Health Policy and Planning Advance Access originally published online on May 3, 2006
Health Policy and Planning 2006 21(4):257-264; doi:10.1093/heapol/czl010
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The cost-effectiveness of a child nutrition education programme in Peru
1Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA and 2Instituto de Investigación Nutricional, Lima, Peru
Correspondence: Hugh R Waters, MS, PhD, Department of International Health, Johns Hopkins Bloomberg School of Public Health, 615 North Wolfe Street, Baltimore, MD 21205, USA. E-mail: hwaters{at}jhsph.edu
 |
Abstract |
|---|
 Top Abstract IntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
This article reports impact and cost results from a health facility-based nutrition education programme targeting children less than 2 years of age in Trujillo, Peru. Key elements of the programme included participative complementary feeding demonstrations, growth monitoring sessions and an accreditation process. Data were collected from six intervention and six control health facilities to measure utilization and costs associated with the intervention. To calculate the unit costs of services, these costs are allocated using activity-based costing. To measure the effects of the intervention, 338 children were followed through household surveys at regular intervals from birth until the age of 18 months.
The intervention had a clear positive impact both on the use of nutrition-related services and on children's growth outcomes. Children in the intervention areas made 17.6 visits to health facilities in the first 18 months of life, compared with 14.1 visits for children in the control areas (P<0.001). This pattern holds true for all socioeconomic groups. The intervention prevented 11.1 cases of stunting per 100 children. In multivariate logistic regression analysis, children in the intervention were 0.33 times as likely to be stunted as the controls (P = 0.002). The marginal cost of the intervention – including external costs, training, health education materials and extra travel and equipment – is US$6.12 per child reached and US$55.16 per case of stunting prevented. The estimated marginal cost of the intervention per death averted is US$1952.
Key Words: nutrition education, behavioural change, cost-effectiveness, economic evaluation, stunting
|
Introduction |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Childhood malnutrition remains an important public health and development problem in low and middle-income countries.1 An estimated one-half to two-thirds of child mortality worldwide is associated with malnutrition (Murray and Lopez 1997
; Caballero 2002). A pooled analysis based on 10 studies estimates that 53% of child mortality is attributable to being underweight; ranging from 44.8% for deaths due to measles to 60.7% for deaths resulting from dehydration due to diarrhoea (Caulfield et al. 2004). The mortality risk worsens for children more severely malnourished. Children who are mildly underweight – with weight-for-age Z-scores (WAZ) between –1.0 and –2.0 – have twice the risk of death of children above –1.0. The relative risk increases to 5 to 8 times for those moderately underweight (WAZ between –2.0 and –3.0) and severely underweight (WAZ<–3.0) (Caulfield et al. 2004).
The Integrated Management of Childhood Illness (IMCI) strategy has emphasized improved breastfeeding and complementary feeding practices to promote child health and nutrition. Within IMCI, nutrition education has been identified as an important mechanism to achieve this goal (Winch et al. 2002). Nutrition education – also known as behavioural change communications (BCC) – seeks to improve caretaker behaviours by motivating appropriate feeding practices.
Research has shown that nutrition education can improve dietary intake and child growth. The evidence shows that programmes promoting infant complementary feeding can improve children's weight and reverse growth retardation (Schroeder et al. 1995; Caulfield et al. 1999). A programme evaluation in Bangladesh found that nutrition education increased dietary intake and improved children's weight gain (Brown et al. 1993). Another evaluation in rural areas of the Philippines showed that nutrition education for mothers improved infants’ diets (Guthrie et al. 1990).
Community-level interventions for nutrition education have also been shown to improve infant feeding practices, growth and health. Bhandari et al. (2003) found that an educational intervention to promote exclusive breastfeeding in India lowered the prevalence of child diarrhoea at both 3 months (odds ratio of 0.64; P = 0.03) and 6 months (odds ratio of 0.85; P = 0.04). In a non-randomized community-based study, Guldan et al. (2000) showed that a culturally appropriate nutrition education programme improved child growth in rural Sichuan in China, with average WAZ and length-for-age Z-scores (LAZ) of –1.17 and –1.32, respectively, for the education group at 1 year of age, compared with –1.93 and –1.96 in the control areas (P = 0.004 for WAZ and P = 0.022 for LAZ).
Nonetheless, very little is known about the costs of nutrition programmes. Existing estimates are not standardized in terms of the content of the intervention or the costing methodologies applied. There is a clear need for additional scientific evidence to provide a credible evidence base showing that these interventions are cost-effective. In particular, there is very limited documented evidence concerning the cost-effectiveness of nutrition education programmes for children.
|
Description of the intervention |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Our study team worked with the regional public health authority in Trujillo, Peru, to develop and implement an infant nutrition education programme. Trujillo is a coastal city of 600 000 people, approximately 400 km north of Lima. The intervention promoted key nutrition messages for young children, improved counselling and introduced participatory demonstrations of the preparation of complementary foods.2 Support for these activities included the development of education materials, including flipcharts and recipe fliers, and promotion of the use of growth monitoring cards. A system of accreditation – carried out by local health professionals and based on previously defined criteria, interviews with health workers and children's caregivers, and record reviews – provided motivation for institutional behaviour change. Existing heath centre staff received training, but no additional personnel were added.
The study design included four stages: preliminary data collection, selection and randomization of facilities, formative research and implementation of the intervention. Preliminary surveys were conducted to identify the intervention site and information on population characteristics, including the prevalence of stunting, feeding practices and perspectives on health service utilization. Within the city of Trujillo, health facilities were first stratified by the type of facility and placed in pairs, and then matched within each stratum by the socioeconomic characteristics of the catchment population. Facilities were randomly designated as intervention or control, prior to the onset of formative research. Pairs were dropped if the randomization resulted in a control centre being directly adjacent to an intervention centre. The intervention is described in detail elsewhere (Penny et al. 2005).
A representative sample of newborns in the catchment area of each health facility – 187 children in the intervention areas and 190 in the control areas – were followed longitudinally from birth to measure feeding patterns, dietary intake and growth. Of those enrolled, 171 and 167, respectively, completed the study through to the age of 18 months. The intervention and control groups are generally similar in terms of characteristics that might influence the outcomes of the study, although there are statistically significant differences in the economic status of the families and the proportion of children whose mothers completed secondary school (Table 1).
|
|
Data and methods |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Measuring child malnutrition
Population-based anthropometric measures of child malnutrition include stunting, wasting and underweight children. Stunting is defined by the child's length given his or her age, and wasting by weight for a given length. Underweight is measured by weight for a given age. To measure children's nutritional status, we use anthropometric standards to calculate children's Z-scores for length-for-age and weight-for-age. The Z-scores are calculated using the reference growth curves developed by the US National Center for Health Statistics and recommended by the World Health Organization for international use (World Health Organization Working Group 1986; Dibley et al. 1987). Z-scores are calculated as the difference between the anthropometric score (length-for-age and weight-for-age) and the standard score, divided by the standard deviation.
Mildly stunted or underweight children are those who are between one and two standard deviations below the WHO-NCHS international standards for the indicator in question – length-for-age and weight-for-age, respectively (–2.0<LAZ<–1.0; –2.0<WAZ<–1.0). Moderately stunted or underweight children are between two and three standard deviations below these standards; severely stunted or underweight children are those who fall more than three standard deviations below. Using the 2000 Peru Demographic and Health Survey, we estimated the nationwide prevalence of moderate and severe stunting together to be 16.5% for children living in urban areas in the coastal region of Peru.
Household survey data
For each child enrolled in the study, the household survey was carried out in nine separate visits. At each visit, the children's weight, length and age were recorded, in addition to data on feeding practices and dietary intake.3 The household survey at 18 months also includes data on the use of nutrition services, users’ out-of-pocket costs related to the use of these services, and the time spent in travelling to health centres and using the services. Cost data include transportation costs and travel time to different provider types, and consultation fees.
We used multiple regression analysis to assess the effects of independent risk factors on the likelihood of the utilization of health services and on the likelihood that a child is malnourished. Utilization is measured as total health facility visits. Logistic regression analysis is used to analyse the effects of independent risk factors on the probability that a child is malnourished, since malnutrition is defined here as a dichotomous outcome. Independent variables include the presence of the intervention, the economic background of the child's household, the child's mother's education level, the presence of the child's father in the household, and the number of children less than 5 years of age in the family. In addition, the regression analysis includes independent variables for the availability of a toilet and latrine, and the type of water supply. An interaction term measuring differential effects of maternal education in the intervention and non-intervention areas is also included in the regression analysis.
To measure household economic status, we used principal components analysis to combine a series of household variables, including type of toilet facilities, type of cooking file, type of floor, and the presence of a television, radio, sewing machine, blender and telephone. Principal components analysis groups together correlated variables to form a composite linear index capturing the underlying groupings. The index assigns values to each household based on their relative possessions and characteristics, creating a rank ordering of all households (Filmer and Pritchett 2001). Based on this index, we divided households into quintiles of socioeconomic status, with the first quintile incorporating the poorest 20% of households and the fifth quintile incorporating the wealthiest 20%.
Health facility data and activity-based costing (ABC)
Data were collected from the six intervention and six control health facilities to assess the resource use and costs associated with the intervention. In each facility, the study gathered information on the average number of child medical consultations, child malnutrition cases and food preparation demonstrations. Additionally, all 12 health centres were surveyed on a monthly basis over a 1-year period, from October 2000 to September 2001, to gather cost data. Monthly cost information was organized by cost category, including personnel, food, transportation, capital costs (buildings and vehicles), telephone, administration, electricity, vehicle fuel, taxes and petty cash.
To calculate the unit costs of services, these costs are allocated to the services provided by the health facilities using activity-based costing (ABC). The services that are costed in this manner include medical consultations (with separate estimates for the unit cost of consultations for children under 2 years old), treatment of child malnutrition cases, food preparation demonstrations (cost per participant), and household visits related to the nutrition education intervention.
ABC essentially defines the principal activities of the individuals who work within an organization, then traces costs, first, to these activities, and then from the activities to products and services. Human and financial resources within a department (production centre) are traced to activities, which are in turn traced to products and services. Allocation of personnel time among the activities becomes the principal means for assigning overhead and other indirect costs (Waters et al. 2001).
Following ABC procedures, we calculated total operating costs for the intervention and control health facilities over a 1-year period. While traditional costing procedures group indirect costs in one pool and allocate these costs to products based on relative production figures, ABC attributes support costs based on time allocation and, in some cases, direct consumption of production inputs (Chan 1993; Cokin 1996). We traced indirect costs using personnel time allocation among activities, based on a series of 42 interviews with health personnel at all of the 12 health facilities in the study. These interviews asked personnel to describe their time allocation among 15 key activities related to nutrition activities, as well as their time commitment to other activities.
|
Results |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Utilization of health services
Overall, during the 18-month period that each child was followed, intervention participants made more total visits to health facilities than control participants; 17.6 compared with 14.1 visits on average (P<0.001). These visits do include care-seeking for illnesses, as well as visits for immunizations, food preparation demonstrations and well-baby check-ups. Multivariate regression analysis confirms that there is a statistically significant difference in utilization after controlling for observable factors. The additional effect associated with the intervention on participants’ total visits to health facilities over the course of the intervention is 1.7 (P<0.05) (Table 2). The results of the main intervention are described in detail elsewhere (Penny et al. 2005).
|
As a subcategory within overall health facility visits, the number of well-baby check-ups was also higher in the intervention areas (9.7 visits per child over 18 months) than in control areas (8.2 visits per child) (P<0.001). Children in the intervention areas were also taken to more food preparation demonstration sessions than children in the control areas; 6.9 sessions over 18 months compared with 5.4 (P<0.001). Multivariate regression shows that this association is highly statistically significant even in the presence of other factors, with the intervention having an average association of 2.7 additional demonstration visits over the course of the intervention (P<0.001). As a result, caregivers in the intervention area were more likely to receive nutrition advice; 51.5% compared with 24.3% (P = 0.02).
Nutritional status
Children in the intervention areas gained more weight than those in control areas and were on average 295 grams heavier at 18 months of age (P = 0.014). From the first wave of data collection at birth to the final wave, at age 18 months, the average WAZ declined for children both in control and in intervention areas, but the decline was smaller in the intervention group than in the control group (0.08 standard deviations compared with 0.28 standard deviations). At the age of 18 months, intervention children had an average WAZ of –0.34 compared with –0.62 for control children. However, multivariate regression found no significant impact of the intervention on the concluding WAZ scores (Table 3).
|
The percentage of stunted children, defined as LAZ<–2.0, was lower in the intervention group at the conclusion of the intervention. At 18 months of age, 15.8% of the children in the control group were stunted, compared with 4.7% in the intervention group. In other words, the intervention prevented 11.1 cases of stunting per 100 children. In multivariate regression, the intervention likewise had a significant impact on the odds of stunting at the end of the intervention; children in the intervention were 0.33 times as likely to be stunted as the controls (P = 0.002).
The intervention has a non-significant positive association with children being underweight (WAZ<–2.0). This association is, however, an artifact of small numbers; at the conclusion of the intervention only eight children were underweight, of which five were in the intervention sites. The intervention shows a positive (also non-significant) association with the average improvement in WAZ scores. At 18 months, children in the intervention areas had gained 295 grams more, on average, than those in the control areas.
Participants' costs
Because nearly all study participants lived in peri-urban areas of Trujillo, travel costs were relatively modest. Average transportation cost to public health centres was US$0.11 for intervention participants, compared with US$0.16 for those in the control group.4 Consultation fees at public health centres were also lower on average within the intervention areas, at US$0.35 compared with US$0.48 for controls. These differences are not statistically significant. There are no charges for immunizations, well-baby check-ups and nutrition rehabilitation visits.
Programme costs
On average, the six intervention health centres had total monthly operating costs of US$673 for child nutrition activities, including treatment of nutrition cases, group demonstrations, well-baby visits, growth monitoring, household visits and extra time dedicated to the implementation of the intervention. The comparable number in the control centres was US$417 per centre per month (Table 4). In addition to these costs, there were costs associated with the intervention that were not incurred at the health facility level, including nutrition demonstration training and other training (which occurred in Lima), expenses for travel related to obtaining accreditation for the health facilities, and producing educational materials. These costs – unique to the intervention – totalled an additional US$77 per facility per month on average. Overall, over the 18 months corresponding to the implementation of the intervention, total invention costs were US$322 higher per facility per month; a total of US$35 859 for the six facilities.
|
We calculated the distribution of costs for nutrition activities in both the intervention and control areas. At the health facility level, the majority (78%) of additional expenditures related to the intervention were for salary, with other extra added expenses related to transportation for household visits (5%) and equipment, including weighing scales (10%). Other intervention activities – including training, time and travel related to accreditation and the production of education materials – were financed by the project budget and were thus external to the health facilities. Combining both facility-based and external expenditures, 60% of the extra costs attributable to the intervention are represented by the salaries of health personnel in the facilities, and an additional 23% by external contributions.
The costs of the intervention can be expressed in terms of costs per case of child stunting averted. A total of 2333 children under the age of 18 months live in the geographic areas served by the six health facilities participating in the intervention. Applying the rate of 11.1 cases of stunting (LAZ<–2.0) prevented per 100 children in this age range, we estimate that 259 cases of stunting were prevented by the intervention in this cohort. The full intervention cost is therefore US$15.37 per child reached and US$138.50 per case of moderate to severe stunting prevented.
|
Discussion – the economic feasibility of a nutrition education intervention |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
This nutrition education intervention had a clear positive impact on the use of nutrition-related services and on children's nutritional outcomes, as measured by stunting – length for age. During the 18 month period that each child was followed, children in the intervention areas made 17.6 visits to health facilities, compared with 14.1 visits for children in the control areas (P<0.001). These visits include preventive and curative visits, as well as monthly growth monitoring educational sessions. The improvement in utilization was distributed equitably; children in all economic quintiles in the intervention group experienced higher total visits, and especially well-baby visits, than their counterparts in the control group.
The intervention also had a strong impact on reducing child malnutrition levels, as defined by stunting. At the conclusion of the intervention, 4.7% of children in the intervention group were moderately stunted, compared with 15.8% in the control group. In multivariate logistic regression, the intervention had a very significant impact on the odds of stunting at the end of the intervention; children in the intervention were 0.33 times as likely to be stunted as the controls (P = 0.002). This effect holds at all socioeconomic levels. There was no significant impact of the intervention on levels of malnutrition as measured by weight for age (underweight) or weight for length (wasting), both of which were low in both the intervention and control groups.
The intervention did not require hiring additional staff in the health centres involved. The cost of the intervention is US$15.37 per child reached and US$138.50 per case of moderate to severe stunting prevented. In addition, the intervention had positive effects on cases of mild stunting (LAZ<–1.0) which are not factored into this cost-effectiveness ratio. Sixty percent of intervention costs are due to health workers’ time and salaries within the health centre. These are fixed costs; health workers are paid whether the intervention is carried out or not. The marginal cost of the intervention – including external costs, extra travel and equipment, and other miscellaneous expenses – is US$6.12 per child reached and US$55.16 per case of stunting prevented.
It is possible to further estimate the impact of the programme in terms of saving lives, and the resulting cost per death averted. Caulfield et al. (1999), following methods introduced by Pelletier et al. (1994), present estimates of the impact on infant mortality rates of improvements in the nutritional status of infants aged 6 to 12 months. At the end of our intervention, the intervention group had, on average, LAZ scores 0.41 standard deviations higher than the controls, and WAZ scores 0.28 standard deviations higher. Given the average LAZ and WAZ scores in the control group, applying the methodology of Caulfield et al. leads to the estimate that 15.2% of mortality in this age range was prevented by the intervention.
Using Peru's 2000 child mortality rate for ages 1–5 (4q1) of 14 deaths per 1000 children as an estimate of the average mortality rate for the 6–18 months age group yields an estimate of 3.15 deaths prevented per 1000 children in an 18-month period.5 Applied to our target population of 2333 infants, the resulting estimate is that 7.3 deaths were prevented by the intervention. The marginal cost of the intervention per death averted is therefore US$1952. Since the child mortality rate is considerably lower than Peru's infant mortality rate of 33.3 deaths per 1000 live births, this estimate of cost per death averted is conservatively high.
In addition to the intervention costs, families incurred minor financial costs for participation. Transportation costs to public health centres were US$0.11 for intervention participants. Multiplied by the additional 1.7 visits to a health facility associated with the intervention over its course, the total cost to families for transportation was US$0.19. These costs do not include some intangible items related to the intervention, such as the extra cost to the family of additional food.
Comparisons with the available literature on the costs of nutrition education programmes suggest that the programme costs recorded here are both relatively high and more comprehensive than those in other studies. Per child reached, the costs of the Trujillo intervention are lower than those recorded by a child malnutrition treatment and education programme in Bangladesh, where a similar package of services offered through the government and through NGOs cost US$24.30 and US$29.89 per child, respectively. The Bangladesh intervention included supplemental food packages, growth monitoring and dissemination of nutrition-related information (Brown et al. 1993).
Other estimates suggest lower costs per outcome for other types of child nutrition interventions. Horton et al. (1996) found that breastfeeding promotion programmes that eliminate formula feeding in nurseries and maternity wards can reduce deaths due to dehydration from diarrhoea for US$100–200 per death averted and reduce the burden of disease for US$2–4 per Disability Adjusted Life Year (DALY). A breastfeeding promotion programme in Ghana has been estimated to cost US$7.80 per DALY (Chee et al. 2002).
The costs of a nutrition education programme in Mali – including breastfeeding promotion and counselling and education on optimal child feeding, prevention of diarrhoeal disease and growth monitoring – were estimated to be US$11 per child reached and US$282 per death averted (Ross et al. 1987). Other estimates of the costs of nutrition education programmes vary from US$2 to US$10 per child, depending on the intensity of nutrition counselling (Fiedler 2003). The US Institute of Medicine (1998) estimates that nutrition education programmes in developing country settings cost US$238 per death averted.
Despite being higher than these estimates, our findings show that the Trujillo nutrition education programme is economically feasible for both health facilities and households, costing the health system US$15.37 per child reached and US$1215 per death averted, with an additional cost to households of US$0.19 per child reached. A separate issue is that of the sustainability of the intervention. Internationally, evidence suggests that behavioural change interventions for health care are often not sustained beyond initial investments by donors or governments (Graeff and Waters 1995). Although it is too early to know with certainty, the intervention described here has characteristics that encourage sustainability. The intervention was designed by the regional health authority in Peru. Implementation did not involve external personnel, whose role was limited to evaluation.
The intervention was effective in reducing child malnutrition and was associated with increased utilization of nutrition services. All socioeconomic groups benefited. Given the continued high prevalence of child malnutrition in low- and middle-income countries, and the scarce resources available for nutrition programmes, implementation of cost-effective child nutrition interventions is a clear priority.
|
Biographies |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Hugh R Waters, MS, PhD, is a Health Economist and Assistant Professor in the Department of International Health at the Johns Hopkins Bloomberg School of Public Health, MD, USA. His areas of expertise are: (1) health insurance and health financing reforms, (2) equity and access to health care, and (3) costing health care interventions.
Mary E Penny, MBChB, is a physician from the United Kingdom. After working in the UK National Health Service, she has worked on research on diarrhoea at the Instituto de Investigación Nutricional in Lima, Peru, where she is currently Director and Senior Investigator.
Hilary Creed-Kanashiro, MPhil, received a Senior Drummond Fellowship from University College London to conduct research at the Instituto de Investigación Nutricional in Lima in 1971, and she has worked there since. Her research has focused primarily on the diagnosis of the nutritional situation of infants and young children in underprivileged populations of Peru, and the development and implementation of nutrition intervention strategies to improve infant, young child and family feeding and nutrition.
Rebecca C Robert holds a MS in Nursing from the University of Minnesota and a PhD in International Public Health from Johns Hopkins University, USA. She lived and worked in Trujillo, Peru during the time of the trial. At present, she is adjunct faculty in the College of Nursing and Health Science, George Mason University, Fairfax, VA, USA.
Rocío Narro qualified in nursing at Cajamarca University, Peru in 1990. She has worked in health education projects and as a coordinator in nutrition projects with the Instituto de Investigación Nutricional since 1991, in Cajamarca, Trujillo and Iquitos.
Jeffrey Willis is a PhD candidate in the Department of International Health at the Johns Hopkins Bloomberg School of Public Health, MD, USA. Mr Willis holds an undergraduate degree in Economics from Amherst College, MA, USA. Mr Willis has work experience at the World Bank, the Harvard AIDS Institute and Health Economic Research, Inc.
Laura E Caulfield, PhD, is a nutritional epidemiologist and is Associate Professor and Director of the Program in Human Nutrition, Department of International Health at the Johns Hopkins Bloomberg School of Public Health. She has conducted research in maternal and child nutrition in the region of the Americas for more than 15 years.
Robert E Black, MD, MPH, is the Edgar Berman Professor and Chair of the Department of International Health and Director of the Institute for International Programs of the Johns Hopkins Bloomberg School of Public Health. He currently has active projects in Bangladesh, Ethiopia, India, Peru, Zanzibar and Uganda.
|
Acknowledgements |
|---|
This article was made possible through support provided by the Office of Health, Infectious Diseases and Nutrition, Global Health Bureau, US Agency for International Development, under the terms of Award HRN–A–00–96–90006–00, the Family Health and Child Survival Cooperative Agreement. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the US Agency for International Development.
|
Endnotes |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
1This article classifies countries by income level using the following categories from the 2003 World Development Report (World Bank 2005
): Low-income – US$745 per capita or less; Lower middle-income – US$746–2975; Upper middle-income – US$2976–9205; High-income – US$9206 or more. 
2Three key messages were developed (paraphrased in English): (1) Offer your child a thick purée to eat rather than soup; (2) Add a special ingredient to your child's meal: chicken liver, egg or fish; and (3) Use patience, love and good humour to help your child to learn to eat.
3The visits correspond to the child's age as follows: 1st visit before 1 month of age; 2nd visit at 3 months; 3rd visit at 4 months; 4th visit at 6 months; 5th visit at 8 months; 6th visit at 9 months; 7th visit at 12 months; 8th visit at 15 months; 9th visit at 18 months. All visits except for the 3rd and 5th included the collection of anthropometric data; the 9th visit also included collection of economic data for the household.
4Throughout this document, monetary values have been converted to 2001 US dollars. The exchange rate applied is that which was in effect at the midpoint of the health facility intervention, in April 2001: 3.54 Peruvian Soles to US$1.00.
5From the Peru 2000 Demographic and Health Survey, available at: [http://www.measuredhs.com/countries/country.cfm?ctry_id=33].
|
References |
|---|
|
TopAbstractIntroductionDescription of the interventionData and methodsResultsDiscussion - the economic...BiographiesEndnotesReferences |
|---|
Bhandari N, Bahl R, Mazumdar S, et al. 2003. Effect of community-based promotion of exclusive breastfeeding on diarrhoeal illness and growth: a cluster randomized controlled trial. The Lancet 361:1418–23.[CrossRef][ISI][Medline]
Brown LV, Rogers BL, Zeitlin MF, et al. 1993. Comparison of the costs of compliance with nutrition education messages to improve the diets of Bangladeshi breastfeeding mothers and weaning-age children. Ecology of Food and Nutrition 30:99–126.
Caballero B. 2002. Global patterns of child health: the role of nutrition. Annals of Nutrition and Metabolism 46:(Suppl 1), 3–7.
Caulfield LE, Huffman SL, Piwoz EG. 1999. Interventions to improve intake of complementary foods by infants 6–12 months of age in developing countries: impact on growth and on the prevalence of malnutrition and potential contribution to survival. Food and Nutrition Bulletin 20:183–200.
Caulfield LE, de Onis M, Blossner M, et al. 2004. Undernutrition as an underlying cause of child deaths associated with diarrhea, pneumonia, malaria, and measles. American Journal of Clinical Nutrition 80:193–8.
Chan YL. 1993. Improving hospital cost accounting with activity-based costing. Health Care Management Review 18:71–77.[Medline]
Chee G, Makinen M, Sakagawa B. 2002. Cost and effectiveness analysis of LINKAGES' breastfeeding interventions in Ghana Bethesda, MD Abt Associates Inc.
Cokin G. 1996. Activity-based cost management: making it work New York McGraw-Hill.
Dibley MJ, Goldsby JB, Staehling NW, et al. 1987. Development of normalized curves for the international growth reference: historical and technical considerations. American Journal of Clinical Nutrition 46:736–48.
Fiedler JL. 2003. A cost analysis of the Honduras Community-Based Integrated Child Care Program Washington, DC The World Bank.
Filmer D and Pritchett LH. 2001. Estimating wealth effects without expenditure data–or tears: an application to educational enrollments in states of India. Demography 38:115–32.[ISI][Medline]
Graeff J and Waters H. 1995. A strategic framework for setting priorities for research, analysis, and information dissemination for behavior change and maintenance for child survival in Africa Washington, DC USAID Bureau for Africa, Office of Sustainable Development.
Guldan GS, Fan HC, Ma X, et al. 2000. Culturally appropriate nutrition education improves infant feeding and growth in rural Sichuan, China. Journal of Nutrition 130:1204–211.
Guthrie HA, Guthrie GM, Fernandez T, et al. 1990. Nutritional intervention: a key to child survival. Asia Pacific Journal of Public Health 4:53–8.[Medline]
Horton S, Sanghvi T, Phillips M, et al. 1996. Breastfeeding promotion and priority setting in health. Health Policy and Planning 11:156–68.
Institute of Medicine. 1998. Committee on Micronutrient Deficiencies, Board on International Health, Food and Nutrition BoardPrevention of micronutrient deficiencies: tools for policymakers and public health workers In Howson CP, Kennedy ET, Horwitz A (Eds.). Washington, DC National Academy Press.
Murray CJ and Lopez AD. 1997. Global mortality, disability and the contribution of risk factors: Global Burden of Disease Study. The Lancet 349:1436–42.[CrossRef][ISI][Medline]
Pelletier DL, Frongillo EA, Schroeder DG, et al. 1994. A methodology for estimating the contribution of malnutrition to child mortality in developing countries. Journal of Nutrition 124:2106S–2122S.
Penny ME, Creed-Kanashiro MD, Robert R, et al. 2005. Effectiveness of an educational intervention delivered through the health services to improve young child nutrition: a cluster-randomised controlled trial. The Lancet 365:1863–72.[CrossRef][ISI][Medline]
Ross SM, Loening WE, Mbele BE. 1987. Breast-feeding support. South African Medical Journal 72:357–8.
Schroeder D, Martorell R, Rivera JA, et al. 1995. Age differences in the impact of nutritional supplementation on growth. Journal of Nutrition 125:(Suppl. 4), 1051S–1059S.
Waters H, Abdallah H, Santillán D. 2001. Application of Activity-Based Costing (ABC) for a Peruvian NGO Healthcare Provider. International Journal of Health Planning and Management 16:3–18.[Medline]
Winch PJ, Leban K, Casazza L, et al. 2002. An implementation framework for household and community integrated management of childhood illness. Health Policy and Planning 17:345–53.
World Bank. 2005. World Development Report 2005: A better investment climate for everyone Washington, DC The World Bank.
World Health Organization Working Group. 1986. Use and interpretation of anthropometric indicators of nutritional status. WHO Bulletin 64:929–41.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||