The report of the Central Bureau of Statistics, enclosed, presents the situation of the "youngest child of 6,000 selected households" as well as findings related to causal factors. The estimated malnutrition rates for children less than seven years in each of the provinces and cities surveyed is given in the enclosed table.
1. Nutrition Assessment Report, Central Bureau of Statistics
2. Three summary tables of estimated malnutrition rates
Pyongyang February 2003
Report on the DPRK Nutrition Assessment 2002
Central Bureau of Statistics, DPRK
November 20 Juche 91 (2002)
The Democratic People's Republic of Korea is pleased to have successfully completed the implementation of the Nutrition Assessment in collaboration with the United Nations Children's Fund(UNICEF) and the World Food Programme(WFP).
With representative of our government, the Central Bureau of Statistics (CBS) carried out the survey in collaboration with the Institute of Child Nutrition (ICN). UNICEF and WFP provided financial and logistical support and the Institute of Child Health, London and Thailand Health Foundation, Bangkok provided technical advice. CBS was directly responsible for preparation of questionnaire, identification of survey target households, enlistment and training of survey team members, organization and monitoring of field survey activities, review of questionnaires and data collection, data analysis and report write-up.
CBS extends its sincere gratitude to the National Co-ordinating committee for UNICEF, DPRK and all the statistics and medical staff concerned for the invaluable assistance provided in the realization of survey. We also express our thanks to UNICEF and WFP and their staff for the logistical and technical support provided during the survey.
Central Bureau of Statistics, DPRK
November 20 Juche 91 (2002)
The Nutrition Assessment (NA), carried out by CBS in collaboration with ICN with financial and logistical support of UNICEF and WFP, was successfully concluded on schedule in October, as planned and agreed in a joint protocol signed in July 2002. The NA aims to serve as a baseline reference to evaluate the future programmes and expenditure of the Government, UNICEF and WFP that are directed at improving the nutrition situation of women and children in DPRK. A second aim of the NA is to improve the capacity of CBS and ICN in carrying out such surveys. The objective of the NA is to assess the nutritional status of a representative sample of children from birth to six complete years of age and of their mothers in seven provinces and three cities.
The survey was carried out in 200 randomly selected Ri and Dong from seven province sand three cities of DPRK. Three different teams with members from both central and provincial level carried out the NA involving over one hundred professionals of CBS, ICN, WFP and UNICEF. Twenty teams of five persons from the central level carried out data collection. They were assisted by provincial level logistics teams that arranged for local preparations and support in the Ri and Dong, and by a provincial level data entry and quality control team.
The survey sample includes 6,000 randomly selected households located in 200 randomly selected Ri and Dong of seven provinces and three cities. The selected households were those that had children aged under seven years of age. In each such household selected, the youngest child was weighed and measured and information collected on household food. In those households that the child was under two years of age, the mother was also interviewed concerning maternal and child health care practices, infant feeding practices, and maternal feeding practices. The mothers of children under two years old children were also weighed and measured, and invited to have a haemoglobin examination.
The survey data was verified and analyzed by CBS with technical assistance from ICH and THF in the third week of November 2002. The data collected and the measurements made are of a high quality and reflects well the great efforts put into ensuring the standardization and data quality checking by the staff of CBS and ICN. Using standard computerised statistical analysis packages, the frequency, rate, cross tabulation and correlation were tabulated and calculated. Fleiss quadratic 95% confidence interval for cluster sampling proportion, relative risk estimated by rate ratio and its 95% confidence interval, and incidence density rates, and adjusted prevalence proportionate to population size were also calculated.
The results of the NA suggest that the nutritional situation of children in DPR of Korea has improved considerably since the first survey was carried out in 1998, and are supportive of the results of the second survey carried out in 2000. The prevalence of child underweight in the surveyed sample of 6000 children is 20.15%, and of stunting is 39.22%, and of wasting is 8.12%. The overall prevalence of severe wasting, defined as less than 3 z-score weight for age, is 2.7%.
The nutrition situation has certainly improved dramatically since 1998, and the results of 2002 assessment are consistent with the improvement observed by the 2000 survey. Although these three surveys are not strictly comparable, they are still all very large surveys, each covering great majority of the national population. Careful interpretation of the results suggests that real improvements have occurred and continue to do so. These improvements are much more for underweight than they are for stunting . The fall in underweight across three surveys from 60% to 28% to 20% shows a consistent trend.
However, the rates of child underweight and stunting are still high and require continued efforts on this part. There is evidence that maternal nutrition is from being optimal with a third of mothers who are measured less than 22.5 cm for MUAC. For future, we will do our best to improve maternal nutrition in order to solve the problem of children malnutrition.
Whilst the food situation surveyed in seven provinces and three cities appears to be good from a quantitative perspective, it must be improved qualitatively. The majority of households reported that they had some food in store, and the presence in the food store of the main starchy staple was universal. The mothers reported that they had all eaten the day before and the great majority had eaten three meals. The consumption of the main food energy sources was almost universal. The consumption of protein rich food like meat, fish, eggs and pulses, and of fruits was not universal and showed provincial variation.
The child feeding practices are good from many aspects. The overall rate of exclusive breastfeeding of infant aged less than 6 months in seven provinces and three cities of DPRK is quite good, with almost 70% of mothers of children aged less than six months giving only breastmilk. The diarrhoea rates is 19.1% and the feeding practices during illness could be improved. The coverage of Vitamin A supplementation to young children is very high at 98.6% and the coverage of post-partum Vitamin A supplementation is 33%.
The food and nutrition situation is not the same across seven provinces and three cities surveyed. The nutrition assessment did not investigate all possible causes of child growth and a more in depth analysis of the possible causes at each provincial level is certainly merited. Results from this survey already suggest however, that in those provinces and cities where foetal and infant growth failure is most common future efforts should consider looking at ways to improve the diet of the mothers before pregnancy, during pregnancy and during lactation.
The design of the protocol and the realization of the Nutrition Assessment in DPRK (2002) was the result of series of discussions and collaborations between the consultant team (CICH and THF), UNICEF and representatives of the National Coordinating Committee for UNICEF, DPRK, the Central Bureau of Statistics (CBS) and the Institute of Child Nutrition (ICN).
The results presented in this report are those of the surveyed children only i.e. the youngest child in each of 6,000 selected households.
The Government of the Democratic People's Republic of Korea has carried out the Multiple Indicator Cluster Survey in 1998 and 2000, and based on the result of these surveys the Government has continued to provide assistance to the population of DPRK in the best interests of their health and development.
With the welcome assistance of UNICEF and WFP the Government is committed to continue to try and improve the welfare of women and children. Because of these shared aims and in recognition of these common interests, a protocol was signed committing the various parties to jointly realize the Nutrition Assessment.
AIMS AND OBJECTIVE OF NUTRITION ASSESSMENT
The first aim of the Nutrition Assessment is to serve as a baseline reference to evaluate the future programmes and expenditure of the Government, UNICEF and WFP that are directed at improving the nutrition situation of women and children in DPRK. The second aim is to improve the capacity of CBS and ICN in carrying out such surveys.
The objective of the NA 2002 is to assess the nutritional status of a representative sample of children from birth to six complete years of age and of their mothers in seven provinces and three cities.
The survey for the Nutrition Assessment 2002 was conducted during the month of October 2002, between 7th and 25th of October. The allotted time for the survey was 10 working days per province plus travel time. Twenty teams worked in seven provinces and three cities; two teams per province and each team completed one PSU per day, with each team completing 10 PSUs.
The DPR of Korea consists of 9 provinces and 3 municipal cities. There are 206 counties in the country. Each county further divided into small administrative units called Ri in rural area and Dong in Urban area. Total population of the DPRK is about 23 million.
The sample of the NA 2002 survey represents the universe of 7 provinces and 3 municipal cities. All Dong or Ri of these provinces/cities served as the sampling frame to randomly select 20 of them, which formed the Primary Sampling Unit (PSU). This selection was performed according to the urban-rural proportion. The total of 20x10200 PSU was selected from seven provinces and three cities. The selection was used by program (RAMDOM-DONG-RI-V4-24-7-2002XIS) on Excel. If a Dong or a Ri was selected that fell into a county that was non-accessible then the next number became the selected number until an accessible Dong or Ri was chosen. In each PSU one nursery was randomly selected and two children from the list of those enrolled was be randomly selected and their families became the index household. These households together with 14 households with children under seven years of age closest to them formed the Secondary Sampling Unit (SSU). Thus 30 households were selected for each PSU, and a total number of 6,000 households to be surveyed in SSUs of seven provinces and three cities. In cases of 2-3 children in each house, the last child was chosen. A total number of 600 households were surveyed in each province.
Therefore a total number of 6,000 mothers and children aged under 7 were respectively surveyed in seven provinces and three cities.
The Different Teams
There were 4 different teams, namely central commander team, data collection team, data entry and quality control team and local logistic team. The central commander team was consisting of CBS/ICN responsible members. The members of the 20 data collection teams were provided from the central level in Pyongyang. Each team consists of one CBS staff, one ICN staff, one international officer from UNICEF or WFP, one national from UNICEF / WFP and a driver. The data entry and quality control teams and logistics teams were from CBS provincial level. The lists of collection teams, data entry and quality teams are shown in Appendix.
Training for the members was given in 5 phases. 1st training (TOT, training of trainers) was done by consultants to the staff of headquarters and the data checking staff from each province, from Sep. 26th to Sep. 28th. The training included theory and practise. The 2nd, 3rd and 4th training for the data collection teams were carried out by CBS and ICN trainers as three batches of about forty participants. The participants received the questionnaires in Korean and English, and guidelines of nutritional assessment in English were distributed to the international staff. TOT trainers gave these trainings in accordance with the same contents and procedures as that given in the TOT from Sep. 29th to Oct. 4th. The 5th training (Oct. 2nd - Oct. 3rd) was given to the provincial data entry and logistic teams. The former 4 trainings took place at Grand People's Study House and the last 5th training at CBS.
The questionnaire, the main instrument of survey, was composed of 4 parts; household questions, maternal and newborn health questions, child anthropometry and maternal anthropometry. Considering the requests of UNICEF/WFP, the questionnaire was based on the standard MICS questionnaire of UNICEF with parts not directly related to Nutrition removed. Additional questions related to food source and intake was included. Maternal and newborn health questionnaire was asked only to mothers with a child under 2, and consists of 6 modules, namely: maternal and newborn health; Vitamin A coverage, breastfeeding; care of illness; immunization; and maternal food frequency questions. Child anthropometry module subdivided into anthropometry (child under 7), anthropometry and haemoglobin (mother with children aged less than 2) modules. Maternal anthropometry module contained the MUAC and haemoglobin test.
Weight was measured to the nearest 0.1kg using the UNISCALE in accordance with manufacturer's instructions. All children aged less than two years of age weighed in arms of their mother. Height/ length were measured using a wooden board especially developed and adapted locally to measure to the nearest 1.0mm. All children under two years were measured lying down, and those over two standing up. Mid Upper Arm Circumference (MUAC) was measured to the nearest 1.0 mm on the mothers of children under two using a plastic insertion tape provided by UNICEF. The mothers of children under two were only measured MUAC. Among mothers of children under two who agreed to have a test, Haemoglobin was measured in the field using the Haemocue method.
Roles and responsibilities of data collection team member
The roles of the data collection team members were very complementary and equally important, such that good teamwork was essential for the success of the survey. The division of responsibilities across the members of the team was as follows; The CBS staff was the principle person responsible for all interviews. The ICN staff was the principal person responsible for the anthropometrical and clinical examinations. The national UN officer in close collaboration with the international UN officer was the main person responsible for doing the random selection of nurseries, and identifying the index child through house to house seeking, identifying and referring of mothers to the central location. Five officers from UNICEF and 15 officers from WFP took part in this survey. The survey was carried out with the interviews and clinical examinations by every collection team.
Before fieldwork began, the anthropometrists from ICN performed a standardization protocol, especially developed based on a model obtained from WHO Geneva that measured their precision and accuracy. Based on a first standardization test, the accuracy of seven anthropometrists was improved by correcting their measurement technique. The result of the second improved anthropometry standardization procedure on all 20 anthropometrists was available for inspection.
Organization of the fieldwork
The Survey teams were based in the provincial capital and where feasible travelled from there to each selected Ri/Dong and returned again at the end of the day. On arrival in the province the two data collection teams met with the provincial survey coordinator and his logistics and data entry team chiefs. The provincial logistics team prepared the lists of nurseries in each of the selected Ri and Dong available for the data collection teams in order to allow the random selection of two nurseries by them for each Ri and Dong. The logistics team advised the local committees where the chosen nurseries are located so that the families surrounding the nurseries could remain in their houses in the morning the data collection teams were scheduled to visit their area.
In each PSU the provincial logistics team had drawn rudimentary maps showing where all the houses of the children in the elected nurseries and prepared the local communities to be at home on the day of the visit by the data collection team. When the data collection team arrived in a PSU, they visited the nursery and randomly selected two children from the register. From the data collection team, the national UN officer in close collaboration with the international UN officer did the systematic search and selection of the 30 households.
The randomisation was carried out as follows: First two children were randomly selected from the register and noted down the names and address. Then from the houses of these children, through a process of "oiling" the 28 households around these two index children's houses where have children aged under seven. In case the mother with a child refused to participate in the survey, the team went to the next household until they found a mother agreed to participate. If the mother said "Yes" to participate, she was given a referral slip and a ribbon tied around her wrist and the wrist of the child, so that they could be identified as part of the selected population sample when they arrived in the central location for examination.
The process of identifying the 30 mothers and their children under seven was completed before mid-day, such that in the afternoon the national officer could also help the CBS person with interviews in the central location. At end of the day the whole team reviewed together the procedures of the day and checked the questionnaires. The team supervisor took the set of 30 questionnaires and handed over to the data entry team leader when they got back to the provincial headquarters.
Data Entry and data quality checking
The data compiled by each collection team was handed over to the provincial data entry/quality checking team on the same day of each fieldwork, and entered to the computer next day. In the province the error checking was done whilst still in the field to guarantee the consistency and the quality of nutritional assessment data. The data entry/checking program, prepared by consultants from Epi Info according to the contents of questionnaire was used to do the data entry. If an error was detected, it was immediately informed to the data collection team supervisor and corrected whilst still in the field by the data collection team.
Data analysis was performed by the staff of CBS together with the consultants from Nov.18th to Nov.25th 2002.
Two clusters in each province and two sample households in each cluster were randomly selected for data entry verification. Each data item from 40 sample households was thoroughly checked both in the data collection forms and in the data files. Of more than 2,000 data items, 6 data item entry mistakes (approximately less than 0.3%) were found, all in non-anthropometrical and haemoglobin data fields. Then, a single database for data analysis totalled 6,000 records was prepared by vertically (combine data from all clusters together) and then horizontally (combined all variables from different forms) merging 4 data files together.
Data processing and data analysis were done by using SPSSPC version 10.5 for Windows and Epi Info 6 version 6.04d. Frequency, rate, cross tabulation and correlation were tabulated and calculated initially by SPSSPC. Fleiss quadratic 95% confidence interval for cluster sampling proportion, relative risk estimated by rate ratio and its 95% confidence interval, and incidence density rate were carried out later from SPSSPC output files using Epi Info 6. Adjusted prevalence proportionate to population size was also done using Excel Spreadsheet.
(pdf* format - 1.51 MB)