2. Defining the objectives of Rapid SMART ... - SMART Methodology

[pic] Guidelines: Rapid SMART Surveys for Emergencies Developed by
ACF - International, SMART Initiative at ACF - Canada and CDC Atlanta
Version 1, September 2014 TABLE OF CONTENTS
Preamble 3 Overview of Rapid SMART surveys 4
1. Limitations of Rapid SMART surveys 5
2. Validity of Rapid SMART results - via Plausibility Check 5 Stages of Rapid SMART surveys 6
1. Deciding whether to do a Rapid SMART or full SMART survey 6
2. Defining the objectives of Rapid SMART surveys 7
3. Defining the geographic area and population group(s) to be assessed 8
4. Informing national and local authorities and coordination 8
5. Resources required to carry out a Rapid SMART survey 9
6. Determining the data collection methodology 10
5.1 Selecting the sampling method 10
5.2 Converting number of children into number of households to select in
case of cluster sampling 11
5.3 Household Selection 12
5.4 Selection of children to be included in Rapid SMART surveys 12
7. Organizing Rapid SMART surveys 12
5.1 Preparing the assessment equipment, supplies, and materials 12
5.2 Assessment teams and training 12
5.3 Management and supervision of Rapid SMART surveys 13
8. Data collection 13
7.1 Main data to collect 13
7.2 Special Cases 15
9. Data record, data entry and data quality check 15
10. Data analysis 16
11. Reporting and Interpretation of the Results 16
10.1 Assessment report 16
10.2 Discussion and interpretation of the results 17
12. Validation procedure and Results Dissemination 19 Preamble Proper data is essential in understanding the extent of nutritional needs,
as well as how to best formulate programs to most appropriately address
those needs. During emergencies and when large-scale responses are needed,
it is important to know where to respond and to concentrate supplies.
Several different methods to rapidly assess a nutritional situation have
been elaborated and endorsed by different humanitarian actors and
scientific bodies. Nevertheless, none of them have been largely implemented
and approved at international level. Moreover, the lack of standardized
method of rapid nutrition assessment hampers the comparison of data in
order to capture significant changes. The nutrition Global Cluster[1] is
calling for a global guidance in Rapid assessments as there is limited time
for preparations and consultations during the immediate response period of
an emergency. The method presented in these comprehensive guidelines is built upon
Standardized Monitoring and Assessment of Relief and Transitions (SMART)
Methodology. It proposes a way to rapidly collect reliable nutrition data
and to address the above mentioned problems. To emphasize on its strong
link with SMART and to differentiate it from the large variety of Rapid
Nutrition Assessment (RNA) methods, it is called a "Rapid SMART surveys". Since 2012, the Rapid SMART method has been field-tested in several
settings, and where possible a full multi-cluster survey was conducted
simultaneously to compare the results. The Rapid SMART methodology was
tested in South Sudan, Afghanistan, India, Myanmar and Madagascar between
November 2012 and July 2013[2]. After each test, the method was readjusted
and further improved, orienting the development of the current Rapid SMART
methodology. The Rapid SMART methodology was technically endorsed by ACF International,
the SMART Initiative at ACF-Canada and by the International Emergency and
Refugee Health Branch at the Center for Disease Control and Prevention in
Atlanta, US.
Overview of Rapid SMART surveys [pic] The main objectives and contexts in which Rapid SMART surveys are carried
out are detailed above. The validity of Rapid SMART surveys is confirmed
only after the representativeness, accuracy and precision of the results
are evaluated (see details in the paragraph: "Validity of Rapid SMART
results"). Assessment should only occur in a clearly delimited zone (e.g.
group of villages, IDP/Refugee camps or settlements, urban slums and
neighborhoods) which population has similar patterns (affected by the
crisis, having equal access to services, having similar culture, same
livelihood zone etc.). For the sake of accurate and precise under-nutrition[3] prevalence
estimation, a full multi-cluster nutrition survey is recommended as soon as
the situation gets stable and the population is accessible. If it is judged that other information such as food security, Infant and
Young Child Feeding (IYCF) and care practices, as well as WASH situation is
more necessary in a given emergency then other adapted studies have to be
prioritized over Rapid SMART surveys.
Limitations of Rapid SMART surveys The testing of the Rapid SMART methodology based on the trials in various
countries aforementioned have confirmed following limitations: . Rapid SMART surveys cannot be used for assessing GAM and SAM
prevalence of large geographical areas like provinces or other larger
zones as population will have large variety of patterns, but rather in
specific and defined administrative units/areas. In such case bigger
design effect has to be planned and therefore much larger sample has
to be studied. This can be done only through a full multi-cluster
SMART survey. . The results of Rapid SMART surveys can neither be extrapolated to
larger than the zone of the assessment nor disaggregated (not
representative of one village out of the administrative unit
surveyed). . Aiming at collecting a lot of additional data slows down the rapidity
of the data collection and its fast analysis and dissemination of
results. . The results of a Rapid SMART surveys cannot be used for long term
programming[4]; it can only inform emergency response.
Validity of Rapid SMART results - via Plausibility Check Based on the same epidemiological science and principles as the SMART
methodology, Rapid SMART results are valid only when data collected from
the field are from a representative sample and good quality measurements.
The overall quality of the data collected is analyzed using the SMART
plausibility check generated by the ENA for SMART software[5].
The respect of the sampling method and the good quality of collected
anthropometric data insures the representativeness of the sample. There
should not be significant difference between sexes and age groups in the
final sample (sex ratio close to 1 and an age ratio of children aged 6 to
29 months old to children 59 months old close to 0.85).
Key to good representativeness is the random selection of a sample among
the target population so that the prevalence estimate is close to the true
population value. By convention, results will use a 95% confidence
interval which means that in absence of bias, once you have the estimate,
you can be 95% sure (level of confidence) that the true population value
of your indicator is within the limits of the interval calculated. The more
representative is the sample and the more accurate are the measurements,
the more certain you can be about the absence of bias and of having the
true population value for the indicator within the confidence interval. Stages of Rapid SMART surveys
For each Rapid SMART survey, specific and context-adapted methodology has
to be rapidly prepared. However, in order to guarantee the quality and the
validity of the data, each Rapid SMART has to be prepared and conducted
strictly following the stages outlined in Annex 1.
1 Deciding whether to do a Rapid SMART or full SMART survey A full two stage cluster sampling survey using the SMART methodology will
always be the best method to estimate acute malnutrition prevalence and all
efforts should tend towards trying to implement such a survey, in the
fastest time possible. Rapid SMART rather than full SMART surveys would be considered in case when
affected population: a) Live in a clear geographically delimited small administrative units
such as IDP/refugee camps, urban slums, settlements or neighborhoods,
group of villages; b) Have similar access to public services and socio-economic patterns[6]; c) Live in an insecure and/or with limited access area where the survey
team cannot spend long time on the ground. And it is combined with at least one of following factors: 1. Limited time: emergency situation due to epidemics, drought,
calamities, displacement, high insecurity in the zone. 2. Rapid information of nutritional situation in a representative and
accurate manner. However, if there is no possibility to organize and conduct Rapid SMART
surveys in a week, than the best choice is a full SMART survey. Rapid SMART
surveys have to remain RAPID as it aims to inform emergency response.
Defining the objectives of Rapid SMART surveys The main objective of Rapid SMART surveys is to quickly assess whether
there is a nutrition emergency or not. Rapid SMART surveys allow a rapid
estimation of the prevalence of GAM and SAM based on WHZ, MUAC and
nutritional bilateral pitting ?dema. Only in situations in high security, GAM and SAM prevalence can be
estimated based on low MUAC and nutritional bi-lateral pitting ?dema with
extremely limited time for data collection due