|Nutrition Guidelines (MSF, 1995, 191 p.)|
|Part II: Rapid Nutrition Surveys|
When dealing with populations in emergency situations (refugees or displaced people), planners, health officers and officers in charge need to evaluate the nutrition situation quickly and precisely. This evaluation is based on surveillance data, demographic indicators, direct observation, advice from experts and in some cases rapid anthropometric surveys. The quick anthropometric assessment survey (measuring the prevalence of malnutrition) is one of the many tools for evaluation of the nutrition situation, allowing the quantification of malnutrition in the population.
ANTHROPOMETRIC SURVEYS: WHY AND WHEN?
· In the initial phase of an emergency, to assess the situation and take short term relief action.
· In the course of a programme, to assess the evolution of nutritional status, to redirect nutritional programmes and to assess their coverage and impact. The survey may also help in assessing the impact of the programme: was the prevalence of malnutrition reduced? Was the target population covered?
· In the course of a programme, when deterioration or amelioration of nutritional status is suggested by various sources of information: clinics, food availability.
However, an anthropometric survey is not always a priority in emergency situations. A rapid screening of children presenting a high risk of mortality may be needed urgently in order to take immediate life saving action. An anthropometric survey only becomes a priority when there is an urgent need for intervention in order to take decisions to implement actions.
OBJECTIVES FOR SURVEYS
The objectives should include notions of time, place and persons.
The objectives of anthropometric surveys may be:
- to quantify malnutrition in a specified population, at a given time, using indicators of malnutrition;
- to identify higher risk groups. These may be a certain age group, newly arrived refugees, nomads as compared to sedentary people, etc...;
- to estimate the number of children who may benefit from a nutrition programme, e.g. how many children should be expected to need treatment in an intensive feeding care unit with an entry criteria based on a given nutrition index value? How many children could be eligible for supplementary rations?
- to assess trends in nutrition status based on repeated surveys;
- to evaluate a programme in comparison with a target objective;
- to compare the nutrition status of refugees and the local population.
Preparing to do a survey
MEET THE PEOPLE IN CHARGE
Most of the time a survey is carried out in response to the needs for information expressed by the people in charge. The objectives have to be precisely specified with them at this point. They need to understand the methodology used, its constraints, the means required and the limitations of the expected results. People in charge may provide a map of the camp or of the region, and a list or a register of the refugees if available.
GATHER AVAILABLE INFORMATION
Before starting the anthropometric survey, all locally available information should be gathered. Have there been any previous surveys? What were the findings and recommendations? Is there a reliable mortality surveillance system? What information related to food resources is available?...
Collecting demographic information prior to the survey is essential when this information is not readily available or reliable.
DEFINE DATA TO BE COLLECTED, THE PLAN OF ANALYSIS AND THE QUESTIONNAIRE
Data to be collected
During an anthropometric survey, the following basic data needs to be collected:
and possibly, according to the specific objectives of the survey:
- measles immunization status
- date of arrival in the camp
- availability of cooking equipment/fuel at the household level.
- availability of food
The additional questions that could be included should be limited in number. Data collectors and families should not be overwhelmed by the number of questions asked. Only variables with direct relation to nutritional status should be considered. A variable must be easy to measure, and its value should be precisely coded: has this child been correctly immunized against measles (Yes or No)? What is the month of arrival of this child in the camp? Only factors for which action can be taken should be included, such as the number of cooking pots per household.
Variables which are not precisely definable or easy to measure should not be included. For example, the presence of anaemia, because clinical assessment is not reliable and will vary from one data collector to another. Vitamin A or C deficiency assessment can only be included if all data collectors have been well trained in detection
Plan of analysis
Once the objectives have been clearly defined, a plan of analysis should be drawn up. The method of expression should be pre-determined for each indicator. The format for presentation of results can be pre-defined, and tables of results should be constructed without filling in the values. Doing this guarantees that appropriate measurements will be taken, appropriate questions asked and hence, that the objectives of the survey will be met.
The plan of analysis is composed of two parts:
· Descriptive analysis
In this first part, distributions according to the different variable data are constructed: how many boys and girls are there in the sample? Are all age groups represented? What is the proportion of oedema in the sample? This allows verification of sample representativeness and orientation of the analysis (see Annex 9).
· Interpretive analysis
Selected variable data is cross tabulated, in order to compare different groups of the sample according to certain characteristics: are younger children more malnourished than older children? Are the newly arrived refugees more malnourished than those currently in the camp?...
The number of cross tabulations should be kept to a minimum, concentrating on relevant information in order to take action for particular sub-groups of the population. (For an example of a plan of analysis, see Annex 9.)
Questionnaire and information collecting form:
The plan of analysis helps to realize what information must be collected. The questionnaire must be designed to clearly collect all the necessary information (see Annex 12). The questionnaire is used in conjunction with a "surveyor manual" which summarizes the questions as they should be asked and answers most of the questions that a data collector will come across during the course of the survey. Age criteria (6 - 59 month), as well as definitions used in coding the variable data should be included in the surveyor manual (e.g. a correct vaccination against measles is a vaccination reported by card and administered after 9 month of age).
The date of the survey will be chosen with the help of community leaders and administrators. Special dates, such as a local celebration or a food distribution day, should be avoided because most of the people will not be found at home on such a day. The schedule of the survey should allow time for preparation, training, community mobilization, collection of data, analysis and writing of the report.
MEET THE COMMUNITY LEADERS
The community should fully understand the objectives of the survey and be included in the planning of the survey - community members should also be encouraged to participate during data collection. Important points, such as why children should be visited in their households and not in the clinics or feeding centres, should be stressed.
A list of the required equipment should be drawn up, including transport, fuel, measuring equipment, paper and pens, per diem, etc... (Annex 22).
SELECTION AND TRAINING OF THE DATA COLLECTORS. PRE-SURVEY
Data collectors currently involved in delivering health services to the population can be selected if their participation in the survey does not disturb health care delivery to the population. Data collectors do not need to be health professionals, anyone from the community can be selected and trained as long as they are able to read and write. In some cultures, women might be needed in order to deal with young children.
A survey team is composed of 3 people: 2 measurers and a supervisor. Two to four teams may be needed according to the size of the area to be covered.
The training of the data collectors is the corner stone in the course of an anthropometric survey. Each and every data collector should undergo the same training whatever his former experience.
The training takes 2 to 3 days, according to the number of people to be trained and the availability of trainers.
- The objectives of the surveys are explained.
- The sampling method and its rationale are detailed, stressing the importance of a representative sample.
- Height and weight measures are demonstrated. At least 20 height and weight measures should be performed by each enumerator. A test of standardization of anthropometric measurement is used in order to assess the performance of each data collector (the procedure for the test can be found in Annex 11). This test evaluates the precision and exactitude of the data collectors and allows misunderstanding of measurement techniques to be detected prior to the survey.
An on-site visit should be a part of the training, to see that all team members understand the sampling procedure, select the right children and correctly measure and question the respondents. The questionnaire should also be assessed for ease of use and suitability. This may mean having to do additional training or alter the questionnaire after the trial. Data collected during this pre-survey should not be included in the actual survey results.
· Define the survey objectives.
· Collect available information.
· Meet the people in charge.
· Define a plan of analysis.
· Draw up the questionnaire.
· Inform the community.
· Draw up a schedule.
· Gather the necessary equipment.
· Select and train the data collectors.
Anthropometry is the measurement of the human body. Body parameters such as weight and height are used to assess nutritional status.
The various anthropometric indicators and the method of measurement
Many body parameters can be used to assess individual nutritional status. The weight, the height and the mid upper arm circumference are the most commonly used, but skin-fold thickness and various other measurements are sometimes used.
A 25 kg hanging spring scale, graduated by 0.100 kg, is used. The scale is hooked to a tree, a tripod or a stick held by two people.
The weighing pants are suspended from the lower hook of the scale, and the scale is readjusted to zero. The child's clothes are removed and the child is placed in the weighing pants. The pants then hang freely from the hook. In cold countries or in certain cultures it might be impossible to undress a child. The average weight of the clothes should be evaluated and deducted from the measure. When the child is steady, the weight is recorded to Adapted from How to weigh and measure children, UN, 1986 the nearest 100 grams - the scale should be read at eye-level.
Figure 6: Weight assessment
If the child is moving and the needle does not stabilize, the weight should be estimated by recording the value situated at the mid-point of the range of oscillations. The measurer announces the value read from the scale, the assistant repeats it for verification and records it on the questionnaire.
Every morning the scale should be checked against a known 10 kg weight. If the measure does not match the weight, the scale should be discarded or the springs must be changed.
Children aged more than 2 years old are measured standing up. Children less than 2 years old are measured lying down. If the age is difficult to assess, children of more than 85 centimetres are measured standing, those less than or equal to 85 centimetres, lying down.
· For children of more than 2 years, the measuring board is set up in a place where there is room for movement. The child's shoes are removed. The child is placed on the measuring board, standing upright in the middle of the board. The child's ankles and knees should be firmly pressed against the board by the assistant while the measurer positions the head and the cursor.
FIGURE height assesment
· The child's head, shoulders, buttocks, knees and heels should be touching the board. The measurer reads the measure to the nearest 0.1 centimetre. The assistant writes down the measurement and repeats it to the measurers to make sure it has been correctly heard and recorded.
· For children of less than 2 years old, the measuring board is placed on the ground. The child is gently placed, lying down the middle of the board. The assistant holds the sides of the child's head and positions the head until touching the foot board. The measurer places his hands on the child's ankles or knees. While positioning the child's legs, he positions the cursor up against the bottom of the child's feet, which should be at right angles. He reads the measure. The remaining procedures are the same as for standing children.
If birth dates have been recorded on a health card or immunization card, determination of age is simple. In such cases, the date of birth is directly recorded onto the questionnaire in order to avoid mistakes in calculating the age. If birth dates are not recorded, a local calendar of events is used. The mother is asked whether the child was born before or after certain major events until a fairly accurate age is pinpointed. If that is not possible, children are selected on the basis of height. Only children more than 65 centimetres and less than 110 centimetres tall should be included in the sample.
In order to determine the presence of oedema, normal thumb pressure is applied to the foot or the leg for three seconds (3 seconds is approximately the time necessary to say one thousand and one, one thousand and two, one thousand and three). If a shallow print or pit remains when the thumb is lifted, then the child has oedema. Nutritional oedema should be found on both feet or legs. Only children with oedema on both feet or legs are classified as having nutritional oedema.
Mid upper arm circumference (MUAC) (see Ref. 14, 15,18, 20)
Mid upper arm circumference is measured on the left arm, at the mid-point between the elbow and the shoulder.
The arm should be relaxed. A special measuring tape is placed around the arm. The measurement is read from the window of the tape without pinching the arm or leaving the tape loose. The mid upper arm circumference is recorded to the nearest 0.1 centimetre (see Annex 2).
Various indices and their meaning
None of these parameters, except mid upper arm circumference, give information about nutritional status when taken alone. They should be related to each other in order to define indices. The weight is related to age: weight/age index; the weight with the height: weight/height index; the height with the age: height/age index.
THE CONCEPT OF A REFERENCE POPULATION
The indices are compared to values for a reference population to see if they are worse than expected from the reference. For the same age, the height or the weight of a child from the sample is compared to the height or weight of the children of the reference population. For the same height, the weight of a child from the sample is compared to the weight of the children from the reference population. Reference tables have been drawn up for both sexes. For field use, sex combined tables have also been drawn up (see Annex 13).
These reference values for the various indices have been calculated from data collected by the National Centre for Health Statistics (NCHS) in the United States of America. This reference population, composed of young Americans, should not be considered as reflecting an <<ideal>> nutritional status, but should be used as a tool which allows comparison of data sets against a standard. It is then possible to compare the nutrition status from samples from two different countries, or the nutritional status of one population over a certain time period. Local reference curves exist in some countries. They can be used locally, but results should also be presented using international NCHS curves in order to allow international comparisons.
MODIFICATION OF THE WEIGHT AND THE HEIGHT
The weight of a child can change substantially in a short period of time. Hence, a child exposed to nutritional stress may lose up to 20% of his body weight within a few weeks. In contrast, height cannot change to the same degree. The height of a child cannot reduced, but the speed of growth may be slowed down. In the same way, a decrease in weight can be corrected rapidly if the nutritional situation improves, whilst the effected height can only be corrected in a small proportion of children. These are the reasons why each index has a different meaning.
MEANING OF THE INDICES
The weight / age index
The weight for age index expresses the weight of a child in relation to his age. However this index does not allow differentiation between two children of the same age and weight, one being tall and thin (wasted), the other shorter but not wasted. This index is mainly used during Maternal and Child Health clinic visits, since it is a good way of assessing the nutritional evolution of a child over time.
The height / age index
The height/age index expresses the height of a child in relation to his age. It reveals stunting at a given age, but does not allow discrimination between 2 children of the same age and height, one being thin (wasted) the other one being heavier. This index reflects the past nutritional history of a child rather than his current nutritional status. It is mainly used to identify chronic malnutrition.
The weight / height index
The weight/height index expresses the weight of a child in relation to his height. It reveals whether a child is thin or not but does not discriminate between 2 children of the same height and weight, one being older than the other, and possibly stunted. It is the index used to measure acute malnutrition called "wasting", meaning current or acute malnutrition at the time of the survey.
Mid upper arm circumference
The mid-arm circumference is almost stable from 6 to 59 month and hence does not need to be related to the age. But it is less reliable to measure and so it is only used for the rapid screening of populations to get an idea of the situation and for entry to nutrition programmes. We will not consider MUAC as a tool to assess nutritional status in this part.
In emergency situations where acute forms of malnutrition are the predominant pattern, the weight for height index (W/H) is the most appropriate index to quantify levels of current acute malnutrition in the population with an assessment of oedema. Furthermore, weight for height does not require the determination of age which is often difficult in these situations. (Ref. 15,18, 20).
Calculation and expression of the indices
Indices can be calculated by using reference tables or by using appropriate computer software.
NORMAL DISTRIBUTION CURVE
For a given height, one can draw the distribution curve of the children according to their weight. This bell shaped curve is called Gauss's curve or the normal distribution. It has some specific characteristics. The curve is symmetrical around the mean weight, the mean weight being the sum of all weights divided by the number of observations. The mean weight is equal to the median) weight, the median weight being the weight which splits the sample in two parts of equal size according to weight. This curve can be defined by its mean weight and its standard deviation.
The standard deviation is the square root of the sum of the squares of the differences between each weight and the mean weight, divided by the number of observations minus one.
Standard deviation = sqrt(sum (observed weight - mean weight)2)/(n-1)
In fact, it is not exactly the case for a distribution according to weight for a given height. The distribution is slightly asymmetrical, because weight variations are greater in the upper part of the distribution. This is the reason why we will deal with the median rather than the mean, since it is a better indication of the distribution for weight for height.
Expression in percentage of the median
This mode of expression requires knowing the median weight of the children of the reference population of the same length/height. The value of the median weight can be found in reference tables for each height by 0.5 cm. Calculation is simple: the observed weight is divided by the median weight and multiplied by 100 in order to be expressed as a percentage of the median.
Weight / height index =Observed weight/Median weight x 100
For example, for a child of 80.5 cm weighing 9.6 kg, reference tables give a median weight of 10.9 Kg. The weight/height index expressed in percentage of median is:
9.6 / 10.9 x 100 = 88.1%
Expression in percentiles
In the reference population, for a given height/length, the weight of children aged between 6 to 59 months is normally distributed. The 50th percentile is the weight which divides the distribution into two equal parts, 50% above, 50% below. It coincides with the median weight. In a similar way one can define the 10th percentile as being the weight under which 10% of the children of the reference population lie (90% being above). In the survey sample, for a given height/length, one can express the weight of a child according to its position in the reference distribution. The various weights corresponding to the various percentiles are shown by the reference tables.
For example, for a child weighing 9.6 kg and measuring 80.5 cm, the tables show weight values corresponding to the 5th percentile, the 3rd percentile as well as the deciles in the reference population. By reading the table, the weight of the child can be expressed as corresponding to the 5th percentile.
For the whole sample, one can determine the number and thus the proportion of children situated below a given percentile. When one says that in the sample 12% of the children were found to have an index below the 3rd percentile, it means that where 3% of the children from the reference population are found below this weight, 12% are found in the sample.
Expression in Z-Scores
The expression in Z-Scores uses the standard deviation of the reference distribution for a given height/length as a unit. The weight/height index expressed in Z-Scores represents the difference between the observed weight and the median weight of the reference population expressed in standard deviation units:
Weight/height index = (Observed weight - Median weight) / Standard deviation
Reference tables give the standard deviation and the median weight for each given height/length. This allows us to calculate, for each child in the sample, the value of his index expressed in Z-Scores. For example, for a child 80.5 cm and 9.6 kg, reference tables show a median weight of 10.9 Kg and a standard deviation of 0.870 kg2. Hence, his index expressed in Z-Scores is:
WHICH MODE OF EXPRESSION SHOULD BE USED?
(9.6 - 11.0) / 0.870 = -1.61 Z-Scores
Expressions in percentiles and Z-Scores have a true statistical meaning, which percentage of the median does not have. A child is more malnourished if the weight/height index is 80% at 6 months than at 59 months. The expression in percentiles does not allow the identification of severely malnourished children since percentiles corresponding to severely malnourished children do not exist in the reference population. Expression in Z-Scores is recommended. However, if people in charge of the refugees or people going to use the information are used to another mode of expression, this should also be used in order to deliver meaningful information.
CALCULATION OF THE INDICES
Calculations do not need to be carried out in the field when collecting the measurements. The main aim of the survey is not to locate malnourished children (screening) but to gather information on the whole population. The major preoccupation is not individuals but rather the condition of the population. If a child is found to be obviously malnourished during the course of the survey, he has to be referred to an intensive nutrition unit in order to seek treatment, but this is not the objective of the survey.
If computer equipment is available, one of the existing anthropometric software packages maybe used for calculation and analysis of the results (see Ref. 21). Data is directly entered into these software packages and nutrition indices are calculated by the programme, thus avoiding mistakes in reading the tables. Computerization of the nutrition indices is more accurate than manual calculation and takes into account the sex of the children.
If computer equipment is not locally available, reference tables can be used (see Annex 13). They are derived from the NCHS reference curves and are valid for both sexes. The height/length is rounded to the nearest 0.5 cm, as shown in the next table.
Weight values corresponding to the different cut off values are read from the table, enabling us to classify a child as belonging to an interval of the percentage of the median. For example a child 80.5 cm tall and weighing 8.6 kg will qualify for the interval of 75% to 80% of the median. Similar tables exist for Z-Score classification (see Annex 13).
We will only focus on the weight/height index in this part, since it is the most appropriate for assessing acute malnutrition and thus in meeting the objectives of a nutrition survey in emergency situations.
Principles of sampling
If all individuals in a given population were surveyed, we would get a precise picture of the nutritional status of this population. An exhaustive survey of this type would be long, costly and difficult to carry out. This is why measurements are only recorded for a sub-group of the population, called a sample, which "represents" the whole population. In fact, only children aged 6 to 59 months (65 -110 cm) are included in the target population, since it is this group which will best reflect the nutritional status of the population. Children in this age group are in a growing period, hence a modification in the availability of food will affect them first. It is from this sub-group that the sample is selected.
If the main objective of the survey is to compare two groups according to their nutritional status, two different surveys, one for each group, are required.
REPRESENTATIVENESS OF THE SAMPLE
The representativeness of a sample is essential. It is the prerequisite for extrapolation of results observed for the sample to the entire population. In order for a sample to be representative of the population, two criteria should be met: each individual should have an equal chance of being selected for the sample, and the selection of one individual should be independent of the selection of another individual.
Whenever a sample is drawn, a probability of error exists, meaning that there is a risk that the sample may not be truly representative of the population. In nutrition surveys, we accept an error risk of 5%. This means that we accept that in 5% of the surveys, results observed for the sample will not reflect the true nutritional status of the population. In other words, whenever an organization carries out 100 nutrition surveys, 5 of them will give a result not reflecting the true situation.
PRECISION, THE CONFIDENCE INTERVAL
By carrying out measures on a sample of the population, we only get an estimation of what the results would be if they were carried out on the entire population. If a second sample is drawn out of the same population, slightly different results may be obtained just because of the variation of the children selected for the samples.
The actual percentage of malnutrition in the entire population lies in a range around the observed value. The upper and lower limit of this range determines the confidence interval of the estimation. For example results will be expressed as follows: malnutrition rate = 13% + 5%, meaning the confidence interval ranges from 8% to 18%. The size of the confidence interval is related to the error risk and the size of the sample.
The sample size is related to three factors:
The expected precision: the greater the precision desired, the more people needed in the sample.
The probability of error chosen: the smaller the probability, the more people needed in the sample. If the whole population is surveyed, the probability is zero. In nutrition surveys, an error risk of 5% is accepted.
The expected prevalence: the nearer the expected proportion of children presenting malnutrition is to 50%, the greater the size of the sample required, for the same absolute precision.
Furthermore, a fourth factor should be taken into consideration:
The available means: the ideal objective in determining the sample size is to have the highest precision for the smallest error risk. The limiting factor is the available means. How many children can reasonably be surveyed in a day? How many data collectors are available?...
In conclusion, measuring malnutrition in a sample gives values affected by a known and accepted margin of error. On the other hand, sampling reduces the workload and allows surveys to be carried out in a short period of time.
Calculation of the sample size
When calculating the size of the sample the three factors previously defined should be taken into consideration. The formula used is the following *:
n = t2 * (p * q) / d2
n = sample size
t = parameter related to the error risk, equals 1.96 or 2 for an error risk of 5%
p = expected prevalence of malnutrition in the population, expressed as a fraction of 1
q = 1 - p, expected proportion of children not presenting malnutrition, expressed as a fraction of 1.
d = absolute precision, expressed as a fraction of 1.
<<t>> is fixed 1.96 (or 2) in this type of survey (corresponding to an error risk of 5% 3).
<<p>> and thus <<q>> (q = 1-p) are estimated from previous surveys. The expected prevalence is always chosen to be closer to 0.5 (50%) than truly expected order to get a bigger sample size. If we have a larger sample size than needed, we are sure of getting at least the desired precision even if the measured prevalence is larger than expected. A short survey of 30 households can give an idea of the expected prevalence if no information is available prior lo the survey
<<d>> is a parameter that can be modified (<<t>> is constant, <<p>> is estimated The factors which are considered in determining <<d>> are: the objectives of the survey the expected prevalence and the available means.
If the main objective of the survey is to demonstrate a moderate difference in the nutritional status between two groups, or over a certain period of time, the precision will have to be high (and therefore, <<d>> very small).
Usually, in nutrition surveys, the expected prevalence ranges from 5% to 20%.
The precision should be proportional to the expected prevalence. For example, 10% precise for an expected prevalence of malnutrition of 10% will give a confidence interval from 0% to 20%. No conclusion can be reached from such results. Refer to the next table in order to see how the precision affects the sample size for digit levels of expected malnutrition.
For example, in a survey where the expected malnutrition rate is 15% (12% from a previous survey), and with a desired precision of 3%, the sample size is:
n = 1.962 * 0.15 * 0.85 / 0.032 = 544
The size of the target population does not usually effect the required size of the sample. This is true when the size of the population is much larger than the size of the sample. However, if the sample size approaches the size of the population, a correction factor can be applied to the formula. It reduces the required sample size needed to get the chosen precision. This correction factor is used whenever the sample size is more than one tenth of the total population. The revised sample size is given by the following formula:
Revised n = n/(1 + (n/N))
In our example, if the total population of children aged 6 to 59 months was 5000, the revised sample size would be:
Revised n = 544 / (1 + 544/5000) = 490
Three main sampling methods are available: random sampling, systematic sampling d cluster sampling.
Random sampling is the best method, when it can be used, since it is the only one meeting the two criteria for representativeness as previously defined. A sampling base should be available which lists every individual in the population and allows you to locate them. The list must be kept up to date with regard to the ages and location of each individual and include all new births. Individuals are randomly drawn from the list using a random number table (see Annex 10). Most of the time, such a list is not available or reliable.
Systematic sampling is a method in which the geographical organization of thee area to be surveyed is used. Every household should have the same chance of being surveyed by a team going across the whole area. Then one household out of X is visited. This technique can often been used in well organized refugee camps where houses are arranged in blocks and lines. In the same manner, if houses are enumerated, it is possible to survey one household out of X number, going across the camp from one extremity to the other.
TWO STAGE CLUSTER SAMPLING
This method is used when the two previous ones are not possible: no register is available and the geographical organization of the area does not permit a visit to all houses. The population is grouped in smaller units for which the population sizes can be estimated. The smallest unit for which the population can be estimated should be chosen as the sampling base. These units maybe villages, city blocks or sections of a camp. Thirty dusters' am randomly drawn (first level of sampling), in each cluster a certain number of children will be selected and surveyed (second level of sampling). The chance for each unit to be selected is proportional to its population size.
This sampling technique does not meet the second criteria for representativeness. The fact that several children are selected within a cluster by proximity means that the choice of a child is not independent from the choice of other children. Within each cluster children will have a tendency to be more similar, as far as nutritional status is concerned. This phenomenon is called the <<Design effect>>. The design effect is taken to account when calculating the sample size by multiplying the result obtained through the formula by 2. It means that when cluster sampling is used, the survey should use a sample size twice as large as for the other two sampling methods to reach the same level of precision.
WHICH METHOD TO CHOOSE?
Whenever a reliable register is available, random sampling is preferred. When populations are living in small, well defined geographical areas, systematic sampling should be chosen. In other instances, a two stage cluster sampling strategy should be applied.
Realization of the sampling
Random sampling implies the existence of a sampling base, such as a register. The steps are as follows:
· calculation of the sample size:
The following information is required:
- expected prevalence of malnutrition: for example p = 0.15
- error risk: 5%, meaning t = 1.96
- precision wanted: p = 0.03 (3%)
The sample size is:
n = 1.962 * 0.15 * 0.85 / 0.032 = 544
· a serial number is given to each child. For example, in a population of 12,481 children, a serial number between 00001 and 12,481 is attributed to each child.
· draw numbers from the list using a random number table
until the required
number of children is selected. For example, the table may generate the following random numbers: 00002, 00006, 00013, 00017, 00023,...,11,872,...
Children corresponding to these numbers are included in the sample.
Systematic sampling is used in relatively small geographical areas. The draw is based on a register of families or on the spatial arrangement of households. The organization of the site should allow one to comprehensively cover all houses. This technique is particularly adapted to well organized refugee camps. The steps for systematic sampling are as follows:
· Determine the number of inhabitants and the number of households: For example let's consider a camp of 50,000 refugees and 11,000 households.
· Determine the number of children between 6 and 59 months of age
The proportion of children between 6 and 59 months is quite stable, usually around 20%. However in certain situations, when a high infanto-juvenile mortality is suspected, this proportion can be smaller. The proportion of children has to be estimated from a rapid survey covering about 30 households selected at random.
In our example we have an estimate of 10,000 children (20% of 50,000).
· Calculation of the sample size
The same calculation as for random sampling is used: n = 544.
· Determine the required number of households
The first step is to calculate the average number of children by household. It is equal to the total number of children divided by the number of households: 10,000/11,000 = 0.9. Therefore, 604 households (544/0.9) will have to be visited in order to complete the sample.
· Determine the sampling interval
This is calculated by dividing the number of households by the number of households required in the sample. In our example: 11,000 / 604 = 18.2. One household every 18 households will be visited and all children (between 6 and 59 months) found in these households are included in the sample.
- Determine the first household to visit
The first household is randomly selected in the first interval, 01 to 18, using a random number, 05 for example.
· Selection of the households
One household is then selected, starting with the fifth one, then the twenty-third (18 + 5), the forty-first, etc.....
If two eligible children are found in a household, both are included in the sample. If no children are found in one household, the closest household (or as found using the sampling interval) is visited. If a child is not present at the time of the visit, the data collectors will have to come back to this very household in order to measure the child.
It is important not to overestimate the proportion of children aged 6 to 59 month when calculating the sampling interval. If this were the case, the sampling interval would be too large and the sample would not reach the desired size.
The steps for 2 stage cluster sampling are as follows:
· Determine the geographical units and their population
Cluster sampling requires the grouping of the population in smaller geographical units The smallest available geographical unit is always chosen as long as its population can be estimated. For each of these units the population of children 6 to 59 months is estimated.
These units can be villages, sections of the camp, or naturally defined geographical areas (river, road,...). In the rest of our example we will refer to these geographical units as sections.
· Calculation of the sample size
The calculation of the sample size uses the same formula as for random or systematic sampling. However, the size of the calculated sample should be doubled to take into account the design effect. A minimum of 30 clusters is always required. In each of these clusters, the number of children to be selected is the sample size divided by the number of clusters. For example, in a survey where the expected prevalence of malnutrition is 20%, the required precision 4%, 2 times 384 = 768 children are required. Hence, in each cluster, 768/30 = 26 children will be included.
· Calculation of the cumulative population
A list of the sections is established, as well as their respective population. In a third column, the cumulative total is calculated by adding the population of each unit to the sum of the population of the preceding sections. In other words, it is as if each section was given a certain amount of points, proportional to its population size.
· Calculation of the sampling interval
The sampling interval, in cluster sampling, is the total population divided by the number of clusters, usually 30. The thirty clusters are selected using the sampling interval. In our example, the sampling interval is: 10,000 / 30 = 333.
· Determination of the location of the first cluster
The location of the first section to appear in the sample is randomly selected within the first sampling interval. If the drawing of the first cluster was done from the beginning of the list, the first section would always appear in the sample, which would not give each section the same chance of being selected. A random number is used (Annex 10), in our example, between 001 and 333. Let's say the random number drawn is 256.
· Selection of the clusters
The sampling interval is added to this random number and the first cluster is selected in the section which includes this number. In our case, section No 1 includes 256, and so is the first section to be included, followed by section No 2: 589 (333+256), section No 4: 922 (589 + 333), section No 4 again: 1,255, etc...
A large section may appear twice - two clusters should be drawn in section No 4. In the same way, a small section (smaller than the sampling interval) may not be selected - section No 3 in our example.
· Selection of children in the clusters
Having identified the thirty clusters, a team of data collectors goes to the centre of the selected section. A random direction is picked by spinning a bottle. The bottleneck indicates the direction. A surveyor goes in that direction, from the centre to the border of the section, while counting the number of households he encounters. The first household to be visited is randomly selected from among these households by drawing a random number. A sketch of these households can be used.
The subsequent households are chosen by proximity. The next nearest household available is selected until the required number of children has been measured.
All eligible children are included and thus should be measured and weighed. If a child is not present when the team passes, he has to be found, or the team must come back later to measure this child. If a child has been admitted to an intensive feeding centre, the team must go to the centre and measure him there.
The analysis is composed of two parts:
· A descriptive analysis, which consists of building distributions according to the variables.
· An interpretive analysis where cross tabulations are used to make comparisons between groups.
The analysis uses the weight/height index values. There are two approaches when analysing and presenting results:
- the first approach estimates the proportion of children with W/H index falling below a cut off value.
- the second approach describes the whole distribution of children according to index values (= standard prevalence).
These two approaches are complementary. If the survey's objectives are to quantify the number of children who may benefit from an intensive feeding programme or from supplementary rations based on a cut-off value of the index, the first approach is the most appropriate.
However, if the objective is to assess the overall impact of a programme on the whole population of children, the second approach is preferred.
In this Part, we will only develop the first approach.
Description of the sample
The first step in the analysis is to describe the sample, by describing the distribution of characteristic variables. This will indicate if the sample is made up from eligible children. For example, a distribution according to age will give the proportion of children less than 6 months or more than 59 months, which should not have been included in the sample. An age pyramid can demonstrate an under-representation of an age group in the sample (see Annex 9).
An under-representation of an age group, such as the 6-17 months group, in the sample may reflect a higher mortality in this subgroup, but may also reflect the fact that these children were not present on the day of the survey.
In the same way, a distribution according to sex allows us to verify that both sexes are equally represented, and hence, that no selection bias has occurred.
A cluster sample survey has been carried out. The expected prevalence was 12%, the desired precision was 4%. The required sample size was 254 x 2 = 508 children, implying 30 clusters of 17 children (510 children). The following distribution according to age and sex was observed.
Expression of results with their confidence intervals
The age classes proposed here are centred around the months representing fullyears: 12 months (6 - 17 months),24 (18 - 29), 36 (30 - 41) and (48 (42 - 32) months. Many ages are mis - reported and age biasing is toward the full years (i.e. the child is 1 year if it is really 10 months). Making age classes around the full year months is meant to evenly distribute peaks of distribution usually seen around these values
In our example the distribution by sex is not statisticaly different from an even distribution.
In our example, W/H and oedema were distributed as follows:
Definition of nutrition indicators
The two main signs of acute malnutrition described in the chapter concerning the antropometric measures are: a decrease in the value of the W/H index and the presence of oedema. The combination of these two signs and a cut off value for the index are used to define 2 classes of malnutrition:
Global acute malnutrition: proportion of children with a weight / height index <- 2 Z-Scores or oedema
Severe acute malnutrition: proportion of children with a weight / height index <- 3 Z-Scores or oedema
Calculation of malnutrition indicators
To start with, the number of children presenting with oedema is calculed (2 in our example). Then the number of children presenting with a weight / height index?<- 3 Z-Scores and not presenting oedema is calculated. These two numbers are added in order to determine the total number of children with severe acute malnutrition (5 in our example).This number is then expressed as a proportion of the total number of children (5/510 = 10%)
Then these steps are repeted using - 2 Z-Scores as the cut off, in order to determine the number of the children who are defined as globally acutely malnourished (39 in our example). Again this result is expressed as a proportion of the whole sample (39 /510 + 7.6%).
In our example,the following figures are obtained:
Index <- 3 Z-Scores
Index <- 2 Z-Scores
Severe acute malnutrition:
Severe acute malnutrition 7.6%
Oedema + index <-3 Z-Scores
Oedema + index <-2 Z-Scores
When calculating the sample size, the notion of precision was introduced. This is the reason why the proportion of children presenting with malnutrition should be expressed with a corresponding precision which determines the 95% confidence interval. The confidence interval is the prevalence found plus or minus the precision achieved. Calculation of the precision uses the formula already used for determining the sample size but in another way. As a matter of fact, when calculating the sample size (n), an expected prevalence (p) was estimated and a desired precision (d) was used. Now that the survey has been carried out, the approach is reversed: the sample size is known, and the prevalence has been measured, what is going to vary is the precision achieved. If the observed prevalence is closer to 50% than the predicted one, the precision will be worse than expected. If the observed prevalence is less than expected, the precision will be better than expected. This highlights the importance of overestimating the expected prevalence when calculating the sample size, in order to be on the safe side when the survey is completed. Formula for calculation of random/ systematic survey precision:
d = t * sqrt(p * q/n)
The above formula concerns random sample or systematic sample surveys. In a cluster sample survey, the exact formula for the precision uses the prevalence observed in each of the clusters in order to take into account the design effect.
Formula for calculation of cluster survey precision:
d = 1.96 * sqrt(sum(pi - p) ^ 2/((k * (k - 1)))
pi: proportion observed in the cluster i
p: proportion observed in the whole sample
k: number of clusters
In fact, unless a computer is available, a more simplified formula is used in estimating precision for cluster surveys, assuming a design effect of 2. Simplified formula for calculation of cluster survey precision:
d = t * sqrt(2 * p * q/n)
The confidence interval for the estimation of malnutrition is the observed prevalence plus or minus the precision as calculated above. The confidence interval has a 95% of chance of including the actual proportion of malnutrition in the whole population.
In our example, calculation of the precision gives the following:
d = 1.96 * sqrt(2 * 0.076 * 0.924/510) = 0.033
Therefore, a 95% confidence interval is:
C.I. = p + d
Results would be presented as follows:
510 CHILDREN AGED 6 - 59 MONTHS
Proportion of children with oedema = 0.4%
Global acute malnutrition defined by an index < -2 Z-Scores or presence of oedema: 7.6%
95% confidence interval of this estimation: 4.8% to 10.9% (7.6 + 3.3%).
Severe acute malnutrition defined by an index < -3 Z-Scores or presence of oedema: 1.0%
95% confidence interval of this estimation (corrected): 0.2% to 3.4%.
From these two proportions, we can estimate the actual number of malnourished children in the population. The proportions can be applied to the number of children in the population. In our example we estimate the total number of globally acutely malnourished children in the population to be between 430 (4.3% of 10,000) and 1090 (10.9% of 10,000); and between 20 (0.2% of 10,000) and 340 (3.4% of 10,000) children are severely acutely malnourished.
These results can also be expressed for a sub-group of the children, based on age, in order to avoid masking malnutrition in this group by including them in a broader group.
252 CHILDREN AGED 6 - 29 MONTHS
Global acute malnutrition defined by an index < -2 Z-Scores or presence of oedema:10.7%
95% confidence interval of this estimation: 5.3% to 16.1%
Severe acute malnutrition defined by an index < -3 Z-Scores or presence of oedema: 0.8%
95% confidence interval of this estimation (corrected): 0.0% to 5.2%.
Expression of results with their confidence intervals
If a correction was used for calculating the sample size because the sample was more than 10% of the size of the total population, the value of n appearing in the above formulae should be the one before the correction in order to calculate the precision.
A corrected formula exists (quadratic method) which should be used for proportions close to 0% or 100%:
Lower limit = ((2 np + t2 - 1(t2 - (2 + 1/n) + 4p(np + 1))1/2)/2(n + t2)
Upper limit = ((2 np + t2 + 1) + t(t2 + (2 -1/n) + 4 p (np -1))1/2)/2(n + t2)
t: error risk = 1.96
n: sample size
p: proportion of children with malnutrition
q = 1 - p: proportion of children without malnutrition
Some variables can be cross-tabulated. For example, nutritional status (defined according to a cut off value of the weight/height index) and the date of arrival in the camp.
Nutritional status according to the date of arrival
Children 6 - 59 months, Region X, Period Y
Such a table only gives an indication of the trend. The sample size will often not allow a stratified analysis.
Interpretation of the results in context
NOTION OF A <<SNAPSHOT>>
Figures obtained through a single cross-sectional nutrition survey only reflect the nutritional status of the population at the moment of the survey, in a certain region. Taken alone, these figures do not give any indication of the trend, whether the nutritional status is improving or deteriorating. Additional information, collected at the preparatory phase, will allow the interpretation of the results in context.
In a cluster sample survey, figures should not be analyzed for each cluster. Malnutrition observed in one cluster is never representative of the section in which the cluster was drawn, but it is representative of many similar clusters that could be drawn in different sections. As a matter of fact, it is the whole sample which is representative of the population. Practically, in order to avoid misinterpretation by people not acquainted with the cluster sampling method, results should not be presented by cluster.
INTERPRETATION IN CONTEXT
The proportion of malnutrition observed in the sample can be compared to malnutrition rates observed in a previous survey. If there is information from different surveys, some idea of the trend in nutritional status can be inferred. One can only conclude that there was a statistically significant difference between two surveys if confidence intervals do not overlap. Hence, in our example a previous survey had shown a rate of malnutrition of 15.4% + 4.4% (11.0% to 19.8%). The current situation is significantly better, since the lower limit of the first survey (11.0%) is greater than the upper limit of the current one (10.9%).
Information about infanto-juvenile mortality may suggest that a number of malnourished children have died in the past few months. The significance of the survey figures will depend on context factors such as the level of mortality and the timing of the harvest. Complementary information will help in interpreting the results. A stable nutritional situation with moderate levels of malnutrition may deteriorate rapidly if there is an outbreak of measles or a high seasonal transmission of malaria.
A <<Standard recipe>> does not exist for interpreting nutrition survey results. Interpretations and recommendations cannot be based on observed figures only.
However, when the level of malnutrition reaches certain values, emergency actions have to be taken. These actions focus on the children who are classed as malnourished by anthropometric indicators as these indicators are associated to high risk of mortality. In a population where more than 2% of children present oedema, action directed towards these children should be taken without delay. In the same way, a lower limit of the confidence interval of the proportion of children presenting an index value < - 2 Z-Scores of more than 10% depicts an alarming situation.
Survey results may be used to assess the current nutritional situation, to see if nutrition programme objectives have been reached and rates of malnutrition have been reduced. However, this measure of the impact of a programme does not allow us to assess the reasons for success or failure in reaching the objective. The anthropometric survey is just a tool which is part of a more global approach.
Writing of the report / recommendations
A report on the survey should be written as soon as preliminary results are available. Such a report should indicate the procedure used in defining the survey.
The first page should include a short summary (1 page), covering the objectives of the survey, the methodology used, the main results and recommendations. It is meant for the use of people in charge of the population in order to make appropriate decisions guided by the findings of the survey.
In the introduction, the context in which the survey was carried out should be described. What population was surveyed, at which period, in which geographic area.
Furthermore, any relevant information concerning the status of the population should also be reviewed. Nutrition programmes, surveillance data, morbidity and mortality information are included.
OBJECTIVES OF THE SURVEY
The objectives of the survey should be clearly stated. What was measured, in which population and why?
· Survey methodology and sampling frame
The choice of a cross-sectional survey is justified by the necessity for getting information in a short period of time. The sampling frame must be indicated as well as all parameters used, such as error risk, expected prevalence and expected precision.
· Variables measured and recording information
The type of measuring instruments used should be noted. For example, the weight was assessed using a Salter spring scale, to the nearest 100 grams.
· Training of the data collectors
The schedule of the training and its duration should be mentioned, including the realization of a standardization test, if it was possible.
Distribution of the sample, according to age and sex is the first stage of the analysis. The mode of expression of indices should be recorded as well as the definition of grades of malnutrition used. Distribution according to the indices is presented.
INTERPRETATION OF THE RESULTS, DISCUSSION
The discussion puts the results back into the context. Comparison can be made with previous surveys, or surveys from a similar situation. Tentative explanations may be suggested at this stage.
A report should always include recommendations. A nutrition survey is meant to promote rational decision making. As an example, an intensive feeding centre might be recommended in the area if the proportion of children with severe acute malnutrition is high. On the other hand, an active screening programme through home visits can be proposed if the number of malnourished children found during the survey is much higher than the number of children currently seen in an ongoing feeding programme. A global approach might be recommended if the whole population is suffering from malnutrition. However, if the situation is demonstrated to have improved, one can recommend the interruption of a "vertical supplementary programme" and the integration of the nutrition programme in the daily MCH activities of the health system.
· Among displaced populations or refugees, evaluation of the nutritional status is essential in planning a relief programme. The measure of the prevalence of malnutrition, through a quick cross-sectional anthropometric survey gives valuable information when making decisions. The results of such a survey often have vital consequences for the community. That is the reason why these results must be reliable.
· The reliability of the results is related to compliance with the protocol. Each step is essential.
· The objectives should be clearly defined from the start, as should the plan of analysis.
· The sample should be representative of the population from which it was drawn. Systematic sampling when possible gives the same precision with half the sample size required for cluster sampling.
· Measures should be reliable. Training of data collectors and assessment of their performance through standardization tests is a corner stone of the survey.
· The report should include a summary mentioning the main findings and the recommendations made. Confidence intervals should be indicated whenever a proportion is given.
· The realization of a survey in the field is an excellent occasion for health workers to appreciate the living conditions of the population. On top of the actual proportion of malnourished children given by the survey, valuable additional information maybe observed. This additional information will be of great help in appreciating a situation.