What Is Known about Efficacy and Effectiveness

The efficacy of health and nutrition interventions in developing countries has been established for decades (for example, Gwatkin, Wilcox, and Wray 1980). Prospective studies in several settings showed that health interventions with or without supplementary foods caused children to thrive and survive better: studies in Narangwal, India (Kielmann and others 1978; Taylor, Kielmann, and Parker 1978); by the Institute for Nutrition for Central America and Panama (Delgado and others 1982); in Jamaica (Waterlow 1992); and in The Gambia (Whitehead, Rowland, and Cole 1976) are examples. ² These studies showed the effect of interventions on growth and (usually) mortality but did not generally factor out the relative contributions of health and nutrition. In fact, results from Narangwal showed similar mortality effects from food or health care; results from The Gambia indicated interaction such that sick children did not grow even with adequate food intake (appetite also playing an important role), and well children did not grow with inadequate food intake (Gillespie and Mason 1991, annex 2).

By the early 1980s, the conclusion, based on data at the experimental level (not from routine large-scale programs), was that better health and better nutrition are both required for child survival and development. This conclusion remains generally agreed on today; furthermore, concern exists that health interventions may become less effective unless nutrition is concurrently addressed (Measham and Chatterjee 1999; Pelletier and Frongillo 2003). In their chapter on malnutrition in the first edition of this book, Pinstrup-Andersen and colleagues (1993) drew largely on efficacy findings, with an emphasis on food supplementation. Those studies are not revisited here, but we can continue to build on their conclusions.

The efficacy studies were followed by a number of national or other large-scale programs in several countries. Some of those were a direct follow-on; for example, the World Bank Tamil Nadu Integrated Nutrition Program (TINP) followed the Narangwal study, which was supported by the U.S. Agency for International Development (USAID). A number of overviews and analyses of these programs have been conducted—for example, Allen and Gillespie (2001); Berg (1981. 1987); Gillespie, Mason, and Martorell (1996; includes a summary of overviews, 60); Gillespie, McLachlan, and Shrimpton (2003); Jennings and others (1991); Mason (2000); Sanders (1999); and Shrimpton (1989). These plus some newer examples provide case studies for this chapter, and the sources for the case studies are included in table 56.2.

Underweight prevalences are improving at about 0.5 percentage points (ppts) per year except in Sub-Saharan Africa, which is largely static (ACC/SCN 1989, 1992, 1996, 1998, 2004). Programs are needed to accelerate this trend. Cost data from an earlier study (Gillespie and Mason 1991, 76), combined with the estimated improvements from large-scale programs, led to the assertion that "there seems to be some convergence on around $5 to $10 per head (beneficiary) per year being a workable, common level of expenditure in nutrition programmes, though not generally including supplementary food costs … effective programmes, with these levels of expenditure, seem to be associated with reducing underweight prevalences by around 1–2 percentage points per year" (Gillespie, Mason, and Martorell 1996, 69–70).

A further important consideration is that the effect is likely to be nonlinearly related to the expenditure, showing the familiar dose-response S-shaped curve. Thus, the first expenditures produce little effect on the outcome, and one needs a minimum input level of resource use before a worthwhile response is achieved (Habicht, Mason, and Tabatabai 1984). This factor generally applies to drawing inferences from cost-effectiveness ratios, which often assume linearity. If the relation is S-shaped, the implication is important: applying too few resources does not simply solve the problem more slowly but does not solve it at all and is a waste. Therefore, program intensity (resources per person) is a critical measure.

Effective interventions must include a range of activities relating to health and nutrition. They should be multifaceted, not just for effectiveness but also for organizational efficiency. The structure needed for community-based programs could never make sense or be sustainably set up for single interventions alone. One often-argued case (for example, by Save the Children U.K. 2003) concerns children's growth monitoring: evidently growth monitoring in isolation from activities that improve children's growth is not going to achieve anything (or worse, considering the opportunity cost); however, weighing children and charting their weight can be a useful part of broader programs (for example, as growth monitoring and promotion).

Community- and Facility-based Programs

Protecting and improving health, especially in poor communities, requires a combination of community- and facility-based activities, with support from central levels of organization, as well as some centrally run programs (for example, food fortification). The place of these activities in a strategy is likely to vary, depending on level of development (of infrastructure, health services, and socioeconomic status) and on many local factors. For the poorest societies, the first priorities are basic preventive services, notably immunization, access to basic drugs, and management of the most serious threats to health, such as some access to emergency care. Moving up the development scale, starting community-based activities may soon become cost effective for prevention, referral, and management of some diseases (notably diarrhea) when coverage of health services is poor. Community-based programs continue to play a key role until health services, education, income, and communications have improved to the point that maternal and child mortality has fallen substantially and malnutrition is much reduced; at this intermediate development level, the needs are less felt, and health services again take on a more prominent role. In this scheme, the widely felt need for better access to emergency obstetric services is problematic, requiring a well-developed human and physical infrastructure, yet arguably being one of the highest priorities.

Facility-based programs can be seen either as linking with the community program (referrals, home visits from clinics, and so forth) or as actually being part of the same enterprise. A distinction is that community-based activities take place outside the health facility, in the home or at a community central point, even if they may be supported by health personnel based in health facilities. The local workers in community-based programs may be drawn from the community itself, may be home visitors from a health center or clinic, or may sometimes be volunteers supervised by these home visitors. Many community-based programs come under the health sector, whatever the exact arrangements with local health services. Regarding specific program components, we return to the relative role of community programs and facilities later.

The integrated management of infant and childhood illness (IMCI) program provides guidance mainly on the curative health aspects and contains a number of nutrition activities (for example, administration of vitamin A capsules). Links to local health facilities are essential for the maintenance of the community activities and for referral in cases of illness (see chapter 63). As the IMCI training and implementation progresses, it should integrate directly with CHNPs (in fact, become part of the same exercise), which will add treatment of additional diseases. IMCI addresses diarrhea, acute respiratory infection (ARI), malaria, nutrition, immunization, safe motherhood, and essential drugs (WHO 1997). The 16 key practices for child survival defined in the context of IMCI (Kelley and Black 2001, S115) are exactly those to be promoted within CHNPs, and most are already included (four are nutritional).

Decentralization should be considered in this context. Although decentralized systems might be thought to be more effective in supporting CHNPs, the evidence for this assumption is scarce. Decentralization can reduce resources available at the local level if it involves devolving responsibility without the concomitant budgetary resources (Mills 1994). For example, in Kenya, decentralization did not accompany devolving authority for raising revenue locally. In other cases (for example, the Philippines), decentralization has involved a shifting of resources, but with priorities set in the local government units by locally elected officials (municipal and city mayors), these resources may be used for shorter-term priorities than under previous, centrally decided, policies.

Success Factors

A number of useful concepts grew in the 1990s in relation to effective community-based programs. The concept of success factors helped sort out complex interactions: when numerous possibilities exist, understanding the successful pathway to effectiveness is more important than trying to disentangle what did not work. Focusing on successful programs helps simplify complexity and identify success factors, only some of which are programmatic (directly under the influence of the intervention itself); others are contextual.

The importance of context, within which programs are initiated and run, thus emerged as crucial, and priority factors were proposed from studies of community-based programs in Asia (Gillespie, Mason, and Martorell 1996, 67; Jonsson 1997). Sanders (1999) described similar concepts under the headings of community participation and political will. This distinction and interplay between context and program factors is helpful in identifying required supporting policies to improve the context to make programs work. Details are in the later section titled "Contextual Factors."

An overall framework (figure 56.2) for causal links to child survival and nutrition, put forward by the United Nations Children's Fund (UNICEF 1990), gave a basis for a common language—even if the details might be questioned—revolving around food, health, and care as proximal causes to be addressed through programs. Improving these factors attacks hunger, disease, and neglect, which are the converse of food, health, and care. Basic causes are, like context, open to influence through policy decisions and acting through directly influencing food, health, and care and by modifying the effect of programs. Here malnutrition is seen as the outcome of processes in society, and direct interventions are seen as both shortcutting the needed basic improvements in living conditions and being dependent on these improvements in the long run for sustainability.

Figure 56.2

Conceptual Framework for the Causes of Malnutrition in Society

Community-based Programs—What Are They?

Community health and nutrition programs are often initiated and run by the health sector, but sometimes a separate ministry (for example, in India and Indonesia) or service (for example, in Bangladesh) is set up. Attempts to use a national coordinating body appear to be less effective in leading to widespread community programs; an example existed in the Philippines until approximately 2000 (Heaver and Mason 2000). This ineffectiveness stems from the tendency of the coordinating body not to have direct authority over fieldworkers or the budget to create a national program with sufficient coverage and intensity to have a measurable effect. In some other cases, the services linked to poverty alleviation and social welfare programs can play this role (for example, the Samurdhi program in Sri Lanka). Involvement of the health services remains crucial, sometimes as the operational agency responsible for the programs and certainly always for referral.

CHNPs have so far been much more relevant to communicable diseases than to noncommunicable diseases in conditions of poverty and where undernutrition is common. (An exception occurs if CHNPs help prevent intrauterine growth retardation with later risks of noncommunicable diseases.) However, in areas where diet-related chronic diseases are developing in conditions of poverty (for example, much of Latin America and the Caribbean) and obesity is rising rapidly, the promotion of behavior change through counseling in CHNPs may become increasingly important. Promoting healthier diets requires access to outlets for fruit and vegetables, often displaced by fast foods, which should be a concern of community activities, as should lifestyle improvements such as use of exercise and recreational facilities.

CHNPs often include activities well beyond direct prevention and behavior change. As envisaged with primary health care, water, sanitation, and other aspects of environmental health are frequently included, as well as agricultural interventions (for example, Zimbabwe in the 1980s). In Thailand, the village programs are part of the "Basic Minimum Needs" approach, which includes housing and environment, family planning, community participation, and spiritual and ethical development.

A diagram of the structure, derived from Thailand's program (figure 56.3), shows the relations between services that provide supervision and contacts with the community ("facilitators") and with community workers, referred to as "mobilizers."

Figure 56.3

General Structure for Community-based Programs, Based on Thailand's Program

The activities undertaken in CHNPs—the program content—are familiar and are described here only briefly. Program components, implemented by village workers or in facilities, come under the following headings, which form a menu, with the actual mix depending on local capabilities and conditions (UNICEF 1998, 84; see chapter 24):

• Prenatal care includes checking weight gain in pregnancy, prepregnancy weight, anemia, and blood pressure; providing multiple micronutrient supplementation and immunization (tetanus); counseling on diet, workload, breastfeeding; and predicting and arranging for delivery.
• Women's health and nutrition entails counseling on health and nutrition and checkups, promoting improved status and resource allocation in home and outside, promoting improved access to health services, and often offering family-planning services (these services may even be an initiating factor for CHNPs, for example, in Indonesia).
• Breastfeeding includes providing knowledge on practices (initial, exclusive, continued); arranging mutual support; building confidence; preventing misinformation and undermining factors; facilitating time for breastfeeding; and providing information along the lines of the infant formula code.
• Complementary feeding includes providing knowledge and counseling (timing of introduction, type, energy density, frequency, and so on); sometimes promoting village or urban area production of weaning foods; sometimes marketing inexpensive food; facilitating mother's time allocation; and promoting technology—storage, preservation, hygiene methods (fermentation, even refrigerators).
• Growth monitoring and promotion requires equipment (scales, charts, manuals); training and supervision; needs training of weigher to interpret charts and counsel mother; and a referral system for problems (for treatment, counseling, or other preventive intervention if growth is faltering). Weighing at birth and monthly weighing should be included, if possible, and adequate weight gain (rather than achieved weight or any gain) should be used for guidance on counseling or other intervention.
• Micronutrient supplementation should include vitamin A for nonpregnant and pregnant women (low dose weekly, preferably as part of multinutrients); for women within one month of delivery (massive dose to protect infant through breast milk); for infants and children (massive dose at nine months immunization contact and thereafter every six months and when medically indicated). It should also include vitamin A—daily or weekly, with immunization campaigns, and so forth—and iron—daily or weekly for women (especially during pregnancy) as well as for children and adolescents. Iron is usually provided together with folic acid and may also be provided as part of multiple micronutrient supplementation. Iodine is usually provided by fortification and can be an infrequent (six-monthly) oral supplement, if necessary, but it should be part of multiple micronutrients for pregnancy.
• Micronutrient fortification is not usually included locally, although it is an important central program, but local monitoring is a coming opportunity, especially of iodized salt testing kits.
• Supplementary feeding, using external supplies may sometimes be appropriate in emergencies and in conditions of extreme poverty (for example, the Bangladesh Integrated Nutrition Program, or BINP), providing 200 to 500 kilo-calories per person per day, but otherwise it is to be avoided as costly, with high opportunity cost, and not very effective; moreover, it can distort programs, which come to be seen largely as a source of free food.
• Supplementary feeding, using local supplies can be useful for complementary feeding (weaning) if carefully organized (which requires some resources). Village community production and processing are useful, if feasible (for example, in Zimbabwe), and the system can move to coupon method (for example, in Thailand).
• Oral rehydration includes highly effective local preparations for dehydration in acute diarrhea, as well as (or better than) oral rehydration salts. These preparations require counseling of mothers and take a lot of parents' time. Persistent diarrhea requires other intervention, especially nutritional. Care of children during sickness—especially continued breastfeeding and other foods—needs to be stressed (applies also to other illnesses).
• Immunization includes informing, referring, and facilitating.
• Deworming requires distribution and dosage supervision of mebendazole every few months, a highly effective nutrition intervention. Distribution methods are an issue.

The relative suitability of community- and facility-based operations for the different components again depends on local conditions, and these operations should be complementary. Community activities are essential for infant and child feeding, other caring practices, environmental sanitation, and the like. Facilities have a key role in immunization, prenatal care, and—of course—referral for treatment. Growth monitoring, micronutrient interventions, oral rehydration, and similar activities may be focused in either. Because it has more regular contact with clients, a community-based program may be more effective in actually reaching mothers and children with the component interventions than one that is facility based. Box 56.1 compares two programs in Honduras that offered the same content but differed in where the programs were based.

Box 56.1

Differential Effectiveness of Community- and Facility-based Programs. Effectiveness is more likely to be possible through community-based programs because contact with care-givers is typically more frequent and consistent. For example, 83 percent of children (more...)

Programmatic Factors

Programmatic factors are considered first in terms of the characteristics of the activities—their population coverage and targeting, how much resources are applied per head (intensity), and the technologies used. Then the needs for initiating and sustaining these activities are discussed—the training needs, supervision methods, and (importantly) incentives and remuneration for field workers.

Coverage, Targeting, Resource Intensity, and Technology

Even effective programs improve the health and nutrition only of those they reach, so achieving as complete coverage as possible of those at risk is a major determinant of the effect. Although variations in the content of programs are seen in different circumstances, most activities are common to most programs. Variations in effect stem from factors such as coverage and adequacy of resources. How have CHNPs fared in reaching large sections of the population with adequate resources—and, indeed, what is the gap that would need to be filled? The achievements of the 14 programs drawn on here as case studies are summarized in table 56.3.

Table 56.3

Characteristics of Selected Programs.

The programs expanded to include most of the communities within the areas targeted. The common evolution was to target select areas and specific biological groups within those areas—generally women and children—but not to give priority to any great extent to poorer or less healthy communities. Screening is sometimes done of individuals for admittance into the programs (a form of targeting), based on nutritional status, as in growth monitoring and promotion, as well as on a onetime basis (for example, thin children in Zimbabwe). Recent thinking suggests that because mortality risk, growth failure, and morbidity are concentrated in children less than two or three years of age, in contrast to an earlier focus on children under five, these younger children should increasingly be a focus of CHNPs. A common policy observed in practice, therefore, is to aim for complete coverage within the areas participating, adding new sites until the entire region is covered. Relatively untargeted expansion to universal coverage may have been at the expense of establishing adequate resources and quality in the areas initially covered. In at least one case (Thailand), having achieved broad coverage and reduced malnutrition, the program became more targeted to areas in which progress was lagging. The coverage figures in table 56.3, although approximate, demonstrate considerable success in initiating and implementing CHNPs on a large scale—usually enough to have a substantial effect if the other factors needed for success were met.

How complete a coverage of the population should one recommend? This factor relates to targeting, to the additional resource requirements to reach the nonparticipants, and to their level of risk. Usually risk is spread throughout the population, although the extent varies considerably—at least a doubling of indicators of risk is usually seen between better-and worse-off areas or groups (for example, see Mason and others 2001, figures 1.4–1.7). The remoter areas—or groups that are hard to include for other reasons—may be more expensive to reach. Clearly the calculations depend on conditions and have to be made on a case-by-case basis. The principle is obvious: only those areas and people included in CHNPs are going to benefit; so wherever need exists, programs are indicated. The implementation strategy, in theory, may need to begin with the most urgent needs, although in practice, programs may expand from the easier, more accessible areas; this practice seems reasonable, provided that the expansion really occurs and leads to equitable use of resources.

The program content is a mix of the components described earlier, varying with local priorities. The most crucial difference is whether extensive supplementary feeding is included. In middle-income countries, supplementary feeding was less prominent, often considered unnecessary, and because expensive, perhaps counterproductive (for example, in Costa Rica; Mata 1991). At the other extreme, such as for the Integrated Child Development Services (ICDS) in India, food distribution became the raison d'être of the program but, alone, was again probably not worthwhile. For some of the intermediate cases, supplementary food played a supporting role, with varying results. Except in the very poorest societies, supplementary feeding seems unlikely to be cost-effective.

The resources used for the programs found in table 56.3 can be expressed per participant (referred to as intensity), as total expenditures, and in terms of personnel; the latter figures may be more generalizable. (The outcomes associated with these resources are shown in table 56.5.) Data such as these have been the basis for estimating that US$5 to US$10 per child per year may be needed for effective programs. The dollar figures vary from less than US$1 to more than US$20. Probably the low end of this range (say, less than US$1 per child per year) does explain low or doubtful effect. Both low coverage and low intensity may explain the unchanged underweight prevalences in the Philippines until 2000. Fund levels in Indonesia are unsure; Rohde (1993) gave a figure of less than US$1, but others gave higher estimates. Most would reckon the intensity in India too low (Measham and Chatterjee 1999) at about US$2 per child per year. Looked at otherwise, the intensity planned for external funding (even if part of such funding is international costs) is in the US$10 to US$20 range (Bangladesh, India—Tamil Nadu, and Tanzania) and is the same as the estimate for Thailand. A level of US$10 to US$20 per participant per year is probably advisable for planning and sustaining effective programs.

Table 56.5

Outcomes and Resources in Selected Programs.

The intensity measures of workers per mother-child and the supervision ratios are relevant in assessing needs. The suggested norms, originating from the Thai experience are 1:10–20 for both. Since then, it has emerged that the full-time equivalence of community workers must be taken into account; the Thai workers are local volunteers, probably devoting 10 to 20 percent of their time. In Honduras, Fiedler (2003) in a careful cost study estimated that each volunteer spent 3.5 hours per week (less than 10 percent of full-time equivalent, or FTE), with a ratio of 1 volunteer to 8 children. The ratio of community health and nutrition workers (CHNWs) to children may, therefore, be as low as 1:200 for FTEs and as high as 1:8 or 1:10 for part-time volunteers. In Jamaica, where the community health aides work full time, the ratio is 1:500 households; in the BRAC program in Bangladesh, it is 1:300, about half-time work (afternoons) (Chowdhury 2003). (An indication of the status of community workers is shown in brackets in the last column of table 56.3.) In any event, these ratios provide some basis for gauging the adequacy of personnel, and it seems that an effective ratio may be about 1:500 for community workers employed full time and 1:10 or 1:20 for local volunteers working part time.

In reality, the ratios of community workers to children are probably—not surprisingly—on the low side. Thailand, which trained 600,000 village workers (1 percent of the population), operated at about 1:20 for part-time volunteers, with similar supervision ratios. The Indonesian program was similar (or better) but had much less supervision. In contrast, the low resourcing of the ICDS in India shows up in a ratio of 1:200 (for part-time anganwadi workers, or ANWs), and in the Philippines, the ratio has until recently been 1:300 (for essentially voluntary workers).

Increased application of technology can contribute to the organization and running of community-based programs. Technology can be applied easily to methods of assessment and monitoring of children's progress; improved weighing scales (or in some circumstances, where rapid assessment in remote areas is important, using arm circumference) can simplify anthropometry. Modern computer technology for recordkeeping could be much more widely used, freeing staff time for home visits (for example, in Jamaica); e-mail, which is being rapidly adopted, has great potential for transferring information, troubleshooting, and consultation. Cell phone use is beginning to transform communications even in the poorest countries, where it is leapfrogging landline installation and use; as coverage expands, it will facilitate referral, for example, for emergency obstetric care, the need for which may first be identified by community workers. Coupled with improved transportation and procedures to allow the use of such transportation in cases of urgent need, modern communications can link communities to centers with advanced knowledge for information exchange and, by facilitating transportation when time is crucial, for referral. Modern communications may also provide more efficient ways of providing training, retraining, and supervision.

Application of current research and resulting technologies can improve many of the other interventions discussed earlier. In the micronutrient field, periodic supplementation (with vitamin A in high doses) can be extended through community programs, and fortified foods and micronutrient "sprinkles" can be promoted (see chapter 28). The prospect of enabling communities to test their salt for iodine content with simple and cheap test kits is intriguing and has often been recommended but has not yet been widely applied. Improved immunization technology should continue to protect health, for which CHNPs' main role is to provide information and to ensure that children are taken for immunization (either to regular clinics or for National Immmunization Days and the like). Periodic deworming can be conducted by community programs (and hookworm vaccines currently under development may soon contribute). Supporting the use of insecticide-treated bednets could be fostered through CHNPs. By far the most potentially important application of technology, certainly in Sub-Saharan Africa, will be the unprecedented effort to provide millions of people with antiretroviral therapy and associated care and support, as discussed later.

Contextual Factors

Community-based programs can work usefully, bringing steady progress; whether they do depends on myriad factors relating to the context. Three different concerns are (a) factors affecting widespread initiation of CHNPs of potentially adequate coverage, intensity, and content; (b) factors that lead to sustainability; and (c) factors that allow activities to be effective in improving health and nutrition—at best, when they, themselves, also contribute to a rapid transitional improvement, as in Thailand, Costa Rica, and Jamaica.

Contextual factors may bring about improvements in health and nutrition without any additional direct action—through improving living conditions, education, and so forth. Often, the changes caused by such nonprogrammatic factors are difficult to distinguish from program effects (current examples are in Bangladesh and Vietnam, both showing rapid improvement in nutrition). Moreover, the same factors (again, such as education) may both produce endogenous change and increase the effect of program activities.

Five contextual factors have been suggested as priorities (in Asia; Mason and others 2001):

• women's status and education
• lack of social exclusion
• community organization
• literacy
• political commitment.

Table 56.4 shows estimates of the positions of countries with case study programs in regard to these factors. The levels of health and administrative infrastructure have been added. The table also shows changes in these factors that may help explain why the CHNPs declined in three cases.

Table 56.4

Context in Which Selected CHNPs Start and Run.

Political commitment can lead to initiating community programs and mobilizing resources. It may also respond to emerging community mobilization, as seems to be the case when programs start after political upheavals, as in Zimbabwe and Nicaragua. Declining political commitment accounts for loss of interest by the government in CHNPs; economic decline undermining resource availability may cause a shift away from financial support of CHNPs (for example, in Tanzania). In table 56.4, estimates of levels of contextual factors are totaled both without and including political commitment (last two columns). The total without commitment may indicate how favorable the context is if commitment is then made. Costa Rica, Jamaica, and Thailand had a favorable context and, with commitment, succeeded. The Philippines had comparable favorable conditions—the position of women is generally good, there is limited social division (exclusion), and so on. However, the necessary commitment (of resources, in particular) was made only recently, with new legislation, adherence to regulations (for example, iodized salt went from 25 to 65 percent coverage), and increased resource allocation and assignment of community workers. This new commitment may well explain the recently resumed decrease in child malnutrition (figure 56.2). In other examples—such as Indonesia and Tanzania—the conditions were moderately favorable, and while political support and finance existed, progress was made. In Tanzania, financial crisis denied the programs sustained support; in Indonesia and Zimbabwe, bureaucratization and centralization of the political process, followed by political turmoil, contributed to a similar outcome (Sanders 1993). The situations in India and Bangladesh have not been very favorable. The position of women and social rifts, amounting to exclusion, have probably inhibited effective programs, even with political commitment. This context may now be changing in Bangladesh, as seen in the activities of BRAC. Finally, this analysis demonstrates the relation of decline in programs to falling political commitment in Tanzania, Zimbabwe, and Indonesia.

If this analysis approximates the truth, the forward-looking policy implications may be important:

• First, investing initially in a favorable context makes sense (as does possibly committing resources preferentially to interventions in the more favorable contexts). Supporting policies can address social constraints—such as improving education for women—and (relatedly) seek to improve human rights. In many cases, human rights may be of overriding importance for health: Farmer (2003) has made a compelling case for rethinking health and human rights as a prerequisite for progress and as a responsibility for those working for health, especially of the poor and of the destitute sick. This investment may be long term and difficult—as in Kerala, India, for instance—but must be seen as integral to the struggle for health (Sanders 1985). Operationally, this commitment to human rights puts greater responsibility on advocates and investors in health to broaden the dialogue and scope for allocating resources and to avoid committing resources regardless of the prospect of success as influenced by the social and human rights context. In health and nutrition, as in other areas, adjustment of policies to support the success of interventions would be pragmatic as well as the right thing to do.
• Second, even if the context is more favorable, genuine political commitment is still essential. Excessive donor input may inhibit this commitment. It is striking that Thailand had to reject donor influence and make its internal decisions before its programs became successful (Tontisirin and Winichagoon 1999), Costa Rica had to fight and overcome a medicalized approach (Muñoz and Scrimshaw 1995), and Indonesia's posyandu system was undermined when treatment displaced prevention (Rohde 1993).
• Third, it is clear that severe economic stress, political pressure, or both have caused unsustainability (Indonesia, Nicaragua, Tanzania, and TINP).
• Fourth, if the context is unfavorable, it might be better to work on improving the context than to commit resources to programs that may not succeed—but, of course, success in improving context itself depends on circumstances, notably political commitment.

Considerations like these should contribute to identifying supporting policies needed for programs to be effective and modifications to interventions in particular conditions. For example, it is often observed that a particular factor—say, access to health services—is more strongly related to improvement among the better off (for example, the educated) population. This interaction of program with context leads to identifying new needs—in this example, perhaps facilitating access for the illiterate. In the longer run, resources or legislation (for example, to combat social exclusion or discrimination against women) may be highlighted as prerequisites before a program can be expected to work. Often failure to take account of context when trying to transfer experiences from a pilot trial ("scaling up") may explain why efficacious interventions prove ineffective in a larger population.

This analysis of contextual circumstances indicates that targeting the poor may not always be cost-effective, and some interventions may not be feasible in certain contexts. An example is when the health infrastructure and services are almost nonexistent; under those conditions, it can be argued that emergency treatment (especially for the diseases addressed by IMCI) should be established and reliable resources put in place first. A similar difficulty, often seen in food security, is that most interventions may not work for the poorest of the poor. For instance, supporting food (or cash crop) production in low-potential areas may not be realistic; nonagricultural employment may be better.

Thus, community-based programs work in a specific time and place: programs may start, work for a time, and then evolve or fade away. Even if they fade away, some useful effect may be achieved: sustainability need not be forever. At the same time, short project cycles (three years for many donors) can act against sustained programs. Some compromise in donor policies to allow assurance of continuity for reasonable periods (such as 10 years) could do a lot to increase the effectiveness of donor support to CHNPs.

The essence of time and place is not fully understood. Werner and Sanders (1997) give examples of favorable times, as when the old order is changing (for example, after internal conflict, as in Nicaragua and Zimbabwe) and when there is renewed vigor and some new organization is in place. Another generalization of a favorable context is when energy and inter-connectedness exist in society. Thailand illustrates both: the Thais needed to change the approach, moving away from donor influence, in order to initiate the successful community programs that helped transform health and nutrition throughout the country, and that worked in part because of cohesive features of Thai society (Tontisirin and Winichagoon 1999).

In these examples, programs that continued on a large scale—either until the problem was largely resolved, as in Costa Rica, Jamaica, and Thailand, or as it was expanding, as in BRAC in Bangladesh or AIN-C (Atención Integral a la Niñez Comunitaria) in Honduras—clearly had supportive context, but their specific common features (and hence how they could be replicated) are elusive. Perhaps one crucial condition for success is that circumstances are such that people and communities begin to have the sense that they can take responsibility for—and control of—their health and quality of life. Responsibility comes with the emancipation of societies from colonial or other repressive conditions and possibly when grassroots attention becomes widespread, as it did in Bangladesh through an NGO that identified with the people. Evidence is growing that, among the poor in the United States, this sense of control is directly related to better health and reduced exposure to HIV and AIDS; Sampson, Raudenbusch, and Earls (1997) call the concept collective efficacy. Cohen and others (2000) show that health conditions improve when communities themselves fix up their environment—the "broken windows" theory. Such ideas may equally apply to poor communities, especially urban ones, in developing countries too.

Results

Indicators of progress in implementation—process indicators—referring to coverage, intensity, and so on, are shown in table 56.3. As discussed earlier, most programs expanded population coverage without much targeting. But usually the level of resource application (intensity) was on the low side. More research is needed on the contribution of CHNPs to health process indicators, such as immunization coverage rates, as well as to nonhealth activities, for instance, in agriculture and community development.

Impact evaluation, which refers to the net effects of interventions on changing health outcomes, is sorely lacking. The efficacy of most of the component parts of CHNPs, when resources are adequate and the problems are correctly identified, is established, but in routinely administered large-scale programs, the changes in outcome that can be ascribed to program actions are less known. Although controlled trials by definition are not applicable, plausible evidence can be obtained by careful attention to research design, measurement, and analysis (Habicht, Victora, and Vaughan 1999). Some form of "with and without" and "before and after" comparisons is needed; for instance, such methods as staggered implementation, natural experiments, and selection of comparison groups with some statistical control can yield valuable information now lacking and should be more widely attempted. In this context, it should be noted that because of the timing and level of effort necessary for the evaluation, the impact evaluation results (changes in outcome ascribed to the program) may be more important for policy decisions on future programs than for the program that has been evaluated. Moreover, not all programs require detailed evaluation. Thus, financial support for such policy-relevant evaluations may come from budgets other than that of the program to be evaluated. The evaluations should also be prospective as far as possible, so decisions on evaluation design and finance are needed earlier rather than later.

Future Applications

The experience so far in CHNPs can be applied more broadly, especially where community organizations can sustain support for CHNWs. CHNPs have worked best so far in Asia and Latin America. However, with the HIV/AIDS epidemic in Sub-Saharan Africa needing high-priority attention, application of CHNP experience to the HIV/AIDS crisis should be explored.

Research Needs

The question of incentives, training, and support for community workers urgently needs research, both from current experience and with prospective designs. The issues include the following:

• What is the CHNW's status, relative to the community or to the government (or NGO) hierarchy?
• How are CHNWs selected and to whom do they report (for example, community health committees, supervisors employed in the health or other system)?
• What educational background and how much training and retraining—and by what methods—are needed for CHNWs?
• What ratios of CHNWs to households are effective (or most cost-effective), both as part-time workers (volunteer or otherwise) and as full-time employees?
• What supervision ratios work?
• What remuneration and incentives are effective?
• How can these efforts be financed?

Enough programs have been in operation for long enough that researchers could base on them much of the needed research on processes of implementation, launching new trials only for processes for which sufficient information does not exist. In contrast, impact evaluation requires new and preferably prospective studies.

A major gap in research is the application of community-based programs to urban areas. Urban communities are conceived differently from the rural areas of most CHNPs, organizations run along different lines, and so forth. Yet population growth is in urban areas, and some problems, notably HIV and AIDS, are worse there.

Finally, the cost-effectiveness analysis results given in an earlier section are based on rather few and approximate results. CHNPs may well provide a viable and cost-effective approach under many circumstances in poor countries, and it may be necessary to demonstrate this viability better and more quantitatively for support to CHNPs to compete with more traditional service delivery interventions. That, too, would constitute worthwhile research.

References

1. ACC/SCN (United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition). 1989. Update on the World Nutrition Situation: Recent Trends in Nutrition in 33 Countries. Geneva: ACC/SCN.
2. ———. 1992. Global and Regional Results. Vol. 1 of Second Report on the World Nutrition Situation. Geneva: ACC/SCN.
3. ———. 1996. Update on the Nutrition Situation, 1996. Geneva: ACC/SCN.
4. ———. 1998. Third Report on the World Nutrition Situation. Geneva: ACC/SCN.
5. ———. 2004. Fifth Report on the World Nutrition Situation. Geneva: ACC/SCN.
6. Administrative Staff College of India. 1997. "National Strategy to Reduce Child Malnutrition; Investment Plan; Final Report." Asian Development Bank, Manila, December.
7. Allen, L. H., and S. R. Gillespie. 2001. "What Works? A Review of the Efficacy and Effectiveness of Nutrition Interventions." ACC/SCN Nutrition Policy Paper 19; Asian Development Bank (ADB) Nutrition and Development Series 5, United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, Geneva; ADB, Manila.
8. Berg, A. 1981. Malnourished People: A Policy View. Poverty and Basic Needs Series. Washington, DC: World Bank.
9. ———. 1987. Malnutrition: What Can Be Done? Washington, DC: World Bank; Baltimore: Johns Hopkins University Press.
10. BINP and UNICEF (Bangladesh Integrated Nutrition Program and United Nations Children's Fund). 1999. "Preliminary Results." UNICEF.
11. BRAC (Bangladesh Rural Advancement Committee). 2004. "BRAC Health Programme," http://www​.brac.net/health_programme.pdf, and "BRAC at a Glance," http://www​.brac.net/aboutb.htm.
12. Chowdhury, M. 2003. "Health Workforce for TB Control by DOTS: The BRAC Case." Joint Learning Initiative Working Paper 5-2. http://www​.globalhealthtrust​.org/doc/JLI%20WG%20Paper%205-2​.pdf.
13. Cohen D, Spear S., Scribner R., Kissinger P., Mason K., Wildgen J. `Broken Windows' and the Risk of Gonorrhea. American Journal of Public Health. 2000;90(2):230–36. [PMC free article: PMC1446134] [PubMed: 10667184]
14. Delgado H. L., Valverde V. E., Martorell R., Klein R. E. Relationship of Maternal and Infant Nutrition to Infant Growth. Early Human Development. 1982;6(3):273–86. [PubMed: 6813101]
15. Engle P. L., Bentley M., Pelto G. The Role of Care in Nutrition Programmes: Current Research and a Research Agenda. Proceedings of the Nutrition Society. 2000;59(1):25–35. [PubMed: 10828171]
16. Ezzati M., Lopez A., Rodgers A., Hoorn S. vander, Murray C. J. L. the Comparative Risk Collaborating Group. Selected Major Risk Factors and Global and Regional Burden of Disease. Lancet. 2002;360:1347–60. [PubMed: 12423980]
17. Ezzati M., Hoorn S. Vander, Rodgers A., Lopez A., Mathers C. D., Murray C. J. L. the Comparative Risk Collaborating Group. Estimates of Global and Regional Potential Health Gains from Reducing Multiple Major Risk Factors. Lancet. 2003;362:271–80. [PubMed: 12892956]
18. Farmer, P. 2003. Pathologies of Power: Health, Human Rights, and the New War on the Poor. Berkeley and Los Angeles: University of California Press.
19. Fiedler, J. L. 2003. "A Cost Analysis of the Honduras Community-Based Integrated Child Care Program (Atención Integral a la Niñez-Comunitaria, AIN-C)." Health, Nutrition, and Population Discussion Paper, World Bank, Washington, DC.
20. FNRI (Food and Nutrition Research Institute). 2004. "Results of FNRI National Nutrition Surveys, 2001 and 2003." Department of Science and Technology, Food and Nutrition Research Institute, Government of the Philippines, Manila.
21. Gillespie, S., and J. Mason. 1991. "Nutrition-Relevant Actions—Some Experiences from the Eighties and Lessons for the Nineties." ACC/SCN State-of-the-Art Series Nutrition Policy Discussion Paper 10, United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, Geneva. http://www​.unsystem.org​/scn/archives/npp10/index.htm.
22. Gillespie, S., J. Mason, and R. Martorell. 1996. "How Nutrition Improves." ACC/SCN Nutrition Policy Discussion Paper 15, United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, Geneva. http://www​.unsystem.org​/scn/archives/npp15/index.htm.
23. Gillespie, S., M. McLachlan, and R. Shrimpton. 2003. Combatting Malnutrition: Time to Act. Washington, DC: World Bank.
24. Guillermo-Tuazon, A., and R. M. Briones. 1997. "Comprehensive Assessment of Nutrition Interventions." Report prepared for Asian Development Bank–UNICEF Regional Technical Assistance project. University of Philippines at Los Baños, Regional Training Programme on Food and Nutrition. UNICEF, Manila. Photocopy.
25. Gwatkin, D., J. Wilcox, and J. Wray. 1980. Can Health and Nutrition Interventions Make a Difference? Monograph 13, Overseas Development Council, Washington, DC.
26. Habicht, J.-P., J. Mason, and H. Tabatabai. 1984. "Basic Concepts for the Design of Evaluation during Programme Implementation." In Methods for the Evaluation of the Impact of Food and Nutrition Programmes, ed. D. Sahn, R. Lockwood, and N. Scrimshaw, 1–25. Tokyo: United Nations University.
27. Habicht J.-P., Victora C. G., Vaughan J. P. Evaluation Designs for Adequacy, Plausibility, and Probability of Public Health Programme Performance and Impact. International Journal of Epidemiology. 1999;28(1):10–18. [PubMed: 10195658]
28. Heaver, R. A., and J. M. Hunt. 1995. Improving Early Child Development: An Integrated Program for the Philippines. Washington, DC: World Bank and Asian Development Bank.
29. Heaver, R., and J. B. Mason. 2000. Making a National Impact on Malnutrition in the Philippines: You Can't Get There from Here—A Case Study of Government Policies and Programs, and the Role of UNICEF and the World Bank. New York: UNICEF.
30. Horwitz A. Comparative Public Health: Costa Rica, Cuba, and Chile. Food and Nutrition Bulletin. 1987;9(3):19–29.
31. INFS (Institute of Nutrition and Food Science) and Department of Economics, University of Dhaka. 1998. "Strategies for Bangladesh." Dhaka Urban Community Health Program. Asian Development Bank, Manila.
32. Jennings, J., S. Gillespie, J. Mason, M. Lotfi, and T. Scialfa. 1991."Managing Successful Nutrition Programmes." ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper 8, United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, Geneva. http://www​.unsystem.org​/scn/archives/npp08/index.htm.
33. Jonsson, U. 1997. "Success Factors in Community-Based Nutrition-Oriented Programmes and Projects." In Malnutrition in South Asia: A Regional Profile, ed. S Gillespie, 161–89. ROSA Publication 5. Kathmandu: UNICEF, Regional Office for South Asia.
34. Kachondam, Y., P. Winichagoon, and K. Tontisirin. 1992. "Nutrition and Health in Thailand: Trends and Actions." ACC/SCN case study, Institute of Nutrition at Mahidol University, Bangkok, and United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition, Geneva.
35. Kavishe, F. P., and S. S. Mushi. 1993. Nutrition-Relevant Actions in Tanzania. ACC/SCN case study. Tanzania Food and Nutrition Centre, Monograph Series 1. Geneva: United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition. http://www​.unsystem.org​/scn/archives/tanzania/index.htm.
36. Kelley L. M., Black R. E. Research to Support Household and Community IMCI. Report of a Meeting, 22–24 January 2001, Baltimore, Maryland, USA. Journal of Health, Population, and Nutrition. 2001;19(Suppl. 2):S111–54. [PubMed: 11503353]
37. Kielmann A. A., Taylor C. E., DeSweemer C., Uberoi I. S., Takulia H. S., Masih N., Vohra S. The Narangwal Experiment: II. Morbidity and Mortality Effects. Indian Journal of Medical Research. 1978;68(Suppl.):42–54. [PubMed: 751920]
38. Mason, J. B. 2000. "How Nutrition Improves, and What That Implies for Policy Decisions." Paper prepared for World Bank–UNICEF Nutrition Assessment, UNICEF, New York, and World Bank, Washington, DC. http://www​.tulane.edu​/~internut/publications​/WB_Bckgrd_Pprs/Narrative​/NarrativethreeMason.doc.
39. ———. 2003. "Philippines Case Study." In Combating Malnutrition: Time to Act, ed. G. Gillespie, M. McLachlan, and R. Shrimpton, 85–101. Washington, DC: World Bank.
40. Mason, J., A. Bailes, K. Mason, O. Yambi, U. Jonsson, C. Hudspeth, and others. Forthcoming. AIDS, Drought, and Child Malnutrition in Southern Africa. Public Health Nutrition. [PubMed: 16236184]
41. Mason J. B., Hunt J., Parker D., Jonsson U. Investing in Child Nutrition in Asia. Asian Development Review. 1999;17(1, 2):1–32.
42. ———.2001Improving Child Nutrition in Asia Food and Nutrition Bulletin 223 Suppl.5–80.
43. Mason, J. B., P. Musgrove, and J.-P. Habicht. 2003. "At Least One-Third of Poor Countries' Disease Burden Is Due to Malnutrition." Disease Control Priorities Project Working Paper 1, Fogarty International Center/National Institutes of Health, Washington, DC. http://www​.fic.nih.gov/dcpp/wps/wp1.pdf.
44. Mason, J., J. Rivera, and C. Helwig, eds. 2005. "Recent Trends in Malnutrition in Developing Regions: Vitamin A Deficiency, Anemia, Iodine Deficiency, and Child Underweight." Food and Nutrition Bulletin 26 (1) Special Issue: 1–110. [PubMed: 15810800]
45. Mata, L. 1991. "National Nutrition and Holistic Care Programme (NNHCP), Costa Rica." In Managing Successful Nutrition Programmes, ed. J. Jennings, S. Gillespie, J. Mason, M. Lotfi, and T. Scialfa, 77–81. ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion Paper 8. Geneva: United Nations Administrative Committee on Coordination/Sub-Committee on Nutrition.
46. Measham, A. R., and M. Chatterjee. 1999. Wasting Away: The Crisis of Malnutrition in India. Washington, DC: World Bank.
47. Mills A. Decentralization and Accountability in the Health Sector from an International Perspective: What Are the Choices? Public Administration and Development. 1994;14:281–92.
48. Muñoz, C., and N. S. Scrimshaw. 1995. The Nutrition and Health Transition of Democratic Costa Rica. Boston: International Nutrition Foundation for Developing Countries.
49. Pelletier, D. L. 1991. "The Uses and Limitations of Information in the Iringa Nutrition Program, Tanzania." Cornell Food and Nutrition Policy Program, Working Paper 5, Cornell University, Ithaca, NY.
50. Pelletier D. L., Frongillo E. A. Changes in Child Survival Are Strongly Associated with Changes in Malnutrition in Developing Countries. Journal of Nutrition. 2003;133(1):107–19. [PubMed: 12514277]
51. Pinstrup-Andersen, P., S. Burger, J.-P. Habicht, and K. Peterson. 1993. "Protein-Energy Malnutrition." In Disease Control Priorities in Developing Countries, ed. D. T. Jamison, W. H. Mosley, A. R. Measham, and J. L. Bobadilla, 391–420. New York: Oxford University Press.
52. Plowman, B., J. I. Picado, M. Griffiths, K. Van Roekel, and V. Vivas de Alvarado. 2002. BASICS II Evaluation of the AIN Program in Honduras. Arlington, VA: Basic Support for Institutionalizing Child Survival Project (BASICS II) for the U.S. Agency for International Development.
53. Reddy, V., M. Shekar, P. Rao, and S. Gillespie, S. 1992. "Nutrition in India." ACC/SCN case study. National Institute of Nutrition, Hyderabad, India. http://www​.unsystem.org​/scn/archives/india/index.htm.
54. Rohde, J. 1993. "Indonesia's Posyandus: Accomplishments and Future Challenges." In Reaching Health for All, ed. J. Rohde, M. Chatterjee, and D. Morley, 135–57. Oxford, U.K.: Oxford University Press.
55. Saenz, L. 1995. "Evolution of an Epidemiological Profile." In The Nutrition and Health Transition of Democratic Costa Rica, ed. C. Muñoz and N. S. Scrimshaw, 119–43. Boston: International Nutrition Foundation for Developing Countries.
56. Sampson R. J., Raudenbusch S. E. W., Earls F. Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy. Science. 1997;277:918–24. [PubMed: 9252316]
57. Sanders, D. 1985. With R. Carver. The Struggle for Health. London: Macmillan.
58. ———. 1993. "The Potential and Limits of Health Sector Reform in Zimbabwe." In Reaching Health for All, ed. J. Rohde, M. Chatterjee, and D. Morley, 239–66. Oxford, U.K.: Oxford University Press.
59. ———.1999Success Factors in Community-Based Nutrition Programmes Food and Nutrition Bulletin 20(3):307–14.
60. Sanders, D., and M. Chopra. 2004. "Child Health." Paper prepared for Save the Children U.K., London.
61. Save the Children U.K. 2003. Thin on the Ground: Questioning the Evidence behind World Bank Funded Community Nutrition Projects in Bangladesh, Ethiopia, and Uganda. London: Save the Children U.K. http://www​.savethechildren​.org.uk/temp/scuk​/cache/cmsattach/666_ThinOnTheGround​.pdf.
62. Shekar, M. 1989. "The Tamil Nadu Integrated Nutrition Programme: A Review of the Project with Special Emphasis on the Monitoring and Information System." Paper prepared for the Rockefeller Foundation and the Food and Nutrition Policy Programme, Cornell University, Ithaca, NY.
63. Shrimpton, R. 1989. "Community Participation in Food and Nutrition Programmes: An Analysis of Recent Governmental Experiences." Paper prepared for the Food and Nutrition Policy Programme, Cornell University, Ithaca, NY.
64. Soekirman, I. Tarwotjo, I. Jus'at, G. Sumodiningrat, and F. Jalal. 1992. Economic Growth, Equity and Nutritional Improvement in Indonesia. ACC/SCN case study. http://www​.unsystem.org​/scn/archives/indonesia/index.htm.
65. Tagwireyi, J., and T. Greiner. 1994. Nutrition in Zimbabwe. Washington, DC: World Bank.
66. Taylor C. E., Kielmann A. A., Parker R. L. The Narangwal Nutrition Study: A Summary Review. American Journal of Clinical Nutrition. 1978;31:1040–52. [PubMed: 102180]
67. Tontisirin, K. 1996.
68. Tontisirin K., Winichagoon P. Community-Based Programs: Success Factors for Public Nutrition Derived from Thailand's Experience. Food and Nutrition Bulletin. 1999;20(3):315–22.
69. UNICEF (United Nations Children's Fund). 1990. "Strategy for Improved Nutrition of Children and Women in Developing Countries." Policy Review Paper E/ICEF/1990/1.6, UNICEF, New York.
70. ———. 1998. State of the World's Children: Progress against Worms for Pennies. New York: UNICEF.
71. ———. 2004. Drought, AIDS, and Child Malnutrition in Southern Africa: Preliminary Analysis of Nutritional Data on the Humanitarian Crisis. Nairobi: UNICEF, Eastern and Southern Africa Regional Office.
72. Vargas, W. 1995. "Development and Characteristics of Health and Nutrition Services for Urban and Rural Communities of Costa Rica." In The Nutrition and Health Transition of Democratic Costa Rica, ed. C. Muñoz and N. S. Scrimshaw, 68–117. Boston: International Nutrition Foundation for Developing Countries.
73. Waterlow, J. C. 1992. Protein Energy Malnutrition. London: Edward Arnold.
74. Werner, D., and D. Sanders. 1997. Questioning the Solution: The Politics of Primary Health Care and Child Survival. Palo Alto, CA: HealthWrights.
75. Whitehead R. G., Rowland M. G., Cole T. J. Infection, Nutrition, and Growth in a Rural African Environment. Proceedings of the Nutrition Society. 1976;35(3):369–75. [PubMed: 829285]
76. Winichagoon, P., Y. Kachondam, G. Attig, and K. Tontisirin. 1992. Integrating Food and Nutrition into Development: Thailand's Experiences and Future Visions. UNICEF/EAPRO and Institute of Nutrition at Mahidol University, Bangkok, Thailand.
77. WHO (World Health Organization). 1997. Improving Child Health. IMCI: The Integrated Approach. WHO/CHD/97.12 Rev 2. Geneva: WHO.

Footnotes

1

Social exclusion refers to the exclusion of groups from the mainstream of public actions: lower castes in India, poorer groups in Pakistan, many indigenous ethnic groups throughout Asia and the Americas, and migrant workers in China and elsewhere; the result for public health is that excluded people do not participate in programs even if they are available.

2

Pinstrup-Andersen and others (1993) provide a more complete list.