Improving the cost-effectiveness of government health spending in Turkmenistan: the design of an essential package of health services

IMPROVING THE COST-EFFECTIVENESS OF GOVERNMENT HEALTH SPENDING IN TURKMENISTAN: THE DESIGN OF AN ESSENTIAL PACKAGE OF HEALTH SERVICES Matthew Jowett Research Fellow International Programme Centre for Health Economics University of York Prepared for the Lukman Project, Turkmenistan, May 1997. Contents 1. Basic packages of health care - the principles .............................................................................. 2 2. Prioritising health services - experience to date. .......................................................................... 3 3. The burden of disease and priorities for the health sector ........................................................... 5 4. Priority health services based on cost-effectiveness analysis ...................................................... 6 5. Improving the cost-effectiveness of government health spending ............................................... 7 6. Financing priority health services ................................................................................................. 9 7. Conclusions. .................................................................................................................................. 9 8. Appendices. ................................................................................................................................. 10 2 1) Basic packages of health care - the principles The idea that governments in low and middle-income countries should direct their money towards a package of essential health services, has gained ground in recent years, in particular since the World Development Report (WDR) in 1993. Variously referred to as essential, basic, minimum or guaranteed packages, the motivation behind such a development is the belief that current government spending in many countries is misallocated, and could be better directed to meet the health needs of populations. By attempting to define a list of health services to be either fully or partially publicly funded, it is argued that a more open rational process is required, something which can only be an improvement on implicit, more emotional decision-making. Rationing takes place in every health care system, so it is argued that this process should be made explicit using epidemiological and economic information, as well as the growing body of evidence-based medicine. Moreover with an explicit process in place, the selected package of health services can be refined as information is updated, and priorities change. The World Bank argues that the choice about which services to prioritise, should be guided first and foremost by cost-effectiveness analysis. Such analysis can contribute important information to the decision about which services to earmark for public funding, and which to leave to private financing. Two sets of information concerning health interventions are required; the first concerns health gain, the second costs. The development of a new measure of health gain, known as a disability adjusted life year (DALY), which combines information on premature mortality and disability to give a clearer picture of gains in the quality as well as quantity of life, is a major improvement on previous measures. For each intervention the number of DALYs gained for a patient is calculated. When compared to the cost of providing the intervention a ‘cost per DALY gained’ ratio can then be calculated. Different interventions can then be compared to each other. On another level, the WDR proposes that certain services be bundled and delivered together. The advantage of packaging services in this way is to achieve efficiency gains, through minimizing delivering and access costs. There are several scenarios where such packaging is particularly relevant. The first is where a range of services tackles a similar disease or risk factor and an integrated, comprehensive approach is required, possibly with referral between service levels. For example the ‘integrated management of the sick child’ combines the treatment of diarrheal diseases and acute respiratory infections for under 5 year olds. The ‘mother and baby package’ combines community-based obstetrics, district-level hospital services, and information, education and communication services. The second situation arises where different priority interventions require the same level of skills in terms of human resources, and types of equipment, and can be delivered at the same health facility. For example where a health centre can perform a chest sputum acid-fast bacilli microscopic examination, it is also likely that a pregnancy test of urinalysis can also be performed. In a one-off calculation for the WDR, the burden of disease across the world was calculated, along with the costs of a selection of key interventions. A cost-effectiveness exercise was then performed which resulted in the selection of a priority package of public and clinical services, which it was argued the governments of low and middle income countries should redirect their funding towards: 3 Box 1: Minimum package of clinical interventions a) short-course chemotherapy for tuberculosis b) management of the sick child c) pre-natal and delivery care d) family planning e) treatment of STDs f) limited care The delivery of the curative package alone is estimated at US$ 8 per person in low-income countries and US$ 15 per person in middle-income countries. Box 2: Minimum package of public health interventions a) immunisations b) school-based health services c) information and selected services for family planning and nutrition d) programmes to reduce tobacco and alcohol consumption e) regulatory action, information and limited public investment to improve household environment f) HIV/AIDS prevention The combined cost of delivering the two packages was estimated at US$ 12 per capita in low- income countries, and US$ 22 per capita in middle-income countries. The difference reflects demographic structures, epidemiological conditions and labour costs in the two settings. In many low-income countries government health spending is significantly lower than $12 per person, for example estimates are put at $US 2 per capita in Vietnam, US$ 1.4 per capita in Uganda, and US$ 1.3 per capita in Nepal (SCF(UK) 1993). The WDR estimates that if it’s recommendations for reallocating resources were put into practice, approximately 50% of current government spending in developing countries would need to be redirected, and around 25% of the current disease burden could be reduced. The implication of such a reallocation is that government spending on discretionary services must be eliminated, and private finance encouraged. In doing so access will be improved for certain groups. In middle-income countries there is again a variation in spending levels around the US$22 estimated as necessary to finance the essential package, although most spend well over this amount. Amongst lower-middle income countries in 1990 (Turkmenistan is classified in this category) Bolivia spends $25 (official exchange rate dollars) per capita, Cameroon $24, Thailand $73, Philippines $14, and Turkmenistan $125. Of the upper-middle income countries South Africa spends $158 per capita on health, Saudi Arabia $322, Mexico $132, the Republic of Korea $377 and Venezuela $89. These comparisons shed light on the context within which reallocating resources would take place. In resource-poor environments shifting money away from already under-financed services would, politically, be far more difficult to achieve than in a relatively resource-rich environment. 2) Prioritising health services - the experience so far Few countries have explicitly defined a list of services, to be guaranteed free of charge to the population, at the exclusion of other services, though it is an idea attracting increasing attention. In Oregon State in the USA, a decision was taken in 1987 to stop using Medicaid funding for soft-tissue transplant, on the basis that the money would be better spent invested in pre-natal care (Daniels 1991). The prioritisation process was extended to consider 709 treatments, which 4 were then analysed in terms of their impact using the ‘Quality of Well-Being’ measure, and following a series of public meetings to establish thirteen basic values, the first 587 treatments were selected for state government funding. The absence of cost data in this process however, meant that prioritisation was based on effectiveness alone, rather than cost-effectiveness (Drummond and Maynard 1993). In the Netherlands a basic package of care has been designed using four criteria. Included interventions must be necessary, effective, efficient and not be left to individual responsibility. For example the Dunning Committee decided, as a result of the exercise to exclude in-vitro fertilisation, adult dental care and homeopathic medicines in the package, and to include sports injuries and care of the elderly in residential homes. Many low and middle income countries are currently calculating DALYs. Mexico was one of the first to complete the exercise, and to combine it with cost information to complete a cost- effectiveness analysis. Colombia and Uruguay have also completed DALY calculations. In Africa, Mozambique, South Africa, Zimbabwe, Kenya and Ghana are currently working on the process, whilst many of the Magreb countries have completed it. Mauritius is slightly further down the line having completed the DALY exercise, and calculated costs for 40 interventions, as has India. The Tanzanian Essential Health Intervention Programme is currently investing 20 million Canadian dollars to complete in two districts, the exercise of designing a package of essential health services. The countries of the former Soviet Union (FSU) are also considering prioritisation of health services and the development of basic packages, though the principles of comprehensive health care systems are still considered of primary importance. In both Kazakstan and Kyrgyzstan guaranteed and basic packages of health care have been designed, effectively consisting of the existing comprehensive health service, rather than a targeted priority list. It is with the development of social insurance systems where the idea of packages of health care are currently attracting most attention. Georgia has advanced further down this route than many countries in the region, having defined a ‘basic benefits package’ comprising the most important health services (UNDP 1996). This approach was adopted as a rational response to the financial crisis in Georgia following the collapse of the Soviet Union. There are two elements to the package, first that at the federal level preventive care plus some curative services, including curative care for children under 1 year old, and the insulin programme will be provided free of charge. At the district/municipal level curative care for children aged 1-14 years, emergency care and drugs for terminal care will be included (World Bank 1996). Certain services within the package face co-payments (e.g. deliveries), although the poor are exempt from such charges, as they are for many services not included in the benefits package. Selective and comprehensive PHC The debate concerning the use of cost-effectiveness analysis to identify priority health services has been on-going for some time. Following the Primary Health Care (PHC) declaration in Alma Ata in the late 1970s, which advocated a comprehensive approach to PHC, a selective primary health care (SPHC) approach was promoted by Walsh and Warren (1979). They argued that a comprehensive approach towards PHC was unaffordable for many countries. Walt and Rifkin (1988) subsequently argued that SPHC was the antithesis of the Alma Ata declaration, which itself was based on the experience in several countries of providing good health at low cost. They argued that SPHC placed control over the inputs and outcomes of health systems firmly in the hands of the medical profession, leaving patients as passive recipients of health care. Comprehensive primary health care it was argued, was a more holistic 5 strategy for health development, with a quite different view of the importance of equity, community participation and the very concept of health itself. In 1985 UNICEF developed a programme known as GOBI-FFF, which followed the selective approach to improving child health. This selection of health services considered to be a priority and cost-effective initially included growth monitoring, oral rehydration therapy, breast- feeding and immunisation. Subsequently female literacy, food production and family planning were added. In many countries wider health sector reform is taking place, involving decentralisation, and the promotion of private finance and provision of health services. Such developments entail the redirection of both capital and recurrent government spending, generally away from tertiary care. For example:  Zimbabwe recently imposed a decade-long moratorium on new investment in central hospitals, and concentrated on improving health centres and other district-level infrastructure.  Tunisia has converted eleven large government hospitals into semi-autonomous institutions with strong incentives for improved performance. By and large the use of economic analysis, and the process of prioritising health services into a basic package is in it’s infancy. The very exercise of collecting better quality information on the relative health gain, and costs of different health interventions, is an important goal itself in the short to medium term 1. 3) The burden of disease and priorities for the health sector Diseases of the circulatory system account for 43% of total morbidity in Turkmenistan (see Appendix 1), followed by diseases of the respiratory system which account for 18%2. The World Bank estimated the burden of disease in Turkmenistan (World Bank 1995), using data collected from the Ministry of Health and Medical Industry (MOHMI), and from national experts. According to these calculations ischemic heart disease is the greatest cause of death and disability in Turkmenistan, accounting for 19.1% of all DALYs lost, with respiratory infections accounting for 17.5%. Perinatal and diarrheal diseases account for 15.1% and 12.7% of total DALYs lost respectively (see Table 1). Table 1: DALY estimates for Turkmenistan Major cause % total DALYs lost Ischemic heart disease 19.1 Respiratory infections 17.5 Perinatal 15.1 Diarrheal disease 12.7 Cerebrovascular disease 8.7 Motor vehicle accidents 4.4 Chronic obstructive pulmonary disease 3.5 1 For an update on developments in the collection of health expenditure data in the developing world see Berman 1997. 2 For a breakdown of the latest crude mortality data by age and sex see Appendix 2. 6 It should be noted that to conduct a comprehensive exercise of DALY calculation requires an enormous amount of information, and would take around one or two years. The data presented in the World Bank report is very much an approximation. Despite this the general patterns of mortality and morbidity in Turkmenistan appear clear, and reflect what is often referred to as the double burden of disease, with both communicable and non-communicable diseases constituting substantial health problems. The masterplan of the Lukman Project states the following as the main goals for improving health in the next ten years (see Box 3): Box 3: Priority health goals for the next ten years  to reduce infant mortality by tackling diarrheal diseases and acute respiratory infections, increasing breast-feeding, greater immunisation and better food hygiene, and improved health education  to reduce maternal mortality through family planning, health education/promotion and care during pregnancy and delivery  to eradicate polio and reduce measles cases through the EPI programme, social mobilisation and health education  to reduce STDs and prevent new cases of HIV/AIDS through health education and promotion, the increased availability of condoms, and the provision of quality health care for STDs  to reduce morbidity and temporary disability from tuberculosis through greater uptake of diagnostic services, short course chemotherapy treatment, improved case management and health promotion  to reduce viral hepatitis through greater immunisation, education and better water and sanitation conditions  to reduce mortality from cardiovascular and cerebrovascular diseases and from oncological diseases through the promotion of healthy lifestyles and early diagnosis and treatment Specific health sector targets are shown in Box 4: Box 4: Health targets for the next ten years  to reduce the infant mortality rate by 30%  to reduce the maternal mortality rate to 50 per 100,000 live births  to reduce STDs by 20% and prevent new cases of HIV/AIDS  to reduce morbidity from tuberculosis by 50% and temporary disability by 60%  to reduce hepatitis A by 30% and hepatitis B by 25%  to reduce mortality from cardiovascular by 20% in under 65 year olds  to reduce mortality from cerebrovascular diseases by 15%  to reduce mortality from oncological diseases by 15% in under 65 year olds 4) Priority health services based on cost-effectiveness analysis The DALY calculation referred to earlier was taken one stage further and combined with cost estimates, to complete the cost-effectiveness analysis. Cost estimates were based on the ideal amount of resources required for each intervention rather than the actual amount, and assumed a system that uses standard international protocols, generic pharmaceuticals, the minimum appropriate number of outpatient visits, and the maximum use of outpatient services rather than inpatient services. Once again it should be noted that this data is very much a rough estimation, and it is important that more accurate data is collected in-country. The following table summarises the World Bank’s calculations resulting in seven priority health interventions, ranked in terms of cost per DALY3. 3 Using 1994 prices, US$ 1 = Manat 88. 7 Table 2: High priority and cost-effective interventions for Turkmenistan - 1994 Intervention DALYs gained Cost per patient $US Cost per DALY gained $US 1 Integrated management of the sick child 0.25 7 29 2 TB short course chemotherapy 6.66 289 43 3 Smoking prevention 0.01 0.40 123 4 Hypertension treatment 0.17 32 188 5 Family planning 0.02 4 208 6 Alcohol prevention 0.01 0.40 212 7 Mother and baby package 0.37 94 257 According to these estimates the integrated management of the sick child is the most cost- effective intervention i.e. it offers the greatest health gain for each dollar spent. In addition to this list it is believed that five other interventions are extremely cost-effective although no analysis was conducted into these by the World Bank. They include an expanded programme of immunisation, clinic-based treatment of STDs, breast-feeding promotion, iodisation of salt and nutrition education. Other areas for prioritisation should include water and sanitation improvements as well as hygiene education. 5) Improving the cost-effectiveness of government health spending Improving the cost-effectiveness of health spending in Turkmenistan is an on-going process which will take many years. Several developments must take place, some of which have already begun. Prioritising health services - what to include in an essential package of health services Defining those services that are essential and those that are not can be as straightforward a process as desired. The use of cost-effectiveness analysis provides very clear and powerful information regarding how to get maximum value for health spending. However it is also important to consider the relative importance of cost-effectiveness and equity (Bobadilla et al 1995(b)), and to decide whether to modify the package in order to preserve equity at the expense of efficiency, in some instances such as care of the elderly. Ideally the decision will result from the opinions of the users, providers and planners of health care. However the decision is made there are a variety of information requirements needed to make informed choices:  better information concerning the epidemiological situation should be collected, for example through a more thorough calculation of DALYs lost.  outcome measurements should also be improved, at least for the most important interventions, in terms of DALYs gained.  costing information for specific interventions must be developed and combined with health gain information, to give accurate local estimates of cost-effectiveness for the key interventions. In the short-term proxies can be used to make estimates, as in table 2, but more comprehensive and accurate information needs to be collected for future decisions. A clear programme is currently being developed for a family physician-led (FP) health system in Turkmenistan. The main tasks of the FP form the basis of an essential package of primary health services (see below), and can be used as a starting point for a package of essential 8 services. This list can be broken down in to more specific tasks, and a cost-effectiveness analysis performed on the key ones.  preventive services  periodic examination of the population  diagnostic and ambulatory treatment services  emergency services  health education The context - political acceptability In Turkmenistan, the Constitution enshrines the principle of free health care for the entire population. The ‘Presidential State Programme of Turkmenistan: Health’ sets out the priority services that the government will tackle. This commitment was further reinforced with the publication of a list of services guaranteed free of charge to Turkmen citizens at state health care institutions as shown in Box 5. Despite this all out-patient drugs are paid for and informal payments are regularly paid to medical practitioners. A recent qualitative study by Ladbury (1997) found that over 50% of people spoken to had made payments to receive medical attention. The same study found that whilst respondents often understood and were sympathetic with the reasons for informal charges (i.e. extremely low wages of health practitioners), the fact that services were officially free was considered one of the most positive aspects of the health service. Box 5: List of health services to be provided free of charge by the state  Emergency medical treatment OUTPATIENT CARE  Chronic disease treatment  Injuries/poisoning/accidents  Home delivery  Ante-natal treatment  FP services / contraceptive supplies  Prevention related dispensing and examination  Epidemic prevention including vaccinations  Home visits for those unable to attend hospitals INPATIENT CARE  Treatment of acute and chronic diseases  Injuries/poisoning/accidents  Obstetrics  Delivery and care of new-born babies  Treatment of infectious diseases  Specialised care (oncology/TB/leprosy/alcohol etc.)  Stomatology care  Sanatoria care  Services at children's homes  Sanitary aviation services  Care during natural disasters/epidemics Given this commitment it is unlikely that explicitly defining a list of services to be covered by the state will be popular either with policy-makers or users of health services. For this reason it may be more appropriate to define a range of services that will not be covered by the state. Excluding services and items is often easier than defining all that should be included. These additional services would be charged for either through user charges or a system of insurance and might include:  drugs with low or unproven effectiveness  non-generic drugs for which a generic alternative exists  hotel services: better quality room, food and facilities such as television, air conditioning or own shower  preferred provider: choice of specialist or hospital  luxury services: cosmetic surgery, coronary bypass surgery, transplant surgery  treatments that are ineffective or considered ‘harmful’  by-pass fee for patients who self-refer 9 Implementing a package of essential health care Some commentators (Vos 1997) argue that the message of the basic package places too much emphasis on making major shifts in resource allocation e.g. taking large amounts of money away from hospital services and redirecting it towards primary services, in order to improve allocative efficiency. More attention could be paid to reducing waste in health services under existing allocations, in doing so saving money and improving technical efficiency e.g. in the purchasing, storage, distribution and use of vaccines. Moreover major changes in the allocation of funds require strong political support, and are likely to face resistance from various interest groups within the health sector. There are however various ways in which money can be spent differently to reduce waste, and to focus on more effective health services. A set of objectives can be identified, each of which improves public spending. Step 1: Reduce/stop spending on interventions that may cause more harm than good. Research and evidence from other countries occasionally expose the harm commonly used interventions cause, such as the use of Eusol for wound- cleaning which it was found destroyed good as well as bad human tissue. The unnecessary admission of patients into hospitals, where many patients actually acquire infections should also be avoided. Step 2: Reduce/stop spending on interventions that are shown to be ineffective. An examples in this category is the use of hyperbaric oxygenation for the treatment of liver diseases such as viral hepatitis. There is little evidence that such treatment is effective, but many hospitals throughout the FSU continue to buy the expensive equipment necessary. The use of low-frequency laser therapy to treat rheumatoid arthritis and bronchial asthma is another intervention for which there is little evidence of effectiveness. Step 3: Reduce/stop spending on those interventions that are shown to be of low effectiveness. The treatment of tuberculosis is an example of one treatment considered to be more effective than another. Long stays in sanatoria have been the traditional approach in Turkmenistan, on the basis that compliance amongst out-patients was extremely poor, and hence ineffective. Step 4: Reduce/stop spending on interventions that are inappropriate. The use of high-tech equipment such as MRI scanners where cheaper X-rays would suffice is an example of inappropriate interventions. Certain procedures, in particular diagnostic ones are often over-supplied due to financial incentives or the fear of litigation. Clearly such inappropriate activity is wasteful of resources. Step 5: Increase spending on interventions proven to be cost-effective. In terms of treating tuberculosis the use of a short-course chemotherapy treatment is now considered as more cost-effective than long, expensive in-patient stays at sanatoria. In terms of pharmaceuticals the use of generic rather than brand drugs is also more cost-effective. The development of an essential drugs lists can help achieve this, as can adherence to treatment protocols using cheaper first-line drugs, and more expensive second-line drugs if necessary. By the same token limiting spending on expensive high-tech equipment for low cost- effective services e.g. through the development of essential equipment lists is another step that can be taken. 10 Step 6: Increase spending on interventions proven by cost-benefit and cost-utility analysis. Such analysis is used for the reallocation of resources away from certain diseases and towards others. For example the interventions outlined in Boxes 1 and 2 and in Table 2, were selected on the basis of cost-utility analysis, namely cost per DALY gained. In the case of Turkmenistan such analysis recommends shifting resources away from coronary artery bypass surgery and the treatment of esophageal cancer, and towards for example a package of interventions aimed at child health, and the prevention of smoking. Once again it must be stated that the use of cost-utility and cost-benefit analysis should be complemented with other information sources, and judged against health sector goals. It is important that services do not fragment into a series of independent vertical programmes, as a result of such analysis. Other action that could be taken to improve spending on health care includes:  limiting financing for the training of specialist physicians to redress the balance between numbers of family practitioners and specialists.  create incentives to attract high quality staff to the primary level. This process has begun with the development of family practitioners who receive higher salaries.  greater emphasis must be placed on the most effective method for treating each disease. The focus should be on treatment protocols rather than rigid normatives. Emphasis should be given to simple interventions that reduce mortality or risk of infection, such as the prophylactic use of antibiotics during cesarean section. The development of treatment guidelines for a range of interventions, in particular midwifery can have a substantial impact on the effectiveness with which limited resources are used.  include in the resource allocation process from central government, a requirement that the velayat provides the defined priority primary and hospital services to a given standard. Penalties can be enforced to ensure this.  the decision about which services to prioritise should be a joint one between government, healthcare providers and users. A first step is to provide information concerning the major causes of death and disability to the community. This can be done through public meetings, posters at the health facility, and through radio and television broadcasts. Information concerning how such health problems can be prevented will also be included. Feedback from the public will be encouraged, in order to involve them in the process of prioritisation. This will ensure the acceptability of changes in the priority given to certain health services in the future.  health professionals should also be more involved in the process of prioritisation. Information concerning the cost-effectiveness of specific interventions should be included in the curricula for the training of doctors and nurses to facilitate this process, and feedback encouraged. 6) Financing priority health services At present many of the services in the government’s priority list are not provided for the entire population or are provided inadequately. One reason for this is that the 1995 level of spending, estimated at US$6 per capita (MOHMI data), is extremely low, and significantly below the estimated $US 22 required (World Bank 1993), to provide a minimum package of essential public health and clinical services in middle income countries. Under the ‘Presidential State Programme of Turkmenistan: Health’ it is projected that spending will increase to just under US$ 40 per capita over the course of the next ten years, 11 both as the economy grows and as a greater proportion of total government expenditure is devoted to the health sector. This is an optimistic scenario, but such an increased commitment to increased funding is important for the improved quality of services provided by the public sector. However increasing finance by no means removes the need for defining a set of priority health services, along the lines suggested in this paper. Priority services should continue to be financed out of government revenue and be available for the entire population. Non-essential services, should be charged for through either a system of user fees or insurance. 7) Conclusions Turkmenistan is currently attempting to provide a comprehensive range of health services, with spending estimated at only US$ 6 per capita. With the lowest life expectancy of all the FSU countries at the end of 1992, the highest infant mortality rate of those countries, and the fifth highest maternal mortality rate (World Bank), in order to meet the needs of the population, government spending should be directed towards more priority areas. A process of prioritisation should be initiated, using epidemiological and economic information, and involving MOHMI officials, health professionals and users of health services. Appendix 1: Main causes of deaths by velayat, Turkmenistan, 1995 Causes of deaths A sh g a b a t C it y A k h a l B a lk a n D a sh k h o w u z L e b a p M a r y T O T A L 1 Infections and parasitic diseases (e.g. acute diarrhoeal disease, TB, hepatitis) 6% 14% 8% 10% 13% 9% 10% 2 Malignant neoplasms 9% 8% 11% 4% 6% 6% 7% 3 Diabetes mellitus 1% 1% 1% 1% 1% 2% 1% 4 Circulatory system diseases (e.g. ischaemic heart disease, cerebrovascular disease) 48% 39% 48% 40% 45% 42% 43% 5 Diseases of respiratory system (e.g. ARI, pneumonia) 6% 18% 8% 31% 15% 20% 18% 6 Diseases of the digestive system 4% 4% 4% 2% 5% 5% 4% 7 External causes injury and poisoning 8% 8% 7% 5% 8% 7% 7% 8 Other causes 18% 8% 13% 7% 8% 9% 10% Source: Adapted from data from the National Statistics Office, Ashgabat 1997. Appendix 2: Main causes of deaths by age groups, Turkmenistan, 1995 Causes of deaths > 5 y e a r s 5 -1 4 y e a r s 1 5 -4 9 y e a r s > 4 9 y e a r s M F Total M F Total M F Total M F Total 1 All causes 4,879 3,735 8,614 426 288 714 3,284 1,821 5,105 8,254 8,717 16,971 2 Infections and parasitic diseases (e.g. acute diarrhoeal disease, TB, hepatitis) 25% 26% 2,227 10% 14% 83 12% 10% 580 2% 1% 302 3 Malignant neoplasms 0% 1% 39 6% 6% 42 7% 14% 487 10% 8% 1,545 4 Diabetes mellitus 0% 0% 1 1% 1% 7 1% 2% 68 2% 2% 313 5 Circulatory system diseases (e.g. ischaemic heart disease, cerebrovascular disease) 1% 1% 111 5% 3% 31 28% 27% 1,400 68% 74% 12,011 6 Diseases of respiratory system (e.g. ARI, pneumonia) 48% 49% 4,209 27% 30% 200 5% 8% 318 6% 5% 933 7 Diseases of the digestive system 1% 1% 56 4% 2% 22 8% 8% 425 5% 4% 756 8 External causes injury and poisoning 7% 6% 546 39% 25% 238 26% 14% 1,114 3% 1% 314 9 Others 17% 16% 1,425 9% 18% 91 12% 8% 713 5% 5% 797 Source: Adapted from data from the National Statistics Office, Ashgabat 1997. References: 1. Investing in Health. World Development Report 1993, World Bank, OUP. 2. Turkmenistan: Rationalising the Health Sector. World Bank, 1995. 3. Berman P. (1997) National Health Accounts in Developing Countries: Appropriate Methods and Recent Applications. Health Economics 6, pp 11-30, 1997. 4. Bobadilla J-L, Cowley P, Musgrove P, Saxenian H. (1995a) The Essential Package of Health Services in Developing Countries: The Minimum Package of Health Services: Criteria, Methods and Data. World Bank. (Sourced from World Bank website). 5. Bobadilla J-L, Cowley P, Musgrove P, Saxenian H. (1995b) The Essential Package of Health Services in Developing Countries: Design, Content and Financing of an Essential National Package of Health Services. World Bank. (Sourced from World Bank website). 6. Bobadilla J-L, Saxenian H. (September 1993) Designing an Essential National Health Package. Finance & Development. 7. Brenzel L. (1993). Selecting an Essential Package of Health Services Using Cost-effectiveness Analysis. Harvard University. 8. Daniels N. (1991) Is the Oregon Rationing Plan Fair ? JAMA, no. 17, pp 2232-2236. 9. Drummond M, Maynard A. (1993) Purchasing and Providing Cost-Effective Health Care. Churchill Livingstone. 10. Jamison DT, Saxenian H, Bergevin Y. (1995) Investing in Health Wisely: the role of needs- based technology assessment. International Journal of Technology Assessment in Health Care, 11(4), 1995. 11. Ladbury S, (1997) Turkmenistan Health Project: Social Assessment Study, report to World Bank / Government of Turkmenistan. 12. Musgrove P. (1994) Cost-effectiveness and health sector reform. Human Capital Development and Operations Policy Working Papers, World Bank. (Sourced from World Bank website). 13. Save the Children Fund (UK) (1993). Investing in Health - World Development Report 1993: The SCF(UK) Perspective. SCF(UK) Policy Paper. 14. United Nations Development Programme (1996). Human Development Report - Georgia 1996. 15. Vos T. Personal communication, London School of Hygiene & Tropical Medicine. May 1997. 16. Walsh J, Warren K. (1979) Selective Primary Health Care - An Interim Strategy for Disease Control in Developing Countries. New England Journal of Medicine 301, pp 967-994. 17. Walt G, Rifkin S. (1988). Why health improves: defining the issues concerning comprehensive and selective primary health care. Social Science and Medicine 23, pp 559-566. 18. Website of the Burden of Disease Unit, Harvard University: HTTP//WWW.HSPH.HARVARD.EDU/ORGANIZATIONS/BDU/ 19. World Bank Supported Georgia Health Project. World Bank 1996.