Humanitarianresponse Logo

Nutrition & Mortality SMART Survey in Logar Province (7 January 2018)

Subject/Objective: 
To determine the nutritional status of vulnerable population mainly under five, pregnant and lactating women living in Logar province.
Methodology: 
As the survey areas was large with a dispersed population, a two-stage cluster methodology was applied as follows: Stage 1: Random selection of clusters/villages was chosen using probability proportional to size (PPS) using ENA for SMART software version 2011 of (9th July 2015). A list of all updated villages amounted into the ENA for SMART software where PPS was applied. The villages with a large population had a higher chance of being selected than villages with a small population and vice versa. 47 clusters were supposed to be surveyed (606/13=46.6 clusters, rounded up to 47 Clusters) as each survey team could complete anthropometric measurements in 13 households a day. 45 clusters were surveyed out of 47 clusters : 2 clusters were missed due to insecurity and as the nonresponse cluster was less than 10 % we did not use the reserve clusters (as the reserve clusters were supposed to be used if 10% or more clusters would have been impossible to reach during the survey as per SMART methodology), see Annex 1. HHS size was calculated from ENA in the planning stage, 606 being the number of households planned to be surveyed. In each selected village, one or more community member(s) were asked to help the survey teams to conduct their work by providing information about the village with regard to the geographical organization or the number of households. For large villages, the village was divided into smaller segments and a segment was selected randomly to represent the cluster. This division was done based on existing administrative units e.g. neighbourhoods, streets or natural landmarks like a river, road or public places like market, schools, and mosques. Stage 2: Random selection of households from updated and complete list of households within a given village. In this case, the actual survey data collection incorporated 606 households randomly selected based on survey parameters calculation for anthropometric Based on total sample size each team could cover effectively 13 households a day. In this assessment, 7 teams were engaged during the assessments, while data collection was done in 9 days. All households were enumerated and given numbers by the survey team. The 13 households were chosen randomly from these enumerated households, by systematic random sampling method to identify the households to be surveyed. The teams were trained in both methods of sampling (simple and systematic random sampling) and they were offered with materials to assist in determining the households during the data collection exercise. All the children living in the selected house aged 0 to 59 months old were included for anthropometric measurements. Children aged <24 months were included for IYCF measurements. To ensure that every child had the same chance to be surveyed, if more than one eligible child was found in a household, both were included, even if there were twins. Eligible orphans living in the selected Households were surveyed. All the selected HH were included in the mortality survey as well as to answer to questions concerning the HH as a whole (ex. water storage, WASH and FSL). Any empty households or households with missing or absent children were revisited at the end of each day in each cluster; any missing or absent children that was not subsequently found was not included in the survey. A cluster control form was used to record all these absent households. The abandoned ones were excluded from the total HHs list at the beginning of the fieldwork. This information was provided to the teams by an elder of the villages. The household was our basic sampling unit. The term household was defined as all people eating from the same pot and living together (WFP definition). In Afghanistan, the term household is often defined and/or used synonymously with a compound – which potentially represents more than one household as defined here. In this case, a two-step process was ensured with the village leaders/community elders, identifying “compound” together with the use of the list of households within the community, asking if there were multiple cooking areas to determine what members of the household/compound should be included in the study.
Key findings: 
The survey findings revealed that the Prevalence of Global Acute Malnutrition (GAM) and severe acute malnutrition (SAM) in children aged 6-59 months based on Weight for Height (WHZ) were at 7.1% (5.2-9.6 95% C.I.) while SAM prevalence is 1.1% (0.5-2.4 95% CI.). This situation is “Poor” according to WHO classification of acute malnutrition. The prevalence of GAM based on MUAC cut-offs was 4.0% (2.2-7.1 95% C.I.) and SAM was at 1.5 % (0.9-2.8 95% C.I.) respectively. The combine GAM and SAM prevalence based on MUAC and WHZ both criteria was (n=73) 9.6% (7.5-11.7 95% CI) and (n=9)1.2 % (0.4-2.0 95% CI) respectively. The prevalence of stunting for children age under 6-59 months was at 30.7 %( 26.4-35.2 95% C.I.) and considered serious public health problem based on WHO classification. Crude Death Rate and Under-five Death rate was at 0.42/10,000/day and 0.30/10,000/per day respectively. The rates are both below SPHERE emergency thresholds. The survey revealed that fever, diarrhea and ARI were major illnesses reported among the under-five, with above 51.8% of children reported to have been sick prior to the survey period.
Sample size: 
770
Assessment Report: 
Publicly Available
Assessment Questionnaire: 
Publicly Available
Assessment Data: 
Publicly Available
Webspace(s): 
Assessment Date(s): 
10 Jul 2018

Level of Representation

District / Province / Locality / County
Status: 
Report completed
Unit(s) of Measurement: 
Individuals
Population Type(s): 
Children
Cluster(s)/Sector(s): 
Location(s): 
Logar