In Chapter 2, the purposes of food consumption and anthropometric surveys are outlined. The types and objectives of surveys are discussed and it is shown how they can be used in different ways to assess the nutritional status of individuals and groups. The importance of making surveys appropriate to different needs and explaining the types of information that can be generated by them is then outlined. Finally, the complementary roles of data from anthropometric and food consumption surveys in the assessment of nutritional status is explained.
FOOD CONSUMPTION SURVEYS
The purposes of food consumption surveys are:
to identify, predict and detect chronic or acute food and nutritional problems;
to target population groups for both short-term relief efforts and longer-term policy and programme development;
to plan and implement food and nutritional programmes and interventions;
to monitor changes and evaluate the impact of interventions and programmes.
Box 2.1 outlines the usual information that can be obtained from food consumption surveys.
ANTHROPOMETRIC SURVEYS
The purposes of anthropometric surveys are:
to characterize overall nutritional status of population groups and variations by demographic and socioeconomic characteristics, e.g. age, sex, location, income;
to monitor trends in nutritional status;
to evaluate the impact of changing health and socioeconomic conditions;
to assess the impact of intervention programmes;
to increase awareness of nutritional problems and define appropriate policies and programmes for addressing them.
Whatever the approach used, it is important to determine what additional information will be needed to interpret the anthropometric data. Nutritional status is influenced by a number of factors, such as age, sex, socio-economic status, culture, household food security, dietary patterns, environmental sanitation etc. Many of these factors have been studied in relation to energy-protein malnutrition in the Caribbean and elsewhere. They may also be important determinants of overweight. Substantial evidence also suggests a relationship between behavioural determinants or lifestyle-related factors and the prevalence of overweight.
When combined with dietary assessment information, anthropometric data can help elucidate the nature of food and nutrition problems identified within a population. For instance, it can guide decisions on whether specific interventions are necessary, the types of intervention and which groups should be targeted. The data can also be used to construct critical indicators in food and nutrition monitoring and surveillance systems, which in turn can provide vital input into policy and programme planning and management.
Box 2.1 Information from food consumption surveys
|
The information generated from food consumption and anthropometric surveys can be used in several policy and programme areas - by policy/programme decision-makers, technical staff or researchers in government ministries, training and research institutions, mass media and other private sector organizations. For example, policy-makers in the Ministries of Finance, Planning and Agriculture can use the information:
to formulate national policies for food and nutrition security;
to mobilize and advocate for resources for targeted interventions to achieve food and nutrition security and poverty alleviation;
to evaluate domestic food production and food imports in relation to recommended food goals;
to evaluate fiscal/trade policies to enhance nutrition and well-being;
to strengthen intersectoral collaboration;
to evaluate local production and food distribution systems for increasing diversity of food supply.
Similarly, users in other government ministries, private-sector agencies and civil society will find specific uses for this information. Annex 1 (adapted from the Guidelines for National FIVIMS) summarizes the application and use of food consumption and anthropometric survey data and provides a detailed list of potential users and specific uses for information generated.
Food consumption and anthropometric surveys can be designed in different ways to facilitate differing requirements. For instance, they can:
cover a specific geographic area, i.e. national, regional, community;
collect consumption data, anthropometric data or both;
be cross-sectional i.e. gathering data over a specified period in a target group, or longitudinal when data are collected from the same target group periodically.
While anthropometric data are always collected from individuals, food consumption data can be collected at different levels - national, household or individual levels.
NATIONAL DATA
At the national level, data can be obtained from food balance sheets or food disappearance data. The data can be used to:
provide information on food and macronutrient availability;
examine trends in food supply;
formulate and evaluate health and agricultural policies for planning food production and distribution;
regulate food imports and exports;
establish priorities for food aid.
They cannot, however be used to identify vulnerable groups in the population or differences in consumption by socio-economic or demographic groups.
HOUSEHOLD DATA
Household level data on food availability can be used to:
make comparisons between different communities and socio-economic groups;
look at trends in diet in the population.
However, these data will not provide information on the distribution of foods within the household.
INDIVIDUAL DATA
Data on food and nutrient intake from individuals are collected using a variety of methods.1, 2 The main ones are:
24-hour recalls;
food records;
diet history;
food frequency questionnaires.
(These methods are discussed in more detail in Chapter 3.)
Box 2.2 TYPES OF INFORMATION USEFUL IN ANTHROPOMETRIC SURVEYS |
|
Undernutrition: · age sex; · mother's educational level; · household monthly income; · household size; · toilet facilities; · access to potable water; · dietary intake; · infectious disease burden; · infant and child-feeding patterns. |
Overnutrition (obesity): · age, sex, ethnicity/race; · educational level of respondent or caregiver; · socio-economic level; · dietary intake; · parity (females); · chronic disease status; · family history of overweight; · smoking habits; · physical activity level; · alcohol consumption; · chronic disease history (CVD, diabetes, etc.). |
It is important to have a clear definition of survey objectives in order to determine what anthropometric information should be collected. Decisions will have to be made as to which information will be most useful and can be most easily obtained. A list of factors known to be associated with or to modify the effect of the prevalence of undernutrition and overweight is shown in Box 2.2. An analysis of the anthropometric data by these variables will help in the identification of risk factors and the design of appropriate interventions.
The effective use of anthropometry will depend on:
sampling procedures;
choosing the most useful measurements for the target group;
selecting appropriate instruments;
using accurate techniques in taking the measurements;
selecting appropriate reference standards with which results can be compared;
interpreting the findings carefully.
In this section we look at how to plan a nutrition survey. We outline the steps needed to plan, implement, analyse and use the results of a food consumption/anthropometric survey in the Caribbean region. (These steps are summarized diagrammatically in Annex 2.) If these general points are addressed early on, it helps to clarify the full extent of undertaking a survey and enables appropriate planning.[1]
The Caribbean experience highlights the importance of involving planners, decision-makers and other end users of the data from an early stage and throughout implementation. In most countries where surveys have been conducted, there has been involvement from relevant ministers of government in the early stages of planning the national surveys. Even more involvement throughout the process may be required to increase the impact of survey findings on policy and programme development.
The steps necessary for planning a survey are to:
1. convene a multisectoral team;
2. assess resources;
3. ascertain the budget;
4. select and train high-quality supervisors and field workers.
These shall be examined in more detail below.
1. CONVENE A MULTISECTORAL TEAM
Step 1 is to convene a multisectoral team, which should be made up of both implementers and end users. This team will in the first instance:
define objectives, research questions and priority areas with users and funders in mind;
determine what information the potential users and funders need;
make sure that users and funders are committed to the project and feel a sense of ownership;
consider the social, economic and other consequences of inaction.
Within the team itself, it is useful to identify a smaller core group to undertake specific tasks that will also help to guide the planning of the survey. This may, for instance, include conducting background work (such as literature reviews) or contacting other Caribbean countries that have carried out surveys to find out their experiences and lessons learnt and possibly to utilize their expertise.
2. ASSESS RESOURCES
One important lesson learned in the Caribbean has been that many surveys have been hampered by insufficient resources - both human and financial. Therefore, a thorough assessment of the resources available to support the survey is vital and should be done at a very early stage of planning. The shortage of human resources is particularly critical in the Caribbean and has had an impact on how surveys have been conducted in the past.
Assessing resource availability will help determine what type of survey is realistic - for example, whether it is to be qualitative or quantitative (or both). It will also help to strike a balance between depth and breadth in survey design, i.e. whether a survey is national or regional, the level of detail that can be collected etc.
3. ASCERTAIN THE BUDGET
Development of an adequate budget to support the final survey design is an important ingredient in determining the final outcome of these surveys. Annex 4 sets out some recommendations concerning resource needs for surveys in the Caribbean.
4. SELECT AND TRAIN HIGH-QUALITY SUPERVISORS AND FIELD WORKERS
It is important to allow adequate time to train and observe selected personnel while in the field and to make changes where necessary. Failure to budget sufficient time and resources for recruitment and training was frequently cited as a problem in the Caribbean. A detailed training manual and an interview manual for data collectors and supervisors should be developed to guide the training (see Box 2.3). (Suggested criteria are set out in Annex 4.)
OTHER ISSUES TO CONSIDER
Other issues that must be considered by the planning team are:
the study design and research questions;
sampling strategy;
obtaining ethical approval;
designing and pre-testing survey questionnaires;
pilot-testing all aspects of the project from data collection to data analysis - this will point out any difficulties before the study begins;
selecting appropriate analysis software in relation to the study objectives and design;
development of a data analysis plan and a strategy for disseminating findings.
Box 2.3 RECOMMENDATIONS FOR RECRUITMENT AND TRAINING OF FIELD WORKERS AND SUPERVISORS Experiences from a number of Caribbean countries point to several issues relating to the recruitment and training of survey staff:
Based on these, it is recommended that:
|
Box 2.4 RECOMMENDATIONS TO IMPROVE DATA QUALITY AND RESPONSE RATES The main problems that were identified from many of the Caribbean countries in carrying out surveys were inadequate supervision in the field and high rates of refusal by subjects from high socio-economic status households. Recommendations to improve data quality and response rates include:
|
The data collection process must be planned carefully: a detailed time line should be developed and the data collection and data entry organized and scheduled before the survey begins (see Boxes 2.4-2.6). Other factors to be considered are:
avoiding collecting data when other studies are taking place, particularly the census or elections. Holiday seasons and other special occasions such as carnival or festivals should also be avoided;
the timing and time frame that will be used to collect data;
recruitment strategy for data entry personnel;
whether to use separate teams to collect anthropometry and dietary data;
keeping track of non-respondents and recording reasons for refusal, if given;
keeping interviewers motivated by providing a reasonable stipend and incentives if possible;
having regular contact with the interviewers so problems in the field can be addressed quickly;
inspecting all questionnaires for completion as soon as possible - payments should only be approved for completed questionnaires
Box 2.5 EXAMPLE OF HIGH RESPONSE RATE AND HIGH QUALITY OF DATA (GUYANA) A high response rate was achieved in Guyana's food consumption and anthropometric survey because of careful selection of interviewers; commitment of interviewers; dedication, initiative and management skills of supervisors and management team; an excellent tracking and storage system; and the use of random field checks. Data quality was high because of:
|
Box 2.6 IMPACT OF OTHER SURVEYS ON RESPONSE RATE (BARBADOS) Barbados carried out a food consumption and anthropometric survey shortly after the national census and there was a lower than expected response rate because many of those sampled thought the food consumption survey workers were from the census bureau. |
[1] Annex 3 provides a
checklist of important considerations relating to the planning and
implementation of surveys. |