Colombia - Health sector review

FiLE COPY Report No. 4141-CO Colombia Health Sector Review December 15, 1982 Population, Health & Nutrition Department FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (as used in this report) Currency Unit = Colombian Pesos (Col$) US$1 = Col$ 61 WEIGHTS AND MEASURES Metric System MAIN ACRONYMS AND TERMS USED IN THE REPORT (See next page) GOVERNMENT OF COLOMBIA FISCAL YEAR January 1 - December 31 i  FOR OFFICIAL USE ONLY MAIN ACRONYMS AND TERMS USED IN THE REPORT ASCOFAME - Colombian Association of Medical Schools BENEFICENCIAS - Charities CABS - Social Security Institute Basic Care Centers CAJANAL - NationaLl Provident Fund CIMDER - Multidisciplinary Research Development Center CONPES - National Council of Social and Economic Policy DEPARTAMENTO - State or Province DNP - National Planning Department DRI - Integrated Rural Development Plan FNH - National Hospital Fund ICBF - Colombian Family Welfare Institute ICFES - Colombian Institute for Advanced Studies INAS - National Health Institute INSFOPAL - National Municipal Development Institute ISS - Social Security Institute MAC/UPA - Primary Care Unit MS - Ministry of Health PAN - National Food and Nutrition Program PHC - Primary Health Care PINDESA - Plan to Accelerate Health Development PNS - Nationm Health Plan SITUADO FISCAL - Proporlion (15%) of the central budget earmarked for education and health SSS - State (departamento) health services TFR - Total Fertility Rate ZMI - Intermediate Cities Program This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   COLOMBIA HEALTH SECTOR REVIEW TABLE OF CONTENTS Page No. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . - vi PART ONE: SECTOR BACKGROUND I. POPULATION, HEALTH AND N1JTRITION STATUS . . . . . . . . . . . 1 A. Population . . . . . . . . . . . . . . . . . . . . . . .1 B. Health .... . . . . . . . . . . . . . . . . . . . . . 2 C. Nutrition .... . . . . . . . . . . . . . . . . . . . . 3 D. General Living Condiitions . . . . . . . . . . . . . . . . 4 E. Summary . . . . . . . . . . . . . . . . . . . . . . . . . 5 F. Focus of this Report ... . . . . . . . . . . . . . . . 5 II. HEALTH SECTOR POLICIES, ]?ROGRAMS AND INSTITIJTIONS . . . . . . 6 A. Historical Background ... . . . . . . . . . . . . . . . 6 B. The Ministry of Health . . . . . . . . . . . . . . . . . 8 C. The Social Security '[nstitute (ISS) . . . . . . . . . . . 13 D. Health Planning . . ..... . . . . . . . . . . . . . . 14 E. Information Systems, Monitoring and Evaluation .15 F. Operational Research ... . . . . . . . . . . . . . . . 15 III. SECTOR PERFORMANCE . . . . . . . . . . . . . . . . . . . . 16 A. Coverage and Infrastructure . . . . . . . . . . . . . . . 16 B. Manpower . . . . . . . . . . . . . . . . . . . . . . . . 18 C. Pharmaceuticals . . ...... . . . . . . . . . . . . . . 21 D. Health Service Utilization . . . . . . . . . . . . . . . 22 E. Service Delivery Constraints . . . . . . . . . . . . . . 25 IV. HEALTH CARE FINANCING . . . . . . . . . . . . . . . . . . . 27 A. General Overview . . . . . . . . . . . . . . . . . . . . 27 B. Public Health Sector Revenues . . . . . . . . . . . . . . 30 C. Public Health Sector Expenditures . . . . . . . . . . . . 33 D. Public Health Sector Investment . . . . . . . . . . . . . 34 E. Social Security Institute (ISS) . . . . . . . . . . . . . 35 F. Summary . . . . . . ..... . . . . . . . . . . . . . . 38 This report was prepared by Dr. Anthony Measham, Mr. Jose Andreu and Mr. William McGreevey (PHN), and Mr. Dieter Zschock (consultant). Mr. Andreu led sector missions in January and March/April 1982. Other members were: Dr. Guillermo Herrera and Messrs. Willy de Geyndt and Dieter Zschock (consultants on both missions); Dr. Jorge Brenes (consultant, second); Mr. Howard Barnum (consultant, first mission); and Dr. Anthony Measham, PHND3 (second mission).  TABLE OF CONTENTS (Continued) PART TWO: SECTOR ISSUES AND POLICY OPTIONS V. SECTOR ISSUES .... . . . . . . . . . . . . . . . . . . . 40 A. Health Service Coverage ... . . . . . . . . . . . . . 40 B. Organizational, Institutional and Manpower Constraints . 42 C. Health Care Financing ... . . . . . . . . . . . . . . 45 VI. POLICY OPTIONS .... . . . . . . ..... . . . . . . . . 49 A. Implementation of the National Health System . . . . . . 49 B. Independent Development of the MS/SSS and Social Security Subsectors ... . . . . . . . . . . . . . . . 51 C. Health Manpower .... . . . . . . . . . . . . . . . . 53 D. Health Care Financing ... . . . . . . . . . . . . . . 55 E. Next Steps .... . . . . . . . . . . . . . . . . . . . 58 TABLES: 1. Population by Age Group, 1965-66 and 1977-80 2. Crude Death Rates, 1970-82 3. Infant Deaths According to 1973 Census, Registered in 1973, and Estimated by C. Pierret 4. Infant Mortality, 1950-82 5. Cause-specific Mortality by Age Group, 1977 6. Principal Causes of Morbidity, 1980 7. Height and Weight of Children under 15 by Sex (1965/66 and 1977/80) 8. Nutrition Indicators, 1965-80 9. Prevalence of Malnutrition in Colombia in 1965 and 1980 10. Consultations for Disease or Accident by Person Consulted, Region, Annual Family Income, Age, Sex and Social Security Affiliation 11. Consultations for Disease or Accident, by Place of Consultation, Region, Annual Family Income, Age, Sex and Social Security Affiliation 12. Relative Magnitudes and Elasticities of Ministry of Health Budgets, 1971-80 13. Sources of Public Health Care Financing, 1975-80 14. Distribution of Public Health Care Allocations, 1975-80 15. Distribution of Public Health Care Expenditures, 1978 16. Foreign Aid Commitments and Expenditures 1973-82, as of 1978 17. Regional Distribution of Health Care Allocations and Population, 1980 18. Public Health Investment Allocations and Expenditures, 1980 19. Investment Expenditures by Purpose, 1980 20. Hospital Resources and Costs, 1979 21. Health Care Facilities Investment Plan, 1978-82 22. ISS Assured Members as a Percentage of Total Population 23. Financing of ISS Deficit, 1978-80 24. ISS Income 1980 25. ISS Expenditures 1980  TABLE OF CONTENTS (Continued) ANNEXES: 1. PAN, DRI and ZMU 2. National Health Plan (PNS) and Program to Intensify Health Development (PINDESA) 3. Ministry of Health's Organizational Structure 4. National Health Institute (INAS) 5. (a) SSS Organization Chart (b) Regional Unit Organization Chart 6. ISS's National Level Organigram 7. ISS's Seccional "A" Organigram 8. ISS's Seccional "B" Organigram 9. Organigram of ISS Local Units 10. Health Sector Investments MAP: Colombia - Infant Mortality Rates by Departamento, 1973 (IRRD No. 16764)   SUMAKRY AND CONCLUSIONS Overview 1. Despite impressive mortality declines and increases in life expectancy in the last two decades, Colombia continues to face pressing health problems, especially high infant and child mortality. Chronic diseases are growing rapidly in importance and will demand a larger share of sector resources over time. There are major regional disparities in health conditions and services, with one quarter of the population lacking adequate coverage. Little progress is evident in developing the National Health System, mainly due to lack of coordination between the Ministry of Health and the social security sector. The Ministry of Health serves about half the population, emphasizing preventive and primary health care, while the Social Security Institute provides mainly curative services to one tenth of the population at four times the per capita cost of the Ministry program. Primary care deserves high priority and can be expected to reduce infant and child mortality. A major challenge will be to act simulta- neously to prevent chronic disease when possible, and to provide affordable treatment when it is not. Colombia in 1978 spent US$46 per capita on health, amounting to 5.4% of GI)P, a relatively high level for a developing country. This figure includes household expenditures on health. Average expenditures per person in 19713 were US$28 in the Ministry of Health program and US$108 for social security programs. The main sector issues are how to: (1) assure universal coverage; (2) devise revenue systems that do not discriminate against the poor; and (3) lower costs, discourage over- use, and foster fee systems that lessen the burden on general treasury resources. Policy options inc:Lude increased coordination or varying degrees of integration of the Ministry of Health and social security sectors, a gradual shift to program budgeting for health, and relating physician supply to projected needs. Population, Health and Nutrition Status 2. Fertility and genera:L and infant mortality have all declined sharply in the last two decades, nutritional status has improved, water and sewerage service has increased, and the annual population growth rate has fallen from over 3% to 2.1% annually. Fertility fell by more than a third in one of the fastest declines recorded anywhere. About half of couples married or in union are current contraceptive users. Maternal mortality remains high, due mainly to inadequate obstetrical care and induced abor- tion. General and infant mortality have declined by about half since the start of the 1960s in response to improved nutrition, water and sanitation, and health care. While the majority are better fed than two decades ago, a significant minority, particularly children, remains seriously mal- nourished. Two-thirds of the population has access to piped water, and over two-fifths have sewerage connections, among the highest levels in Latin America. The population growth rate continues to constrain socioeconomic development desp-Lte its recent decline. 3. Health conditions vary markedly with infant mortality more than twice as high in the poorer departamentos as in the more developed areas. Health conditions are worst on the Pacific coast, and to a lesser extent on  - ii - the Atlantic coast, in the tropical regions, and among the rural and peri- urban poor. The most pressing health problems are high infant and child mortality resulting from malnutrition and infectious disease, especially diarrheal and respiratory disease; accidents and homicides in the 15-44 age group; and chronic diseases (especially high blood pressure, heart disease, cancer and stroke) in those 45 and over. The six leading causes of death for all ages combined are cancer, ischemic heart disease, other heart disease, diarrheal disease, stroke and pneumonia, in that order. With mortality of children under five now one third of total deaths (versus one half on average in low income countries), and the increasing prevalence of chronic diseases, accidents and homicides, Colombia faces a double challenge in the health sector. First, how most effectively to deploy its resources for primary health care (PHC) aimed at reducing the toll from acute infectious disease, especially the diarrheas. Second, and simulta- neously, Colombia must focus its sector strategy on preventing chronic diseases where possible, and treating and controling them where it is not. There is a particular need to mount preventive campaigns (against smoking and high blood pressure, for example) and organize sector resources to avoid the costly emphasis on curative, high technology care that character- izes most developed countries. Sector Policies and Performance 4. In 1975, the Government created the National Health System to coordinate and eventually integrate sector institutions. The 1979-1982 National Integration Plan emphasized expansion of coverage to vulnerable groups (mothers, small children) and areas (rural and peri-urban). It set targets for morbidity and mortality reduction, and coverage with services, for example, immunizations. The extent to which these targets were met is unclear. 5. The main health service providers are the Ministry of Health (MS) and departamento health services (SSS) (45% of population), the Social Security Institute (ISS) (10%), about two hundred other public and private social security agencies (5%), and the private sector (15%). One quarter of the population lacks effective access to health care services. Overall, the physical infrastructure, with exceptions in the poorest, mostly rural departamentos, is sufficient to achieve universal coverage. 6. Colombia has an adequate supply of physicians and nursing personnel in urban areas, two few in the countryside. Physician produc- tion is likely to exceed absorptive capacity within 3-5 years. Colombia urgently needs to review its physician supply to determine whether current subsidies to medical schools are warranted in the light of impending over- supply. 7. The private sector provides more medical consultations than either the Ministry of Health (MS/SSS) or the social security agencies. MS/SSS provides a third more consultations than the Social Security Insti- tute (ISS). ISS beneficiaries average more than three outpatient visits a year, versus less than one visit a year for the average of MS/SSS clients. Two-thirds of hospitalizations occur in MS/SSS facilities, about one in five in private institutions and one in seven in ISS hospitals. A decline in average length of hospital stay and an increase in patient discharges  - iii - per bed-year indicate increasing efficiency in hospital use over the past decade. However, occupancy rates vary widely, from high at the tertiary (university) level to inefficiently low at the secondary (regional) and primary (local) level. Hospital utilization rates are high for the more affluent and ISS beneficiaries, and low for the poor. 8. Immunization rates are relatively low in Colombia, ranging from about one third to one half of children under five for BCG, DPT, polio and measles. Just over half of births are institutionalized, about one half are attended by physicians and one quarter by traditional birth attendants. 9. A number of problems are evident in the health services delivery system. The MS/SSS primary health care (PHC) program is understaffed, underfinanced and falling short of its ambitious coverage expansion targets. Inadequate training, supervision and logistic support are additional problems and the effectiveness of the PHC model is unproved. Moreover, lack of staff, equipment and supplies at regional hospitals result in patients bypassing this secondary level in favor of the tertiary care level. This means that tertiary care facilities provide PHC services at high cost. 10. ISS until recently provided mainly high cost curative services and had an excessive rate of hospitalization. A crisis in 1977 led to financial and administrative changes and development of a PHC program to increase efficiency. ISS contracts hospital beds and out-patient services from MS/SSS but there is limited coordination between the institutions. The Superintendency of Health Insurance, created in 1979 to foster coordi- nation, has not been effective. MS/SSS and ISS facilities are planned independently and often serve overlapping geographic areas resulting in underutilization. 11. The Ministry of Health has nominal responsibility for the entire National Health System. Its financial authority extends only to affiliated institutes and the departamento health services, which account for less than one third of total sector expenditures. The social security sector spends over one third of health-care expenditures. The private sector accounts for the remaining third. Average expenditures per person in 1978 were US$108 for social security programs and US$28 for MS/SSS. The MS/SSS share of GDP declined over the past decade but coverage increased while the average number of visits per person remained constant, indicating an increase in efficiency. 12. Underfinancing of the recurrent cost of health care is the domi- nant issue facing MS/SSS. Hospital care accounts for 69% of the MS/SSS budget but chronic deficits nevertheless persist. The recent facilities investment plan has not been matched by the necessary increased allocation for operating costs. Health Care Financing 13. Federal funding provides about two-thirds and departamento sour- ces about one-third of the MS/SSS budget. Federal sources are the regular central government budget, the situado fiscal - a proportion of the central  - iv - budget earmarked for education and health, and internal and external loans. Departamento sources include revenues from lotteries, beer and liquor taxes, and sales of services. The frequent recourse to extra- ordinary central budget allocations, and uncertainty regarding the amount and timing of departamento revenues, result in serious flow-of-funds pro- blems. 14. ISS health services are financed by a 7% wage levy (12% in the case of full family coverage), two-thirds paid by the employer and one third by the employee. Regular coverage is comprehensive only for the employee but covers maternity and care for children under one. Expendi- tures of US$113 per beneficiary (versus $108 for all social security agen- cies combined) among this relatively healthy population implies high costs and inefficiency. Issues and Options 15. The overriding health sector issues concern equity, efficiency and effectiveness: Equity - how to redress regional disparities in health status and health services to achieve universal coverage by 1990 and substantial reductions in mortality and morbidity in the poorer departamentos by that same date; Effectiveness - how to deploy the organizational, institutional and manpower resources available to achieve overall sector goals; Efficiency - how to achieve sector goals within the resource constraints that limit further increments of the share of GDP or of central government revenues devoted to health care expenditures. These issues in turn imply the need for policy decisions to refine and extend the primary health care model(s), to organize health-care management and finance to provide incentives to effective resource use, and to recon- sider the currently excessive earmarking of both revenues and expenditures which undermines health-sector management, financial control and the search for efficient modes of service delivery. 16. The principal policy options facing Colombian authorities relate to when and how to implement the National Health System. The major altern- atives are to integrate the MS/SSS and social security or to coordinate the two systems. While integration is the more difficult course, at least in the short run, it seems the only way to assure a national health system with universal coverage and uniform service standards. The practical dif- ficulties of integration, however, are formidable. The two systems have quite distinct administrative structures, delivery modes and methods of financing. Moreover, there is an urgent need to increase efficiency and effectiveness in both systems prior to integration. Integration could begin either in the major cities where both systems are well developed, or in the intermediate cities where MS/SSS has an extensive infrastructure.  - v - 17. Several concrete steps to improve resource utilization are possible regardless of the ultimate decision regarding integration. They are: (a) introducing program budgeting at all levels; (b) joint planning by MS and the social security agencies; (c) increased contracting of services between the two systems; (d) introduction of a standard salary schedule for health workers; (e) introducing user charges or co-insurance in the social security system; (f) faster expansion of ISS based on an actuarial analysis of new beneficiaries; (g) increased coordination within the MS, especially with FNH and INAS; and (h) joint planning at departamento level between the SSS's and beneficencias. 18. Improved resource utilization would also result from integration within the social security sector itself. This might start with ISS and other private agencies, or with the integration of all public sector agen- cies. Another option would be phased integration of the whole social secu- rity sector. ISS coverage could rise to 30% or more by including all family members and other segments of the labor force. Its impact on equity could be enhanced by subsidizing coverage for the poor, but the lack of social security infrastructure places a limit on expansion outside of cities. Experience elsewhere suggests the wisdom of separately integrating the public and private social security agencies before proceeding to inte- grate the entire social security sector. 19. MS/SSS efficiency and effectiveness would increase with greater decentralization of authority to the departamento health services. Decen- tralization could be achieved under the PINDESA model of the advanced departamentos providing technical assistance to their less developed neigh- bors, or with the Ministry itself performing this function. 20. Without increasing resources available to the health sector, it is feasible to extend primary health care coverage to all Colombians. Several models of such coverage are currently being tried with some success in Colombia (PINDESA, CIMDER, for example). By adopting a system with low costs per patient served, the Government can attain a first-stage goal of providing services to all. A second-stage goal is to achieve equal quality service to all, i.e., equivalent services to those served by MS/SSS, by ISS and by the private sector. 21. Currently, MS/SSS has a health-care responsibility without effec- tive budgetary authority over semi-autonomous institutes and the govern- ment's several social security systems. Either a re-definition of respon- sibilities will be necessary, cr some central budget-making system, with authority over all health-sector activities, must be put in place of the divided MS/SSS and ISS systems. It is, however, premature to alter organi- zational relations. First, there must be put in place cost-control proce- dures designed to reduce unit costs in the social security system and to improve the amount and quality of primary health care coverage in MS/SSS. Further, the utilization of MS/SSS hospital facilities can be markedly improved even with reduced budgets for in-patient care. 22. The potential efficiency of the system is undermined by current earmarking procedures. Certain tax revenues - the situado fiscal at the federal level and alcohol taxes and lottery receipts at departamento level  vi - - are earmarked for health in such manner as to reduce program budget authority at central planning levels. Central authorities thus have little discretionary power with which to reward efficiency and punish waste through withholding of budget support. On the expenditure side, local revenues are almost entirely devoted to specific uses in the support and maintenance of hospitals which, while seemingly laudable, undermines the necessary shift to primary health care outside the hospital system. Federal expenditures are de facto also earmarked because of the demand for support for tertiary-level hospitals. A gradual shift over several years to health care financing on the basis of well-defined program budgets offers the best hope for achieving health-care coverage within the stringent budgetary limits faced by the Colombian government. 23. The following examples illustrate immediate actions that might be undertaken to rationalize resource utilization regardless of decisions on broader policy options: (a) Establishment of a strong inter-agency mechanism for plan- ning, coordinating and implementing changes in the National Health System; (b) A freeze on tertiary level facilities construction and mandatory joint MS/SSS and ISS planning of new facilities; (c) Halting of Social Security Institute basic care centers (CABS) construction pending decisions regarding expansion of health service coverage; (d) Early decisions regarding priority geographic areas if coverage is to be expanded; (e) Cost-effectiveness studies to inform policy choices; (f) Analysis of the private social security funds; (g) Projection of future needs and supply of physicians.  PART ONE: SECTOR BACKGROUND I. POPULATION, HEALTH AND NUTRITION STATUS A. Population 1.01 Colombia's estimated population for mid-1982 is 27.7 million. Recent demographic change can *be divided into two distinct periods: rapid growth between the early 1950s and 1964, and rapid fertility decline between 1964 and the present. The crude death rate fell from 22.4 per thousand for the intercensal period 1938 to 1951, to 17.4 per thousand for the period 1951 to 1964, and an estimated current level of 7.7 per thousand. Fertility began to decline in the mid-1960s, with the crude birth rate falling from 47.2 per thousand for the period 1951 to 1964, to 33.1 per thousand in 1973, and to 28.9 per thousand in 1980. The total fertility rate fell from 7.04 in the early 1960s to an estimated 3.92 in 1978. The population growth rate has declined, therefore, from over three percent annually between 1951 ;nd 1964, to a current level of 2.1 percent. Despite one of the fastest decLines in fertility recorded anywhere, the current population growth rate represents a significant impediment to economic development and improved welfare. 1.02 Age Structure. The declines in mortality and fertility have pro- duced shifts in the age structure as shown in Table 1. The dependency bur- den peaked in about 1964 and since then the 0-15 age group has declined from 46.6% to 41.4% in the late 1970s, while those 60 or over increased from 5.0% to 5.9%. The median age of the population increased to 17.9 years while the average family size declined to 5.4 persons. Life expectancy at birth rose from 55 to 62 years between 1970 and 1980. 1.03 Spatial Distribution of the Population. High rates of rural- urban migration have doubled the share of the urban population (defined as concentrations of 2,500 or more) from 30.9% in 1938 to over 60% currently. Urban areas grew at an annual eate of 4.2% between 1964 and 1973, while rural areas increased at below 1% annually. Colombia has experienced balanced urban growth. The four largest cities (Bogota, Cali, Medellin and Barranquilla), each with more than a million residents, account for about one quarter of the population. The small (10,000-50,000) and intermediate (50,000-500,000) size cities account for an additional quarter of the popu- lation. 1.04 Fertility. The crude birth rate fell about 35% from the early 1960s to the late 1970s. A better measure is the total fertility rate (TFR) which is independent of ithe effect of changing age structure. The TFR declined by 44% between the early 1960s and 1978. The proportion of single women increased during the 1970s. Fertility is much higher in rural than in urban areas, by a factor of over two in 1978 (TFR = 6.05 versus 2.94). Fertility is also much higher among less-educated women and those in low-income households. The total fertility rate is more than three  -2- times as high in the poorest as in the richest group (TFR about 6.0 versus 1.9). The higher fertility among poorer women is accompanied by shorter birth intervals, higher infant mortality, and generally poorer health among both mothers and children. The 1980 National Contraceptive Prevalence Survey indicated that 49% of women in union were using a family planning method, 54% in urban and 37% in rural areas. 1.05 Mortality. Table 2 provides recent estimates of the crude death rate from 1970 to 1982. The estimates are lower than that of the National Statistical Office (DANE) for the 1973 census year (8.8 versus 10.7 per 1,000), and lower than Bank projections, which are 8.3 for the 1975-1980 period and 7.7 for the 1980-1985 period. It appears doubtful that the crude death rate has fallen as far as indicated in Table 2, but no firm judgement can be made given the unreliability of mortality data. Table 2 shows substantial differences between the 23 departamentos, but the infant mortality estimates in Table 3 and the map demonstrate these disparities even more clearly. These data, derived from the 1973 census, suggest that health conditions are particularly poor on the Pacific coast and to a lesser extent on the Atlantic coast. National estimates of infant mortali- ty, 1950-1982, show the rate has fallen from 140 to 64 (see Table 4). Maternal mortality is estimated to be about 2-3 per 1000 live births. Complications resulting from induced abortion are believed to account for about one quarter of maternal mortality. B. Health 1.06 Main Health Problems. The most important causes of illness and death in children under five are gastrointestinal and respiratory infec- tions, perinatal diseases and malnutrition. Mortality is lowest in the 5-14 age group in which accidents are the leading cause of death. The leading causes of death among adults 15-44 are accidents and homicide. Chronic degenerative disease accounts for the bulk of deaths in those over 45, principally resulting from heart disease, cancer and stroke. Table 5 provides cause-specific mortality by age group for 1977. For all ages combined, the six leading causes of death in 1977 were cancer, ischemic heart disease, other heart disease, enteritis and other diarrheal disease, stroke, and pneumonia, in that order. 1.07 Infant and Child Mortality. In 1964, children under five made up 17.6% of the population and accounted for 49.8% of deaths. In 1976, this age group represented 14% of the population and incurred 32.8% of deaths. In spite of major declines in infant and child mortality during the past two decades, diarrheas, respiratory infections and malnutrition still account for about half of mortality (see Table 5). Improved water and sanitation, hygienic practices and nutrition can be expected to greatly reduce deaths from these causes. Neonatal deaths (occurring within one month of birth) accounted for 47% of all infant mortality in 1976, but a substantial proportion is preventable by improved maternal health and nutritional status, better obstetrical care and childspacing. Measles and diphtheria fell from the six leading causes of death in children 1-4 between 1964 and 1976, reflecting improved nutrition and protection by immunization. The overall picture suggests that substantial further  - 3 - reductions in infant and child mortality are possible through proven public health measures such as improved maternal and child care, family planning, health education, nutrition, water and environmental sanitation. 1.08 Adult Mortality. The data in Table 5 demonstrate that violent death from accidents and homicide is the prime cause of death in the 15-44 age group. This reflects the social pathology that has attended the rapid urbanization process, and the high accident rates on Colombia's highways. The mortality profile for those over 45 indicate Colombia's transition to middle-income status, with the elimination of tuberculosis and other infectious diseases from the five leading causes of death. Colombia now faces the same challenges as industrialized countries in preventing and treating heart disease, stroke, cancer and high blood pressure. Chronic degenerative diseases can be expected to make increasingly heavy demands on the health system in the years ahead. 1.09 Morbidity. The morbidity data presented in Table 6 were derived from clinical histories and physical examinations conducted on 12,000 indi- viduals in the 1977-1980 National Health Survey. Chronic respiratory infection is the most prevalent cause of morbidity, affecting 12%. of the population. The most important cause is probably smoking since 40 percent of Colombians over 15 are cigarette smokers. Helminth parasite infestations, combined with amebiasis and other intestinal infections, form the category of gastrointestinal disease responsible for 13.3% of morbidity. (Viral and bacterial diarrheas were probably underdiagnosed and parasitic infestations overdiagnosed.) The findings nevertheless confirm the primacy of gastrointestinal disease potentially preventable by environmental sanitation measures and good hygiene, or treatable with rehydration or drug therapy. Overall, the principal causes of morbidity indicate the high incidence of respiratory and gastrointestinal disease and malnutrition noted as major causes of infant and child mortality. The high prevalence of hypertension correlates with the heavy death toll from heart disease and stroke, and signals the need for a major campaign to prevent or control high blood pressure. 1.10 Morbidity data differentiated by geographic area are not avail- able to demonstrate the different epidemiologic patterns that occur in the cooler highland and lower tropical and coastal regions. Malaria has increased in recent years to an incidence of 155 cases per 100,000 popula- tion in 1980, and threatens four million people or 15% of the population. Yellow fever and dengue are significant problems in endemic areas. Hookworm is also prevalent in tropical areas on the Pacific coast and elsewhere. The National Health Survey data will provide a powerful health planning tool when analyzed by region. C. Nutrition 1.11 The nutritional status of the Colombian population has improved significantly since the mid-1960s. Anthropometry (measurement of weight and height classified by age) was used to characterize the nutritional status of children in representative sample surveys conducted in 1965-1966 and 1977-1980. The risk of morbidity and mortality increases significantly  -4- among children classified as severely malnourished by any of several anthropometric criteria. In 1965, below normal average height for age was documented and corresponding substandard weight for age (Table 7). Average height for age had increased significantly by 1980 and there was a corres- ponding increase in weight for age (Table 8). There was a significant reduction in the prevalence of malnutrition as defined by three commonly used criteria (Table 9). Moderate and severe protein/calorie malnutrition, according to the Gomez Classification, had declined by half among children under five and by a third among those between five and ten. 1.12 The reduction has been more marked for moderate than for severe malnutrition. In 1965, 1.7% of children under five were severely mal- nourished, and this was still the case in 1980 in two of five regions of the survey, the Pacific and Eastern areas. This finding suggests that in 1980 there was still a subset of the population with severe nutritional deficiencies. In summary, nutritional status has improved since 1965 with the exception of a small but significant group of seriously malnourished chilAdren. This secular trend reflects improved diet and living standards, the declining incidence of infectious disease, vaccinations and better health care. D. General Living Conditions 1.13 Colombia's GDP expanded at a cumulative average annual rate of 5.1% during the 1960s and 6.0% in the 1970s, slowing somewhat in recent years. Per capita income rose to a level of US$1010 equivalent in 1979. About 25% of Colombian households have income too low to satisfy basic human needs. This quarter of the population lives in rural huts or urban hovels, the latter mostly in the periphery of major cities. The three largest metropolitan areas (Bogota, Medellin and Cali), and major regions within some departamentos -- Antioquia, Caldas, Valle, Cundinamarca -- enjoy much better living standards than many departamentos bordering on the Pacific and Atlantic and their hinterlands, such as Choco, Cauca, Narino, Cordoba, Sucre, Bolivar, Magdalena, Cesar and the Guajira. The sparsely-populated Eastern plains, intendencias and comisarias are also much poorer than more developed areas of the country. 1.14 Colombia's adult literacy rate is about 80%. While school enrollment is increasing, less than half of the population has completed primary education. Primary education grew less rapidly than secondary or higher education between 1965 and 1975, in terms of availability per child. Moreover, of rural children who began primary school in 1969, more than 85% dropped out before completion. 1.15 Availability of water supply and sewerage services in Colombia ranks among the highest in Latin America. Piped drinking water is avail- able to about three-quarters of the urban population and almost one half of the rural population, averaging 65% overall. Sewerage connections are available to about 60% of the urban and 12% of the rural population, over- all coverage being about 42%.  E. Summary 1.16 In summary, fertility and general and infant mortality have all declined sharply in the last two decades, and the population growth rate has fallen from over 3% to 2.1% annually. Despite important advances in health and nutrition status, and in the availability of piped water and sewerage services, serious heaLth problems nevertheless persist. There are significant regional disparities in health status, with the Pacific coast and to a lesser extent the Atlantic coast exhibiting the worst statistics. For the country as a whole, the most important problems are high infant and childhood mortality resulting fErom the combination of malnutrition and infectious disease, especially diarrheal and respiratory disease; accidents and homicide in the 15-44 age group; and chronic disease (especially high blood pressure, heart disease, cancer and stroke) in those 45 and over. With under five mortality now one third of total deaths (versus one half on average in low income countries), and the increasing prevalence of chronic diseases, accidents and homicides, Colombia faces a double challenge in the health sector. First, how mosi: effectively to deploy its resources for primary health care (PHC) aimed at reducing the toll from acute infectious disease, especially the diarrheas. Second, and simultaneously, Colombia must develop a sector strategy to prevent chronic diseases where possible, and to treat and control them where it is not. There is still time to mount preventive campaigns (against smoking and high blood pressure, for example) and organize sector resources to avoid the costly emphasis on curative, high technology care that characterizes most developed countries. F. Focus of this Report 1.17 The remainder of this report will focus mainly on the health pro- blems identified above, health sector issues and policy options. Colombia is addressing malnutrition through a major inter-sectoral National Nutri- tion Program (PAN) which the Bank is supporting through Loan 1487-CO. PAN began in 1976 and data are not yet available to judge its impact on malnu- trition, especially since geographic coverage was only gradually expanded. It was not possible to include a review of nutrition issues and policy options in this report. A major review of recent demographic change resul- ted in a late 1979 Bank report entitled Fertility Decline in a Developing Country: The Case of Colombia. The water and sanitation sector is curren- tly under review and a Bank mission visited Colombia in July 1982.  -6- II. HEALTH SECTOR POLICIES, PROGRAMS AND INSTITUTIONS A. Historical Background 2.01 During colonial times and the first century following indepen- dence, health care in Colombia consisted of services provided by tradi- tional healers and private physicians trained first in Europe and later in national medical schools. These physicians served the elite and practiced curative medicine exclusively. Health care of the indigent, orphans, and the mentally ill was at first the domain of charity institutions, largely run by the Catholic church. As the population increased, orphanages, shelters, and municipal and community hospitals, usually staffed by religious orders, emerged throughout the country. Political pressures and local initiative, rather than assessment of regional needs, determined the size and kind of health facilities built and operated. The resulting distribution of hospital beds and services in the country was haphazard. Each facility was conceived and developed as a free-standing unit which in time became progressively more dependent on a complex array of funding sources. With the advent of modern high cost technology, this approach led to wasteful duplication of services and a major escalation in investment and particularly in operating costs. 2.02 The first government action in the health field was the creation in 1913 of the "Consejo Superior de Salud", later renamed "Directorio Nacional de Higiene". The Ministry of Public Health (MS), mainly as it is today, was established in 1953. Government programs were initially small, geared exclusively to control communicable diseases through reduction of environmental hazards, provision of water and sewerage facilities, and garbage disposal. Vaccination campaigns were attempted, as was the isolation of patients with contagious diseases. At first, there was no relationship between these government activities and hospital care. Rural health care was virtually non-existent and reliance on traditional practitioners was almost universal until the 1950s. In 1945, the National Provident Fund (CAJANAL) was created to provide prepaid health services and other benefits to government employees. In 1946, the Colombian Institute of Social Security (ISS) was organized under the Ministry of Labor to provide life and disability insurance, a pension plan, and a health program for employees in the modern private sub-sector. The ISS health system grew rapidly and independently of both municipal hospitals and the MS. In addition, a number of smaller prepaid health programs have been organized more recently for railroads and telecommunications workers, the police, armed forces, and other employees not protected by ISS or CAJANAL, or dissatisfied with their services. These social security and family welfare funds now number approximately two hundred. 2.03 The health sector can therefore be divided into three main sub-sectors: (a) the government (public health or "official") sub-sector, which comprises the Ministry of Health (MS), its five autonomous specialized agencies, and the departamento health services (SSS); (b) the social security sub-sector, which comprises importantly the Social Security  Institute (ISS) for private employees, CAJANAL for public employees, and the smaller funds for specific population groupsl/; and (c) the private sector. The historical background provided above explains the separate paths taken by the public health and social security sub-sectors, which have resulted in substantial differences in philosophy, objectives and modus operandi; and their present lack of integration. 2.04 Dispersed responsibilities were the norm rather than the excep- tion in the MS in the 1950s; inadequate funds from multiple sources were spent without reference to priority needs or national plans. Inadequately trained personnel subject to patronage appointment and removal lacked professional guidance and supervision. In addition, reliable data on even the most prevalent health problems were practically non-existent. 2.05 In 1963, ASCOFAME and MS conducted a national study of health manpower,. medical education, and morbidity. The findings of this study, which included social, economic and demographic data, served as the basis for an MS reorganization and decentralization plan. Decentralization included establishment of the SSS (departamento health services)-as autonomous agencies of the MS. A National Health Council chaired by the Minister of Health attempted to unify government health programs and financing, but met with little success. 2.06 The main objective of MS programs from the late 1960s on was extending coverage to those not protected by organized health services. Priority was given to rural areas, to the poor, marginal urban population, and to maternal and child care. Primary health care, largely provided by paramedical personnel, was the principal instrument for achieving this objective. 2.07 The Government undertook a major review of the health sector in 1974 which led to the development of a National Health System. The prin- cipal findings included the following: - 36% of the population lacked adequate health services, while MS and ISS combined covered 49% of the population and the private sector 15%; - two out of three rural residents lacked adequate health care, piped water and sewerage services; - the sector was characterized by lack of coordination and duplication of effort; - a large proportion of deaths occurred in infants and children under five; - diarrheal diseases were the number one cause of death, and malnutrition and infectious disease constituted major problems; 1/ Only the ISS will be analyzed in this report due to lack of basic data on the other institutions.  -8- - lack of balance between the quantitative and qualitative pro- duction of health manpower and the country's needs; - major problems with hospital services, due to administrative deficiencies, excessive demand, labor disputes, and excessive increases in operating costs. These findings formed the basis of policies designed to: (a) Improve health conditions, eliminating inequalities and pro- viding access to all; (b) Develop a coordinated National Health System composed of all health institutions; (c) Use the National Health System and a National Health Plan to increase services to the entire population, giving priority to the rural, marginal urban and inaccessible population; (d) Give priority to population groups with a high proportion of children under 15, to mothers and children generally, and to the labor force; (e) Assure government control of health care financing; (f) Regionalize health care under the criteria of policy centra- lization and administrative decentralization. Finally, health would constitute a major component in integrated develop- ment efforts, such as the Integrated Rural Development Program (DRI), the National Food and Nutrition Program (PAN), and the Intermediate Cities Project (ZMI). The Bank is participating in DRI, PAN and ZMI and a summary evaluation is provided in Annex 1. 2.08 The policies outlined above formed the basis for the National Health System created in January 1975. Five decrees established the legal and organizational bases of the system. Of particular significance was legislation to the effect that all health institutions except those of the Ministry of Defense would form part of the system. Another decree established a personnel statute for health workers. Equally important was legislation to put in place six sub-systems in MS: planning, information, personnel, supplies, investments and research. The exhaustive review of problems and policy options did not result in major shifts in policy direc- tion, but did lead to important organizational changes and the creation of the National Health System. B. The Ministry of Health 2.09 Main Policies and Objectives. The Government's stated develop- ment strategy is directed toward increasing the rate of growth of the economy, lessening the constraints to longer term growth, and avoiding balance of payments difficulties in the late 1980s, as a means of expanding  -9- employment opportunities for the growing labor force, improving the distri- bution of income and providing greater welfare to all Colombians. The overall health sector goal is to extend coverage to all Colombians through expanded and improved health services. 2.10 The 1979-1982 National Integration Plan, the overall government development plan, continued to emphasize expanding coverage to the most vulnerable groups (mothers and children under five) and areas (rural and peri-urban), in recognition of the disparities in health status among regions and population sub-groups. The National Health System was viewed as the major instrument to achieve the goal of expanded coverage and conti- nued emphasis was placed on increased coordination between MS, ISS, the provident funds and family welfare funds. Specific coverage targets were identified, including 80% for infants, 100% immunization of children under five, piped water and sewerage to 78% of the urban population and 79% of the non-dispersed rural population, and a 15% increase in prenatal care. The ultimate goals were to reduce infant mortality by 15%, child (1-4) mortality by 25%, and various kinds of morbidity by given percentages. 2.11 There did not appear to be an operational plan translating the above targets into clearly specified activities expected to lead to their accomplishment. Nor were data available to judge the extent to which targets have been met. Such evaluation efforts are hampered in two major ways. First, coverage hes not been defined so that it can be readily measured. The definition is more in terms of the availability of physical facilities and personnel than in the provision of a given quantum of services. Moreover, the complementarity of coverage between MS/SSS, ISS and other institutions is not spelled out. Second, the health information system does not yield data from which conclusions can be drawn about cover- age. Existing evidence indicates that the water, sanitation and immuniza- tion targets have not been met. Progress in increasing coverage and reducing mortality is also probably less than targeted. Furthermore, a reliable estimate of progress achieving the target of increasing coverage by 9 million (one million a year) is not available. Estimates range from 3 million to 8 million. Finally, progress has been slow in moving towards greater coordination between the various elements of the National Health System. 2.12 In 1981, Government established a Plan to Accelerate Health Development (PINDESA) (Annex 2) grouping SSS in six nuclei centered on the six most developed, or lead, departamentos, which will have responsibility for helping develop their less well-favored neighboring departamentos. Specifically, lead SSS are charged with providing technical assistance, coordination, data processing and monitoring services, supervision and evaluation of programs in their nuclei. The central nucleus in Bogota is to act as overseer, and is charged with the normative development of the system. PINDESA is just beginning to be discussed with SSS and other agen- cies and represents the latest government effort to extend coverage through decentralization and coordination of field activities. 2.13 Organization and Management. The MS/SSS structure has four levels: national (MS and its autonomous agencies), departamento (SSS), regional (general hospital), and municipal (local hospital and dependent health centers, health posts and village workers). MS' organizational  - 10 - structure is given in Annex 3. The line of command extends from the Minister through a vice-minister, a secretary general and nine operating divisions. The most important divisions are those dealing with human resources, medical care, and operations control and supervision. Of the five staff offices (administration, technical, legal, international rela- tions and planning), the planning office is the most important, having acquired substantial responsibilities in the past four years. These include carrying out surveys and data processing; analysis of organiza- tions, procedures and administrative models; evaluation of all official health sub-sector activities; and budgetary planning and preparation of annual budgets for MS and SSS. MS' organizational structure and allocation of responsabilities have been found wanting on occasion, however, as in the case when a special task force was attached to the office of the secretary general to supervise the health components of the PAN program. 2.14 MS has five autonomous national agencies under its jurisdiction. They are: the National Institute of Health (INAS - Annex 4) which is responsible for water supply and sewerage in towns with up to 2,500 inhabitants, and for production and control of vaccines; the National Municipal Development Institute (INSFOPAL) which handles water supply and sewerage for larger towns2/; the National Hospital Fund (FNH) which allocates funds for, and approves and supervises, hospital investments and maintenance; the Family Welfare Institute (ICBF) which is responsible for a national pre-school nutrition program, child-care facilities, and for the protection of minors; and the Cancer Institute. In addition, the Superintendency for Health Insurance which supervises the ISS is technically under MS supervision. INAS, INSFOPAL, and FNH budgets are approved by the Minister of Health, but the agencies are managed rather autonomously without formal coordination with MS staff or line units. The lack of coordination between INAS, INSFOPAL and other sections in the Ministry results in less joint planning of primary health care and water and sanitation improvements than desirable. ICBF is the most autonomous of the five decentralized agencies. 2.15 At the SSS level, overall policy is vested in a departamento health board, the members of which include the governor, health-related agency representatives and citizens. Policy matters and overall internal management are the responsibilities of the SSS director, appointed by the governor and supported in day-to-day management by a technical coordinator who is generally a career officer. The three operational units are for medical care, environmental sanitation and administration. A typical SSS organization chart is given in Annex 5a. A similar structure (Annex 5b) operates at the regional and local levels, which are managed respectively by regional and local hospitals designated as headquarters. 2.16 Efforts to provide primary health care (PHC) services in Colombia predate international initiatives and the Alma Ata Declaration by almost twenty years. The earliest experiments, carried out in the late 1950s by 2/ INAS and INSFOPAL cover only part of the total water supply and sewerage systems. A number of systems in large towns, cities and regions (especially the more developed ones) are managed either locally or by departamento agencies.  - 11 - Valle University in Candelaria, included most of what have come to be accepted worldwide as the essential components of primary health. Nation-wide PHC activities beg;n in the late 1960s when the MS endeavored to extend services to all those in need. The main features of the model that was adopted were: (a) a regionalized, integrated service pyramid in which care would be provided al: the lowest and least expensive level possible; (b) referral of cases of increasing complexity up the pyramid from the local level (community health workers, health post, health center, local hospital), to the regional, and finally to the university hospital level; (c) delegation of functions to paramedical personnel; and (d) deployment of a cadre of female community health workers known as health promoters. Although this basic model has been modified and refined several times, its essential features have remained intact for more than a decade. 2.17 The most innovative elements of the Colombian PHC model are the services to be provided by promoters and auxiliary nurses. The promoter is responsible at the community level for: (a) basic health care, including dispensing drugs; (b) nutrition and health education; (c) immunization and epidemiologic surveillance; and (d) promotion of safe water supplies, sanitary excreta disposal, solid waste disposal, and improved housing. The model calls for one promoter for every 200 households (1,000 people) in rural areas, and one for every 400 families (2,000 people) in urban areas. The basic health care nucleus :Ls the health post staffed by two nurse auxiliaries, six promoters and a part-time environmental health assistant, which serves a population of 6,000 (more in urban areas). Health posts receive weekly visits by a physician doing his or her year of compulsory social service, and periodic visits by a dentist. A health center has a full-time physician, two auxiliaries, an environmental health promoter, a pharmacy aide, an ambulance dr:iver and six promoters. It serves 8,000 to 20,000 people. The health cenl;er, but not the health post, has observation beds, rooms for minor surgery and deliveries, and two obstetrical beds. The auxiliaries provide basic preventive, curative and emergency care on an out-patient basis and refer complicated cases to a health center or local hospital. They are responsible for supervising each promoter at least once a month and accept referrals from them. Promoters and auxiliaries both have important responsibilities in maternal and child care, family planning and nutrition. Each is expected to screen pregnant women for high risk; attend normal deliveries and make post-natal visits; inform couples about family planning and either provide services or refer them to a physician. Both are expected to provide nutrition education. Promoters select and enroll under-fives and mothers in the Food and Nutrition Program (PAN) where it is available, monitor food consumption, and promote vegetable gardens and breeding of domestic animals. Auxiliaries monitor child growth, enroll those moderately or severely malnourished in a rehabilitation program, and distribute food coupons to beneficiaries enrolled by the promoter. 2.18 The PHC model described above is the result of several changes made during the last decade. Initially, health promoters were part-time village volunteers who received three months training and an honorarium for their services. Beginning in L977, they became regular government health workers paid the official minimaum wage. In 1976, the Government introduced the MAC (extended coverage module) whose acronym was subsequently changed to UPA (primary care unit). A MAC I or UPA I consists of a health post,  - 12 - staffed by two auxiliary nurses and visited weekly by a physician, and six health promoters assigned to a specific geographic area. A MAC II or UPA II consists of a health center with six promoters. The introduction of the MAC/UPA system defined more clearly the functions of promoters and other members of the PHC team, the population to be covered, and the supervision and evaluation system. Another change in the model was the gradual increase in delegation of functions to paramedical personnel including the use of basic drugs by auxiliary nurses and health promoters. 2.19 The next level of complexity is the local hospital, an acute care facility staffed by general practitioners, auxiliary nurses and nursing aides. The local hospital provides in-patient care in the four basic specialties (internal medicine, surgery, pediatrics and obstetrics- gynecology) and ambulatory care (emergency and out-patient clinics), including dental treatment. Local A hospitals are distinguished from local B hospitals by not performing elective surgery and being smaller in size (about ten beds on average versus 25-30 for B hospitals). The local hospital represents the limit of the peripheral or local level of the decentralized health care system that is expected to satisfy 95% of the demand for services. 2.20 The local hospital relates functionally and administratively to the regional hospital which is usually located in a large town or city. The regional hospital offers the four basic specialties plus more specialized services, such as orthopedics, ear/nose/throat and pathology. It usually has between 50 and 100 beds (more in major cities) and includes medical specialists on its staff, although they are frequently in short supply. The regional hospital is expected to meet 4.5% of the service demand -- the more complicated cases. The remaining 0.5% of demand -- the most complex cases -- is to be met by the university and other specialized hospitals which usually have 300-500 beds and all medical specialties represented. The system of treating cases at the lowest possible level with upward referral of only those cases beyond the capacity of a particular level, is designed to expedite processing and reduce costs by resolving health problems as close as possible to where the patients live or work. The most prominent roles are assigned to general practitioners and paramedical personnel. 2.21 The SSS office is at the apex of the service pyramid just described. Its functions are to: (a) adapt national health policies to the special circumstances of the departamento; (b) supervise all health institutions in its jurisdiction, including those of Social Security and the beneficencias; and (c) coordinate health with other sectors. The SSS offices also coordinate specific programs, for example, vaccination campaigns. SSS's attempt to coordinate their activities with those of ISS, the provident funds and the beneficencias have been largely ineffective. 2.22 Activities are recorded and aggregated at each level of the health system and information is sent to the next higher level. The SSS produces an annual statistical report of several hundred tables containing a detailed account of activities performed and services rendered. Unfortunately, the data are often not reliable and this information system does not meet management's need for data for planning, decision-making, problem-solving and evaluation purposes. The general management system  - 13 - outlined above varies with the stage of development and sophistication of the SSS. Antioquia, Valle and Caldas carry out most of the responsibilities entrusted to them by the MS. Other SSS's still need central support and often cannot cope with their operational duties. 2.23 Government's National Health Plan (PNS) was designed primarily to extend primary health care in rural and urban areas. An evaluation (Annex 2) disclosed important shortfalls in the operation of MS and SSS services. Lack of operating funds, staffing and training difficulties, administrative and management weaknesses all limited the coverage and effectiveness of expansion plans. Remedial action has been taken to strengthen these problem areas during the past two years under the secretary general's task force. C. The Social Security Institute (ISS) 2.24 Main Policies and Objectives. ISS has progressively expanded its health services during the last two decades. In addition to complete coverage for the worker, ISS provides maternity care to spouses and care to children under one. ISS recently embarked on a program to extend full health coverage to the families of contributors (spouses for life and children up to 18 years of age) and to rural agricultural workers. About one out of ten families now has full coverage. These policies reflect management's belief that total ISS coverage of the population--currently 10%--is low and should be expanded. 2.25 An illegal physicians' strike in 1976 prompted a government investigation of ISS that revealed serious financial and management problems. The fundamental pro'blem was the concentration of expensive curative services of uneven quality on a small proportion of the population. Reforms were legislated in 1977 and the main efforts sought to: a) give priority to prima:ry care, community, and occupational health, and to foster prevention and self-care; b) focus on concentrations of beneficiaries in defined geographic areas; c) develop basic health care teams with a physician, dentist, auxiliary nurse and auxiliary social worker for every 3000 beneficiaries. Some headway has been made in implementing reforms but overaLl progress has been slow. 2.26 Organization and Management. ISS's organizational structure is given in Annexes 6 (national level), 7 and 8 (seccional level), and 9 (local level). The Administrative Council, composed of representatives of the Ministries of Labor and Health, sets policy, objectives and programs, authorizes new investments and oversees operations. The director general handles day-to-day responsibilities through a general secretary (usually a civil servant) and a sub-secretary for general health and maternity (EGM), aided by the personnel/administration, physical plant and financial sub-secretariats. These national level units provide the legal, policy, normative and technical guidance, as well as much of the funding and other support for the field agencies at seccional and lower levels, which carry out the bulk of operations in a very similar fashion to MS/SSS. The seccionales type "A" (for the i'our largest departamentos) exactly replicate the organizational structure at: the national level from the director  -14- general down. The less developed seccionales have a similar, but simpli- fied organization and the local units are organized around the four basic health care services: obstetrics-gynecology, pediatrics, internal medicine and general surgery, with supporting out-patient and emergency services. 2.27 ISS's organizational structure is well geared to satisfy ISS's responsibilities and is based on the principles of functional need and decentralization of operational responsibilities to the lowest possible level. The present structure has permitted the ongoing extension of servi- ces to the peripheries of large cities and towns, one of ISS's most import- ant programs. 2.28 ISS first developed a network of clinics and hospitals emphasiz- ing curative services and specialist care. General physicians and para- medical personnel played a relatively minor role until recently and primary care received little emphasis. Patients were usually admitted to secondary or tertiary care facilities when hospitalization was required. ISS is now progressively moving towards decentralization based on basic health care units (CABS), which will provide primarily out-patient care to cover 70% of the needs of groups of 40,000 beneficiaries in the periphery of urban centers. CABS will be staffed by twelve physicians and an equal number of nurses. A second level of primary health care will be provided by local hospitals, which should satisfy an additional 15% of needs. Regional hospitals will cover another 10%, leaving no more than 5% of cases for the specialized hospitals. General practitioners rather than paramedical staff will be emphasized, in response to the higher expectations of ISS benefi- ciaries compared to the poorer clients of MS/SSS. A recently approved ISS information system is being set up but is unlikely to be fully operational in less than two years. D. Health Planning 2.29 Each level in the MS/SSS system consolidates information and plans from the area of its jurisdiction and reports to the next highest level. In each departamento, the SSS has ultimate responsibility for elaborating the annual plan of investments and operations. Each SSS in turn submits a plan to the MS, which reviews it to ensure internal consis- tency and compatibility with national policy. National health sector policy is the joint responsibility of the MS Planning Office and The National Planning Department's (DNP) Health Division. Major policy decisions are reviewed by the National Council of Social and Economic Policy (CONPES). In practice, the process is somewhat more centralized than this description implies, since the delegation of responsibility for planning has not been carried out even to SSS level in some departamentos because of their lack of planning capability. 2.30 ISS develops an annual investment and operations plan independently of MS/SSS. The National Health System calls for close coordination of the two plans and responsibility for this coordination lies with the Superintendency of Health Insurance. It appears, however, that little coordinated planning occurs. This topic will be addressed further in Chapter 5.  - 15 - 2.31 DNP is the governmen1: unit in charge of researching the issues and preparing policy options for discussion, first by CONPES and then by the Congress. There have been changes in emphasis, but also clear evidence of continuity in concept between one administration and the next. Public sector planning is well developed, and in the health field there is generally good coordination between DNP and MS. However, there is still much to be done to establish an adequate data base, and to initiate perio- dic reporting, monitoring and evaluation. The present overloading of MS planning staff with routine administrative tasks and the rudimentary plan- ning capability at SSS and regional levels constitute major constraints. Upgrading staff quality through training, making higher level jobs avail- able to planning staff, and establishment of a system to contract specific studies and technical assistance would help alleviate these problems. Once these tasks are sufficiently advanced, a progressive transfer of planning responsibilities to lower levels could be carried out to introduce planning from below within the general framework of sector strategy and priorities. E. Information Systems, Monitoring and Evaluation 2.32 The MS/SSS organization for data gathering and processing is in place in broad outline. The MS Planning Office Data Processing Unit has been set up and its operations are being computerized. At this point, however, there is no coherent and standard information system which would: (a) select the data to be gathered based on agreed procedures; (b) collate and cross check data; and (c) determine who should receive the data for analysis and decision-making. At present there is no way to verify data authenticity and reliability, which complicates substantially sector and sub-sector analyses. 2.33 The more advanced SSS's have established systems for data collec- tion and analysis, but the types of data collected and their uses vary, making inter-SSS comparisons difficult. There is no provision for periodic health surveys and ad hoc studies are few and far between. The National Health Survey is a commendable recent effort to obtain information on health indicators and service utilization. 2.34 Monitoring and evaluation of MS projects and programs is largely ad hoc and incomplete. Internal audits are formal exercises of relatively little utility, as are external audits by the Controller General's office. 2.35 The ISS information system is less well developed than that of MS. Information from the field is of questionable validity, often arrives late and is frequently too detailed and so organized that it cannot be put to managerial use. F. Operational Research 2.36 The MS conducts little operational research, with the important exception of INAS's National Health Survey. Operational research outside of MS has developed rapidly in Colombia and made important contributions to  - 16 - public health policies and programs. The main research activities are carried out at the universities of Valle, Antioquia and Javeriana, often through autonomous institutes. Operational research units at the University of Valle include the Program for Research on Operational Health Models (PRIMOPS), which has worked on urban health models; the Multidisciplinary Research Development Center (CIMDER) which designed a PHC model being tested in six departamentos; and the Simplified Surgery Group (CIS) which analyzed and costed surgery procedures by levels of complexity. At the University of Antioquia, the School of Public Health is working on management of health services, and at Javeriana University the Faculty of Interdisciplinary Studies (FEI) is reviewing health systems management and the relationship of health with other sectors. The Association of Medical Schools (ASCOFAME) has conducted analyses of the role of general practitioners, medical education, and manpower resources. 2.37 Health sector operational research is coordinated in principle by the MS Research Division, but in practice very little coordination is carried out, and this mainly with the help of ASCOFAME. Studies are conducted without much knowledge or regard for what is being done elsewhere in the country. A proposal to regionalize MS research has not been adopted and may be unwarranted given the paucity of research activities at MS. 2.38 More operational research is urgently required to guide sector planning. There is a pressing need to evaluate the impact of the MS/SSS and ISS service delivery models on health status, and to compare their cost-effectiveness. Research on the financing of the family welfare funds and other elements of the private sector is also necessary for implementation of the National Health System. Studies of the causes of chronic hospital deficits would shed light on possible savings at the secondary and tertiary levels. In short, operational research deserves high priority in sector development. III. SECTOR PERFORMANCE A. Coverage and Insfrastructure 3.01 A critical problem facing the Colombian health sector is how to achieve more complete coverage of the population at an affordable cost. Currently, only three-quarters of the population have ready access to services. The distribution of health-care coverage is as follows: Sub-sector Population (%) With coverage 75 MS/SSS 45 ISS 10 Other social security 5 Private sector 15 Without coverage 25  - 17 - 3.02 The MS/SSS are charged with providing health care to the entire population not covered by the Private sector or social security agencies. While the extent of ISS coverage is known precisely, and that of the private sector is probably accurate, MS/SSS and National Planning Department estimates, and data from the National Health Survey, all indicate that about one quarter of the population lacks health services. This appears to be a more reliable estimate than that of 36% made in 1974. However, the adequacy of MS/SSS coverage in relation to need is open to question, as will be discussed below in more detail. ISS currently covers about 40% of private sector employees, but provides full services to only 10% of their families. 3.03 MS/SSS has by far the largest service network of health facili- ties with 617 hospitals and 2000 health centers and posts. ISS owns only 28 hospitals, but contracts beds in 155 MS/SSS and private-sector hospitals, and has a network of about 50 basic care centers (CABS). Eighty-five percent of ISS beneficiaries are in the four largest cities. The distribution of hospital beds is as follows: No. of Beds % of Total MS/SSS 32,750 68 ISS 6,095 13 Other social securit:y 2,200 4 Private sector 7,400 15 Total 48,445 100 The MS/SSS hospital network is made up as follows: Type of Hospital No. No. of Beds % of Total Beds Local 453 10,900 33 Regional 101 9,850 30 University and 63 12,000 37 specialized Total 617 32,750 100 Overall, Colombia has 1.75 beds per 1000 population. This ratio is similar to that of Peru and Bolivia, and lower than that for Brazil, Chile and Costa Rica, all of which have almost 4 beds per 1000 population. 3.04 The most complete coverage is achieved in the three largest cities - Bogota, Medellin and Cali - where ISS and the private sector are most extensive. Moreover, the population of these three cities is better fed, housed and educated than average for the country, and water supply, sewerage and garbage disposal coverage is 95% for Bogota and 80% for Medellin, also clearly above average. Second highest in coverage are the  - 18 - departamentos of Antioquia and Valle (the capitals of which are Medellin and Cali) and the departamento of Caldas. Coffee-rich Antioquia and Caldas benefit from financial contributions to health services by the Coffee Growers' Federation. Next in coverage are the departamento capitals (with the exception of Barranquilla where demand for health services outstrips supply). At the bottom are the poor peri-urban areas (except Bogota, Medellin and Cali), rural nuclei with between 500 and 2,500 inhabitants, and the dispersed rural population. B. Manpower 3.05 Colombia enjoys a laudable record of innovation and accomplish- ment in training and deploying human resources to expand health care coverage. The main institutions responsible for these achievements are the Directorate of Manpower in the Ministry of Health; the Institute for Advanced Studies (ICFES); the SSS; the universities; and ASCOFAME. In 1977, Government created the National Council of Human Resources, composed of the Ministers of Health, Education and Labor, the directors of ICFES and ASCOFAME, and representatives of DNP, ISS, the Colombian Association of Nursing Schools, and the Pan American Health Organization. The MS Director of Manpower acts as secretary to the Council, which has met over ninety times during its five year existence. The Council provides overall policy direction for manpower development and deals with such questions as minimum requirements for the establishment of schools of medicine, dentistry and nursing. In addition, it concerns itself with the training of non- professional personnel, and with changes in curricula for health personnel at all levels. The Council has provided a useful coordinating mechanism and has been well attended except by ISS. The National Council has a counterpart at the departamento level. 3.06 Physicians. The number of physicians graduating each year has recently increased sharply. Annual output was approximately 400 from 1950 to the early 1970s, rising to 650 in 1978, 1153 in 1980 and about 1345 in 1981. About 1450 are expected to graduate in 1982, and 1600-1700 per year through 1985. This increase has occurred notwithstanding a total attrition rate of 58% of students entering first year medicine. Demand for medical school places has undoubtedly increased but the fundamental reason for the rising production is the proliferation of medical schools, which increased from 9 in 1973 to 21 at present. The number of physicians in practice has more than doubled since 1966 to a current level of about 16,000. Colombian authorities are unanimous that within 3-5 years the health system will be unable to absorb the 1,500 graduates a year currently projected. A number of problems are already evident: a shortage of places for the internship year; increased competition in private practice; and a growing scarcity of positions in ISS and the SSS, which together provide the bulk of physician employment. While the overall physician/population ratio is on the order of 1 to 1,700, this average conceals a range from 1 to 800 in major cities to 1 to 10,000 in isolated, rural areas. About 70% of physicians practice in the cities and departamento capitals where about 40% of the population resides, while the remaining 30% of physicians serve 60% of the popula- tion.  - 19 - 3.07 Physician salaries for eight hours a day range from approximate- ly US$750 in the SSS to US$1,350 per month in ISS. The gap between SSS and ISS seems to be widening, which makes ISS employment still more attractive, but it is significantly easier to obtain work with the SSS. Sixty-five percent of all registered physicians are employed part-time by the SSS while about 20% and 10% work for the ISS and family welfare funds, respectively. Many physicians have contracts with two or even more employers, in addition to their private practices. This arrangement makes it impossible for them to meet the terms of their contracts. There has been no systematic effort to project the need for physicians or specific kinds of medical specialists and to train the requisite numbers. Nor is information available regarding the number or geographic distribution of specialists. It is clear, however, that there is an oversupply in several specialties, especially the surgical sub-specialties. 3.08 Despite longstanding efforts by ASCOFAME, the Ministry of Health and ICFES to promote the training of primary health care physicians with a strong community orientation, today's graduate is primarily oriented towards individual health services in a sophisticated medical environment. Major questions on medical manpower are whether Colombia can afford to invest scarce resources in producing more physicians (especially specialists) than the market can absorb, and whether the current level of production is needed to address the country's most pressing health problems. These questions will be addressed in Part Two of the report. 3.09 Professional Nurses. Nurses with a degree or licenciate, requiring 3 or 4 years of training, have always been in short supply in, Colombia. A decade ago, there was only one nurse for every five physicians, i.e., one per 10,090 population. In recent years, the number of nursing schools has risen from 9 to 23, with a corresponding increase in production. Nevertheless, only approximately 5,000 (one per 5,500 population) are thought to be practicing. Although their training still emphasizes clinical and bedside care, most occupy administrative positions in large hospitals. Few are engaged in patient care and even fewer are trained in midwifery. 3.10 About 2,000 nurses work for the SSS, while ISS employs about 800 and the family welfare funds approximately 300. With the advent of the obligatory year of social service, nurses are now more frequently found in regional and local hospitals. Most nurses, however, prefer to work in the cities, with the result that auxiliaries are the de facto nurses in most peripheral health installations. Professional nurses are paid between US$300 and US$650 per month. 3.11 The nursing profession has never found a proper niche in the Colombian health system. The trend away from bedside care, and lack of training in midwifery, make the nurse suitable mainly for duty in specialized hospitals. Auxiliaries' training appears more relevant to primary health care, emphasizing patient care, MCH and family planning. In addition, they are less likely to migrate or leave nursing, and more inclined to live in rural areas. There will continue to be a need for professional nurses, especially in specialized hospitals, but the auxiliary nurse represents a more available and cost-effective resource.  - 20 - 3.12 Auxiliary Nurses and Nursing Aides. There are about 20,000 auxiliary nurses and approximately 15,000 nursing aides, about one per 790 population. The number of auxiliary nurses has more than doubled since 1970. Applicants to auxiliary nursing schools must be female, between 17 and 30, and have completed two years of high school. The training program lasts 18 months, with 12 months of theory and practice in school followed by six months in a hospital. Auxiliaries are the mainstay of MS rural and urban health posts and centers, and also provide most of the nursing care in local, regional and university hospitals. Three functions of the auxiliary deserve particular emphasis. First, the auxiliary is the key person in the rural health post since a physician is available only one day a week. Second, she is responsible for supervising the health promoter who works in the community and performs key functions in the MS effort to expand coverage. Third, thie auxiliary is trained in midwifery. 3.13 Auxiliaries are paid US$150-250 per month, approximately half the salary of a professional nurse. The highest salaries are offered by ISS and the National Provident Fund, with the added attraction of work in large cities. The unfortunate result is that auxiliaries trained by the SSS often leave their serviice, in effect subsidizing the social security and 'private sectors. However, approximately three-quarters of auxiliaries work in the SSS. There is general agreement that the auxiliary plays a key role in the health system and that more of this category of health worker are required. Financial constraints at the departamento level, however, hinder expansion of auxiliary manpower. 3.14 Nursing aides are a residual manpower category of clinical nursing assistants usually without formal training. Their numbers are indeterminate though perhaps around 15,000 persons, predominantly women. Many of them acquire a variety of nursing skills and become increasingly valuable with experience and in-service training. A small proportion eventually become nurse auxiliaries. Experienced aides not infrequently conduct deliveries and carry out other important tasks. 3.15 The Health Promoter. These women are trained for three months, assigned to villages surrounding their homes, and supervised by auxiliary nurses. Results have been mixed depending on the quality of training, supervision and support. There have been problems with the selection of candidates (for example, many were too young to be credible), and with attrition. The number of promoters has consistently lagged behind target, in part due to the high attrition rate. Promoters receive about US$80 per month. 3.16 Government figures show there are about 5,000 promoters but the actual figure may be lower. There are not enough to cover the nine million rural and urban populatior. who are the target of efforts to expand coverage. Furthermore, the number of promoters has not increased for about three years as a result of attrition and financial constraints. 3.17 The Multidisciplinary Research Development Center (CIMDER) rural health care model relies even more heavily than MS/SSS on the work of the promoter, on home visits as opposed to fixed facilities, and on community participation. Promoters receive intensive training and are responsible for nutritional surveillance, immunizations, health, nutrition and sanita- tion education, and deliveries. Promoters operate from their own homes and  - 21 - from the homes of villagers, doing away with the need for health posts, and spend most of their time making field visits. They travel with kits containing instruments, tapes and materials which enable them to check the development status of infants and children, do basic blood and urine analyses, and provide simple treatment, including basic drugs. A schedule of visits and records of actions taken are made systematically and cross- checked by supervisors. Operational costs have been tentatively estimated at US$7 per beneficiary, or about one-half of those of the MS/SSS PHC approach in rural areas. CIMDER originated in Valle and is now being tested in six other departamentos. Preliminary reviews indicate it is being enthusiastically accepted bv rural populations. 3.18 Traditional Practitioners. There are an unknown number of several types of practitioners in Colombia: teguas, curanderos, espiritistas, inyeccionistas, and comadronas (traditional birth attendants). Comadronas assist 23% of all births according to the National Health Survey. There have been few attempts to incorporate this manpower in the modern health system. They represent an important potential resource currently competing rather than cooperating with MS efforts to increase coverage. C. Pharmaceuticals 3.19 There are 350 pharmaceutical laboratories in Colombia, but 80 of these account for about 90% of annual production. About four-fifths of these 80 are at least partly owned by foreign interests. Market share are widely spread, with the largest laboratory supplying only 5% of the drug and chemical needs of the country. This is the result of very strong competition regarding quality, prices and availability. Domestic production of about 8,500 certified drugs satisfies 95% of Colombia's needs and leaves a surplus for export mainly to Ecuador and Central America. 3.20 Domestic production of basic chemicals (such as acetylsalicylic and citric acid, aluminum hydroxide, glycerine, and alcohols), however, has developed more slowly, accounting for only 40% of domestic demand. A similar market share prevails for complementary products, such as bottle and carton containers, lubricants and plastics. The Andean Pact countries are studying the possibility of increasing domestic production of basic chemicals and complementary products as a series of joint ventures. Additionally, a list of generic drugs has been compiled and government approval will be sought in order to limit the number of drugs supplied to government agencies and government-financed services. This is likely to reduce costs substantially. 3.21 Industry regulation, based among other things on a national drug registry including quality control, storage, distribution and pricing is carried out by MS at the national level and by SSS at lower levels. Purchases by official sub-sector entities above US$4,000 must be based on domestic bidding procedures ani on an approved drug list. Physicians are supposed to order generic drugs (although in practice this rule is often not followed) and use brand names only exceptionally. Purchases are made directly from laboratories.  - 22 - 3.22 The legislation to promote adequate distribution and consumption of basic drugs of good quality at reasonably competitive prices is in place. Effective implementation of the law is now needed specifically to limit official sub-sector purchases to generic drugs based on the already approved list. Examples of tighter controls in this direction are beginning to appear, e.g. the ISS PHC services (basic care centers or CABS) use only 80 basic drugs and in some ISS clinics medical committee authorization is required before brand name drugs can be purchased. 3.23 Further information is required regarding drug procurement, storage, distribution and utilization in order to complete the analysis of these critical resources. D. Health Service Utilization 3.24 The 1977-1980 National Health Survey interviewed a representa- tive sample of 12,000 households on health status, use of health facilities and recent morbidity. One third of respondents said they had been ill during the two weeks prior to the interview. Three-quarters of these individuals felt the need for a medical consultation, and 11.5% of the population actually consulted someone. The distribution of consultations by provider is given in Table 10. Physicians accounted for almost three- fifths of all consultations. Social security system beneficiaries were more likely than non-beneficiaries to consult a physician (79% versus 53%). Residents of the Pacific and Central regions and those with lower incomes were less likely than others to see a physician. Physicians provided 80% of the care in Bogota but only about 50% in the Pacific and Central regions. The most frequent reason given for not having a medical consultation was economic. 3.25 The private sector plays a major role in health service deli- very (Table 11). Almost 31% of consultations took place in a private office (usually a physician's), while the MS/SSS system accounted for only 20% and the social security system (ISS, CAJANAL and the family welfare funds combined) for 10%. Those with higher incomes were less likely to visit MS/SSS facilities and more likely to go to social security or a pri- vate physician. A significant proportion of social security beneficiaries consulted private practitioners and MS/SSS facilities. These data high- light the pluralistic nature of Colombia's health system. 3.26 Forty-five percent paid less than US$2 for a medical consulta- tion and three-quarters paid less than US$4. Three times as many indivi- duals purchased drugs as had a medical consultation, nearly a half paying less than US$2 for the drugs received. Overall, the survey data demonstrate a high level of effective demand for curative health services. 3.27 ISS Out-patient Services. ISS beneficiaries receive a larger quantum of these services than MS/SSS recipients. During 1980, ISS-insured beneficiaries (including dependents) averaged 3.2 physician visits. The total number of out-patient visits for 1980 exceeded 8.5 million.  - 23 - 3.28 MS/SSS Out-patient Services. Recent MS/SSS out-patient service data are as follows: No. of initial Total No. No. per Year physician visits of visits % Coverage hour 1980 5,713,553 11,155,498 28.8 2.5 1981 5,862,360 11,108,629 28.8 - MS/SSS target coverage is based on 70% of total population; less than three-tenths of the target was seen by a physician. MS provided about 0.5 physician visits per person per year. Moreover, MS/SSS provided only 11 million physician visits overall compared to 8.5 million for the 10% of the population covered by ISS. Adding patients seen by nurses and promoters does not greatly change this comparison, since in 1981 they accounted for only 4 million visits (2 million each). The total of 15 million visits is less than double the number for ISS, which serves only about one fifth the population. MS/SSS promoters averaged only 0.3 visits per hour. ISS and MS/SSS physicians saw 2.6 and 2.5 patients per hour, respectively. 3.29 In-patient Services. National Health Survey data show 5.7% of the population was hospitalized in a one-year period. Sixty-five percent of hospitalizations occurred in MS/SSS hospitals, 18% in private institu- tions, and 14% in ISS facilities. These findings fit with ISS data showing 295,000 hospitalizations for 1979 and the MS/SSS figure of 933,000 for that year. Twenty-six percent of those hospitalized were operated on, and deli- veries accounted for another third of hospitalizations. 3.30 There are indications of increasing efficiency in the Colombian hospital system over the past decade, but the potential for further econo- mies remains high. The number of MS/SSS and ISS hospitalizations has grown little in the last five years, despite population increase. The average length of hospital stay declined from 8 to 7 days between 1973 and 1979. Increased utilization of facilities shows up in the increase in discharges per bed-year from 22.9 in 1970 to 29.8 in 1979. Discharge rates, however, differ widely among institutions. In 1980 they were 48 for the ISS; 32 for the SSS; 23 for other social security funds, and 34 for the private sector. Part of the explanation for the differences may be that the poorer SSS clientele is less likely to receive adequate home care, thereby lengthening hospital stay and reducing the scope for patient turnover. However, discharge rates vary among SSS hospitals, being 36 patients per bed-year in local and regional. hospitals, 31 in university, and only 18 in other specialized hospitals irt 1979. Rates should be highest in local hospitals which attend the less complicated cases and decline progressively in line with the degree of specialization. 3.31 Hospital bed occupancy tends to decrease with shorter average stay and to increase with higher patient turnover. In general, hospital bed occupancy is low in Colombia, especially for regional and local hospitals. This underutilizat:ion of costly facilities was recognized by Colombian health authorities who greatly restricted hospital construction until 1979. There has been a marginal increase in capacity (less than 10 percent) since then. Patients bypass mediocre secondary hospitals in  - 24 - search of a wider range of services and better quality treatment. This habit leads to high occupancy rates, of between 80% and 85%, for university hospitals in the large cities and to low occupancy rates, less than 40%, for regional and local hospitals. Data from Caldas and Bogota illustrate the problem: Type of hospital Occupancy rate (%) Average stay (days) Caldas: Local A 39.2 3.2 Local B 53.8 4.2 Regional 60.1 5.1 University 79.5 - Bogota: Local 13.6 All 66.8 3.32 Low utilization stems mainly from low demand among the poor. In Bogota, there were 180 private in-patients, 75 ISS in-patients, and only 52 MS/SSS in-patients per 1,000 target beneficiary population of each group in 1979. The high rate for private-hospital patients reflects a relative abundance of hospital facilities, ability to pay for services and leave their jobs for the days required, and more awareness of disease among the affluent. The ISS beneficiaries perhaps over-use hospitals. The popula- tion served by the MS/SSS may have more difficulty gaining admission and affording the time for hospitalization. Other important factors leading to relatively low levels of utilization are poor census management - (admitting patients on Monday and discharging them on Friday); and lack of adequate management, personnel, equipment and supplies at lower levels which encourages patient self-referral to specialized centers. 3.33 Other important health service indicators are immunizations and institutionalized delivery coverage. Immunization coverage for children under five was as follows: BCG DPT Polio Measles 1978 47.3% 33.6% 30.8% 40.0% 1979 49.5% 30.1% 35.3% 46.0% About 55% of births are institutionalized in Colombia. The main providers. of obstetrical services (wherever conducted) are as follows: Physician 47.6% Traditional birth attendant 23.4% Nurse 14.4% Health promoter 1.0% Others 13.6%  - 25 - E. Service Delivery Constraints 3.34 MS/SSS Primary Health Care Model. The 1975 National Health Plan (PNS) was to extend health care to the nine million Colombians who lacked access to services. By 1979, 700 MAC/UPA units were successfully instal- led, but performance by departamento varied greatly. A maximum of 1.9 million of the three million target, 63%, had been reached. Each rural promoter was able to cover a maximum of 500 persons of the 1,000 originally planned, while urban promoters reached up to 2,000. There was significant understaffing--an average of only 2.6 of the six promoters needed for one MAC/UPA were actually in place. A shortage of operating funds resulted in decisions not to replace promoters who resigned. The cost per capita per annum was US$13 for delivering preventive, promotive and curative services by promoters. Whether this expense resulted, as planned, in a progressive reduction of morbidity and mortality is unclear. 3.35 The effectiveness of health promoters as primary care providers remains unproved, except in the case of CIMDER. Supervision was deficient at all levels according to the 1980 evaluation. Three out of five promoters said they had difficulty carrying out their duties. Half of them lacked a procedures manual., 42% did not receive supplies regularly, and 60% cited lack of transport: facilities. Only 22% provided drugs to the sick and 70% stated they were not authorized to do so although this is standard protocol. In spite oi these drawbacks, two-thirds of families requested services from the promoter, 98% of whom received them, of whom 99% were satisfied with the services. Four out of five persons interviewed said they felt better attended due to the presence of the promoter. 3.36 Half of the nurse auxiliaries interviewed reported supervising each promoter every two months or less but 58% said they had problems providing supervision, 70% citing lack of transport. Almost half had not received refresher training for the MAC/UPA system and three out of five lacked a procedures manual. A third of auxiliaries said they did not receive supplies regularly; over half in the case of drugs. Half of the auxiliaries did not dispense drugs and 58% said they were not authorized. Finally, only three-quarters of' the auxiliaries were themselves supervised. 3.37 The 1980 evaluation provided valuable clues regarding how the system could be strengthened. Effective supervision systems already exist in Valle and Caldas, where mob:ile physician and nurse teams supervise local hospitals and PHC units. Second, the effectiveness of auxiliaries and promoters would surely increase with a stronger support system -- procedure manuals for all, timely availability of drugs and supplies, and transporta- tion facilities. Third, evidence from the evaluation and from elsewhere indicates the need to strengthen promoter and auxiliary basic and in-service training. 3.38 The evaluation did not provide definitive answers about the effectiveness of the MS/SSS PHC model. The results demonstrated that the community used and appreciated the services provided by the promoter. On the other hand, neither the auKiliary nor the promoter was allowed to dispense basic drugs in the majority of cases, despite the fact that this was planned and is a vital factor in PHC. Another problem with the system  - 26 - is the diversion of funds appropriated for MAC/UPA units to local and even regional hospitals, to cover their own chronic deficits. Strengthening training, supervision and support is very likely to increase effective- ness. The cost-effectiveness of the model, in the absence of cost account- ing and program impact data, remains unclear. Since 1980, MS has streng- thened regional supervision teams in the expanded DRI program, and included a community participation component to monitor the quality of health care services. 3.39 ISS Health System. Before the 1977 reform, ISS provided cura- tive services of mixed quality at high cost, leading to frustration of both providers and beneficiaries. Administrative flaws resulted in low produc- tivity combined with continuous pressure to increase appropriations. Many health problems resulting in hospitalizations could have been handled effectively and at lower cost on an ambulatory basis. Medical specialists devoted considerable time to problems that could have been delegated to paramedical personnel. Excessive hospitalization was evident: 208 admis- sions per 1000 beneficiaries in 1975 and double that rate in one departamento. Beneficiaries complained of impersonal and fragmented care. Finally, escalating costs had produced a financial crisis that was expected to worsen. 3.40 Considerable progress has been achieved in overcoming the pro- blems described above. The institution is once again on a sound financial footing. ISS now has fewer hospital beds, higher bed occupancy rates, and the average length of stay declined from 5.9 to 5.4 days between 1977 and 1981. The CABS program is an important initiative aimed at delivering pri- mary care and reducing costs. Finally, there has been a continuous decline in the number of consultations per beneficiary, from 3.9 in 1977 to 3.2 in 1980. Much remains to be done in implementing the numerous administrative reforms, but a good start has been made. 3.41 Coordination of MS/SSS and ISS. The different origins, target populations and operating principles have resulted in little coordination between the two main institutional providers. The principal means of coo- peration is ISS contracting beds in MS/SSS hospitals and providing out- patient services through the MS/SSS infrastructure. If an area has fewer than 1,500 beneficiaries, ISS contracts services. If there are more bene- ficiaries, ISS contracts private physicians directly. ISS facilities have usually been constructed solely on the basis of its beneficiaries' needs. This practice results in ISS hospitals and out-patient clinics often being located close to MS/SSS facilities. Under-utilization of both results for lack of effective demand and in turn leads to significant economic losses for both systems. 3.42 The issue of duplication and overlapping service areas has become more serious in view of ISS plans to develop a network of CABS. Some of those already constructed cover the same areas as MS/SSS facili- ties.  - 27 - 3.43 The Superintendency cf Health Insurance, created in 1979 to increase coordination between ISS and other elements of the National Health System, has not been effective. The small staff lacks operational support, adequate funding and the power to enforce its recommendations. 3.44 MS/SSS Referral System. The MS/SSS referral system does not function effectively. Physicians often refer patients directly to tertiary level hospitals because of the lack of specialists at regional hospitals. Patients often seek care at the most specialized institution available, bypassing lower level services even if they have to travel long distances. This pattern results in tertiary (and secondary) care facilities providing a high volume of primary health care services, especially in cities where distance is not a major factor. This in turn leads to under-utilization of lower levels of the service pyramid and high costs when secondary and tertiary care facilities provide PHC. For example, the cost of a normal out-patient consultation and an emergency consultation at a health center are estimated at only 20% and 60%, respectively, of the cost of one local hospital bed-day. The latter in turn would cost only 40% and 10% of the bed-day cost at a regional and specialized hospital, respectively. IV. HIRALTH CARE FINANCING A. General Overview 4.01 Public sector and social security expenditures for health care inevitably are affected by a government's overall economic policies and by general economic conditions. Over the past decade, the Colombian govern- ment has pursued economic policies oriented more toward achieving stabili- zation than promoting development, prompted largely by the need to control strong inflationary pressures. Central government expenditures have varied little from the comparatively low level of 10-11% of GDP throughout the seventies. The social security sector also has shown very little relative expansion, although it recovered from a mid-1970s economic recession. About 2% of central government expenditures were deficit-financed in 1981, with half the deficit covered by external borrowing. The government's deficit, as well as its debt service burden, are projected to increase over the next several years. 4.02 MS has nominal responsibility for the entire National Health System, but exercises financial control over only 25% of health care expenditures. MS concerns itself primarily with the budgets of its own, directly-administered health-care programs and those of the SSS. MS financial control is very limited over the two largest of its affiliated institutes, Family Welfare (ICBF) and Municipal Development (INSFOPAL) which account for 20% of the MS budget (see Table 15). MS effectively controls only the National Hospital Fund (FNH), the National Health Institute (INAS), and several minor entities. Lack of budgetary control hampers MS efforts to improve efficiency through program budgeting.  - 28 - 4.03 Health services are a relatively minor function of the ICBF, which has independent revenues from a payroll tax. INSFOPAL operates municipal water and sanitation projects in medium-size communities, a task obliquely related to health-sector activities. Notwithstanding these reservations concerning ICBF and INSFOPAL, the following analysis of public health sector revenues and expenditures will report aggregate data for public health care in accordance with the budgetary definition of this sector in Colombia (i.e., the 'sector oficial'). 4.04 Public health services provided by MS/SSS and affiliated institutes account for approximately one-third of total health sector expenditures in Colombia, as the tabulation below shows: Composition of Estimated Total Health Sector Expenditures, 1978 As % of GDP As % of Sector Public sector 1.7 31.5 MS/SSS 0.7 13.0 Institutes 1.0 18.5 Social Security 1.9 35.2 Private Sector* 1.8 33.3 Total 5.4 100.0 Source: MS, Gasto en Salud, 1978 * Estimate, based on preliminary data from National Health Survey, 1977-1980. The MS and SSS directly administered about half of that proportion in 1978. The institutes accounted for about half of all public sector financing. Social security health programs, moreover, generate somewhat more services than does the health ministry. Payments for private health care, other than those provided through public health and social security institutions, also account for about one-third of total health sector financing. 4.05 In 1978 (the last year f or which comprehensive expenditure data are available), average health care expenditures per person for the population covered by social security programs were roughly US$108, about four times the average expenditure of US$28 per person dependent on public health care, as shown below:  - 29 - Population Coverage and Average Health Care Expenditure per Person in 1978 Population Covered Expenditure Total Percent per Person (millions) Public health (MS/SSS) 13.4 52.3 $ 28 Social security 3.8 15.0 $108 Private services 2.6 10.0 * Sub-total 19.8 77.3 $ 47 Sub-total without coverage 5.8 22.7 * TOTAL 25.6 100.0 $ 46 Sources: Based on GDP and population estimates in World Bank, World Development Report, 1980 and above health expenditures as % of GDP for 1978. * Cannot be separately estimated. The tabulations also suggest that 5.8 million Colombians were without coverage by modern health services in 1978. This estimate, equivalent to about 23% of the population, probably reflects reality more closely than higher estimates used by MS in 1975. 4.06 The 1979-1982 National Integration Plan repeatedly refers to the "harmonious development" of the National Health System through the "effective coordination" of its two sub-sectors, public health and social security. However, neither the statements on health sector objectives nor the analysis of sector performance over the past decade explain the failure to implement the National Healt:h System or suggest what specific financial measures should be taken to brling about the coordination of public health and social security. The healt:h sector priorities refer to population segments and areas of the count:ry served primarily by MS/SSS. The "Plan de Inversiones y Gastos de Funcionamiento del Sector Salud 1979-1982," which is part of the Plan's statement: of objectives, also refers exclusively to MS expenditures. 4.07 An analysis of MS expenditures over the past decade, however, shows that its share of GDP declined from 1.06% in 1971 to 0.67% in 1980. Its share of the Central Government (CG) budget varied between 6.7% and 9.3% (Table 12). An analysis suggests that the MS share is determined not by the current but by the previLous year's level of CG revenues. MS expanded coverage, 1971-80, wilh the result that average expenditure per person in real terms has declined over the past decade, as the tabulation below illustrates:  - 30 - Ministry of Health Expenditures, 1971 and 1980 Index 1971 1980 1971=100 MS expenditures* 1,435 1,717 119.7 (as percent of GDP) (1.06) (0.67) Population covered** 8,587 13,826 161.0 (as percent of total) (40.0) (52.5) Per capita expenditure 167 124 74.3 Source: Based on Table 13 source data, deflated by Banco de la Republica government services price index. x Expenditures (in millions) of 1971 constant pesos. ** Population (in millions) covered based on informal MS estimates. With a 20% increase in expenditures, MS expanded health services to cover five million more people in 1980 than in 1971. The reduction in per capita expenditure implies an increase in efficiency since the number of medical visits per person covered has remained constant. 4.08 Chronic hospital budget deficits and underutilization of many existing facilities suggest that underfinancing of the recurrent cost of public health care is the dominant issue facing MS/SSS. Any future expansion in health facilities will have to take into account the financial burden implied in staffing and maintaining the facilities. B. Public Health Sector Revenues 4.09 Public health care is financed from both federal and departa- mento sources. The former account for about 65% of total revenues and the latter for 35% (Table 13). Two-thirds of total revenues derive from sources whose funds are earmarked for specific purposes, as explained below. 4.10 Tables 13 and 14 represent a consolidated balance sheet for public health care in Colombia 1975-80. These tables are based on the Minister of Health's report to Congress for the 1979-80 fiscal year. The data present actual revenue and its distribution through 1978 and budgeted amounts for 1979 and 1980. They are not expenditure data. 4.11 MS, after internal budget hearings, prepares an annual budget reqaest that consolidates the SSS and institute requests with its own request for federal funds. MS then receives an initial allocation after  - 31 - review by the Ministry of Finance and passage of the budget law by Congress. This allocation typically is based on conservative projections of government revenues. In 1980, the initial MS allocation represented about 83% of its eventual total, non-earmarked allocation from CG. The difference represents extraordinary allocations distributed by the CG once its actual receipts exceed projected revenues. In order to receive extra- ordinary allocations, however, any government agency must submit special requests with specific justifications. This process becomes a major preoccupation of senior MS officials as programs run short of funds. The extraordinary allocations varied greatly over the years 1975 to 1980. Their size depended on resources available at the CG level, on competing requests by other agencies, and on MS success in submitting its requests in a persuasive manner. 4.12 The second major source of MS revenues is the 'situado fiscal', the earmarked allocation of a proportion of the CG's total ordinary revenues to primary education and primary health care. In accordance with Law No. 46 of 1971, the Ministry of Education receives 11.1% and the MS 3.9%, for a combined share of 15%. (The law permits the share to be increased-up to 25%, but this is not being contemplated). The CG uses its initial conservative projection of ordinary revenues as the base for calculating the size of the 15% share. There may be other limitations of the base total. The situado fiscal allocation to MS in 1980 represented about Col$4 billion. 4.13 MS also receives internal and external credits from the CG covering 2.6% to 11.1% of total revenue in the six years shown in Table 13. There has been a shift from domestic to international credit sources, but foreign aid programs no longer represent the major source of inter- national credit. Instead, the Colombian government is now using commercial loans from consortia of foreign. banks to finance investment projects and operating expenditures, including many in the public health sector. 4.14 Departamento sources of revenue for health are channeled directly to the SSS's or institutes, depending on the uses for which they are earmarked. Revenues from lotteries and alcoholic beverage taxes are generally intended for hospital services under SSS control. Service revenues include user charges in medical care facilities, which are usually forwarded to the SSS. User charges are collected and kept by INSFOPAL and other institutes. Considering the potential significance of user charges, their inadequate management is a major cause of concern. 4.15 The growth in user charges has partially offset the decline of revenues from other local sources, such as alcoholic beverage taxes, trans- fers from lotteries, and direct contributions from departamento and munici- pal governments. These declines since 1971 are generally attributed to the introduction of the situado fiscal. The net impact of earmarking a portion of CG ordinary revenue for health has been to transfer the financing burden from local to national government. The MS is currently studying how to increase the yield from lotteries and alcoholic beverage taxes. Congress, at the instigation of MS, in 1981 legalized a nationwide numbers game with an expected net revenue of Col$2 billion annually, and introduced a civil registry tax expected to yield Col$300 million. Thus some Col$2.3 billion  - 32 - in additional revenues will be earmarked for MS/SSS. MS expects to use this new revenue, an increment of about 4% to health sector resources, to finance hospital operating deficits rather than to expand PHC coverage. Given the distortionary effects of revenue earmarking, this new revenue source will not contribute much to health sector resources and could actually reduce sector productivity and efficiency by supporting high-cost care. 4.16 Legislation to increase lottery income allocated to MS/SSS was defeated in Congress. Lotteries are run by the departamentos to support public health services and, more broadly, social welfare activities. Funds allocated to SSS's do not suffice to support existing health care facili- ties adequately. Most SSS support now comes from the federal level and makes them partially immune to state-level political influence. The new federal levies may raise expectations for more federal support while undermining the tax effort which departamentos and municipios should make to contribute to the expansion of primary health care coverage. 4.17 Two other new sources of revenue for MS/SSS deserve mention. One is the special revenue fund created by the Coffee Growers Federation. The Federation is financed by a 3% tax on coffee exports. When its income surged,during the boom in coffee prices in the mid-1970s, the Federation in 1976 invested a surplus of Col$1.6 billion in Electric Utility Bonds. The 16% interest these bonds pay yields an annual income of Col$220 million which is distributed to coffee growing regions in proportion to their share of total coffee production to improve and expand local public health facilities. Project implementation is decided upon in cooperation with regional SSS authorities, based on a formal contractual agreement between the Federation and MS. The expenditure of funds, however, is earmarked for investment purposes, leaving the responsibility for future increases in recurrent costs with the MS/SSS. 4.18 Another major new source of public health sector revenue was the 1974 law creating a 2% payroll tax for the support of ICBF youth welfare services. This potentially increases ICBF revenues to a level higher than total current MS/SSS revenue from all other sources, although this is far from being reached. The current level of income appears only among 'other sources' in Table 13. In Table 14, ICBF is shown as receiving Col$5 billion, equivalent to 16% of total MS/SSS allocations. Data supplied by ICBF indicate that 60% of this income came from the 2% payroll levy, an amount of about Col$3 billion in 1980. 4.19 It is less important today in Colombia to increase resources available to the health sector than it is to rationalize what has become a complex and unmanageable revenue system. The combination of ordinary and extraordinary allocations, an earmarked situado fiscal, earmarked local revenues from beverages and lotteries, earmarked payroll taxes in ICBF, earmarked Coffee Federation revenues and fresh earmarked revenues from a national lottery and registry tax add up to a revenue base unresponsive to modern program budgeting techniques. Central managers cannot effectively re-program the health sector to shift from hospital-based tertiary care to community-based primary care if they cannot control and program the financial resources of the sector. It would of course be unwise to attempt to change the revenue base overnight; it must, however, be recognized that  - 33 - the current financing scheme is itself a major source of inefficiency in the health sector. As a beginning, the emphasis on earmarking of revenue sources for health purposes needs to be reexamined. Of particular concern will be a determination as to whether the health sector can, by reason of effective presentations to higher budget authorities of proposed deployment of resources and the benefits to be derived therefrom, justify its planned programs of expenditures. MS, in using five to ten percent of central government revenues, and a significant share of local resources as well, would in a program-budgeting environment have to explain the pattern of expenditures and overall qualitative benefits to those officials charged with assuring the best use of available resources for all purposes. 4.20 Because of the changes in the MS/SSS revenue base which are anticipated for 1982 and 1983, it has not proved feasible for this report to prepare projections of health sector revenues for these and future years. Planning authorities in Colombia may wish to make such projections a high-priority task for the near future. C. Public Health Sector Expenditures 4.21 The expenditure side of public health care is presented in Tables 14 (1975-80) and 15 (1978 only), between which substantial differences exist and cast some doubt on the overall reliability of expenditure estimates during this period. Table 15 shows that in 1978, Col$15 million were spent on public health care, whereas Table 14 shows an allocation for that year of only Col$12 billion. The MS balance sheet in Table 14 may not capture all flows of funds. The greatest discrepancy between Tables 14 and 15, at the SSS level, shows allocations of Col$5.4 billion (Table 14) and Col$9.4 billion (Table 15). The SSS probably have sources of income that do not appear in the MS balance sheet. 4.22 The distribution of allocations shown in Table 14 indicates that MS/SSS account for about 59% of the total and the institutes for 41%. These proportions have varied considerably over the period 1975-80, largely because recorded allocations to the SSS were stagnant in absolute terms while declining in relative terms during all of the previous administra- tion. In 1979 and 1980, SSS allocations increased sharply, apparently in response to the increasingly serious shortages of operating funds at the departamento level. 4.23 Operating expenditures represented 82% of the total for 1978 and investment 18%. Somewhat more than half of the former were personnel costs, with the remainder used for materials and supplies, e.g., drugs and food. 4.24 The three economically most advanced departamentos (Antioquia, Atlantico, and Valle) and the capital district (Bogota) in 1980 received a combined share of 33% of MS/SSS expenditures, considerably less than their population share of 43%. Their local sources of revenues were more generous than in the rest of the country, perhaps because they include the four major cities (Medellin, Baranquilla, Cali and Bogota). These more developed regions rely on public sources for only 25-30% of their total  - 34 - health care expenditures. The balance is accounted for by social security and private financing. In contrast, the MS/SSS accounts for 40-85% of total health care expenditures in the remaining 21 departamentos of Colombia.3/ 4.25 As with the revenue side of the health accounts, the expenditure side is heavily earmarked. The large share devoted to physical infra- structure in 1978 (28%), an investment share far larger than most other countries, may have arisen because excessive funds were earmarked for investment rather than recurrent costs. Similarly, locally-collected reve- nues are virtually always assigned to local projects, sometimes only for construction and often only for hospital-based services. Thus earmarking on the expenditure side also inhibits effective action to make the health system more efficient by promoting primary health care models. As with earmarking of revenues, planning authorities in Colombia may wish soon to re-evaluate current financing and expenditure practices in the light of the need to achieve more efficient use of health-sector resources. Effective health planning requires projections for health-care expenditures for the period 1982-86, identifying needs at primary, secondary and tertiary levels of care. D. Public Health Sector Investment 4.26 The investment component of MS/SSS allocations rose sharply during 1979-80, as noted above. This increase, however, may be due to a growing tendency in the public health sector to classify all expenditures by certain programs as investments on grounds that they build human rather than physical capital. Most notable is the ICBF allocation for 1980, shown as Col$5 million in Table 14. There is no justification within convention- al accounting practice for listing all ICBF expenditures as investment. The same questionable accounting practice is used for the national malaria program (SEM), showing all of its expenditures as investment, and for cer- tain programs within the MS, such as MCH, vaccination, and categorical disease programs. Only about one-third of all expenditures shown in Table 18 are physical investments as conventionally defined. If this ratio is applied to the investment data shown in Table 14, then physical investment is reduced to about 13% of total allocations in 1980, still a high invest- ment rate compared to other countries. 4.27 The remaining investment allocation of close to Col$5 billion in 1980 (see Table 19) includes primarily INSFOPAL, INAS, and FNH. About Col$3 billion was allocated to INSFOPAL and INAS for water and environ- mental sanitation projects. Three-fourths of this total, in turn, went to the larger towns, with only about Col$800 million allocated to the rural areas. All of the latter funds, administered by INAS, were actually spent, suggesting that this program does not suffer from inadequate absorptive capacity. 31 Eduardo Wiesner, et al, Finanzas Intergubernamentales en Colombia (Bogota: Departamento Nacional de Planeacion, 1981), Table IX-1, p. 155. (The "Wiesner Report".)  - 35 - 4.28 FNH was able to spend only 71% of its allocated funds in 1980. Hospital and health center construction and equipment projects are parti- cularly cumbersome in their financial and administrative requirements. Project proposals typically originate at SSS level. While MS has a detailed facilities investment plan (see below), the projects proposed are not necessarily or even typically those envisioned in the plan. Rather, they emerge from regional and local demands which are often politically motivated. Many construction projects, for example, are written directly into the budget law by members of Congress obligating MS and FNH to carry them out. 4.29 Once a project is approved by the MS Planning Division, based on project specifications and cost estimates prepared by the FNH, it is then published to solicit bids. This process takes time and is also subject to political influence. The actual construction and acquisition of equipment involve further difficulties, particularly since time lags cause cost increases beyond the approved estimates. 4.30 MS/SSS in 1979 were financially and administratively responsible for 617 hospitals with close to 31,000 beds (Table 20). The operating cost of these facilities was about Col$8 billion, about one-third of the ministry budget and 70% of the MS/SSS share in the total. Hospital operating costs represent the major financial obligation of the public health sector. Hospitals have no common standards of fees they charge, even though these are a significant source of revenues at the SSS level. Each hospital purchases drugs and other expendable supplies on the open market. Hospital administrators are left largely to their own judgement as to fees they should charge and prices they should pay. 4.31 The 1978-1982 MS health care facilities investment plan placed high priority on the PHC level. In its first three years, most of the 44 new health centers and 282 health posts planned were constructed. An even larger number of existing health centers and posts were expanded and remodeled (Table 21). Only about a third of new local hospitals planned were built during the first three years, and expansion and remodeling of existing local and regional hospitals was also behind schedule. Investment expenditures for equipment heavily favored the upper levels in the referral pyramid. 4.32 Assuming marginal operating cost of 50-75% of new facility costs, the net increase would be about Col$1 billion annually. Based on MS/SSS 1979 total operating costs, this would represent a necessary in- crease of 12.5% in real terms. The combined MS/SSS allocation (Table 14) increased by 27% in 1980 over 1979, which was roughly equivalent to the rate of inflation. Thus, there was no real increase in the allocation for operating costs. The MS/SSS health care facilities investment plan has not been matched by the necessary increase in funds to cover operating costs. E. Social Security Institute (ISS) 4.33 ISS directly provides health care for salaried employees of pri- vate sector enterprises that are sufficiently large and formally organized to permit regular payroll deductions and employer contributions. For 1980,  - 36 - the fraction of all salaried employees working outside the public sector who were covered by ISS was 41% or 1.8 million workers. Coverage has expanded relatively slowly in recent years (Table 23), and ISS officials estimate that, under the institute's present design, the maximum feasible enrollment is limited to 50% of private sector salaried employees, or 2.2 million workers in 1980. The following tabulation shows ISS coverage of the total labor force in 1980: i. Economically active population 7.7 million ii. Salaried labor force component 5.3 million iii. Salaried employees public sector 0.9 million iv. Salaried employees private sector 4.4 million v. Employees covered by ISS 1.8 million (v. as % of iv = 41%) 4.34 ISS expenditures on health care amounted to Col$14.4 billion (US$304 million) in 1980, or US$113 per capita (Table 23). Personnel expenditures represented between 50-60% of total expenditures during the 1974-1980 period, while supplies accounted for about 30%; other operating expenditures were 10% and investments less than 5% of total expenditures annually. Construction of basic health centers (CABS) may have increased the investment share after 1980. In 1980, ISS health expenditures per beneficiary were four times higher than MS/SSS expenditures per beneficiary. 4.35 ISS is comprised of three separate risk coverage programs. The program of central interest for this report, with the acronym EGM, provides coverage for non-occupational illness, maternity and infant care. A second program, IVM, provides benefits and pension coverage for disability, old age and death. The third program, ATEP, covers work accidents and occupa- tional illness. To gain insight into the finances of the EGM program it is necessary to also consider the other two programs because of transfers of resources between ATEP, IVM and EGM (Table 23). 4.36 EGM ordinary income is derived from a tax of 7% on wages (12% if the family system is used), two-thirds of which is paid by the employer and one-third deducted from employee salaries. IVM ordinary income is derived from a tax of 4.5% on wages that is also borne two-thirds by employer and one-third by employee contributions. ATEP ordinary income is derived from a 1.5% tax on wages, entirely paid by the employer. Additional non- operational (non-recurrent cost) income is derived by IVM and ATEP from earnings on permanent investments. Other operational income comes from sales of services (negligible), national budget contributions and temporary investment of funds. In addition to these sources of income, there are transfers to EGM operations from IVM and ATEP incomes. 4.37 Administration and operational expenses of the EGM program include services provided to ATEP and IVM beneficiaries. ATEP and IVM do not administer service facilities. The ATEP contribution to EGM is set by mutual agreement between ATEP and EGM at 12% of total EGM operational expenses. The 12% figure was arrived at after a survey of expenses demonstrated that between 10% and 15% of incurred EGM costs are for ATEP beneficiaries. The IVM contribution to EGM operations is set at 3.5% of IVM revenues.  - 37 - 4.38 National budget contributions are made to funds for each of the three programs of ISS. The funds are established with the following sources of revenue: (a) Fondo Promocion y Desarrollo (special programs for improvement of health). Funded by 1% of EGM payroll taxes plus an equal contribution from the national budget. (b) Fondo Salud Industrial (Industrial Health Fund). Funded by 3% of ATEP payroll contributions plus an equal contribution from the national budget. (c) Fondo Servicios Complementarios (programs for the benefit of pensioners). Funded by 3% of IVM payroll contributions plus an equal contribution from the national budget. Total government contributions to ISS are designated first for these fund- ing sources and then, if any excess remains, for free destination at ISS discretion. The government has no legal obligation to cover ISS deficits in operating costs. 4.39 Two alternative views of the extent of EGM dependency on transfers from ATEP and IVM and on ordinary tax revenues from the national treasury appear to be supportable. The first view, presented in the Wiesner report, is that the costs of providing health services exceed EGM payroll contributions and that the resulting deficit in health operations has required massive transfers of funds from the other divisions and the national budget to maintain the delivery of ISS health services. Table 23 illustrates the deficit from ISS health operations and the required transfer of funds from IVM, AIEP and national general revenues for the period 1978-80. For 1980, the operational deficit was Col$3.4 million with internal transfers from the pension and work accident divisions of Col$2.4 million and national treasury contributions of Col$0.6 million. 4.40 The Wiesner report points out that the Fondo de Solidaridad was created in 1973 to alleviate the shortage of funds for ISS health operations. The fund provides for the transfers, described above, from pension and work accident payroll deductions. From 1973 to 1979, the amount of internal transfers through this fund reached the equivalent of 40% of the reserves that the ISS held at the end of 1979. The conclusion drawn by the Wiesner report is that the cost of health services has reduced the reserves accumulated to cover other ISS benefits. The report concludes that the Fondo de Solidaridad should be replaced by a 1.8% additional payroll deduction and that the administration and financing of pensions should be clearly separated from health services. 4.41 An alternative view of ISS budget problems is that the internal transfers are, largely, compensation for the cost of health services extended to pensioners and for work accidents, and that transfers from national revenues are payments into the three ISS programs. Tables 24 and 25 give revenues and expenditures for the 1980 executed budgets of the ISS and each of its programs.  -38- 4.42 With regard to ATEP and IVM contributions to EGM, application of the 12% (ATEP) and 3.5% (IVM) rules described above would lead to a transfer of Col$2.23 billion to cover the cost of health services for beneficiaries of the pension and work accident programs.4/ This may be compared with total ATEP and IVM transfers for 1980 of Col$2.46 billion. Thus, only Col$0.23 billion are an excess transfer to cover an EGM operating deficit. In fact, even this small amount may arise from a mis- understanding of the definition of the funding basis used for the calcula- tions in footnote 5. With regard to the national budget, an application of the rules for national contributions to ISS division funds would lead to a total transfer of Col$387 million in 1980.5/ Thus, only Col$24 million, of the total transfer of Col$411 million, represents an excess transfer of funds that may be used to finance a deficit in health operations. 4.43 These two views do not differ with respect to an evaluation of the efficiency of ISS health operations. A more complete analysis is need- ed of departamento-level studies by the Social Security Superintendency, but an examination of the costs per beneficiary supports the contention of the Wiesner report that costs are high and the use of health services is inefficient. The two views do differ, however, with respect to the implications of the high costs for the internal finances of ISS programs. Further financial analysis of the ISS operations is required to resolve the issues raised by the Wiesner report. 4.44 This review of health-sector financing focused on the internal efficiency of the MS/SSS and did not delve deeply into the microeconomics of the hospital systems financed by social security. The high average unit costs per beneficiary seems to suggest inefficiency; nonetheless, the system has achieved in recent years some reductions in unit costs per patient visit, an awareness of possible efficiencies arising with a primary health care approach and a willingness to consider coordination of service delivery with MS/SSS as a means to reduce costs. Further, the nature of the product delivered by social security health programs has not been analyzed in depth. Treatment of chronic diseases among the mature male beneficiaries of social security may in fact be especially costly. F. Summary 4.45 In the future, sectoral analyses, particularly those to be carried out by the National Planning Department on behalf of the Colombian Government as a whole, will have to include microeconomic analysis of the 4/ ATEP Transfer = .12 x 16 x 109 = 1.92 x 109 IVM Transfer = .035 x 8.9 x 109 = .31 x 109 Total = 2.23 x l09 5/ Fondo Promocion .115 x 109 Fondo Salud .203 x 109 Fondo Servicios .069 x 109 Total .387 x 109  - 39 - social security system with a view to identifying possible cost reductions and opportunities to limit what appears to be excessive use of health services. 4.46 The differences in the mode of financing MS/SSS and the social security system make them very different organizations. MS/SSS is financed out of general and specific revenues widely regarded as appropriate to alleviation of poverty. Hence, an objective of achieving equity in availability of health services through MS/SSS would be widely shared in Colombia. In contrast, social security is financed through specific pay- roll taxes that many workers regard virtually as a user charge entitling them and other contributors to health services which are not to be shared with non-contributors. Few participants would favor using the social security health system to provide services to all Colombians regardless of whether or not they contribute to social security. Though the purposes and clientele of the two systems differ, it is logical that some coordination, if not integration, be sought between the two. Such a step would have to be preceded by achievement of some of the efficiency-yielding changes in the financing system alluded to in earlier paragraphs. Because the two systems deliver different services to different clienteles, their expenditure patterns would also require intensive study prior to steps to integrate their activities. Because major cost reductions might be achieved from more effective coordination and specialization of functions between these two major components of the health sector, it would be useful to begin analysis, first of financing, then of microeconomic operations, as soon as possible.  - 40 - PART TWO: SECTOR ISSUES AND POLICY OPTIONS V. SECTOR ISSUES 5.01 Colombia faces issues of equity, effectiveness and efficiency in planning future development of the health sector. The equity issue is how to redress regional disparities in health status and health services to achieve universal coverage by 1990 and substantial reductions in mortality and morbidity in the poorer departamentos by that same date. The effectiveness issue is how to deploy the organizational, institutional and manpower resources available to achieve overall sector goals. The efficiency issue is how to achieve sector goals within the resource constraints that limit further increments of the share of GDP or of central government revenues devoted to health care expenditures. A. Health Service Coverage 5.02 The Ministry of Health, through its regional offices (MS/SSS), is the principal instrument for expanding coverage. The role assigned to ISS and other social security funds in the National Health System is indeterminate. How they will share in the burden of expansion needs to be specified soon. If health service coverage is extended to all ISS family members, its share of population covered would rise from 10% to about 30% of the population. It is not yet clear whether this option will be followed. It is clear that the choice will greatly affect the scope and duties of the Ministry of Health and its regional offices. The implica- tions of alternative scenarios for the division of labor between MS/SSS and ISS are reviewed below. 5.03 Health professionals in Colombia are unsure about the degree to which MS/SSS facilities truly cover the thirteen million Colombians presumed to have access to ministry facilities. There is, for example, no clear definition in terms of initial and subsequent physician visits in a given catchment area, of current coverage in the various regions of the country. There is no specification of levels or indicators of the thoroughness of coverage which could permit the analyst to distinguish barely adequate coverage from very good coverage. There are no indicators of the kind normally associated with primary health care such as number of household visits by paramedical personnel for MCH or other services. In short, MS/SSS has not yet established feasible indicators to monitor its PHC coverage, its adequacy and its expansion through time. These monitor- ing features will have to be put in place to permit assessment of how well the ministry succeeds in moving toward its goal of covering not fifty percent but three-quarters of the population by 1990. Because the ministry succeeded in widening (if not deepening) its coverage in the 1970s, perhaps even more can be done in widening it further in the 1980s. Revised procedures affecting health-care financing could free up resources for more efficient use and enable MS/SSS to achieve adequate coverage for its target group with no increase in real resources in the sector. The cost-efficient PHC approach offers the best medium to achieve that objective.  - 41 - 5.04 The MS/SSS service delivery model stresses primary health care, disease prevention and a major role for paramedical personnel. This strategy appropriately addresses the infectious and parasitic diseases that constitute the major health problems of children under five who account for one third of all deaths. In addition, the MS/SSS service pyramid provides a potentially sound structural basis for efficient resource utilization. 5.05 The MS/SSS service delivery model contrasts sharply with the curative, high technology, clinic and hospital-based ISS approach. While ISS is now stressing primary care through CABS, the two systems clearly have different standards of care reflected in their respective costs. These differences pose a number of problems for the development of a national health system. First, the widely differing approaches to health care delivery are difficult to reconcile with the concept of a health system that would assure access to some minimum standard of care. Second, consumers are likely to opt for the curative model if given a choice. Third, and most important, the major differences between the systems - in level of service, technology, modes of financing, and consumer expectations - are incompatible with the development of a unified system, at least in the short run. A related issue is the extent to which the MS/SSS and ISS systems provide an appropriate response to the current and evolving disease profile in Colombia. The MS/SSS model is well designed to combat infectious disease and provide maternal and child health care. The ISS model, on the other hand, neglects the preventive approach but is more effective in treating chronic disease. Neither system, however, offers the broad preventive approach to what will soon be the predominant causes of morbidity and mortality -- high blood pressure, respiratory disease, heart disease, stroke and cancer. Colombia still has time to prevent a significant proportion of this toll by imaginative efforts to reduce smoking and obesity, promote exercise and balanced diets, and avoid the known environmental causes of cancer. Such an effort would also provide substantial financial savings, since preventive care is a fraction of the cost of treatment for chronic disease. 5.06 The 1979-1982 National Integration Plan included quantified targets defined in terms of morbidity and mortality reductions and service coverage, but it lacks a feasible plan to measure achievement. Nor are there data on the unit costs of delivering specific health services. These data will be essential to monitor MS/SSS achievements in the future. 5.07 The proximity of ISS and MS/SSS facilities in the major cities may be a cause of excess capac:Lty and underutilization of facilities. Moreover, when two clinics or hospitals are nearby and offer a substantial- ly different quality of service, it becomes obvious that equity in service availability is not being achieved and that, hence, National Health System objectives are being honored only in the breach. The practice of a benefi- ciary attending an ISS facility while his wife and children seek care at an MS/SSS facility is also difficult to reconcile with the concept of a national system. There is an equal need to map the catchment areas of other social security funds in assessing the extent of service duplica- tion. Separate subsystems are not likely to be equal; as such, they are not consistent with national policy.  - 42 - 5.08 Resolution of outstanding issues regarding coverage and National Health Service implementation requires a strong institutional mechanism and continuous dialogue, involving planning authorities, MS/SSS and ISS. The Health Insurance Superintendency, set up to coordinate the MS/SSS and social security sub-sectors, has been unable to carry out this role. There is an urgent need, therefore, to develop ways to translate the National Health System goals into operational plans and programs. 5.09 Continuing the current pluralistic approach reinforces an inequitable dual service standard. Forging a unified system with national standards entails compatibility between sub-systems. Either choice will require increased coordination and a clearer division of responsibility between MS/SSS and the social security system. If MS/SSS is to continue as the health system for the majority, equitable treatment requires the definition of minimum standards of care to meet their health needs. If the social security system is to expand significantly, stringent cost control constitutes a prerequisite. In either case, hard choices are necessary regarding the future shape of the National Health System. A series of policy options appear in the next chapter. B. Organizational, Institutional and Manpower Constraints 5.10 Organization of Health Services. The MS/SSS system succeeds only partially in meeting health needs at the bottom of the service pyramid. These organizational problems remain: (1) Insufficient delegation of functions at the PHC level; (2) Bypassing the secondary level of care and hence underusing these hospital facilities; (3) Poor management and chronic deficits at the tertiary level of hospital care; (4) Half of the promoters and auxiliaries in PHC do not provide drugs as policy dictates, due perhaps to physician opposition; (5) Shortages of personnel and drugs hamper PHC program effectiveness; (6) Low population coverage by promoters due to lack of transport and deficiencies in supervision and logistic support. The system works well in some areas, for example Valle and Caldas; it should be possible to build on these efforts to-Amprove performance elsewhere. 5.11 Promoter-delivered PHC in the MS/SSS costs US$13 per person per year, double the US$7 for the experimental CIMDER model. CIMDER demonstrated substantial impact of promoter services on morbidity and mortality. An evaluation showed that the CIMDER promoter can successfully  - 43 - manage two-thirds of morbidity. A comparable analysis of the MS/SSS model would show how effective it can, be. Based on experience in the six departamentos where CIMDER is being tested, successful elements could be adapted to the MS/SSS service model. 5.12 To stop the MS/SSS secondary-level bypassing requires service improvements: A full staff complement at each hospital, including medical specialists and provision of adequate equipment and supplies. Some regional hospitals may have to be closed, others upgraded. Neither the PHC nor the tertiary level can function adequately if current low capacity utilization continues at the secondary level. 5.13 High costs at the tertiary level are a major constraint on sector development. There are no easy solutions. Preventing chronic deficits is a high-priority objective. Achieving it is unlikely without well-trained hospital administrators and the institution of cost accounting and planning systems. Tertiary (and secondary) facilities must decrease on-site PHC, perhaps barring PHC except for emergencies. 5.14 It is too early to judge the impact of the ISS CABS program on utilization of higher level facilities, costs, and consumer/provider satisfaction. This PHC model is clearly more resource-intensive than that of the MS/SSS and the orientation is more curative. Careful studies are recommended to establish unit costs, referral patterns and health impact. 5.15 CAJANAL currently operates independently of other public sector provident funds, while the private family welfare funds operate independently of one another and of the public sector funds. The issue here is how to bring some or all of these institutions together in the interest of increasing efficiency and effectiveness. The existence of over two hundred separate funds is difficult to justify. This number of independent funds is incompatible with the planning requirements of a national health system. Coverage, costs, complementarity and duplication among these institutions must be a topic of continuing analysis. 5.16 Decentralization. Colombia needs to decentralize health services administration. (ISS never became as centralized as MS/SSS despite the growth of its central bureaucracy. ISS regional offices enjoy a substantial degree of autonomy and financial independence.) MS/SSS must delegate to the SSS financial authority matching their responsibility for the provision of health services. The departamentos divide roughly into three groups, based on their p:Lanning and financial management capability: (a) the most advanced -- Antioquia, Caldas, Cundinamarca, Valle; (b) the intermediate -- Cauca, Huila, Quindio, Risaralda, the Santanderes and Tolima; and (c) the least deve:Loped -- Atlantico, Bolivar, Boyaca, Cesar, Choco, Cordoba, La Guajira, Magdalena, Narino, Putamayo, the intendencias and comisarias. 5.17 Three models for decentralization have some pilot experience and support in Colombia. The PINDESA model (see Annex 2) groups departamentos geographically and charges the most developed of the group with providing technical assistance services to its neighbors. This approach builds on regional strength and common socio-cultural traditions. There are two potential drawbacks: reluctance on the part of the recipients to accept  - 44 - the profferred assistance and inability of the lead departamento to provide the assistance required. PINDESA accomplishments to date must be evaluated to assess accomplishments. The "central" model would have MS providing technical assistance to backward departamentos as needed. This model increases the MS workload and diverts resources from other tasks. A compromise might combine the two models with PINDESA operating where the lead departamento possesses the capability and the assistance was welcome, and MS serving the rest. A third, the "natural model", follows the contour of natural alliances that exist, such as the collaboration between Atlantic coast departamentos. This is likely to be a slow process unless combined with one of the other approaches described above. Financial management and accounting, planning, monitoring and evaluation, and information systems should be delegated to the SSS regardless of which model is followed. 5.18 Manpower and Training. A cost-effective PHC system must depend largely on non-physician manpower. This will require: (a) control of the imminent physician oversupply problem; (b) increased delegation of functions to paramedical personnel; and (c) improved training, supervision and teamwork at all levels. Expansion of coverage will have major implica- tions for personnel development. The MS/SSS delivery system requires the physician to be a PHC team manager. The ISS will need more family medicine practitioners. Matching production and training to the evolving needs of the National Health System will be necessary. 5.19 The imminent physician oversupply problem has many dimensions. First, training physicians represents a substantial drain on the public purse both before and after graduation. Medical training is heavily sub- sidized in the state universities. The cost of this subsidy per physician graduate is amplified by the high academic attrition rate. Physicians working for the SSS (or ISS) are costly in terms of salary and the expen- sive support system they require to practice effectively. Moreover, physi- cians may not be the most suitable personnel resource to provide many primary health care services which the under-served rural and peri-urban population need. Primary care can be delivered effectively by nurses, nurse auxiliaries and health promoters, providing they are well trained, supervised and supported. Physicians are needed only to manage the primary health care team, to treat the few complicated cases, and for restricted- access secondary and tertiary care. Physician oversupply leads to these concomitant problems: When there is not enough work, physicians band together to create supplier-induced demand, thereby lowering productivity and efficiency. In addition, under- and unemployed physicians represent a powerful interest group that can be politically and socially disruptive. 5.20 Colombia can avoid many of the problems posed by excessive supply of physicians by placing a moratorium on medical-school expansion and limiting entrants. Analysis of the physician labor market deserves high priority. 5.21 Professional nurses could usefully provide more patient care and less administration by expanding rehabilitation and emergency care, two neglected areas. More patient contact and PHC training are essential. The supply of nurses is probably sufficient if the emphasis remains on adminis- trative work in large hospitals. A greater role outside the hospital would require a clear division of labor between nurse and auxiliary nurse, and appropriate changes in nurse training.  - 45 - 5.22 There are at least two important issues relating to the auxiliary nurse's role in the MS/SSS system. First, why auxiliaries often do not provide drugs and act more as practitioners is not clear. To what extent is this due to physician opposition, lack of preparation or confidence, or simply under-reporting? Second, why are auxiliaries frequently unable to properly supervise promoters? Understaffing, lack of transport and insufficient training are all cited as reasons. These issues need to be resolved and reflected in changes in auxiliary training, which by common accord requires strengthening. More auxiliary nurses will probably be required regardless of the outcome of decisions on how to expand coverage. Projections of supply and demand are required for this and other categories of health personnel. 5.23 The health promoter is controversial in Colombia mainly because of lack of agreement about her effectiveness. Some argue that the promoter is too expensive given the small number of visits she makes. Yet the CIMDER experience suggests that promoters can successfully deliver a substantial volume of curative and preventive services at reasonable cost. Weaknesses in the promoter's training, support and supervision undermine her effectiveness. In any event, the promoter represents to some a visible symbol of rising health costs because she is frequently the best paid worker in the village. PHC relies heavily on community health workers without whom the system would revert to a traditional facilities-based approach. The appropriate course, therefore, is to devise ways to increase the promoter's effectiveness, perhaps by drawing on the CIMDER training and supervision model. C. Health Care Financing 5.24 Health care financing methods must meet criteria of equity, efficiency and effectiveness. Some major financing issues are the follow- ing: (1) Equity -- how to assure parity in health care availability to all groups in Colombia, regardless of region, rural-urban residence or income class, and prevent financial hardship; (2) Efficiency -- how to devise revenue systems that keep costs down, that discourage over-use of health services, that avoid distortion caused by earmarking revenues for specific uses, and do foster fee systems that lessen the burden on general treasury resources; (3) Effectiveness -- how to devise revenue systems that are easily administered and generally acceptable which avoid earmarking and hence facilitate program budgeting and which complement equity goals by not depending on revenues which inevitably discriminate against lower-income earners. 5.25 Equity Issues. These facts stand out in considering equity issues:  - 46 - (1) One-fourth of the Colombian population accounts for about two-thirds of total health care expenditures; (2) Half of the population accounts for the remaining one-third; (3) One-fourth of the population is without financial resources needed to have effective access to modern health care; (4) Four departamentos, with one-third of the country's population, account for about two-thirds of total health care expenditures. (5) MS/SSS PHC services, intended to cover at least half of the population, receive only about five percent of total financing. Health care is not available to all Colombians in equal degree even though primary health care is intended to be geographically and financially accessible to all. Complementary referral and supervisory 'services (secondary and tertiary health care) are supposed to be accessible when needed. Expansion of PHC necessarily means assuring that adequate secondary and tertiary care is also available. Many hospitals are over- crowded, others are underutilized and most operate with chronic financial deficits. Defining coverage in operational and financial terms remains a prerequisite to addressing the equity issue. Most Colombians are close to health services -- an hour and several kilometers for the vast majority of the population. Yet more than a quarter do not have effective access in light of their non-use of facilities. Future analysis must focus on what barriers -- lack of knowledge, time, money or confidence -- keep people, especially the poor, out of Ministry of Health clinics and hospitals. 5.26 Another dimension of the equity issue is the relationship between financial contributions to the health care system and access to its services. The allocation of resources to health has tended to favor the larger urban concentrations and thus the economically more advantaged population segments. Wage-based levies for social services typically are not viewed by the beneficiaries as a tax. Much of the employer's share, at least, is passed on to the consumer and hence may be regressive. In part because of its more ample revenues, the social security sector tends to bid up the wages of health personnel and cost of facilities making them less readily available to MSS/SSS. By expanding social security coverage at lower unit cost the inequitable resource distribution between the two sectors can be minimized over time. In the near term, ISS should begin to pay the full, and not merely the marginal, cost of public health services which it leases for its beneficiaries. Government subsidies for health care under social security, since they benefit well-to-do urban workers and not the rural or marginal urban poor, are inequitable also. 5.27 Another major equity issue is the regional distribution of health sector resources in Colombia. The four economically most advanced departamentos receive a lesser share of total public sector financing than the proportions of the population residing there. This implies some redistribution in favor of the less developed departamentos. Social  - 47 - security support for health care, however, is concentrated in the four leading departamentos as are private sector payments. Overall, health care resources are very unequally distributed in relation to population composition among the departamentos. 5.28 Efficiency Issues. Inequity may strike outside observers as the overriding failure of Colombian health policy. Colombian officials, in contrast, view inefficiency as the main health sector problem. Chronic deficits plague most hospitals. The virtual absence of cost controls is readily acknowledged. A better data base would help, but the main causes of inefficiency probably lie elsewhere. 5.29 The inefficiency of using multiple sources of health-care financing goes unremarked. Indeed, the introduction of two new sources of funds is regarded as a positive step in improving the financial base of the MS/SSS. Coordinating flows of funds from a dozen different departamento and national sources, however, has proved to be very cumbersome. Delays in transfer of revenues from earmarked taxes and lotteries are common. Uncertain supplementary allocations require special justification, and are a major contributing cause to inefficient resource utilization. There is a generic flow of funds problem. With earmarking of funds, all but the general revenue allocation is targeted by law for specific components of the health care sytem, primarily hospital services. Even the situado fiscal, which is earmarked for PHC, is in fact used extensively to finance local hospitals. A large part of the general tax revenue allocation to MS is absorbed by its administrative infrastructure. Supplementary alloca- tions finance deficits at the secondary and tertiary levels of the system. The existing PHC services, currently operating at about half the capacity intended, thus remain highly inefficient due to insufficient financial resources. Whether PHC coverage can be expanded through a reallocation of financial resources from secondary and especially tertiary levels to PHC; or whether the aggregate level of support needs to be increased to cover deficits and increase capacity at all levels, is not clear. Expenditures at the two upper levels could be reduced if they stopped providing PHC ser- vices, reduced expensive equipment purchases, rationalized pharmaceutical purchasing and contractual services, and carefully controlled personnel costs. Strengthened management: capability would yield increases in operating efficiency at all levels, including the MS. Improvements in efficiency will generate savings that could be reallocated to PHC. Since the upper levels themselves also need to be expanded and improved, a conscious and rigorous budget control to redeploy resources to cost- efficient uses will be essentiaLl. 5.30 High per capita costs in ISS provide prima facie evidence of inefficiency throughout the system. The relatively high utilization rates are partly responsible, as is the heavy reliance on physicians, frequent hospitalization, and extensive use of costly equipment and supplies. However, ISS was able to increase the level of financing by raising wage levies. Raising levies from 7%' to 12%, to convert from the limited cover- age under the classical system to the more comprehensive family system, may be insufficient to meet the increased cost of expanded coverage. ISS is seeking to reduce costs by expanding the number of basic health care centers (CABS) to cover large population segments, while limiting the expansion of hospital care. This approach has some potential to reduce unit costs but remains unproved.  - 48 - 5.31 A dominant efficiency issue in the health sector generally is the largely unplanned and very costly expansion and upgrading of hospital facilities and equipment. The FNH, while nominally responsible for planning and executing all hospital construction, remodeling and equipment purchases in both the public health and social security sectors, does not appear to exercise this function effectively. FNH does not have complete plans for most of the projects that it supervises. The political and practical viability of this centralized hospital construction authority is in doubt. Decentralizing this authority to the SSS level might be more efficient. 5.32 Planning authorities currently fail to link investment plans to the implied increases in recurrent costs. Thus, a major cause of inefficiency is the lack of funds to operate physical facilities when they are completed. This problem appears to be especially acute at the PHC level where the recent government, partly with funds from the Federation of Coffee Growers and PAN/DRI, largely completed a four-year investment program. There was apparently no provision to provide the funds needed to operate these facilities. 5.33 Effectiveness Issues. The current complex financing scheme diverts the funds generated from the intended sector objectives to less effective uses. The extensive earmarking of virtually all departamento and even most national and international sources of health-care financing has this unintended consequence. The programs and facilities currently being funded are not maximally effective in meeting the predominant health needs of the population. Most available health-care funds are being used at the secondary and tertiary levels. MS/SSS has not yet been able to implement effective PHC services to the extent planned for, and much less to the extent these services are needed to cover all of the population. 5.34 There is no effective coordination between the MS/SSS and social security sectors. The practice of ISS contracting MS/SSS services for its beneficiaries in certain areas shows potential for increasing coordina- tion. Very little of this potential is being realized, however, as the two sectors continue to plan their services separately and continue to diverge in their approaches, particularly in providing PHC services. The effective implementation of Colombia's National Health System certainly requires that the legal mandate be given increased political support and an operational plan for implementation.  - 49 - VI. POLICY OPTIONS 6.01 The framework for future policy choices is Colombia's commitment to extending health care coverage to the entire population by 1990. Cover- age is currently estimated at 75%, but the amount, type and quality of services received varies wideLy for different geographic areas and popula- tion segments. Since 1975, the National Health System has constituted the principal mechanism for achieving Colombia's health goals. This system divides responsibility for providing health care between the Social Security Institute (ISS) and other funds (for private employees), CAJANAL and other funds for public employees, and the Ministry of Health and departamento health services (MS/SSS) for all remaining Colombians not covered by the private sector., The system was intended to function in a unified fashion. However, scant progress has been made towards the implementation of a coordinated, much less an integrated, system. ISS and CAJANAL coverage has increased very slowly and MS/SSS is still far short of achieving effective and complete coverage of the population assigned to it. Moreover, the two major systems function largely independently of one another, with considerable overlap and duplication resulting. The principal policy choices facing decision-makers, therefore, relate to when and how to implement the National Health System. The choices made and priorities assigned to various options have different implications for the equity, efficiency and effectiveness objectives implicit in the overall health goal. The following options are not mutually exclusive and questions of phasing and proportional allocation of resources are as important as the choices themselves. A. Implementation of the National Health System 6.02 A national health system should have the basic objective, as it does in the Colombian case, of efficiently directing available physical, human and financial resources to specific health needs, and to providing health care for all who need it, regardless of ability to pay. There are obviously many alternative institutional mechanisms for achieving this objective. In the Colombian case, at least two major alternatives are available -- integration of the MS/SSS and social security systems, or coordination of the two systems. Continuing the present uncoordinated approach is a third option but would not constitute a national health system. 6.03 Integration of MS/SSS and Social Security. This option is clearly the more difficult course, at least in the short run. It is not entirely clear from the National Health Service legislation what degree of integration between the sectors was intended. In any event, this approach offers several advantages. It is the most equitable option because it lends itself both to providing a uniform standard of service to the entire population, and because its financing will not require that revenue from particular sources be applied to the benefit only of certain population segments. To the extent that it would eliminate duplication and permit economies of scale, this approach is potentially more efficient than coor- dination. Similarly, a unified system able to organize health services  - 50 - based on the needs of the entire population should more effectively meet its health objectives. Against these long run advantages in terms of equity, efficiency and effectiveness must be balanced the political, administrative and financial barriers to integration. The two systems have developed and operated independently for almost four decades and the difficulties of integration are obviously great. One major consideration is that social security beneficiaries contribute directly to the cost of their health services. This must be weighed against the argument that these costs are passed along to the general public, and against equity considerations in favor of the poor. Not the least of these considerations is that those most in need of health services are usually least able to afford them. 6.04 The practical difficulties of integrating the two systems are formidable. They have different administrative structures, service delivery modes and methods of financing. Moreover, the MSS/SSS emphasis on prevention and PHC contrasts with the curative medical approach of the social security system. Perhaps the strongest argument against integration, however, is the prior need for changes within both systems to increase efficiency and effectiveness. In the case of MSS/SSS, this implies stringent cost controls and management strengthening to deal with hospital deficits, and increasing the effectiveness of the PHC system. For the social security sector, reduction of high per capita costs and greater emphasis on preventive care are required. It would nevertheless be useful for planning sector development to have a clear decision whether the eventual goal is integration or coordination of the two systems. 6.05 Numerous approaches to integration are possible. Two main alternatives appear most promising in the short run and would address the overlapping and duplication of the two systems: (a) Integration would begin in the principal population concentrations where social security coverage is relatively extensive, but where the two systems are currently operating independently and both are expanding. The principal objective here would be to improve the system's efficiency although by reducing MS involvement in these areas, resources could possibly be reallocated to increase MS/SSS allocations in the lesser developed departamentos and thus also serve equity. Two alternative approaches are conceivable: (i) ISS would manage existing MS/SSS facilities, with MS financing services for those not covered by ISS; (ii) MS would manage ISS facilities, with ISS paying full cost (rather than merely marginal cost) of coverage for its beneficiaries. (b) Integration would begin only in the intermediate-size population concentrations where the MS/SSS system predominates but where the social security system is  - 51 - emerging but could still be prevented from developing separately. As in (a) above, both efficiency and equity objectives would be served. 6.06 The above illustrative examples separate the management of the health system from its financing. Choices among the alternatives, or other options, involve difficult decisions in several areas. For example, ISS is financially stronger than MS/SSS and therefore better able to finance expansion of coverage through the MS/SSS infrastructure. However, expan- sion of coverage to all ISS or all social security family members would serve equity less than a comparable expansion in MS/SSS which serves the poorer segments of society. The main point is that integration of the two systems offers most in terms of rationalizing the use of scarce resources. B. Independent Development of the MS/SSS and Social Security Subsectors 6.07 Alternatively, MS/SSS and social security could be encouraged to continue developing separately. Policy options for the social security sub-sector would include the following: (a) ISS and CAJANAI. could continue to develop and expand separately. In! this case: (i) ISS could continue to develop separately or integrate with other private sector social security funds; similarly, (ii) CAJANAL could continue to develop separately, or integrate with other public sector social security funds; or (b) the government could seek to integrate the entire social security sector. 6.08 ISS coverage could rise to 30% or more from the current 10% by including family members and expanding coverage to certain independent and agricultural segments of the labor force. Coupled with continued efforts to control costs and to achieve economies of scale, this would increase the efficiency of ISS. Its impact on equity could also be enhanced by some publicly-financed extension of coverage for the less well-off. However, limitations of the ISS infrastructure would restrict its ability to reach the marginal rural populations. CAJANAL coverage could increase from its current level of 20% of public sector employees to close to 100%, making health care delivery in this sub-sector more efficient. 6.09 Combining all the public sector social security funds, or all those in the private sector, would yield high dividends in terms of increased efficiency and effectiveness. Colombian officials recognized a decade ago the inherent inefficiency resulting from the proliferation of social security funds. However, efforts to merge the funds will engender opposition from their managemenat, beneficiaries and other interest groups. Integration of the social security sector in other countries has succeeded  - 52 - usually on a phased basis. This suggests the wisdom of first integrating separately the private and public sector funds, before proceeding to inte- grate the entire social security sector. 6.10 The social security sub-sector is at a disadvantage vis-a-vis MS/SSS in assuring equitable access to health services for two principal reasons. First, its coverage is by definition confined to employees in public and private enterprises who do not constitute the neediest groups. This problem could be overcome by subsidies to the poor, but that would not overcome the second reason, which is lack of access. Social security faci- lities are available mainly in the large and intermediate size cities. As a consequence, this infrastructure is not available to the population of most towns and villages, or to the dispersed population. 6.11 MS/SSS would have the following policy options, among others, if it and the social security sytem were encouraged to continue their separate development: (a) MS could aggressively decentralize and under the PINDESA model expect the advanced departamentos to take over responsibility for the expansion of coverage and improvement of health care in the less advanced departamentos; or, (b) MS could leave the advanced departamentos to function semi- autonomously, i.e. with most functions and authority delega- ted to their SSS and municipal levels, and concentrate MS supervision and technical assistance on the intermediate and least developed departamentos with delegation of functions and authority geared to their capacities to accept them. 6.12 The first option would seek to build on the tradition of placing responsibility with the most advanced departamentos for sharing their expe- rience and resources with neighboring lesser developed departamentos, although there is only limited evidence that this strategy has been accept- able and feasible logistically. The second option would seek directly to improve services for those most in need and currently least supplied with health care. While this strategy in the short run may contradict efforts at decentralization, it would nevertheless reflect current reality of MS relations with the lesser developed departamentos and offers advantages in terms of equity and efficiency. Moreover, MS/SSS has built a substantial infrastructure and amassed considerable experience over more than two decades in attempting to provide services to those most in need. 6.13 Several concrete steps to improve resource utilization are possible in the near term regardless of the ultimate decision about integration or coordination of the MS/SSS and social security systems: (a) Joint planning by MS/SSS and the social security system is an urgent need. Joint facilities planning to prevent under- utilization deserves especially high priority. Maintaining a virtual freeze on hospital construction would contribute to controlling unaffordable increases in recurrent costs. Decisions regarding coverage expansion also require maximum collaboration between the two systems. An assessment of future demand for health care is the first requirement here.  - 53 - (b) There is additional scope for ISS to contract MS/SSS ser- vices instead of developing a parallel infrastructure. ISS should pay MS/SSS the full cost for contracted services. There may also be instances where MS/SSS should contract services from social security agencies. (c) Introduction of a standard salary schedule for health workers would greatly facilitate sector coordination and rationalize resource use in this major expenditure category. The National Health System personnel statute legislation could provide the vehicle for this. (d) The introduction of user charges or co-insurance in the ISS system, in addition to the cost controls described above, would reduce over-utilization and increase efficiency. (e) There is an urgent need to project the cost of ISS expansion based on an actuarial analysis of the proposed new beneficiaries (agricultural and other independent workers, and the families of workers already covered). This would permit a much faster expansion of coverage under ISS than has occurred to date. (f) Immediate steps are possible to increase coordination within the MS/SSS. A closer relationship between FNH, the Ministry and the SSS's should improve resource utilization for facilities construction. Similarly, more coordination is needed between INAS and MS in planning the construction of rural water supply and sewerage systems in the 16 departamentos where INAS has this responsibility. (g) More joint planning is required at the departamento level between the SSS's and the beneficiencias which provide health and welfare services. This would ensure complementa- rity and prevent duplication of effort. All of these actions appear feasible in the short term and all would contribute to a more equitable, efficient and effective National Health System. C. Health Manpower 6.14 Policy options here should obviously reflect decisions taken regarding the future shape of the National Health System. There are, however, a number of pressing issues that will require early resolution regardless of broader decisions about how to expand and improve coverage. 6.15 Physicians. A decision regarding future production of physicians requires a projection of the current stock based on estimates of emigra- tion, retirement and death, and the numbers and types needed by the MS/SSS, social security and private sectors. The main options in terms of produc- tion of physicians are to let the market decide the issue or to intervene  - 54 - to control the number of graduates. Experience elsewhere in the continent indicates that the market is an inefficient regulator of physician produc- tion. Uncontrolled production usually results in under- and unemployment rather than a better geographic distribution and other benefits of increased competition. But the alternative path of restricting the number of medical school entrants is politically difficult. However, too many physicians will bias the health care system toward curative, high tech- nology medicine, the cost of which is incompatible with universal health service coverage. 6.16 Colombia's health goals require physician PHC team leaders in under-served towns and rural areas. Three options for assuring this are: (a) requiring all new graduates to serve where needed for three years or more; (b) offering incentives to all physicians to locate in under- served areas; (c) continuing with the current requirement of one year of social service. There is general agreement that the current system (c) is not effective because of its brief duration and the inexperience of new graduates. Either of the first two options would strengthen the National Health System, and both have worked successfully elsewhere. Since there are obvious advantages to having experienced as well as newly-graduated physi- cians locate in under-served areas, both options might be adopted. 6.17 Nurses. Policy options regarding nurses depend on decisions regarding their future role. If the nurse is to deliver more patient care, be assigned a major role in PHC,-or specialize in new areas (e.g. rehabili- tation or emergency care), additional numbers will be required. If little change in the status quo is expected, the current number and output level is probably adequate. Changes in the nurse's role would obviously neces- sitate a revision in curriculum. 6.18 Auxiliary Nurses. Decisions regarding auxiliary nurses depend on the role of the professional nurse and vice versa. One urgent matter is to decide whether to implement fully the delegation of authority to treat patients with defined illnesses. If the auxiliary is not to perform this function as envisaged in the MS/SSS PHC system, it will fall to the profes- sional nurse and/or physician. This will naturally increase costs. Another urgent matter is how to strengthen the auxiliary's supervision of the health promoter. Once the questions of role definition have been resolved, these can be reflected in auxiliary training. 6.19 Health Promoters. Policy options here also depend on decisions regarding the National Health System, delegation of functions, and the division of labor between nurse, auxiliary nurse and promoter. One critical decision is whether or not to ensure that all promoters provide patient care and basic drugs in the community. If promoters are not to perform this function, who will provide the PHC outreach? The ability of  - 55 - the community health worker to provide simple curative care has often been a decisive factor in the success of programs like CIMDER and similar models in other countries. Other options will follow from decisions about the promoter's role, i.e., how to strengthen training, supervision and support sytems. A reaffirmation of the promoter's role will require training addi- tional numbers. D. Options in Health Care Financing 6.20 The multiple sources of health care finance, and the earmarking of revenues to particular purposes no longer consistent with the most effective use of resources, are major sources of inefficiency in the sector. A first step which planning and health authorities can take to improve the revenue base is to document, in a thorough analysis, the costs and administrative bottlenecks associated with the current system. Such an analysis would have the further advantage of identifying optional financing plans. A program budgeting approach, for example, would start with the identification of health sector objectives (with respect to population coverage, physician visits provided, hospitalizations necessary, among the service-provision goals, and morbidity, mortality and infant mortality rate reductions among the outcome goals) and estimation of the current and capital expenditures necessary to achieve those objectives. In a parallel fashion, health planners could provide estimates of the health impact of alternative expenditure plans, so that central authorities can assess the merits of augmenting or reducing health expenditures to make investments in that sector yield benefits on a par with those in other sectors which have legitimate claims on public resources. Until health planners can demons- trate that they are using currently-available funds in the most efficient manner, central authorities may not wish to provide more resources to the sector as a whole. Once program budgeting is introduced into the sector, resource claims may have greaiter general support than they now receive, given the lack of comprehensive overall planning and linkage between expenditures and results. The achievement of specific goals, e.g., full population coverage, should not be costed out separately from other health expenditures, e.g. maintenance of expensive, specialized hospitals, since it may prove to be desirable to reduce hospital allocations to increase primary health care. In that sense, sector resources as a whole must come under review to find the best way to turn financial resources into better health. ISS, as well as MS/SSS, must be included in a comprehensive analysis of how best to spend health funds. Current earmarking of receipts and expenditures stands in the way of such a system of health-care financing. Colombia has a long history of using specific resources such as liquor taxes for health purposes. The willingness to pay a tax on the part of a population group is itself an asset not to be set aside lightly. Nonetheless, some more rational link between financial resources and health outcomes has to be forged to make a convincing case that Colombians are getting their money's worth in the health sector. 6.21 Major changes in MS/SSS and ISS operations will be required to increase efficiency. Such changes are a prerequisite to consideration of proposals to increase the National Health System's share of GDP.  - 56 - Increasing the allocation of financial resources (i.e., in real terms or share of GDP) would be premature under the current approach to health care financing. A first step is to introduce program budgeting. 6.22 Certain payroll and local taxes which are akin to user fees and social-insurance schemes could perhaps be augmented to good effect. Payroll taxes were recently increased to 12% of wages to pay for full coverage of family members in the ISS health-care program. ISS expansion can therefore begin to relieve MS/SSS of some of its responsibilities in urban areas and at the tertlary-care level, making possible a reallocation of resources within the public sector to expand primary-care coverage even with a continuing limitation on MS/SSS resources. At the departamento level, options for increasing health revenues include improving the collection and transfer of beverage and tobacco taxes, expansion of lotteries (allocating all profits rather than only a share to the health sector) and augmenting user charges in the SSS. Even without a detailed analysis of tax effort at the departamento and municipio level, it seems clear that local governments could do much more to pay for local health care. 6.23 MS/SSS expenditures for the past decade have remained at less than one percent of GDP and hovered around 7% of the central government budget. Nonetheless, general tax revenue support from the central government budget has been unstable, primarily due to variations in the supplementary allocations. Stabilizing these supplementary allocations could improve the efficiency of financial management in the ministry. Variations in both internal and external borrowing for the health sector contribute to the unevenness of expenditures. A more predictable and stable investment strategy, based on long-term planning, would yield better results in terms of health services. 6.24 Departamento sources of revenue have declined in relative magni- tude during the last ten years, i.e. since the introduction of the situado fiscal. Earmarked sales taxes and net revenues from lotteries continue to account for about one-third of total public health care financing. Sales taxes, however, have been a troublesome source of revenues. Their collec- tion and transfer to the SSS's has lagged by a year or more in several departamentos and collection rates are reportedly low. Growth in revenues from these taxes, which account for about 20% of departamento-level revenue, is inelastic in relation to overall growth in GDP. The national and departamento governments must decide the extent to which they wish to emphasize this source, weighing its inefficiency and inequity against the benefits in limiting alcohol and tobacco consumption and its unequivocal earmarking for health care. Sales of services account for over half of all departamento health care revenues. They include user charges and receipts from the ISS and other users of MS/SSS facilities. User charges should be on a sliding scale to be equitable, but means tests are difficult to administer systematically and fairly. Contractual payments should be increased where they meet only the marginal rather than the full cost of the services provided. 6.25 Redistribution of central government funding to MS/SSS from richer to poorer SSS's represents a different policy option. The four richest departamentos receive slightly smaller proportions of national  - 57 - level funding than their population shares. Still, they account for disproportionately large health care expenditures by departamento, when social security and private sector sources combined are added to MS/SSS expenditures. Borrowed funds and non-earmarked general tax revenues could be allocated in greater proportion to the poorer SSS's. The situado fiscal allocation formula could be legally redefined to consider need as well as population distribution. 6.26 Expansion of coverage through the ISS represents an altogether ,different strategy. The recent inclusion of family members of beneficia- ries marks a step in social security expansion. There is some coverage for agricultural workers already. Policymakers justifiably fear more rapid expansion on grounds that revenues will fall short of cost increases, as has occurred in other countries. A significant reduction in average cost of coverage is a prerequisite for the expansion through ISS. Basic health care centers (CABS) are designed to provide 70% of all services at low average cost. The success of tlhis new scheme remains uncertain. Joint funding and use of existing MS/SSS facilities and programs has not yet been given serious consideration. It could be a lower-cost alternative to the extensive implementation of CABS. ISS could also expand coverage to lower-income groups through unreimbursed service delivery and a subsidy from general revenues. 6.27 The organization responsible for physical facilities construction and equipment purchases in both the public health and social security sectors, is the National Hospital Fund (FNH). The FNH has not effectively exercised control over all physical investment in the National Health System, nor is it being creditedl with efficiently executing the many proj- ects that it has approved, funded and carried out. Long delays and large cost overruns appear to be typical rather than exceptional. The justifica- tion for many of these projects is highly questionable. On the positive side, however, the FNH has over the past four years been able to execute most of the previous government's health investment plan. The FNH role, needs to be redefined. Its financing and expenditures should be carefully reviewed with reference to any newly emerging health sector policy. Without a comprehensive analysis of the state of the public sector's over 600 hospitals, the ISS's medical care facilities and their respective equipment inventories, no basis exists for reliably determining the investment resources needed for facilities and equipment. 6.28 Central authorities uwgently need a consolidated flow of funds projection for the health sector covering at least the years 1982 through 1986. It would include a comprehensive identification of all sources of revenue for MS/SSS, ISS, and other ministries, e.g., Education, with health-sector responsibilities, as well as details permitting identifica- tion of possible sources of efficiency. It would identify expected expenditures by several alternative breakdowns including personnel and material costs; recurrent and capital costs; primary, secondary and tertiary levels of care and their costs; costs for care of specific dliseases as a basis for identifying the cost of impact of disease prevention; costs associated with different skill levels of personnel in order to assist future human resource planning; and where possible costs aLssociated with specific fits as in the case of disease reductions through  - 58 - vaccinations. One way to proceed is to establish a health planning group to coordinate data inputs and program proposals from the several components of the health sector, making it responsible to central authorities rather than to any specific operational component of the National Health System. The flow of funds analysis could be complemented, as was suggested above, by a detailed inventory of physical and human-resource facilities, i.e., the assets of the health sector, with attention to their periodic deprecia- tion and costs of maintenance. Tables 13, 14 and 15 in this report merely scratch the surface of what needs to be done in terms of detail necessary for central authorities to monitor system performance and plan for more effective use of resources. Financial planning can then complement the human resource planning and facilities planning suggested elsewhere in this report. E. Next Steps 6.29 This report has focussed on a broad range of issues, many of which may involve actions over the medium and long-term that entail subs- tantial system-wide changes. This final paragraph suggests illustrative examples of immediate actions that might be undertaken to rationalize resource utilization and enhance sector development: (a) The development of sector policy woud be enhanced by the establishment of a strong inter-agency mechanism for planning, coordinating and implementing changes in the National Health System. This committee might be headed by the director of the National Planning Department and be composed of representatives of interested parties, including the ministers of health, labor and education, the directors of ISS and CAJANAL, the private social security funds and ASCOFAME. A full-time technical secretariat would be necessary to address major sector issues. (b) Several actions could be taken immediately to control costs at the tertiary care level. A freeze on construction of new facilities might be considered while the causes of chronic hospital deficits are sought and solutions implemented to reduce this major drain on sector resources. In addition, it would be useful to require joint MS/SSS and ISS planning of new facilities to ensure adequate utilization. (c) At the primary health care level further construction of ISS CABS might be halted until clear policy decisions have been made regarding the extent, geographic distribution and institutional mechanisms to be used to expand health service coverage. (d) If it is decided to expand coverage, early decisions are necessary regarding the priority geographic areas for these efforts.  - 59 - (e) Cost-effectiveness analyses could provide a firm basis for choosing between policy options. These analyses could either be conducted under the auspices of groups within each major institution, e.g. MS/SSS and ISS, or be the responsi- bility of a sub-cornmittee of the proposed planning, coordi- nating and implementing committee. (f) There is an urgent need to conduct an analysis of the sector role of the private social security funds with a view to planning their future place in the National Health System. (g) In the area of health manpower, a projection of future needs and supply deserves high priority. An oversupply of physi- cians will result in increases in sector costs and limit the ability to pay for other, less expensive categories of health workers.  - 60 - COLOMBIA HEALTH SECTOR REVIEW Table 1: POPULATION BY AGE GROUP 1965-1966 and 1977-1980 1965 - 1966 1977 - 1980 Age Group % % Accumulated % % Accumulated 0 - 4 17.6 17.6 13.5 13.5 5 - 14 29.0 46.6 27.9 41.4 15 - 24 18.2 64.8 21.2 62.6 25 - 34 12.5 77.3 12.6 75.2 35 - 44 9.4 86.7 9.2 84.4 45 - 59 8.3 95.0 9.7 94.1 60 or more 5.0 100.0 5.9 100.0 TOTAL 100.0 100.0 1977 - 1980 Mean = 23.2 Males: 12,083,958 - 49.0% Median = 17.9 Females: 12,577,584 - 51.0% Source: National Health Survey, 1977-1980.  -- 61 - COLOMBIA HEALTH SECTOR REVIEW Table 2: CRUDE DEATH RATES, 1970 - 1982 REGIONS 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 Atlantica Atlantico 7.1 6.8 6.5 6.3 6.0 5.8 5.6 5.4 5.2 5.0 4.9 4.7 4.5 Bolivar 10.9 10.5 10.0 9.6 9.2 8.8 8.3 7.9 7.7 7.4 7.0 6.7 6.4 Cesar 11.4 10.9 10.5 10.0 9.6 9.1 8.6 8.2 8.0 7.6 7.3 6.9 6.6 Cordoba 12.5 12.0 11.4 10.9 10.4 9.9 9.3 8.8 8.5 8.1 7.7 7.3 6.8 La Guajira 19.3 18.3 17.4 16.5 15.6 14.7 13.7 12.8 12.3 11.6 10.8 10.1 9.5 Magdalena 11.5 11.1 10.6 10.1 9.7 9.2 8.7 8.3 7.8 7.4 7.0 6.5 6.1 Sucre 12.1 11.6 11.1 10.6 10.1 9.6 9.1 8.6 8.0 7.5 7.1 6.6 6.1 Antioquena Antioquia 8.5 8.4 8.2 7.8 7.6 7.4 7.1 6.8 6.6 6.4 6.2 6.0 5.8 Caldas 9.9 9.6 9.4 9.1 8.6 8.0 7.8 7.4 7.2 6.8 6.6 6.1 5.8 Quindio 9.6 9.5 8.8 8.6 8.4 7.8 7.8 7.7 7.2 6.9 6.6 6.3 6.0 Risaralda 8.8 8.5 8.2 7.9 7.5 7.3 7.2 6.8 6.6 6.3 6.1 5.9 5.6 Oriental Norte Santander 10.8 10.1 9.6 9.2 8.6 7.8 7.6 7.1 6.6 6.2 5.8 5.4 5.0 Santander 10.3 9.7 9.3 8.9 8.2 7.7 7.3 7.1 6.5 6.1 5.7 5.3 5.0 Centro--Oriental Bogota 6.4 6,2 6.0 5.9 5.8 5.5 5.5 5.3 5.2 5.1 5.0 4.9 4.8 Boyaca 10.9 10.5 10.0 9.6 9.0 8.6 8.0 7.8 7.1 6.6 6.2 5.7 5.2 Cundinamarca 10.6 10.2 9.8 9.4 8.9 8.4 8.2 7.8 7.3 6.9 6.5 6.1 5.7 Meta 11.7 11.2 10.8 10.4 9.9 9.5 9.0 8.6 8.3 8.0 7.7 7.4 7.1 Centro-Occidental Rutla 10.7 10.1 9.8 9.2 9.0 8.8 8.6 8.3 7.9 7.6 7.3 7.0 6.7 Tolima 11.0 9.4 9.2 8.9 8.5 7.8 7.8 7.0 6.8 6.5 6.1 5.7 5.4 Pacifica Cauca 12.9 12.3 11.7 11.1 11.0 10.2 9.6 9.2 8.7 8.2 7.8 7.3 6.9 Choco 19.3 18.3 17.4 16.5 15.6 14.7 13.7 12.8 12.1 11.3 10.5 9.7 8.9 Narino 10.7 10.1 9.5 9.2 8.7 8.7 8.2 7.9 7.4 7.0 6.7 6.3 6.0 Valle 9.1 8.6 8.2 7.9 7.8 7.4 7.0 6.8 6.6 6.4 6.1 5.9 5.7 Intend. y Comis 26.0 24.7 23.4 22.1 20.8 19.5 18.2 16.8 - - - - - TOTAL 10.1 9.6 9.2 8.8 8.4 8.1 7.7 7.3 7.0 6.7 6.4 6.1 5.8 Sources: National Health Survey and Alberto Bayona and Marda Ruiz, La mortalidad en Colombia: 1970 - 1982, 1982.  - 62 - COLOMBIA HEALTH SECTOR REVIEW Table 3: INFANT DEATHS ACCORDING TO 1973 CENSUS, REGISTERFD IN 1973, AND ESTIMATED BY C. PIERRET Census Departamentos Births Infant Infant Regis- Estimated Mortality Deaths tered by Rate C. Pierret TOTAL 63,998 38,213 69,570 Antioquia 101,073 107 10,814 5,565 8,525 Atlantico 27,810 56 1,557 1,760 3,218 Bogota 76,791 46 3,532 4,021 5,104 Bolivar 30,198 68 2,053 1,223 4,008 Boyaca 46,519 72 3,349 1,909 4,245 Caldas 22,026 104 2,291 1,923 2,360 Cauca 27,192 127 3,453 1,965 2,393 Cesar 15,424 115 1,774 861 1,674 Cordoba 29,043 88 2,556 247 3,224 Cundinamarca 33,906 69 2,340 2,020 3,910 Choco 9,829 198 1,946 575 908 Guajira 8,332 71 592 90 811 Huila 19,711 104 2,050 1,211 1,832 Magdalena 23,446 81 1,899 658 2,641 Meta 10,965 114 1,250 779 1,102 Narino 34,901 125 4,363 1,691 3,030 N. de Santander 25,634 78 1,999 1,636 2,608 Quindio 10,288 77 792 697 1,083 Riseralda 13,362 100 1,336 1,046 1,524 Santander 39,053 85 3,320 2,121 4,362 Sucre 15,677 70 1,097 242 1,770 Tolima 34,887 84 2,930 2,028 3,486 Valle 72,882 92 6,705 3,945 5,752 Source: DANE, Monthly Statistical Bulletin #305, December 1976  - 63 - COLOMBIA HEALTH SECTOR REVIEW Table 4: INFANT MORTALITY, 1950 - 1982 (DEATHS PER 1,000 BIRTHS) Maximum Minimum Average Year Estimates Estimates Estimates 1950 144 136 140 1960 116 103 110 1965 105 89 97 1970 94 77 86 1975 85 67 76 1980 76 58 67 1982 73 55 64 Source: National Health Survey, 1977 - 1980.  CODL&BIA HEALTH SECTOR REVIEW Table 5: CAUSE-SPFCIFIC MIRTALIIY BY ACE GOLUP, 19i7 1 z 1 - 4 % 5 - 14 % 15 - 44 x 45 - 64 x 65+- Enteritis and Enteritis and Heart Heart Other Diarrheal Other Diarrheal Accidents 26.7 Accidents 23.0 Disease 25.9 Disease 32.4 Diseases 22.3 Diseases 22.0 Perinatal Influenza ani Influenza and Conditions 22.0 Pneumonia 14.7 Pneinnnia 6.7 Hanicide 17.8 Cancer 19.2 Cancer 13.2 Bronchitis, Enteritis ani Influenza and Enphysena and Other Diarrheal Heart Pneumonia 15.7 Astima 7.5 Diseases 5.9 Disease 8.8 Strcke 10.3 Strcke 11.1 Emphysema and Influenza andi Asthma 6.7 Accidents 7.0 Cancer 4.8 Cancer 7.8 Accidents 7.2 Pneumnnia 4.2 Complications Nutritional Nutritional Heart of Pregnancy Hanicide 3.1 Accidents 2.5 Deficiencies 4.3 Deficiencies 6.1 Disease 3.3 and Childbirth 4.4 Source: Pan American Health Organization, Health Conditions in the Amricas, 1982.  - 65 - COLO4BIA HEALTH SECTOR REVIEW Table 6: PR:INCIPAL CAUSES OF MORBIDITY, 1980 Population Diagnosis - Percent Affected 1. Respiratory Infections (Other than Acute) 12.4 3,051,385 2. Ophthalmological Refraction and Accommodation Problems 12.3 3,041,988 3. Helminthiasis (Parasitic Infestation) 10.3 2,531,301 4. Acute UJpper Respiratory Infections 9.6 2,363,247 5. Varicose Veins (Lower Extremities) 6.9 1,712,797 6. Nutritional Deficiencies 6.2 1,518,341 7. Other Ophthalmological Problems 5.9 1,442,461 8. Pterigion 5.8 1,421,898 9. Umbilical Hernia 5.0 1,236,027 10. Hypertension 4.6 1,197,345 11. Dermatosis and Dermat:itis 4.3 1,060,853 12. Arthropathies and Problems of the Lower Back 3.6 884,343 13. Cataracts 3.5 863,913 14. Anemia 3.7 913,435 15. Obesity 2.6 637,381 16. Mycosis 2.6 632,438 17. Congenital Malformations 2.5 622,098 18. Infections of the Urinary Tract 2.3 555,638 19. Neurosis 2.1 510,068 Source: National Health Survey, 1977 - 1980. 1/ Not grouped by seriousness of condition.  - 66 - COLOMBIA HEALTH SECTOR REVIEW Table 7: HEIGHT AND WEIGHT OF CHILDREN UNDER 15 BY SEX, 1965/1966 AND 1977/1980. Weight (Kg.) Height (Cms.) Females Males Females Males 1966 1980 1966 1980 1966 1980 1966 1980 Average 53.4 54.4 59.1 60.3 152.4 153.4 164.5 165.4 Median 52.4 53.3 58.0 59.1 152.4 153.3 164.5 165.7 Standard Deviation 10.3 10.2 9.4 9.8 6.4 6.7 7.6 6.9 Sources: National Morbidity Survey, 1965/1966 and National Health Survey, 1977/1980.  - 67 - COLOMBIA HEALTH SECTOR REVIEW Table 8: NUTRITION INDICATORS, 1965 - 1980 (Males) Age in 1965 1980 Years Mean + 1 Standard Mean + 1 Standard Deviation Deviation Height as Percent of Standard 0-4 95.1 + 06.7 95.9 + 05.5 5-9 91.3 + 05.6 93.7 + 05.2 10-14 89.2 + 05.4 91.7 T 05.0 Weight for Age as Percent of Standard o-4 89.8 + 15.0 90.1 + 12.9 5-9 82.9 + 12.4 85.7 + 12.8 10-14 73.5 T 13.2 77.0 T 13.0 Weight for Height as Percent of Standard 0-4 97.8 + 10.3 97.3 + 09.7 5-9 99.6 T 09.4 98.1 T 09.2 Sources: National Morbidity Survey, 1965/1966 and National Health Survey, 1977/1980.  - 68 - COLOMBIA HEALTH SECTOR REVIEW Table 9: PREVALENCE OF MALNUTRITION IN COLOMBIA IN 1965 AND 1980 Age Group 1965 1980 I. Defined as (80% Weight for Age (both sexes) 0 - 4 25.8 19.5 II. Defined as <2 S.D. Weight for Age O - 4 20.9 16.8 III. Defined as (75% Weight for Age, i.e. Gomez Classification Grades II and III 0 - 4 16.0 8.0 5 - 9 29.0 19.0 Sources: National Morbidity Survey, 1965/1966 and National Health Survey, 1977/1980  ClflT n)Ift A T11 AT l H T4f OU l flnITTf ~~~LL ~ El " IVU~ L OIkI. Ul"aL AI\V.V iIW Table 10: CONSULTATIONS FOR DISEASE OR ACCIDENT (EXCLUDING DENTAL DISEASE), BY PERSON CONSUILTED, REGION, ANNIUAL FAMILY INCOME, AGE, SEX AND SOCIAL SECURITY AFFILIATION (Rates per 100 persons) Person Consulted Total Physician Nurse Pharmacist Promoter Traditional Traditional Number of Practitioner Birth Attendant Other % Persons Country 59.8 3.3 7.7 0.3 3.7 0.4 24.9 100.0 4,138,290 Region: Atlantica 66.9 1.5 3.5 0.1 2.9 0.5 24.5 100.0 814,279 Oriental 63.8 4.7 13.2 0.8 4.3 0.3 13.0 100.0 575,853 Bogota 80.4 0.8 2.0 - 6.2 0.1 10.5 100.0 523,-920 Central 50.3 2.9 9.2 0.1 .6 0.6 34.2 100.0 1,219,377 Pacifica 52.4 5.8 8.9 0.5 4.1 0.1 28.2 100.0 1,004,861 Annual Family Income (Pesos) O Less than 60,000 45.9 5.4 7.6 0.2 5.0 0.8 35.0 100.0 1,076,576 60,000 - 119,999 58.2 2.7 7.7 0.1 4.6 0.1 26.5 100.0 1,301,491 120,000 and above 72.6 2.0 6.6 0.3 1.7 0.1 16.6 100.0 1,251,196 No information 61.6 3.5 10.1 0.8 3.7 0.6 19.7 100.0 509,027 Age - Sex Less than 4 years 65.6 4.4 10.5 0.7 3.6 1.1 14.1 100.0 706,861 5 - 14 years 46.1 3.0 6.5 0.6 3.8 0.0 40.1 100.0 636,203 Males 15-44 years 58.0 3.4 9.8 0.1 4.4 0.1 24.1 100.0 727,986 Females 15-44 years 64.5 2.7 6.7 0.2 2.3 0.4 23.3 100.0 1,147,742 45 and over 60.3 3.3 5.8 0.1 5.0 0.2 25.3 100.0 919,498 Social Security* Affiliated 79.3 1.7 3.5 0.1 2.4 0.0 13.1 100.0 1,023,705 Not Affiliated 53.4 3.8 9.0 0.4 4.2 0.5 28.8 100.0 3,114,585 * Affiliates of ISS, provident funds, and family welfare funds. Not affiliates of any social security system and affiliates of -private medical insurance. Source: National Health Survey, 1977-1980  COLOMBIA HEALT11 SECTOR REVIFW Table 11: CONSULTATIONS FOR DISEASE OR ACCIDENT, BY PLACE OF CONSULTATION, REGION,ANNUAL FAMILY INCOME, AGE, SEX AND SOCIAL SECURITY AFFILIATION (Rates per 100 persons) Place of Consultation Total Provident Welfare Private Number of ISS Fund Fund Office Home MS/SSS Other Persons Country 7.8 1.4 0.8 30.9 30.3 20.1 8.8 100.0 4,120,877 Region: Atlantica 4.3 0.8 0.6 37.6 32.0 19.6 5.2 100.0 810,383 Oriental 8.2 1.5 0.3 35.8 16.3 28.3 9.6 100.0 572,025 Bogota 12.1 3.8 2.9 47.2 12.9 16.0 5.0 100.0 522,336 Central 5.7 1.6 0.3 25.9 39.7 18.1 8.7 100.0 1,214,475 Pacifica 10.7 0.2 0.7 20.1 34.6 20.3 13.4 100.0 1,001,658 Annual Family Income (Pesos) 0 Less than 60,000 4.6 0.2 0.1 23.9 41.1 22.7 7.3 100.0 1,074,605 60,000-119,999 8.1 1.3 0.9 28.8 30.2 21.6 9.3 100.0 1,295,409 120,000 and above 11.6 2.3 1.5 39.0 22.2 14.9 8.4 100.0 1,244,148 No information 4.5 1.6 0.3 30.8 27.5 23.4 11.8 100.0 506,715 Age - Sex Less than 4 years 4.5 0.4 1.0 37.6 19.1 27.9 9.9 100.0 705,383 5-14 years 1.0 0.7 2.1 25.6 42.7 17.0 10.9 100.0 631,694 Males 15-44 years 17.7 1.3 0.1 27.4 30.1 13.0 10.5 100.0 725,377 Females 15-44 years 7.9. 2.0 0.6 31.6 27.8 22.4- 7.7 100.0 1,140,537 45 and over 7.0 1.9 0.5 31.1 33.7 18.9 6.8 100.0 917,886 Social Security* Affiliated 31.5 5.5 3.2 26.0 14.7 11.1 8.0 100.0 1,020,379 Not Affiliated - - - 32.5 35.4 23.1 9.1 100.0 3,100,498 * Affiliates of ISS, provident funds, and family welfare funds. Not Affiliates of any social security sytem and affiliates of private medical insurance. Source: National Health Survey, 1977-1980  - 71 - COLOMBIA HFALTH SECTOR REVIEW Table 12: RELATIVE MAGNITUDES OF MINISTRY OF HEALTH BUDGETS, 1971-1980 (PERCENTAGES) Ministry of Health Central Gvt. Ministry of Health Year GDP GDP Central Gvt. 1971 1.06 11.4 7.3 1972 0.90 10.5 6.8 1973 0.92 10.2 7.4 1974 0.95 9.5 8.9 1975 0.86 10.9 6.7 1976 0.95 10.5 9.3 1977 0.76 9.9 8.6 1978 0.70 10.8 6.9 1979 0.83 10.7 7.3 1980 0.67 9.9 6-.7 Source: Percentages calculated from expenditure data for 1971-79, shown in Contraloria General de la Republica, Informe Financiera (annual). Indices for 1980 are calculated from initial budgets as shown in Congreso Nacional, Ley de presupuesto, 1980. Data compiled by Financing Division, MOH.  COLOMBIA HEALTH SECTOR REVIEW Table 13: SOURCES OF PUBLIC HEALTH CARE FINANCING, 1975 - 1980 (IN MILLIONS OF CURRENT PESOS) Sources of Revenues 1975______ 1976 1971 1978 1919 1980 Total Percent Total Percent Total Percent Total Percent Total Percent lotal l'ercent Central Government Budget Allocation 3a553 48.5 4.842 60.2 6.068 56.4 6,579 54.8 89 38.0 13,842 43.5 1. Ordinary allocation 1,140 15.6 2,130 26.5 2,684 24.9 2,723 22.7 3,286 13.9 6,185 19.4 2. Extraordinary allocation 733 10.0 196 2.4 1,316 12.2 825 6.9 210 0.9 1,320 4.1 3. Situado fiscal 1,022 13.9 1,619 20.2 1,792 16.7 2,270 18.9 2,939 12.4 3,968 12.5 4. Domestic credit 387 5.3 452 5.6 276 2.6 732 6.1 360 1.5 150 0.5 5. Foreign credit 271 3.7 445 5.5 -0- -0- 29 0.2 2,203 9.3 2,219 7.0 Regional Sources of -I Revenues 2,228 31.4 2.694 32.2 3,126 29.1 3.791 33.6 7,457 31.4 9,312 29.2 6. Sales of Services 870 11.9 1,139 14.3 1,328 12.4 1,594 13.3 4,397 18.5 5.525 17.4 7. Lotteries 6 Charities 505 6.9 588 7.2 743 6.9 858 7.1 1,366 5.8 1,790 5.h 8. Liquor tax 389 5.3 433 5.4 590 5.5 680 5.7 860 3.6 968 3.0 9. Beer tax 260 3.5 308 3.8 349 3.2 454 3.8 570 2.4 841 2.6 10. State & local allocations 204 2.8 226 2.8 116 1.1 205 1.7 264 1.1 188 0.6 Other Sources 194 2.6 204 2.5 348 3.2 400 3.3 2,913 12.3 1L9 68 6.O Carry-over 1,358 18.5 301 3.8 1_221 11.3 1,240 10.3 4,336 18.3 6,774 21.1 TOT'AL 7,333 100.0 8,041 100.0 10,763 100.0 12,010 100.0 23,704 I(t.O 3t,1846 IOt.u Source: Adapted from Alfonso Jaramillo Salazar, Kinistro de Salud, Informe al Honorable Congreso de la Republica de Colotthia,1979-198U p. 83, Table 26. L971t-  COL4MBIA HFALTH SECTCR REVIFW Table 14: DISTRIBUrION OF PUBLIC HEALTH CARE ALLCATIONS, 1975-1980* (IN MULLIONS oF CURRENT WSOS) Distribution of Experditures 1975 1976 1977 1978 1979 1980 Total Percent Total Percent Total Percent Total Percent Total Percent Total Percent A. By institution 1. Ministry of Health 225 3.1 260 3.2 301 2.8 779 6.5 2,007 8.5 2,649 8.3 2. Regional Health Services 4,167 56.8 4,185 52.0 4,652 43.2 5,350 44.5 9,578 40.4 16,082 50.5 3. ICBF 1,158 15.7 1,795 22.4 2,727 25.3 2,587 21.5 6,371 26.8 5,001 15.7 4. INSFOPAL 465 6.3 826 10.3 830 7.7 1,208 10.1 3,217 13.6 3,345 10.5 5. IN 191 2.6 256 3.2 341 3.2 532 4.4 790 3.3 1,773 5.6 6. FNH 246 3.4 335 4.2 364 3.4 465 3.9 1,185 5.0 1,929 6.1 7. SEM 129 1.8 156 1.9 186 1.7 228 1.9 487 2.1 997 3.1 8. Other* 752 10.3 228 2.8 1,362 12.7 861 7.2 69 0.3 70 0.2 TOMAL 7,333 100.3 8,041 100.0 10,763 100.0 12,010 100.0 23,704 100.0 31,846 100.0 B. By Purpose 1. Operations* 4,594 62.6 5,651 70.3 8,467 78.7 9,472 78.9 17,242 72.7 19,437 61.0 2. Investment 2,739 37.4 2,390 29.7 2,296 21.3 2,538 21.1 6,462 27.3 12,409 39.0 TCOAL 7,333 100.0 8,041 100.0 10,763 100.0 12,010 100.0 23,704 100.0 31,846 100.0 Source: Same as Table 13. * This table includes the "extraordinary allocations" shown in the source in footnote (1), and in Table 13 (abow) as item 2. These extraordinary allocatins, for the most part, are channelled to the regional health services and the hospitals they administer (Item 2 in this table). Hbwever, details of how extraordinary allocations were distributed during the years 1975-78 are not awilable; starting in 1979, extraordinary allocations are included in the institutional allocations.  COLOMBIA IIEALTH SECTOR REVIEW Table 15: DISTRIBUTION OF PUBLIC HEALTH CARE EXPENDITURES, 1978 (IN MILLIONS OF PESOS) A. By Institution Total Percent B. By Budget Category Total. Percent 1. Hlnistry bf llealth 606 4.0 1. Operating Expenditures 12,403 82.2 2. Regional llealth Services a. Salaries vages 6,492 43.0 a. Regional & local hospitals 6,938 46.0 . . b. Drugs 1,200 8.0 b. Other regional services 2,449 16.2 c. Food 1,009 6.7 3. ICBF 1,806 12.0 d. Other 3,702 24.5 - 4. INSFOPAL 1,120 7.4 dO r0 2 2. Investment expenditures 2,688 17.8 1 5. INS 379 2.5 a. Construction & equipment 1,866 12.4 6. FNII ,5S 0.4 b. Training 669 4.4 7. SEM 267 1.8 c. Research 153 1.0 8. Others 1,471 9.7 TOTAL 15,091 100.0 TOTAL 15,091 100.0 Source: Ministry of llealth, Gasto en Salud, 1978, forthcoming.  - 75 - COLOMBI]A HEALTH SECTOR REVIEW Table 16: FOREIGN AID COMMITMENTS AND EXPENDITURES 1973-1982, AS OF 1978, IN MILLIONS OF DOLLARS Type of Aid Commitments 1973-82 Expenditures Total Percent 1972-78 Bi-lateral loans 136.2 33.9 87.3 Canada 25.5 6.3 5.4 Europe 64.6 16.1 39.6 U.S . 46.1 11.5 42.2 Multi-lateral loans 152.7 38.0 71.3 IBRD 84.3 21.0 33.9 TDB 68.4 17.0 37.4 Grants 113.3 28.1 94.0 FOREIGN AID, TOTAL 402.2 100.0 252.6 Source: Ministry of Health, Analisis de Fuentes de Financiamiento del Sector Salud, 1978, Tables 2.3  COLOMBIA HEALTH SECTOR REVIEW Table 17: REGIONAL DISTRIBUTION OF HEALTH CARE ALLOCATIONS AND POPULATION, 1980* (IN MILLIONS OF PESOS; POPULATION IN THOUSANDS) National Regional Combined Distribution Distribution Regional Health Services Sources Sources blocations of Allocations of population Total Percent Total Percent Total Percent (percent) Total Percent Antioquia 677.6 52.9 603.2 47.1 1,280.8 100.0 11.8 3,812 14.5 Atlantico 261.3 55.0 214.0 45.0 475.3 100.0 4.4 1,228 4.7 Bogota, D.E. 642.1 67.3 311.8 32.7 953.9 100.0 8.8 3,627 13.8 Valle 494.2 56.9 374.6 43.1 868.8 100.0 8.0 2,616 9.8 Sub-total 2,075.2 58.0 1,503.6 42.0 3,578.8 100.0 33.0 11,283 42.8 All others 4,556.4 62.8 2,702.2 37.2 7,258.6 100.0 77.0 15,053 57.2 TOTAL 6,631.6 61.2 4,205.8 38.8 10,837.4 100.0 100.0 26.136 100.0 Sources: hlealth care expenditure data from Division of Finances, HOlt; population estimates from DNP. * These data refer to allocations for operating expenditures, only.  - 77 - COLOMBLk HEALTH SECTOR REVIEW Table 18: PUBLIC HEALTH INVESTMENT ALLOCATIONS AND EXPENDITURES, 1980 (IN MILLIONS OF PESOS) Personal Environmental Physical Total Investment Inscitution Health Care Health Services Infrastructure Total Percent Rec'd 570 70 306 946 6.4 MOH Spent 579 47 232 858 7.5 Percent 101;6 67.1 76.6 91.0 Rec'd 8,241 -0- -0- 8,241 54.9 ICBF Spent 5,291 -0- -0- 5,291 46.5 Percent 64.2 -0- -0- 64.2 Rec'd -0- 44 2,509 2,553 17.0 NSFOPAL Spent -0- 75 2,158 2,233 19.7 Percent --