Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Carl E. Taylor
Child growth is the most sensitive and readily measured indicator of health and nutrition in individual children and in population groups. Growth is potentially an even more useful general index of health in a community than mortality because it is dynamic and reflects positive change. Growth of children conveys a sense of hope as the quality of life of the future generation improves. It can be measured at any time, because children are continually present, while deaths need to be counted as they occur. Standard measures of national development have focused on economic indicators such as GNP, but by monitoring child growth greater attention can be focused on social development.
For this discussion the definition of surveillance is "watching over" a series of activities in order to identify problems early and to initiate prompt action. A great deal has been written about growth monitoring to improve the nutritional status of individual children. Nutritional surveillance at the national level has been developed by J.-P. Habicht, professor of nutrition, and his colleagues at Cornell University mainly as a means of focusing attention and information in order to provide early warning of impending food shortages and to guide policy, programming, and advocacy in improving nutrition. Growth-monitoring data provide a principal source of information for nutritional surveillance along with data on household food availability. Some of the best national nutritional surveillance programmes, such as that in Malawi, are now beginning to use features of the community surveillance approach defined in this paper to guide general nutrition and primary health care activities.
In this paper a streamlined community-based methodology is decribed that brings nutritional surveillance to the local level. The primary purposes are (a) to target services to the children in greatest need, (b) to adapt nutritional interventions and procedures to local conditions most cost-effectively, and (c) to provide a framework for community action.
A clear distinction needs to be made between growth monitoring of individual children and community-based growth surveillance because of intrinsic differences in objectives, methods, quality of data collection, population groups covered, and analytic procedures. The two approaches are complementary and require appropriate phasing. General implementation of growth monitoring should not be delayed while setting up a system of community surveillance. Complementarity is greatest when both are done in parallel.
Problems in growth monitoring as it is generally applied
Growth monitoring is the first of the GOBI interventions (growth monitoring, oral rehydration, breast feeding, and immunization), but it has been the hardest to implement. With considerable progress being made around the world in the expanded programme of immunization and some in oral rehydration, concentrated effort for growth promotion should now receive greater priority. A growth-promotion movement is needed that will involve mothers in routine weighing of children and, more importantly, teach them what to do when growth faltering occurs.
Many local projects and a few large-scale programmes have demonstrated benefits from growth monitoring when it leads to follow-up action. Most babies born into conditions of poverty and high mortality do not die from a single acute disease episode but from a descending spiral of synergistic infections and malnutrition. A potential method of breaking the chain of causation is to identify children as they start this insidious and silent downward progression by growth monitoring. Parents should be made aware that a child who is not growing is probably a sick child.
In most situations growth monitoring of individual children has not fulfilled expectations in improving the health and nutrition of children. It has been claimed that growth monitoring is simple and low cost, but this does not mean that implementation will be easy. In spite of considerable effort, in many mass programmes the coverage rates have not reached the levels needed for significant impact. The following generic problems need attention.
The first problem is that routine growth monitoring does not usually reach the most needy children. Setting up neighbourhood weighing days almost automatically tends to favour the elite because the places chosen are usually in the more affluent areas of a village. A trickle-down effect usually does not follow unless deliberate efforts are made to involve the families in greatest need. Periodic community weighings may become social events where mothers who have leisure time dress up and compare observations on how well their children are doing. This provides positive reinforcement for children who are doing well, but does not necessarily encourage a poor and overworked mother to bring her sick and malnourished child out for community attention. Even if growth monitoring reaches 90% coverage, the remaining 10% may include the most severely malnourished children.
A second problem is that faithful growth monitoring does not necessarily lead to appropriate action. The use of growth charts for individual children has the limitation that weighing, by itself, has no health benefit. It differs from most other child-survival-and-development interventions because the screening is supposed to trigger appropriate action rather than directly improving health and nutrition. The actual practice of growth monitoring seems to assume that, once growth faltering is identified, the mother will spontaneously improve feeding practices. Sometimes this works, especially when nutrition education and food supplements are provided. It is not unusual, however, for mothers to have the impression that there is magic in the weighing ritual. Growth monitoring may take so much effort and crying babies generate so much attention that there is little time left to work out the necessary interventions with the mother, who leaves with little understanding of what she is now supposed to do.
A third problem for national programmes of growth monitoring has been a tendency to apply set procedures rigidly and uniformly to varying local conditions. Efforts to develop an international model for growth monitoring will probably not work, because what seems good in one place will probably have to be modified in other situations. The places where growth monitoring has worked best have been where a pervasive community infrastructure supports family action and health services provide good local backstopping to educate and support the efforts of parents.
One reason pilot projects have tended to succeed while mass implementation has been less successful in most places is that local projects have taken the time to define objectives flexibly in accordance with local priorities and have used procedures and methods acceptable to the people. Instances where larger programmes have been successful have usually been those where, by intuition or by chance, appropriate local decisions were made. A local learning process was worked out through close communication to draw on the wisdom of mothers and village leaders.
A fourth problem, which applies mainly to the process of moving from simple growth monitoring to community or regional action, is the tendency to pay inadequate attention to defining the most important local causes of growth faltering. Epidemiological analysis is needed of the local mix of causal factors, which may include the incidence and prevalence of common infections synergistic with malnutrition, lack of food or of particular nutrients in the home because of inadequate production and distribution or poverty, limited knowledge of appropriate weaning and feeding practices, incorrect methods of food preparation, socio-cultural constraints and beliefs, high parity, discrimination against girl babies, lack of time and energy on the part of the individuals responsible for child care, inadequate fuel or facilities to prepare foods for children, specific nutrient deficiencies, etc. Growth faltering may vary according to the energy and nutrient content of local staples such as rice, wheat, or tubers and the way they are prepared. The bulk and digestibility of weaning foods directly influence growth, and most staples require supplements with concentrated sources of energy and protein.
Great geographical variation in the mix of causal factors has been found in several well-documented community studies. In areas of chronic food shortage such as Bangladesh there is extreme seasonal variation in mortality depending on food availability . Causal factors also include common infections such as diarrhoea and low birth weight. In Guatemala also both low birth weight and common infections were important .
In areas of food surplus such as the Punjab, where mothers are well nourished and breast-feeding is common, the impact of nutrition on mortality varies at different ages . The Narangwal field trials showed the special importance of iron and folic acid supplementation during the prenatal period. Infection control measures were the most important influence during the first year of life, when death rates are highest, and a synergistic combination of infections and malnutrition were equally important from one to three years of age.
Once the balance in the mix of local causal factors is understood, then standardized packages of interventions for that region can be worked out to be taught to health workers and mothers. The complexity of the choices that a mother or community volunteer faces in deciding what to do about growth faltering has not been adequately recognized. Professional expertise and local field data are needed to work out a locally relevant package of interventions, which can then be taught to mothers and volunteers in a simplified algorithm.
In some situations an international model of growth promotion is already well accepted; people are using charts kept at home; neighbourhood weighing days are sponsored by local women's organizations; and there is good understanding of the meaning of "road to health." Obviously in such situations the greatest effort should be devoted to social mobilization to extend coverage. The education of mothers should ensure that appropriate interventions are used when screening identifies growth faltering.
Most programmes, however, do not focus attention on the most needy families. Community surveillance should be designed specifically to identify those households where malnourished children are concentrated. Timely mechanisms for feedback should guide community responses. These clusters of families at greatest risk are where public funding and preventive services should concentrate their resources.
Where appropriate food for children is available and growth is generally good, it may still be desirable to monitor growth and development to focus on factors such as specific infections, anaemia and other micronutrient deficiencies, or psychomotor development. Under such circumstances there will be need for adaptation of forms and indicators to measure the factors that are important locally. Then support systems and eductional messages can give priority to these topics. It may be necessary to adapt feeding practices to improve the use of locally available foods. In choosing what to teach families about weaning foods, consideration should be given to the adult diet that the children will have to adapt to.
In some situations it may be decided that educating parents through charts kept at home is a less important objective than the use of growth charts by the health system as an efficient means of supervising health workers and volunteers. Where supervisors have trouble visiting homes, their contacts with local workers might be facilitated by keeping the growth charts in clinics. By reviewing these charts, the supervisors can arrange to visit homes where problems are evident rather than spot-checking homes randomly to review charts kept at home. The main educational messages to parents are transmitted in any case during clinic visits or community weighing sessions, and some studies show that mothers seldom look at charts in between. The very successful use in some countries of growth charts kept at home does not necessarily mean that this is the right method for all situations. We know little about the influence of local culture and literacy on the ability to understand the implicit messages of growth curves.
Where growth monitoring is done primarily to provide data for national nutritional surveillance, it is important to assure rapid reporting and good-quality measurements. Some means of aggregating information from the charts is needed. One possibility is to have health workers plot on a single community growth chart all the weights of children seen at each weighing session to get an easily read visual summary of growth status in the community. The quality of data may be improved by using specified sentinel sites for nutritional surveillance.
The rationale for community growth surveillance
Many international development programmes in the last two decades have tried to reach the "absolute" and "relative" poor using direct poverty-reduction methods such as land and income redistribution. Most have encountered the constraint that local leaders see such programmes as opportunities to improve their own status and secure benefits for their families and friends. Effective leaders are usually able to subvert any new flow of resources from outside to increase their own political and economic position. Efforts to establish controls on community leaders so as to ensure equitable distribution are usually approached cautiously because they seem to be contrary to the objective of promoting self-reliance and decentralization. A community surveillance network has the possibility of providing a direct means of focusing development resources on children in greatest need.
Community growth promotion is probably unique among poverty-reduction measures in its potential for gaining acceptance by most local leaders, because, once sick and malnourished children have been identified, it is very difficult for anyone to interfere with efforts to improve their care. Local leaders can usually see the political advantage of promoting activities showing that they care about the poor.
Perhaps the most attractive feature of parallel activities for individual growth monitoring and community-based surveillance is that this provides a potentially powerful means of ensuring equity in health and nutrition care. By using child growth as an indicator, much can be learned about the characteristics of children in greatest need and the cumulative impact of factors influencing the quality of life. Standardized measurement approaches can make data from different geographical and socio-economic groups comparable. The ages at which growth faltering occurs will provide insight into what interventions might be most effective at what times, with special attention to the weaning period and prenatal influences on birth weight.
Combined surveillance of growth and causes of mortality may be our best way of defining probable risk from the various causes of mortality. If we had better comparative data on causes of mortality, we would be able to quantify the relative impact of different health and nutrition interventions. This approach proved to be the basis of the dramatic impact on both growth and mortality achieved in the Narangwal studies  One benefit was that, once pockets with a high prevalence of malnutrition in villages were identified, it became difficult for village leaders to ignore the evidence.
Systematic charting by caste or socio-economic class has an effect on community leaders equivalent to making a mother aware of the silent progression of malnutrition in her own child. Such awareness of focused nutritional needs in a particular community group provides a triggering mechanism for a variety of local actions, such as the collection of supplemental food by the village at harvest time, the provision of day-care centres during busy seasons, education about feeding practices, controlling infection by improving water systems, and even improving employment opportunities. What is shown to work at the local level can then be tried at the regional or national level.
There is great need now to link activities at the national and the individual level by developing a methodology for community and district data gathering and action. As with primary health care, the next stage in making nutritional programmes work seems to require attention to the district as the most peripheral level at which all management activities can be co-ordinated in order to improve community action.
Practical differences between programmes for growth monitoring in individuals and community nutritional surveillance
A general problem has been that multiple competing objectives have frequently been lumped together without distinguishing between growth monitoring and nutritional surveillance . For any programme, realistic understanding of local priorities rather than hopeful rhetoric based on isolated successes in other places should guide the definition of a limited group of complementary objectives. The following lists of issues illustrate the differences between objectives for growth monitoring in individuals and for community nutrition surveillance. These issues should be considered when planning a local situation analysis for nutrition programmes.
Issues to be considered in bring to improve growth monitoring for individual children
Issues to be considered in community growth surveillance
This list of issues might seem so complex as to discourage any reasonable effort to organize simplified field activities. The issues will not go away, however, if they are simply ignored. In each situation special studies are needed to gain understanding of local causal patterns in order to select out the few items that should receive priority attention. A general approach can now be defined.
Organizational framework for community-based surveillance
A parallel process should be developed in which phased implementation of surveillance as part of general services develops concurrently with systematic field studies. A special research and evaluation unit will probably be needed to conduct field studies which are neither expensive academic studies nor short-term pilot projects. This research unit should do a situation analysis and conduct longitudinal studies to define local priorities, taking into account questions such as those raised above. These studies should solve problems as they emerge in the implementation of mass programmes and test potential solutions for rapid application.
Concern may be expressed about the additional cost of supporting a field unit to do special studies. Such expenditures will, however, be minimal when compared with the extremely large investments required for large-scale nutrition programmes, which are usually implemented without local adaptation. Nutrition interventions tend to be the most expensive of all primary health care activities .
One such team could serve a small nation with fairly homogeneous characteristics. For larger countries a network would be more appropriate, with a subunit in each province or region to make adaptations to local conditions. These units should probably be based in academic or research institutions. They should work with communities to learn how a self-generating process can be started to help local groups solve their own problems in a phased local learning process.
Access to such community involvement can be arranged most readily in selected field demonstration areas attached to each institution, where a general data base is available and where the details of a surveillance system appropriate to a region can be initially worked out. Systematic expansion in larger networks of relationships can make national goals of achieving coverage realistic at very low marginal cost, because what is found by a local unit can be applied through extension services to whole provinces or regions sharing the same characteristics. This extension process should receive as much attention as the original studies.
Guidelines for community nutritional surveillance
In China, starting in 1982, a UNICEF-supported model county project for maternal and child health (MCH) attempted to apply the principles outlined above. The general approach is defined in detail in two WHO publications [6, 7] and in progress reports from the UNICEF office in Beijing. The national MCH network included 30 counties in 17 provinces. It is now being extended to the 200 poorest counties in the country. Each local unit in the network included three components: the country, a school of medicine and public health, and the provincial MCH department. Together they conducted situation analyses, reorganized services in accordance with local priorities, and retrained staff. For specific problems subsidiary networks of six to ten counties were set up for field research co-ordinated by a core group in a leading national research centre. Three of these research networks were on nutritional problems: stunting, which is found in about 40% of Chinese children; anaemia, found in about a third of children; and rickets, found in about a third of children in northern provinces. The others dealt with infectious diseases and management problems.
The experience of developing the above network led to definition of the following seven steps, which can be implemented in any sequence or concurrently. Anyone developing a community-based nutritional surveillance system should at least think about each of these steps and whether they apply to the local situation.
1. Situation analysis
Community nutritional surveillance should start with a situation analysis based on available data supplemented by surveys. A simplified weight- and height-for-age cross-sectional survey should include all socio-economic groups in the region. If resources are limited, appropriate sampling should pay particular attention to including groups that might be at greatest risk. The age groups studied should be decided according to the expected distribution of nutritional problems. Where kwashiorkor is found, children up to five years old might be surveyed, but, where only marasmus is common, measuring children under three years old should usually be sufficient. Measurements should include information on birth weights. Where resources are limited, arm-circumference measurements can provide simple screening since precision is less important than representative sampling.
A carefully worded questionnaire should include simple data on feeding patterns, food availability, socioeconomic information, cultural constraints and preferences, and patterns of health care. A recent methodology that might help is the use of rapid assessment procedures for nutrition and primary health care .
All pertinent information should be brought together in a sociological map that identifies the pockets of greatest need according to defined family characteristics. This will provide a framework for selective targeting of continuing surveillance of families at greatest risk of growth faltering. Causal influences affecting the various local patterns of malnutrition should then be studied to provide a cross-sectional perspective on the relative priority of factors influencing growth. Surveys repeated annually or at seasons of high risk should provide regular contact with statistically valid samples of all children and up-date information on distribution patterns of malnutrition and of growth improvement.
2. Monitoring plan
Because nutrition tends to fall between ministries and agencies, some co-ordinating mechanism or committee will be needed. The first stage in planning, whether at the national, regional, or local level, should be to get agreement on clearly defined objectives and priorities. General objectives should be set nationally but specific objectives adapted locally. Availability of local resources, time, and competing priorities will determine how much effort can be allocated to growth monitoring and who should be involved. Nationally designed and tested growth charts should include other development indicators, depending on local priorities. For instance, in China high priority was given to psychomotor development indicators because of the great interest of parents in the progress of their one child. Initiative should be taken by the unit responsible for field research or any other group having good community outreach.
3. Social mobilization
Publicity about findings from the situation analysis should help create community awareness. The data can be presented so as to increase a sense of social responsibility for the care of sick and malnourished children. The families with available resources can be encouraged to take responsibility for monitoring their own children. Health workers and volunteers should jointly provide intensive monitoring for needy groups. National leaders should regularly receive information on growth surveillance, infant and preschool mortality rates, and immunization rates, as indicators of progress in child survival and development.
A pattern of work that may prove appropriate is to have neighbourhod weighing days organized and run by the community. In parallel, the high-risk groups identified in the cross-sectional survey should have systematic and intensive monitoring at home. The foci of special need should be periodically redefined by surveys repeated about once a year. Co-operation between health workers, community volunteers, and parents will ensure continuity.
4. Community-based definition of interventions
Field studies are needed to define the most cost-effective mix of interventions to be applied selectively according to local need. Review of alternative control measures should start with field epidemiology to understand causal factors influencing growth faltering at various ages. Field testing of interventions should produce standardized packages of preventive and early treatment procedures for use by parents and health workers. These studies should specify requirements for personnel, training, supervision, the logistics of supplies, transport, and communications. A great deal of focused training will be needed to reach parents, community volunteers, and personnel from health and other sectors. Specialists in paediatrics need especially to learn simple and relevant methods of solving local problems because they will probably be involved in training and supervising other workers.
5. The management framework as part of primary health care
The process of implementing community-based nutritional surveillance needs to be routinized within local organizational and support systems for primary health care. It should be integrated with other activities for long-term sustainability. An effective method is to identify entry points where various primary health care interventions can be linked to strengthen each other. Nutritional improvement requires special mechanisms for co-operation between multiple disciplines and services. Local mobilization for growth surveillance can help in promoting community organization for other activities.
6. Evaluation and incremental improvement
No growth-surveillance activity can be expected to work optimally as first designed. In each region the learning process to improve child health and nutrition should be based on evaluation and incremental change. The fact that the process starts with systematic measurement as part of the situation analysis should provide a quantitative base for evaluation. Periodic longitudinal measurements of growth status should record the progress of the children in greatest need. Since data would always be population-based, they can be used to guide decisions relating to cost-effectiveness and programme impact.
7. Extension process
An important part of research and demonstration activities in primary health care is to have a systematic process of extension to general services. In the model county project in China, provincial leaders were responsible for making sure that any lessons learned were promptly built into service routines in other counties. The model counties were used for training workshops for health workers from other counties to convince them that they could implement the new interventions.
A new approach to community-based nutritional surveillance has potential for improving programmes for growth promotion by focusing on the children at greatest risk and increasing the capacity for appropriate action. First, cross-sectional surveys can help to identify where malnutrition is distributed in the community so that high-risk groups can be targeted for intensive monitoring. Second, field studies can be conducted in parallel with general implementation to help define causal factors influencing local patterns of growth faltering and guide selection of an appropriate mix of interventions and methods to suit local conditions. This information can provide a better basis for training mothers, volunteers, and service personnel.
Cultural, ecological, and economic constraints need to be identified as part of stimulating self-reliant community action. Demonstration of locally relevant and simplified procedures by a field research unit in each region should be linked with systematic extension to all parts of that region. These field research units should themselves be linked in a mutually supportive national and international network. Feedback of information from community-based surveillance can assist policy and administrative decisions for programme correction. These methods may provide our most direct means of introducing and measuring "adjustment with a human face."
1. Chen LC et al. Epidemiology and causes of death among children in a rural area of Bangladesh. Int J Epidemiol 1980;9:25-33.
2. Mata L. The children of Santa Marķa Cauque: a prospective field study of health and growth. Cambridge, Mass, USA: MIT Press. 1978.
3. Kielmann A et al. Integration of health, population and nutrition: the Narangwal experiment. Vol. 1. World Bank Research Monographs. Baltimore, Md, USA: Johns Hopkins University Press, 1983.
4. Taylor CE et al. Integration of health, population and nutrition: the Narangwal experiment. Vol. 2. World Bank Research Monographs. Baltimore, Md, USA: Johns Hopkins University Press, 1983.
5. Growth monitoring as a primary health care activity. Workshop proceedings. Yogyakarta, Indonesia, 2024 August 1984. Jakarta: Foundation for Indonesian Welfare, 1985.
6. Taylor CE. The uses of health systems research. Public health paper no. 78. Geneva: WHO, 1984.
7. National health development networks. WHO offset publication no. 94. Geneva: WHO, 1986.
8. Scrimshaw SCM. Hurtado E. Rapid assessment procedures for nutrition and primary health care: anthropological approaches to improving programme effectiveness. Tokyo: United Nations University; Los Angeles: UCLA Latin American Center, 1987.
Contents - Previous - Next