WFP/IFAD Qinling Mountains Area Poverty Alleviation Project (China) Summary Report: Baseline Survey Nutrition Assessment December 12, 2000 Report Prepared by: Seán Kennedy IFAD Technical Advisory Division, Rome Han Zheng Wang Li WFP Vulnerability Analysis and Mapping Unit, Beijing
25
Embed
Summary Report: Baseline Survey Nutrition Assessment · Summary Report: Baseline Survey ... Health Services. The training consisted a limited amount of classroom-style discussion,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Wang Li WFP Vulnerability Analysis and Mapping Unit, Beijing
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
I. Introduction 1. WFP/IFAD Qinling Mountain Area Poverty Alleviation Project in Shaanxi and Hubei provinces of China is a six-year integrated rural development project, combining food aid and long-term loans. The goal of the project is to improve food security and the living standards of the rural population in that area. 2. The nutrition assessment of children under 5 years old is a key part of the baseline survey, which will facilitate the measurement of changes in food security and living standards. According to field investigations carried out by the Vulnerability Analysis and Mapping (VAM) Unit and a national nutrition expert in 1998 and 1999, malnutrition was prevalent among pre-school children1 due to limited dietary variety, limited nutrition and childcare related knowledge, and insufficient access to health facilities and technicians. However, a more precise assessment of the situation was considered imperative for the project monitoring and evaluation process. The indicators generated from this survey will serve as benchmarks and will also be included in project mid-term and final evaluations, providing an important measure of project impact.
II. Methodology 3. The sample size required for the survey was calculated to be 6002 children under the age of five years per Province3, providing a total sample size of 1,200 children for the combined project area. Sampling frames were prepared and 30 clusters (Administrative Villages) were randomly selected in each Province, with 20 children weighed and measured per cluster. 4. As most Administrative Villages consist of six or more Natural Villages, a further random selection was made on the spot to determine the starting point for the data collection.
1 During a previous VAM mission to the project area, Mid Upper Arm Circumference measurements (MUAC) of pre-school children were taken. It was found that the MUAC of approximately one-third of the children measured was below 12.5cm. 2 FAO, Conducting Small-Scale Nutrition Surveys: A Field Manual (Rome: 1990). Formula used: n = D x t2 x p(1-p) m2
D = design effect (standard value of 2 for nutrition survey) n = sample size t = confidence level at 95% (standard value of 1.96) p = estimated prevalence of chronic malnutrition in the project area (est. 25%) m = margin of error at %5 (standard value of .05) Calculation: n = 2 x 1.962 x .25(1 - .25) n = 576 .052 3 According to VAM township data analysis, project areas in the two provinces have different characteristics; hence separate sampling and assessment can provide more information and better benchmark for measuring project impact in the two provinces respectively.
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
Each of the Natural Village was assigned a number, the numbers were written on slips of paper and one of these was drawn at random. Most villages had fewer than 20 children aged less than five years. In those cases all eligible children were weighed and measured, and a second Natural Village was chosen from the paper slips. When there were more eligible children than required, households were selected according to the “random walk” method.4 5. Anthropometric equipment for the survey was purchased jointly by WFP/IFAD. This included 10 locally constructed measuring boards, and 10 Chinese-made digital scales accurate to 0.1 kilogram. The equipment performed well under field conditions. Following completion of the survey, the equipment will be kept by local Project Management Offices (PMOs) for future surveys in the project area or in other WFP/IFAD projects. 6. Approximately 45 persons were trained in survey methodology and anthropometric techniques. Participants included the Monitoring and Evaluation Officers from the PMOs of Shaanxi and Hubei Provinces, and staff from the County PMOs and Maternal and Child Health Services. The training consisted a limited amount of classroom-style discussion, with larger amount of practical activities such as role playing, field testing and coaching. 7. The training and data collection in Shaanxi Province was facilitated by a nutrition and public health specialist from the IFAD Technical Advisory Division in Rome. All survey work in Hubei Province was coordinated and carried out by trained WFP/IFAD project staff without any continued external assistance. 8. Data collection for the 60 villages was completed between 19 – 30 October. Data collectors worked in pairs of two, at least one of whom was a woman. The pairs were divided between the counties according to the number of villages to be surveyed. The fieldwork took slightly longer in Shaanxi than was expected, as rain and cold weather made the dirt roads hard to travel (some villages could only be reached by walking). Despite the difficulties involved, the teams successfully completed the data collection in a relatively short period of time. 9. Data entry and analysis were carried out at the VAM Unit in Beijing. Data entry was done in an Excel worksheet, which was then imported into Epi Info 6.04 for the analysis. Epi Info is a shareware product developed by the WHO and the Centers for Disease Control (CDC) for use in public health and nutrition surveys. The software can be downloaded from the CDC website at no charge.5
III. Results 10. There are three primary indicators of malnutrition (also called anthropometric indicators). Each is briefly described in the following:
• Chronic Malnutrition (sometimes called “stunting”) is a measure of height relative to age. It is generally the result of months or years of food insecurity and nutritional deprivation. Chronic malnutrition is considered one of the best indicators of the overall wellbeing of a community.6 It is a key indicator for the monitoring and evaluation systems of WFP/IFAD assisted projects.
4 UNICEF, A Practical Handbook for Multiple Indicator Surveys (New York: 1995). 5 See www.cdc.gov.epiinfo 6 WHO, Physical Status: The use and interpretation of anthropometry (Geneva, WHO Expert Committee on Physical Status: 1995).
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
• Acute Malnutrition (sometimes called “wasting”) is a measure of weight relative to height. Acute malnutrition tends to result from shorter-term events, such as a drought or natural disaster, or a resent illness.
• Underweight is a measure of weight relative to age. Underweight is most
commonly used to monitor the status of individual children, such as in UNICEF growth promotion programs.
11. The results reported below are based on Z-scores (standard deviation units). According to WHO standards, children who have fallen more than two standard deviation units below the reference population are classified as suffering from malnutrition. Table 1: Indicators of Malnutrition—Combined Project Area (Shaanxi and Hubei). Indicator
[Source: Qinling Mountains Baseline 2000] Note: more detailed data tables and statistics are located in appendix (pages 16-24). Key Finding: Approximately 35 % of children in the project area (Shaanxi and Hubei
Provinces) suffer from chronic malnutrition. Table 2: Indicators of Malnutrition—Shaanxi Province. Indicator
[Source: Qinling Mountains Baseline 2000] 12. There are small differences in the prevalence of malnutrition when the indicators are disaggregated by gender. In most instances, the levels of malnutrition were slightly higher for girls than boys. However, those differences are not considered statistically significant. Figure 1: Chronic Malnutrition (height-for-age).
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
05
101520253035404550
Shaanxi Hubei Combined
GirlsBoysBoth Sexes
[Source: Qinling Mountains Baseline 2000] 13. The levels of chronic malnutrition were significantly higher in Hubei than in Shaanxi. That difference was approximately 10 percentage points, for both girls and boys, indicating a worse nutrition status in the Hubei Province project area. Figure 2: Underweight (weight-for-age).
05
101520253035404550
Shaanxi Hubei Combined
GirlsBoysBoth Sexes
[Source: Qinling Mountains Baseline 2000] 14. Note that the indicators for both Provinces follow the typical pattern, with levels of chronic malnutrition being the highest, underweight being in the middle, and acute malnutrition being the lowest. Figure 3: Acute Malnutrition (weight-for-height).
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
05
101520253035404550
Shaanxi Hubei Combined
GirlsBoysBoth Sexes
[Source: Qinling Mountains Baseline 2000] 15. The most recent available data for China (1992 national survey) show the levels of malnutrition among children less than five in the project area to be significantly higher than national levels.7 It should be kept in mind that the national data are somewhat dated, include both urban and rural areas, and represent the country as a whole, making direct comparison somewhat problematic. Therefore, national data are presented only as a point of reference. Figure 4: Indicators of Malnutrition by Project Area and National Data.
05
101520253035404550
ChronicMalnutrition
Underweight AcuteMalnutrition
Project (2000)National (1992)
[Sources: Qinling Mountains Baseline 2000, WHO Global Database on Child Growth and Malnutrition]
IV. Discussion and Recommendations
7 WHO, Global Database on Child Growth and Malnutrition (WHO website: www.who.org).
Qinling Mountains Area Poverty Alleviation Project in Shaanxi and Hubei Provinces
16. Initially, there was resistance to the idea of using a random sample of villages for the survey. Many of the participants felt that it would be preferable to select villages that were more accessible, and that such a sample would be considered “representative.” This is not an acceptable method for several reasons. Most importantly, remote or marginal communities are the populations most IFAD/WFP projects are intended to reach. Isolated villages tend to have higher levels of malnutrition and excluding them from the survey would invalidate the results. Also, underestimating the true level of malnutrition would make it more difficult for a project to demonstrate the impact of its activities.
• Recommendation: The importance of the sampling methodology should be consistently emphasized in future survey preparation. Survey reports should briefly describe how the sample size was calculated, how the sampling frame was constructed and how the villages were selected.
17. Participants were concerned that parents would not consent to the weighing and measuring procedures. Due to unseasonably cold weather, there was considerable reluctance to undress the children before weighing them, which is necessary for valid weight measures. To solve this problem, the survey team used blankets (appropriately decorated with teddy bears to appeal to the children). During the fieldwork, each mother was first weighed while holding an empty blanket. The mother then undressed the child, wrapped the child in the blanket, and was weighed a second time holding the child in the blanket. The first weight was then subtracted from the second to obtain the weight of the undressed child.
• Recommendation: This procedure should be used in other surveys where appropriate. The survey team should not weigh children who are dressed.
18. Family members most often recalled the date of birth of the child according to the Chinese lunar calendar. For this reason, both the date of the interview and the date of birth were recorded in lunar calendar format. It was thought that there were software programs able to convert the dates to the Western calendar as batch files, although the survey team could not locate the appropriate software. As a result, all dates were converted by hand at the VAM office in Beijing, which was a tedious but necessary step in the analysis.
• Recommendation: In areas where family members can accurately recall dates according to the Western calendar, the survey form should be adapted to record dates in that format. Where lunar calendar dates are used, the conversion should be done at the PMO level, not by survey workers in the field.
19. Seasonal changes in food supply, income levels and disease patterns can significantly influence several of the indicators of malnutrition. Seasonality can be a confounding factor in follow-up surveys (mid-term review, final evaluation) unless it is taken into account. All follow-up surveys should take place during the same season as the baseline. For nutrition assessments it is not required that the survey take place during a specific season, but it is important that follow-up surveys match the season of the baseline.
• Recommendation: The mid-term review for the Qinling Mountains project should be scheduled for October of 2003. Other projects should take seasonality into account when scheduling surveys.
20. A suggested format for reporting on changes in key indicators in follow-up surveys (mid-term and final) is provided below: Table 4: Estimation of Impact