS REPORTING CAPTOPRIL AVAILABILITY Province Proportion of Puskesmas reporting availability of captopril Total Rural West Sulawesi 50.0% 51.4% Maluku 63.1% 59.3% Papua 66.6% 63.1% East Nusa Tenggara 70.4% 70.6% North Maluku 74.0% 68.5% Central Sulawesi 74.1% 72.1% Indonesia 83.8% 83.1% Source: Rifaskes (2011) Additional data on medications was extracted from HFCS in order to provide a picture of drug availability for other cardiovascular diagnoses. With regard to congestive heart failure (there is some overlap or dual-use of drugs with hypertension), data shows that furosemide was available in 84 percent of Puskesmas, isosorbide dinitrate in 60 percent, digoxin 72 percent, furosemide 83 percent, thiazide 79 percent, nifedipine 60 percent, isosorbide dinitrate 60 percent, propanolol 52 percent, and diltiazem 22 percent. Regarding hypercholesterolemia, HFCS data shows that simvastatin is available in 36 percent of Puskesmas and higher in urban (50 percent) than rural Puskesmas (27 percent). It is more available in the Java-Bali (42 percent) region than non-Java-Bali regions (32 percent). Although the ability to conduct cholesterol screening was not included in either the WHO’s SARA guidelines or in Indonesia’s national cardiovascular guidelines, it is notable that nationally only about 35 percent of Puskesmas report being able to conduct cholesterol screening tests (31 percent of rural Puskesmas and 44 percent of urban Puskesmas reported being able to do so). Less than 10 percent of Puskesmas in the provinces of Maluku, Gorontalo, East Nusa Tenggara, Southeast Sulawesi, and Papua reported being able to test for cholesterol. 4.5 Chronic Respiratory Conditions Chronic respiratory diseases – asthma, chronic obstructive pulmonary disease (COPD), and other diseases of the airways and lung structures – are another growing NCD challenge; Asthma and COPD accounted for an estimated 1.0 percent and 2.6 percent of all DALYs lost in 2010 in Indonesia (up from 0.8 percent and 1.7 percent, respectively, in 1990) (Table 9). Riskesdas estimated the national prevalence of asthma to be 3.5 percent in 2007, with the provinces of Gorontalo, Central Sulawesi, West Papua, and South Kalimantan having prevalence rates in excess of 5.0 percent of the population. TABLE 9: SHARE OF DALYS LOST FROM CHRONIC RESPIRATORY DISEASES IN INDONESIA, 1990-2010 Disease/condition Share of DALYs lost 1990 2000 2010 Chronic respiratory diseases 3.4% 4.2% 4.7% Asthma 0.8% 1.0% 1.0% COPD 1.7% 2.2% 2.6% Source: IHME (2013) Supply-Side Readiness for Universal Health Coverage 37 Several risk factors for chronic respiratory diseases – including smoking as well as household air pollution – are prominent in Indonesia (Table 2). The prevalence of smoking among males in Indonesia is particularly high: the latest 2011 Global Adult Tobacco Survey (GATS) estimated that two out of every three adult Indonesian males smoked tobacco (versus only 2.7 percent of adult females) and this ratio has been increasing over time.35 The prevalence of smoking was generally higher in rural areas: 37.7 percent of all adults smoked in rural areas versus 31.9 percent in urban areas. Kretek cigarette smoking was especially popular in the country, and is the dominant modality of tobacco consumption in both urban and rural areas. Exposure to second-hand smoke was also a large problem in the country: 51.3 percent of adults who worked indoors and 78.4 percent of adults at home reported being exposed to second-hand smoke.36 Table 10 summarizes the supply-side implications for provision of chronic respiratory disease- related care at the primary level derived from SARA and the Puskesmas guidelines. Facility data were available to assess the availability of stethoscopes, salbutamol, prednisone or prednisolone, and oxygen in Puskesmas across the country. TABLE 10: SUPPLY-SIDE IMPLICATIONS FOR PROVISION OF CHRONIC RESPIRATORY DISEASE-RELATED CARE AT THE PRIMARY LEVEL Puskesmas guidelines for Indicators used for Domain WHO SARA guidelines CRD (2011) & essential assessment drugs list (2011)6 Staff & Training • At least one staff received training Minimum competency level • At least one doctor on in the diagnosis and management defined. staff at Puskesmas. of chronic respiratory diseases in the last two years • National guidelines available for the diagnosis and management of chronic respiratory diseases. Equipment • Stethoscope -- • Stethoscope. • Peak flow meter • Spacers for inhalers. Diagnostics -- • Hemoglobin. -- Medicines & • Salbutamol inhaler • Salbutamol • Salbutamol Commodities • Beclomethasone inhaler • Prednisolone cap/tabs • Prednisolone • Prednisolone cap/tabs • Epinephrine injectable • Prednisone • Hydrocortisone cap/tabs • Oxygen • Amenophylline • Epinephrine injectable • Terbutaline • Oxygen. • Oxygen concentrators or cylinders. • Aminophylline • Bronchodilators. 35 GATS 2011. 36 Ibid. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 38 In general, at the national level, the availability of the selected indicators appeared to be good: 99 percent of Puskesmas reported having stethoscopes; and over 75 percent of all Puskesmas reported having salbutamol, prednisolone, and oxygen.37 Again, some provincial-level deficiencies were notable: for example, in the provinces of Central Sulawesi, West Sulawesi, and Papua, less than 75 percent of Puskesmas reported availability of salbutamol and prednisolone. HFCS data shows that prednisolone was available in more than 94 percent of Puskesmas. In addition salbutamol was available in 76 percent of Puskesmas, but aminophylline (another bronchodilator) was available in 90 percent of Puskesmas. 37 Oxygen was relevant only for Puskesmas with beds, with 80.7 percent of 5,639 Puskesmas with beds reported availability. Supply-Side Readiness for Universal Health Coverage CONCLUSIONS 40 5. Conclusions Indonesia is currently in the midst of major reforms of its health system that, have resulted in the institutionalization of one of the largest single-payer UHC programs in the world. At the same time, the country is undergoing a rapid epidemiological transition and NCDs are now the dominant share of the overall disease burden in the country. Indonesia faces several challenges to effective implementation of UHC in terms of expanding the breadth, height, and depth of coverage, especially in addressing NCDs that are generally chronic in nature and require careful disease management over time. Given this backdrop, this policy note has outlined a framework for assessing the “effective depth” of coverage in terms of supply-side readiness to diagnose and manage key tracer NCD conditions at the public primary care level: DM, cardiovascular conditions, and chronic respiratory conditions. Regarding supply-side readiness, although DM is the eighth (and rising) cause of the overall disease burden in Indonesia, the service readiness for basic diagnosis and basic treatment of DM is limited. For example, only 70 percent of urban Puskesmas (where the prevalence of DM is higher) could conduct a blood glucose test for the diagnosis and monitoring of DM. In rural Puskesmas, this readiness indicator declined to 51 percent nationwide and in some provinces, such as Gorontalo, Papua, Southeast Sulawesi, Maluku, North Sulawesi, and West Papua, less than 20 percent of Puskesmas fulfilled this indicator. Basic pharmaceutical treatment for DM is more encouraging, with 90 percent of Puskesmas stocking glibenclamide but only 48 percent stocking metformin. As over 95 percent of Puskesmas reported the availability of functioning blood pressure apparatus and over 99 percent reported the same of stethoscopes, most Puskesmas have the necessary equipment to diagnose hypertension. The availability of medical treatment for hypertension using captopril, a commonly used antihypertensive medication in Indonesia, is fair. Nationwide, 84 percent of Puskesmas stocked this medication, although deficiencies were notable in rural Puskesmas in provinces such as West Sulawesi, Maluku, and Papua, where captopril was available in less than 70 percent of Puskesmas. Furthermore, in the case of hypercholesterolemia, only a third of Puskesmas nationally reported the ability to conduct cholesterol tests and only 36 percent reported the availability of simvastatin, a common treatment for hypercholesterolemia. In the case of chronic respiratory diseases such as asthma, over 75 percent of all Puskesmas reported the availability of basic treatments such as salbutamol, prednisolone, and oxygen. There are geographic variations, with less than 75 percent of Puskesmas in Central Sulawesi, West Sulawesi, and Papua reporting availability of salbutamol and prednisolone. A basic index of NCD service readiness indicators can be constructed by combining all available NCD-related service readiness indicators. To construct this, the mean availability (or “score”) of all available NCD-related service readiness indicators was calculated. Next, the mean of all these scores was calculated for all facilities. Each indicator and each facility was weighted equally, as there is no available valid method at the moment for meaningfully weighting each individual indicator, and there is not enough information for the population catchment coverage of each facility to generate population weights Supply-Side Readiness for Universal Health Coverage 41 (as is done for Service Delivery Indicators). Using this basic methodology, provincial-level NCD service readiness can be mapped (Figure 11), revealing that in the eastern part of the country and representing 15 percent of the country’s total population and almost a fifth of all of Indonesia’s poor and near-poor residents (in North Sumatera, Bengkulu, East Nusa Tenggara, North Sulawesi, Central Sulawesi, Southeast Sulawesi, Gorontalo, West Sulawesi, Maluku, North Maluku, West Papua, and Papua) there are substantial deficiencies (using an index of 75 as the cut-off) in NCD service provision. Of these, provinces with the lowest levels of NCD supply readiness were Papua, West Papua, and Maluku (Annex B). FIGURE 11: NCD SERVICE READINESS INDICATOR INDEX, BY PROVINCE, 2011 NCD supply-side readiness index, by province NCD index [0,65] [65,75] [75,85] [85,100] Source: Rifaskes 2011 FIGURE 12: PREVALENCE OF DIABETES, BY PROVINCE (RISKASDES, 2007) Diabetes prevalence, by province Diabetes prevalence (%) [0,5] [5,10] [10,15] [15,20] Source: Riskesdas 2007 Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 42 NCD supply readiness is correlated to economic indicators: richer provinces with generally lower poverty rates also tended to have better NCD supply readiness, although there are some exceptions to this (Figure 12). Provinces with relatively poor NCD supply readiness (NCD index < 75%) also tended to be those where distances to hospitals was higher than the national average and the availability of HRH was low, indicating broader problems related to service delivery and not just for NCDs. FIGURE 13: NCD SERVICE READINESS INDICATOR INDEX VS. SELECTED ECONOMIC INDICATORS (PROVINCIAL-LEVEL) 100 100 DI Yogyakarta DI Yogyakarta East Java 90 90 East Java Central Java Central Java NCD supply-readiness index NCD supply-readiness index 0=Lowest 100=Highest 0=Lowest 100=Highest 80 Indonesia 80 Indonesia Bali Bali DI Aceh DI Aceh Bengkulu Bengkulu Gorontalo East Nusa Tenggara 70 East Nusa Tenggara 70 Gorontalo North Sulawesi North Sulawesi West Papua West Papua Maluku Maluku 60 Papua 60 Papua 50 50 50.00 60.00 70.00 80.00 90.00 100.00 0.00 10.00 20.00 30.00 40.00 Annual per capita consumption (’000s IDR) Poverty rate (%) Source: NCD supply: Rifaskes,2011; Economic Indicators: Susenas 2010; Poverty rate: BPS (Central Bureau of Statistics), 2010 Deficiencies appear not to be a problem at the hospital level across all provinces in Indonesia (Annex B), at least based on the indicators that were available for NCD tracers from the hospital census. In addition, it is important to note that this analysis is limited to government facilities, but in reality, private primary care providers play a large and increasing role. There is no systematic information about service readiness in the private sector. Although it may not be representative of all private primary care providers, analysis of the profiles of family doctors both in the Askes and Jamsostek programs would help to provide some understanding of the service readiness of private providers. Supply-Side Readiness for Universal Health Coverage POLICY IMPLICATIONS 44 6. Policy Implications The analysis reported in this policy note underscores that in order to attain UHC there needs to be a focus not just on increasing the breadth and height but also in ensuring that effective depth of coverage exists, especially in the rural and remote areas of the country and at the primary care level. It is not enough to specify a comprehensive benefit package on paper if facilities do not have the basic supply-side readiness conditions to be able to provide key services.38 In addition to identifying where and along what dimensions the key supply-side deficiencies exist, it is also important to better understand why this is the case. Some policy implications to ensure this are summarized below. Understand Why Supply-Side Deficiencies Exist. The analysis above has utilized Rifaskes and other facility data to demonstrate where some of the deficiencies in supply-side readiness exist at the public primary care level in Indonesia and looked at some broad province-level correlates. One key next step would be to gain a better and more systematic understanding of why these deficiencies exist and what can be done to overcome them. Are supply-side deficiencies a result of data collection and reporting deficiencies, lack of financing, information flow issues, and/or other forms of implementation challenges? In this regard, proposed provincial consultations and analytical work related to assessment of public financial flows at different levels of government (central, provincial, district) and health facilities could provide an excellent opportunity to help fill in some of the gaps. Leverage Demand-Side Financing to Improve Service Readiness at Facilities. At the primary care level, where the predominant provider payment method for health facilities is capitation, this payment should be linked directly or indirectly to the attainment of minimum standards analogous to the accreditation of private facilities. Although the expediency with which the government is expanding geographic coverage using public facilities is understandable, the insurance administrator could assess public facilities in order to signal to local governments where investments are required. Initial standards could start at a basic level and rise as the program develops, with the credible threat that facilities could be deregistered from the program if standards are not met, thus allowing and incentivizing facilities to upgrade their service readiness. As financing gradually shifts from supply-side to demand-side, an appropriate level of autonomy for health facilities, coupled with enhanced managerial capacity to manage revenues, would need to be found for public health facilities. At the secondary level, DRG-related payments could be made conditional on the adequacy of services provided in order to encourage investments in improving service readiness. As the health system evolves, additional measures to mitigate negative incentives inherent in capitation systems – such as over and inappropriate referral to secondary care and undertreatment – are likely to be required. 38 Indonesia is not alone in this regard. In general, there are clear deficiencies in the supply-side readiness of primary care facil- ities in providing NCD-related care across most low- and middle-income countries. See: WHO. 2010. Global Status Report on Non-Communicable Diseases. Geneva: World Health Organization. Supply-Side Readiness for Universal Health Coverage 45 Clarify Supply-Side Implications of Benefit Package. As the analysis reported in this policy note indicates, it is not enough to specify a benefit package on paper; in moving forward to implementing UHC, Indonesia needs to ensure that there are clear supply-side implications that are derived from the specification of the unified BPJS Kesehatan benefit package, in terms of what equipment, training, diagnostic capabilities, and medicines are to be provided at different levels of care and to clearly specify accountabilities for this provision. In this regard, BPJS may want to reconsider its plans for automatic accreditation of public facilities and to consider an independent accreditation process for public facilities as the program develops. Improve Socialization of Benefit Entitlements. Effective implementation of UHC will also require patients to be clear as to what benefits they are entitled to, and providers to be clear as to the minimum requirements for provision of services. For NCDs, Indonesia does have comprehensive guidelines in place; the issue is more of ensuring that these guidelines are widely disseminated to and internalized by providers. An electronic form of disseminating the latest clinical guidelines for NCDs (and other conditions) may hence be useful. Regular Independent Monitoring and Evaluation of Supply Readiness. The MoH should consider institutionalized collection of regular and relevant facility-level data (from a sample of facilities, including private sector facilities, as the private sector contributes to a large proportion of health service provision) and ensure that the data collected reflects national guidelines and norms and can help shed light not just on where the deficiencies lie but also why they exist. Providers of traditional, complementary, and alternative therapy should also be considered for inclusion in the collection of health facility data, in order to capture the unique nuances of such providers, and to explore interrelations between these providers and allopathic providers. It is understood that preparations are currently underway by the MoH to conduct a follow-up health facility survey, which will complement a further Public Expenditure Tracking Survey of health facilities. Facility data collection should be independent and, ideally, separate from routine administrative data monitoring. If possible, data should also be collected from a sample of beneficiaries to ensure that UHC service provision is occurring as intended and that patients are receiving the care they are entitled to. This will be especially critical over the period of expansion of UHC from 2014 to 2019. Invest in Service Readiness for NCDs. In order for investments in the expansion of the breadth and height of UHC to be fully realized, the depth of coverage for NCDs needs to be addressed: to maximize efficiency gains from the prevention and prudent management of these chronic diseases and conditions by minimizing their costly and harmful complications. Findings from this health facility survey and the additional surveys mentioned earlier should shed light on the appropriate balance of investments flowing from the supply side versus financing received from the demand side. However, regardless of the exact mechanism of financing used, it should be noted that public health spending in Indonesia is comparatively low considering its economic status, and an expansion in public health financing generally would be helpful, not only to ensure the long-term sustainability of BJPS Kesehatan, but also to improve the health and productivity of the Indonesian population. Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia 46 Address Risk factors for NCDs Holistically. Although the scope of analysis provided in this paper is focused on facility service readiness, many of the risk factors for NCDs (summarized in Table 2), such as dietary risks, smoking, indoor air pollution, and physical inactivity, should be addressed through public health interventions beyond the confines of health facilities and indeed, beyond the health sector alone. Various cross-sectoral interventions will need to be considered – including taxation policy for tobacco, interventions to promote physical activity at schools and in the workplace, the designs of urban spaces, and environmental and energy policy. Supply-Side Readiness for Universal Health Coverage ANNEXES 48 ANNEX A Rifaskes HFCS Year of data collection 2011 2010-11 Sampling National census of all Puskesmas and Nationally representative survey government general hospitals (stratified random sampling) in 15 provinces, 30 districts Government hospitals 685 120 Private hospitals - 80 Number of Puskesmas, of which 9,005 (8,981 used for analysis) 234 (217 used for analysis) Urban Puskesmas 2,364 82 Rural Puskesmas 6,617 135 Puskesmas with beds 3,052 90 Puskesmas without beds 5,929 127 Provincial7 DI Aceh 311 North Sumatera 506 West Sumatera 248 Riau 195 Jambi 171 South Sumatera 298 Bengkulu 173 Lampung 265 Bangka Belitung 57 Riau Island 65 DKI Jakarta 336 West Java 1031 Central Java 861 DI Yogyakarta 121 East Java 949 Banten 206 Bali 114 West Nusa Tenggara 149 East Nusa Tenggara 302 West Kalimantan 233 Central Kalimantan 176 South Kalimantan 217 East Kalimantan 213 North Sulawesi 167 Central Sulawesi 163 South Sulawesi 406 Southeast Sulawesi 233 Gorontalo 74 West Sulawesi 81 Maluku 161 North Maluku 101 West Papua 104 Papua 294 Supply-Side Readiness for Universal Health Coverage ANNEX B NCD Province Equipment Diagnostics Medicines & Commodities index At least Blood Blood Glucose Stetho- Adult Measur- Salbu- Predniso- Puskesmas one doc- pressure glucose Urine test injectable Captopril Oxygen Average scope scale ing tape tamol lone tor apparatus test solution DI Aceh 98.7% 94.9% 100.0% 98.4% 41.8% 53.7% 49.2% 83.0% 81.4% 78.6% 85.5% 85.2% 77.4% North Sumatera 95.1% 89.7% 98.4% 94.5% 44.4% 33.8% 16.0% 76.7% 85.0% 73.6% 73.1% 80.0% 69.6% West Sumatera 98.8% 98.8% 100.0% 96.0% 60.9% 82.7% 68.5% 75.0% 91.5% 88.5% 89.1% 93.5% 85.9% Riau 100.0% 95.4% 100.0% 96.4% 59.5% 60.5% 56.4% 88.2% 91.8% 91.4% 85.6% 88.1% 83.0% Jambi 97.7% 96.5% 98.8% 98.2% 53.8% 52.6% 60.8% 86.5% 84.7% 86.4% 61.2% 94.1% 79.4% South Sumatera 91.9% 94.3% 100.0% 97.7% 57.4% 43.6% 46.6% 82.2% 91.9% 86.6% 85.9% 78.5% 78.6% Bengkulu 96.5% 98.3% 100.0% 99.4% 54.9% 27.2% 16.2% 83.2% 79.2% 82.5% 81.5% 85.0% 73.4% Lampung 100.0% 91.3% 99.2% 98.9% 64.4% 49.8% 43.0% 80.8% 87.5% 75.0% 76.6% 92.8% 78.1% Bangka Belitung 100.0% 100.0% 100.0% 96.5% 57.9% 57.9% 56.1% 80.7% 87.7% 90.0% 82.5% 93.0% 82.0% Riau Island 100.0% 93.8% 100.0% 98.5% 49.2% 64.6% 60.0% 95.4% 92.3% 83.3% 86.2% 95.4% 83.5% DKI Jakarta 98.8% 100.0% 100.0% 100.0% 64.4% 76.1% 65.2% 76.1% 82.6% 71.7% 80.4% 91.3% 82.5% West Java 98.0% 97.5% 99.9% 99.1% 66.0% 53.5% 52.3% 77.2% 87.3% 84.3% 82.9% 95.1% 81.4% Central Java 99.4% 99.4% 100.0% 99.9% 71.0% 88.2% 72.1% 76.2% 85.9% 93.9% 86.5% 94.6% 88.0% DI Yogyakarta 100.0% 100.0% 100.0% 100.0% 78.5% 100.0% 100.0% 81.8% 96.7% 97.7% 87.6% 98.3% 94.6% East Java 99.2% 99.3% 100.0% 99.8% 84.3% 83.5% 77.1% 83.8% 89.7% 91.8% 82.2% 89.3% 89.2% Banten 98.1% 99.0% 100.0% 99.5% 66.0% 66.5% 35.4% 80.1% 85.9% 95.9% 86.8% 77.9% 81.2% Bali 100.0% 99.1% 100.0% 98.2% 43.0% 82.5% 44.7% 67.5% 85.1% 70.0% 84.2% 94.7% 79.0% West Nusa Tenggara 98.7% 100.0% 100.0% 99.3% 49.0% 72.5% 79.2% 91.3% 87.2% 79.5% 83.8% 89.2% 84.6% Assessing the Depth of Coverage for Non-Communicable Diseases in Indonesia East Nusa Tenggara 95.4% 95.7% 99.7% 98.3% 55.8% 16.6% 21.5% 93.4% 70.4% 67.2% 63.3% 84.7% 69.7% West Kalimantan 86.3% 94.8% 100.0% 98.3% 51.3% 60.5% 49.4% 82.8% 75.1% 85.9% 79.0% 86.3% 78.5% Central Kalimantan 94.3% 97.2% 100.0% 98.3% 59.4% 39.2% 31.8% 90.3% 84.7% 80.0% 75.0% 86.9% 76.6% South Kalimantan 98.6% 99.1% 100.0% 99.5% 47.2% 86.6% 72.8% 87.6% 86.6% 91.5% 83.4% 91.2% 86.0% East Kalimantan 100.0% 97.2% 99.5% 99.1% 51.7% 68.5% 47.4% 76.5% 79.3% 84.6% 90.1% 92.0% 80.5% North Sulawesi 97.0% 89.8% 99.4% 94.6% 49.7% 16.2% 5.4% 69.5% 82.0% 61.5% 79.0% 89.2% 66.9% Central Sulawesi 89.0% 96.3% 98.8% 97.5% 48.1% 29.4% 26.4% 72.4% 74.1% 78.7% 56.5% 69.1% 67.9% 49 50 NCD Province Equipment Diagnostics Medicines & Commodities index South Sulawesi 95.3% 97.8% 99.8% 98.5% 41.4% 59.1% 61.6% 80.0% 80.8% 78.3% 73.4% 88.2% 78.1% Southeast Sulawesi 90.6% 94.4% 97.4% 93.1% 48.7% 14.6% 6.4% 82.8% 84.4% 65.6% 73.2% 84.0% 67.7% Gorontalo 93.2% 97.3% 100.0% 98.6% 58.1% 12.2% 2.7% 87.8% 90.5% 71.4% 59.5% 79.7% 68.9% West Sulawesi 95.1% 97.5% 98.8% 98.8% 39.2% 33.3% 37.0% 77.8% 50.0% 75.7% 59.5% 63.7% 66.5% Maluku 90.7% 78.5% 95.5% 90.4% 42.4% 15.5% 7.5% 78.0% 63.1% 50.9% 63.9% 79.1% 60.4% North Maluku 85.1% 93.1% 96.0% 95.0% 57.3% 30.7% 12.9% 86.1% 74.0% 58.6% 75.0% 86.0% 69.5% West Papua 83.7% 84.6% 90.4% 93.3% 34.0% 16.3% 16.3% 81.7% 77.9% 46.5% 76.0% 84.6% 63.8% Papua 67.3% 82.9% 90.1% 87.4% 43.0% 13.3% 10.5% 69.3% 66.6% 41.9% 63.9% 72.9% 58.3% Urban Puskesmas 98.4% 97.0% 99.5% 97.9% 60.2% 63.5% 57.3% 66.3% 85.8% 74.5% 83.8% 88.9% Rural Puskesmas 94.8% 95.4% 98.9% 97.6% 58.8% 51.0% 43.0% 84.0% 83.1% 81.9% 77.2% 87.4% Puskesmas with beds 98.0% 97.5% 99.5% 98.2% 58.2% 65.2% 59.1% 85.3% 83.2% 80.7% 77.9% 86.7% Puskesmas without 94.6% 95.0% 98.9% 97.4% 59.6% 48.6% 40.4% 76.3% 84.1% - 79.5% 88.3% beds All Puskesmas 95.7% 95.8% 99.1% 97.7% 59.2% 54.3% 46.8% 79.4% 83.8% 80.7% 79.0% 87.8% 78.8% All hospitals 100.0% 100.0% - 97.8% - - 93.6% - 92.7% 100.0% 93.1% 95.9% - (Footnotes) 1 Ministry of Health Decree 416 (2011) on Health Service Tariffs for Members of Askes 2 Ministry of Health Decree 40 (2012) on Guidelines for the Implementation of the Jamkesmas Program 3 Presidential Decree No. 12 Year 2013 on Health Insurance 4 BJPS Kesehatan members register at a primary care facility of their choice, which plays a gate-keeper role for referrals to secondary care. The system can only be bypassed in the case of emergencies (including to non-empanelled facilities) or if the member is outstation. Non-empanelled facilities are reimbursed on ‘replacement costs’ for emergency treatment of BPJS Kesehatan members. 5 The Ministry of Health is mandated to conduct Health Technology Assessments (HTAs) and provide clinical guidelines 6 In the Puskesmas guidelines (2007) the role of Puskesmas is to diagnose “clinical COPD”, while its severity defined at higher level using spirometer. Thus, Puskesmas will give simple treatment, if symptoms persist the patient or clinical assessment shows it is severe COPD the patient should be referred. 7 HFCS data is representative at national and strata level and not representative at province level. Supply-Side Readiness for Universal Health Coverage