A coverage survey using the Simplified Lot Quality Assurance Sampling Evaluation of Access and Coverage or SLEAC methodology was carried out across 71 Local Government Areas (LGAs) implementing CMAM program in 11 states in the north of Nigeria. The SLEAC used a two-standards, three-class classifier with 20% and 50% as the thresholds to determine low, moderate and high coverage classes (i.e. 20% or less, between 20% and 50% and greater than 50% respectively).

Of the 71 LGAs surveyed, more than half have moderate coverage (40 LGAs) but only 4 have high coverage. There are 27 LGAs with low coverage. At the state level, 8 of the 11 states surveyed have moderate coverage and only 3 states with low coverage. However, no state has achieved high coverage. Except for Adamawa and Kebbi state, coverage in all other states is heterogeneous. Overall, coverage in the northern states of Nigeria is moderate with an overall estimate of 36.6% (95% CI: 32.3% – 40.9% ).

The key barriers to service uptake and access for those children who were not in the program are: 1) ignorance of malnutrition; 2) no knowledge of the program; 3) ignorance of how the program works; 4) constraints and responsibility of the mother; 5) service delivery problems; and 5) geographical access issues.

Based on the levels of coverage achieved and the barriers identified, the following actions are recommended to improve coverage:

  1. Strengthening of the program’s community mobilization strategy with a strong emphasis on raising community awareness regarding malnutrition, its causes and manifestations and available treatment through the program. Community mobilization should be aimed at the whole community including community leaders;

  2. Strengthening the integration of CMAM into the activities of the health center (e.g. EPI, consultations, etc.);

  3. Develop, trial and institutionalize alternative service delivery mechanisms that aim to increase beneficiaries’ access to the program’s services with particular attention to those who live far from health centers or health posts providing the service. These alternative mechanisms may include mobile treatment centers that would cater for most distant villages or fortnightly follow-ups for beneficiaries who live far from treatment sites or who face significant opportunity costs related to the standard weekly follow-up visits;

  4. Setting up the management of moderate acute malnutrition component of CMAM; and,

  5. Perform a focused SQUEAC in a selection of the LGAs implementing CMAM which builds on the findings of the SLEAC particularly with regard to the spatial distribution of coverage in each of the LGA and within the states. This would entail more detailed mapping of the results of the SLEAC that would inform how improvements to the program can be implemented particularly in terms of positioning of new treatment sites (if deemed necessary) or alternative service delivery mechanisms mentioned in item 3 above.



In Nigeria, acute malnutrition of children under 5 years is a major public health concern. Nigeria is ranked third in the world for absolute number of children under 5 years with severe acute malnutrition (SAM) particularly in the country’s northern states where the risk of acute malnutrition is highest1.

In response to this situation, the Federal Ministry of Health (FMOH) supported by partners has been implementing community-based management of acute malnutrition (CMAM) program with the aim of treating children afflicted by the condition and hence preventing mortality caused by malnutrition. CMAM implementation was first piloted in 2009 in 3 LGAs of Gombe state and 5 LGAs of Kebbi state. By 2010 the CMAM program was expanded to other states namely Sokoto, Zamfara, Katsina, Kano, Jigawa, Bauchi, Adamawa, Yobe and Borno with a total of 378 CMAM sites2.

Programme coverage, defined as the proportion of children 6 to 59 months old with SAM who receive therapeutic care, is a key indicator of CMAM program impact. Measuring program coverage is therefore a critical step in assessing program performance. It is for this purpose that a SLEAC has been performed in the 11 states across 71 LGAs in the north of the country that are implementing CMAM.

This report describes and details the process and the outcomes of the SLEAC conducted.


The objectives of the SLEAC survey are:

  • To assess the impact of the CMAM program in 75 LGAs across 11 states (i.e. Adamawa, Bauchi, Borno, Gombe, Jigawa, Kano, Katsina, Kebbi, Sokoto, Yobe and Zamfara) in the north of Nigeria;

  • To train in-country survey team on the SLEAC; and,

  • To raise awarenss and understanding on program coverage and build capacity of partners involved at different levels (Ministries, NGOs, UNICEF)


SLEAC sample design

SLEAC was used as a wide-area survey method in order to classify coverage at the Local government Area (LGA) level.

SLEAC is a low-resource method for classifying and estimating the coverage of selective feeding programs. SLEAC surveys classify coverage at the level of the service delivery unit. It provides also limited data (i.e., reasons for non-attendance collected from a single informant type using a single method with a small sample size) on barriers to service uptake and analysis. This varies with the scale of the program.

SLEAC survey uses a two-stage sampling design.

Stage 1 sampling: Selection of PSUs
This is the sampling method that is used to select the villages or settlements to be sampled in the survey. The primary sampling unit (PSU) used in the SLEAC survey was village or settlement. Complete village lists per LGA organized by ward were provided by the National Bureau of Statistic (NBS). The structure of LGA-level samples are shown in Figure 1.

Figure 1: Structure of samples in LGAs


A target of $ n = 40 $ cases was used in each LGA. This is a standard SLEAC sample size for large populations.

The number of PSUs needed to reach the target sample size in each LGA was calculated using the following formula:

$$ n_{psu}=\frac{\text{Target sample size (n)}}{\text{median village population all ages}\times{\frac{\text{percentage of population 6 – 59 months}}{100}}\times{\frac{\text{SAM prevalence}}{100}}} $$

The percentage of the population aged between 6 and 59 months was estimated as 18%. SAM prevalence rates by MUAC were taken from results of SMART surveys conducted in 2011, 2012 and 2013. The SAM prevalence rate used in the formula for each state was adjusted from the results of the prevalence surveys taking into account the difference in seasons between the dates that the prevalence surveys were conducted and the dates that the SLEACs were going to be performed. More importantly, adjustments were made to the SAM prevalence rates with the underlying aim of ensuring that enough villages or settlements were surveyed in order to reach the target sample size of 40 per LGA. This generally meant having to underestimate the prevalence of SAM per state. Table 1 presents the SAM prevalence rates per state used to obtain the number of LGAs to be surveyed using the formula specified above.

Table 1: SAM prevalence rates per state used in the SLEAC survey


It should be noted that the SAM prevalence rates used were state level estimates and not LGA prevalence results. The recommendation for SLEAC surveys is that wherever possible, local prevalence rates (i.e. rates estimated for the local areas in which the SLEAC survey is being conducted) should be used. For the case of this SLEAC, there were no SMART surveys done at the LGA level. Hence, we assumed that SAM prevalence is homogenous within each state and assigned the SAM prevalence of the state to each of the constituent LGAs to be surveyed.

A minimum of 25 PSUs was set in cases were the calculated number of villages / settlements per LGA needed to be sampled was less than 25. This was done to ensure as much as possible an even spatial spread of the stage 1 sample.

A systematic sampling approach was used to select the PSUs to be sampled. This was done through the following steps:

Step 1. The list of villages was organized by LGA and then by ward.

Step 2. A sampling interval was calculated using the following formula:

$$ \text{sampling interval}=\frac{\text{number of villages in LGA}}{n_{psu}} $$

Step 3. A random starting PSU from the top of the list was selected using a random number between 1 and the sampling interval. The random number was generated using Microsoft Excel® software.

Stage 2. A within-community sampling method
Stage 2 sampling involves finding the target population in the selected PSUs. The target population was:

  • Child aged 6 to 59 months old; and,
  • MUAC < 115 mm; and / or,
  • With nutritional oedema; and / or
  • In the CMAM program


The PSUs selected in stage 1 were sampled using a case-finding method designed to find all or nearly all SAM cases in the particular PSU.

In rural PSUs, an active and adaptive case finding approach was used. This method involved developing a case-finding question appropriate to the location and context from the base question of:

Where can we find children who are sick, thin, have swollen legs or feet, or have recently been sick and have not recovered fully, or are attending a feeding program?

This question was adapted and improved using information collected from key informants to include local terms (in all local languages) and local aetiological beliefs regarding wasting and oedema.

In urban PSUs, house-to-house, door-to-door case finding was implemented. This is based on experience in conducting coverage surveys in urban areas where the use of key informants and context-specific case finding questions was not effective in finding all SAM cases.

Once found, SAM cases were then assessed as to whether they were covered in the CMAM program or not.

  • A SAM case was assessed to be covered by the program if the child meets the criteria of the target population stated above and is enrolled in the program verified by the presence of RUTF or the OTP card
  • A SAM case was assessed to be not covered by the program if the child meets the criteria of the target population and is not enrolled in the program.
  • A recovering case is a child aged 6 – 59 months who is enrolled in the program and waiting to be discharged with a MUAC > 115 mm and no nutritional oedema.


A tally sheet was filled for each village sampled and each child who fulfilled the criteria of the survey was recorded. When a SAM case was not covered a standard questionnaire regarding the barriers to access and coverage was administered to the caregiver.

Coverage Estimator

There are two coverage estimators that have been developed: point and period coverage.

Point coverage
This estimator uses data for current cases (i.e. those children who are still currently SAM) only. It is calculated using the following formula:

$$ \text{point coverage}=\frac{\text{number of current cases attending program}}{\text{number of current cases}} $$

The point coverage estimator provides a snapshot of program performance and places a strong emphasis on the coverage and timeliness of case-finding and recruitment.

Period coverage
This estimator uses data for both current and recovering cases. It is calculated using the following formula:

$$ \text{period coverage}=\frac{\text{number of current and recovering cases in the program}}{\text{number of current and recovering cases}} $$

The period coverage estimator includes recovering cases. These are children that should be in the program because they have not yet met program discharge criteria.

Depending on program context, either point or period coverage should be used to describe program coverage. In general, the recommendation is that if the program has good case-finding and short lengths of stay, period coverage is likely to be appropriate. On the other hand, if the program has poor case-finding and recruitment and long lengths of stay due to late presentation and / or late admission, point coverage is likely to be appropriate.

In the case of this SLEAC, it would have been ideal that an appropriate, LGA-specific coverage estimator be chosen to best capture the coverage situation in each LGA. However, there was limited program information and context available to the survey team to fully decide which estimator to use per LGA.

Despite this limitation, there are general characteristics of CMAM programming in Nigeria and the coverage achievements of some LGAs where surveys have been conducted before that provide insight and guidance as to which coverage estimator is most appropriate albeit for all the LGAs surveyed and not LGA-specific.

In general, the following observations about CMAM programming and coverage in Nigeria can be made:

  • CMAM programming in Nigeria is challenging mainly due to the high caseloads and the geographical spread of the need for the services. This manifests as large numbers of cases coming in for the first time or for follow-up. In such context, case-finding for enrollment to the program is always a difficult task.

  • Defaulting is a significant issue in Nigeria. Whilst some of the programs implemented in particular states and LGAs (especially those in which previous coverage surveys has already noted this problem and have provided recommendations for improvement) may have improved on this problem already but for majority of the LGAs surveyed this is most likely still a significant problem.

Given these observations and based on knowledge of previous coverage surveys done in Nigeria, point coverage was chosen as the more appropriate coverage estimator for reporting coverage estimates for the following reasons:

  • Timely case-finding will always be an important factor to achieving good coverage and in the context of CMAM programming in Nigeria this factor is also the hardest to get right. Using the point coverage estimator will further emphasise the importance of timely case-finding towards increasing program coverage. As will be noted later in the report, it is very likely that the LGAs with moderate to high point coverage are more likely to have good and timely case-finding.

  • Point coverage will most likely have less bias than period coverage in the context of high defaulting. A recent study into defaulting in Gombe state showed that whilst death was shown as the main outcome of those who have defaulted, recovery (i.e. MUAC > 115 and no oedema) was the third most common outcome after defaulting. Period coverage doesn’t take into account cases who are recovering but are not in the program (such as those defaulters who have recovered) and as such tend to overestimate coverage. Point coverage, on the other hand, is not susceptible to this bias as it doesn’t account for recovering cases. Whilst this may not be the case for some LGAs with no problems with defaulting, it is very likely that for most LGAs that face the complex mechanism of defaulting, period coverage may give a distorted assessment of coverage.


Data analysis

Coverage classification
The simplified lot quality assurance sampling (LQAS) classification technique was used to analyse the data. A two-standard (20% and 50%), three-class (low, moderate, high) classifier was used to classify the coverage in each LGA. The three classes were defined as follows:

  • Low coverage: 20% or below
  • Moderate coverage: greater than 20% and up to 50%
  • High coverage: greater than 50%


Figure 2: Two-standard, three-tier classifier

The standards were used to create decision rules using the rule-of-thumb formulas:

$$ \begin{align*}
d_{1} &= \lfloor \ {n \times p_{1}} \ \rfloor = \left \lfloor \ n \times \frac {20}{100} \ \right \rfloor = \left \lfloor \ \frac {n}{5} \ \right \rfloor \\[0.5cm]
d_{2} &= \lfloor \ {n \times p_{2}} \ \rfloor = \left \lfloor \ n \times \frac {50}{100} \ \right \rfloor = \left \lfloor \ \frac {n}{2} \ \right \rfloor \\[0.5cm]
&\text {where} \\[0.5cm]
n &= \text{sample size achieved by the survey} \\
p_{1} &= \text{lower coverage threshold} \\
p_{2} &= \text{upper coverage threshold}

Coverage in each LGA and state was classified using the algorithm presented in Figure 3.

Figure 3: Algorithm for a three-class simplified LQAS classifier

Coverage estimation
An aggregate estimate of coverage was calculated for each state and for all the LGAs combined using standard weighted estimation of proportions techniques used for a stratified sample as described in the FANTA SQUEAC and SLEAC Technical Reference. Chi-square testing was performed to assess whether coverage results were homogeneous within each state. Whether or not coverage is homogenous within each state allows for the contextualisation of the state level and overall coverage estimates calculated3.

Survey implementation

The SLEAC coverage survey was carried out in two blocks:

  • West block: Kebbi, Sokoto, Katsina, Zamfara and Kano States
  • East block: Gombe, Bauchi, Jigawa, Yobe, Borno and Adamawa States

Due to security reasons, it was not possible to perform the survey in 4 Bama, Ngala, Mobbar and Monguno LGAs of Borno State hence survey was cancelled in these LGAs.

The SLEAC survey in the West Block was held from 1 October 2013 to 20 December 2013 and in the East Block from 9 December 2013 to 19 February 2014.

Training of interviewers
Two training sessions on the SLEAC methodology (4 days for each training) were conducted for each of the two groups of investigators from two blocks. One was held in the Kebbi State from 2 to 6 October 2013 and another in Gombe State from 9 December to 12 December 2013. In the spirit of capacity building, 8 managers from NBS, one manager from the MOH and 15 surveyors from NBS and MOH belonging to the West Block survey team were trained. For the East Block, 8 managers from NBS, one from the MOH and 20 surveyors (11 from MOH and 9 from NBS) were also trained.

The theoretical component of the training covered an introduction to coverage, the objectives of the SLEAC survey, SLEAC methodology, the procedure for active adaptive case-finding method, anthropometric measurement and tools of the survey. Practice focused on anthropometric measurements (the standardization of anthropometric measurements), research local terminology used to describe malnutrition, and active case finding.

Only LGAs with CMAM program were surveyed. Table 2 shows the number and proportion of LGAs of different states which had CMAM programs hence surveyed.

Table 2: LGAs with CMAM program

State Number of LGAs with CMAM Number of LGAs with no CMAM Total LGAs LGA Geographic Coverage
Kebbi 10 11 21 48%
Sokoto 7 16 23 30%
Zamfara 6 8 14 43%
Katsina 10 24 34 29%
Kano 6 38 44 14%
Gombe 3 8 11 27%
Jigawa 12 15 27 44%
Bauchi 3 17 20 15%
Adamawa 3 18 21 14%
Borno 6 21 27 22%
Yobe 9 8 17 53%
Total 75 184 259 29%


Table 3 summarizes the number of LGAs, wards, villages, the number of SAM cases and recovering cases found during the investigation in the 11 states.

Table 3: Description of SLEAC sample by state

State LGAs Wards Settlements SAM cases Recovering cases
Kebbi 10 111 329 840 14
Sokoto 7 69 360 884 19
Zamfara 6 54 270 765 57
Katsina 10 108 317 823 243
Kano 6 58 173 334 110
Gombe 3 32 138 215 114
Jigawa 12 130 376 751 264
Bauchi 3 46 83 241 35
Adamawa 3 31 110 240 48
Borno 2 23 56 132 14
Yobe 9 23 56 707 75
Total 71 910 2929 5947 993


On the list of villages covered by the CMAM program provided by NBS, some villages sampled were replaced by others during the implementation of the survey. The reasons for replacement were 1) sampled settlement/village was not included in the LGA covered by CMAM program; 2) the village was uninhabited or was no longer there; 3) population has migrated; and, 4) in some villages (Njibiri and Wafi in Borno, Jaji gurawa in Yobe state) surveyors were rejected because population thought survey was about EPI (Expanded Program on Immunization) program of polio; and, 5) due to insecurity. The new village chosen was selected for its proximity to the old village.

Data collection
Data collection was carried out for the Western Block by 15 investigators (7 from NBS and 8 from MOH) who were grouped into 6 teams and data collection of the states of Eastern block was carried out by 20 investigators (NBS 11 and 9 MOH) who were grouped into 10 two-person teams.

During the investigation, supervision of activities was provided by Valid International consultants. After completing the survey in a state a debriefing session was organized to provide instant results of the state to the authorities of MOH, NBS and other partners involved in the CMAM activities.

Quality control of data
To ensure data quality, the following measures were put in place:

  • Regular field supervision of surveyors in each state except for Yobe and Borno due to insecurity;
  • Random field data checks were performed by supervisors to confirm case finding and also recording of data on tally sheets; and,
  • In the state of Yobe where supervisors were not able to go due to insecurity, some LGAs like Geidam, Nguru and Potaskum were done two times by different teams of surveyors to confirm the result found.

Difficulties encountered during the survey

  • Some villages were inaccessible by vehicles due to lack of roads and in some situations investigators were obliged to walk to reach the villages or used motorcycles, canoes or walk on foot to cross a river to reach the village. Some roads were too sandy;
  • Administrative procedures with local authorities before starting the survey in a state and LGAs were cumbersome which contributed to the slowness in the implementation of the activities of the investigation;
  • The scheduled dates of debriefing in some states did not suit or fit in very well with the agenda of local authorities and in some situations, the SLEAC survey team was obliged to go back from one state to the previous state in order to perform a debriefing session and this situation slowed down the activities of the survey; and,
  • Insecurity in some states especially Yobe and Borno states. Two investigators withdrew from the survey because of the insecurity in these two states. Due to insecurity, 4 LGAs in Borno State, namely Bama, Ngala, Monguno, Mobbar, were not surveyed and some villages were replaced.


Profiles of SAM children

We profiled the SAM cases found during the survey by their age and their MUAC. The median age was for all cases found in the survey and the median MUAC was for all the uncovered cases because it gives us information about the risk of mortality of SAM not covered in the community. Table 4 summarises the characteristics of age of all cases of the survey and MUAC of uncovered cases found.

Table 4: Profile of SAM cases and uncovered SAM cases

State Age (months) MUAC (mm)
Median Mode Median Mode
Kebbi 16 24 109 110
Sokoto 16 12 109 110
Zamfara 16 24 108 113
Katsina 16 24 109 113
Kano 17 24 110 114
Gombe 18 24 112 113
Jigawa 17 24 110 114
Bauchi 18 24 108 113
Adamawa 17 24 108 114
Borno 13 12 108 113
Yobe 14 12 110 112


This age profile indicates that the sample of SAM cases found during the coverage survey was what expected. SAM was expected to be more prevalent in the younger age group of children between 6 to 24 months, as they are the most susceptible to various known causal factors of malnutrition.

Regarding the MUAC, except for Kano, Gombe, Jigawa and Yobe states, the medians of uncovered cases found of others states are less than 110 mm indicating a high risk of mortality. This gives information about time of the SAM cases spent in the community without being identified by active case finding of volunteers. As the median was less than 110mm, (not close to 115 mm) these showed that SAM cases have spent more time in community without being detected by volunteers, therefore in recommendation much effort should be done to catch SAM cases very early and to recruit them into the program for these states. Annex 9 presents the histogram of age of all cases and the histogram of MUAC of uncovered cases

Furthermore it should be noted that of the total of 5947 cases of severe acute malnutrition, 237 cases had oedema or 4.5%. Table 5 presents the number of edema cases found for each state and their degree.

Table 5: Characteristics of oedema cases

State Oedema Cases
n + ++ +++
Kebbi 2 1 1 0
Sokoto 8 7 1 0
Zamfara 5 2 1 2
Katsina 55 42 8 5
Kano 29 23 5 1
Gombe 27 17 7 3
Jigawa 62 20 28 14
Bauchi 5 2 2 1
Adamawa 6 3 3 0
Borno 3 1 2 0
Yobe 35 23 11 1
Total 237 141 (59.5%) 69 (29.1%) 27 (11.4%)


Coverage classification

Coverage classification at the state level
Table 6 presents the point coverage classification results for all states. The point coverage has been used in all states. The coverage in the states of Kebbi, Sokoto and Zamfara is low. The coverage in the rest of the states is moderate.


Table 6: Classification of coverage at the state level

State SAM cases found Covered SAM cases Decision rule 1 $ c > d_{1} \text ? $ Decision rule 2 $ c > d_{2} \text ? $ Coverage classification
n c $ d_{1} = \left \lfloor \frac{n}{5} \right \rfloor $ $ d_{2} = \left \lfloor \frac{n}{2} \right \rfloor $
Kebbi 840 101 168 No 120 No Low
Sokoto 897 63 179 No 448 No Low
Zamfara 766 131 153 No 383 No Low
Katsina 824 314 162 Yes 407 No Moderate
Kano 334 141 66 Yes 167 No Moderate
Gombe 215 45 215 Yes 107 No Moderate
Jigawa 751 248 150 Yes 375 No Moderate
Bauchi 241 115 48 No 120 No Low
Adamawa 240 114 48 Yes 120 No Moderate
Borno 132 37 26 Yes 66 No Moderate
Yobe 707 187 112 Yes 281 No Moderate
Total 5947 1496 1189 Yes 2973 No Moderate

Coverage classification at the LGA level
Table 7 presents classification of coverage for each LGA surveyed.

Of the 71 LGAs surveyed, more than half (40 LGAs) have moderate coverage and 27 have low coverage. Only 4 LGAs have high coverage. All LGAs in Kebbi and Sokoto had low coverage. All LGAs in Adamawa have moderate coverage and in Katsina and Kano, all LGAs have moderate coverage except for one for each state which has high coverage. The coverage classification per LGA gives an indication of the homogeneity of coverage results in each of the states. Adamawa, Kebbi and Sokoto whose LGAs all have the same coverage classification most likely have quite even distribution of coverage with Adamawa having an even moderate coverage while Kebbi and Sokoto with an even low coverage. This should be taken into consideration when interpreting the results of the aggregated classification presented above in Table 6.

Figure 4 presents a map of point coverage classification for the 71 LGAs surveyed. Figure 5 presents a map of period coverage classification for the 71 LGAs surveyed.


Table 7: Coverage classification per LGA

State LGA SAM cases found (n) Covered SAM cases (c) Decision rule 1 $ c > d_{1} \text ? $ Decision rule 2 $ c > d_{2} \text ? $ Coverage classification
Kebbi Arewa 66 5 13 No 33 No Low
Argungu 129 11 25 No 64 No Low
Augie 124 18 24 No 62 No Low
Birnin Kebbi 108 18 21 No 54 No Low
Bugudo 65 8 13 No 32 No Low
Kalgo 110 9 22 No 55 No Low
Koko Besse 64 8 12 No 32 No Low
Sakaba 50 3 10 No 25 No Low
Shanga 59 9 11 No 29 No Low
Suru 65 8 13 No 32 No Low
Sokoto Tangaza 83 9 16 No 41 No Low
South Sokoto 46 6 9 No 23 No Low
Illale 73 8 14 No 36 No Low
Gude 129 25 25 No 64 No Low
Goronyo 180 1 36 No 90 No Low
Saborn Birnin 138 9 27 No 69 No Low
Garda 248 5 49 No 124 No Low
Zamfara Bungudu 78 10 15 No 39 No Low
Birnin Magaji 108 11 21 No 39 No Low
Bakura 111 34 22 Yes 55 No Moderate
Maradun 336 48 67 No 168 No Low
Shinkafi 77 20 15 Yes 38 No Moderate
Tsafe 56 8 11 No 28 No Low
Katsina Mashi 130 31 26 Yes 65 No Moderate
Mani 81 22 16 Yes 40 No Moderate
Daura 66 31 13 Yes 33 No Moderate
Dutsi 74 32 14 Yes 37 No Moderate
Zango 115 40 23 Yes 57 No Moderate
Baure 134 68 26 Yes 67 Yes High
Sandamu 59 20 11 Yes 24 No Moderate
Ingawa 48 16 9 Yes 24 No Moderate
Batsari 81 37 16 Yes 40 No Moderate
Kaita 36 17 7 Yes 18 No Moderate
Kano Bichi 53 14 10 Yes 26 No Moderate
KMC 43 20 8 Yes 30 No Moderate
Madobi 61 20 12 Yes 19 No Moderate
Sumaila 38 17 7 Yes 19 No Moderate
Tsanyawa 58 24 11 Yes 29 No Moderate
Wudil 81 46 16 Yes 40 Yes High
Gombe Gombe 50 2 10 No 25 No Low
Dukku 81 22 16 Yes 40 No Moderate
Nafada 84 21 16 Yes 42 No Moderate
Jigawa Babura 74 37 14 Yes 37 No Moderate
Birnin Kudu 78 17 15 Yes 59 No Moderate
Birniwa 49 10 9 Yes 24 No Moderate
Guri 64 26 12 Yes 32 No Moderate
Jahun 88 14 17 No 44 No Low
Kaugama 45 16 9 Yes 22 No Moderate
Kazaure 54 13 10 Yes 27 No Moderate
Kiyawa 43 7 8 Yes 21 No Low
Roni 50 18 10 Yes 25 No Moderate
Gwiwa 105 64 21 Yes 52 Yes High
Yankwashi 45 14 9 Yes 22 No Moderate
Maigatare 56 12 11 Yes 28 No Moderate
Bauchi Damban 91 43 18 Yes 45 No Moderate
Katagum 65 44 13 Yes 32 Yes High
Kirfi 85 28 17 Yes 42 No Moderate
Adamwa Song 103 50 20 Yes 51 No Moderate
Guyuk 62 28 12 Yes 51 No Moderate
Mubi North 75 36 15 Yes 37 No Moderate
Borno Biu 61 9 12 No 30 No Low
Askira Uba 71 28 14 Yes 35 No Moderate
Yobe Damaturu 95 29 19 Yes 47 No Moderate
Fika 61 28 12 Yes 30 No Moderate
Fune 65 15 13 Yes 32 No Moderate
Geidam 86 7 17 No 43 No Low
Machina 51 15 10 Yes 25 No Moderate
Nguru 145 47 29 Yes 72 No Moderate
Postikum 43 10 8 Yes 21 No Moderate
Yusufari 81 16 16 Yes 40 No Moderate
Yunusari 80 20 16 Yes 40 No Moderate


Figure 4: Map of point coverage classification across the 71 LGAs surveyed
Point Coverage


Figure 5: Map of period coverage classification across the 71 LGAs surveyed
Period Coverage


Coverage estimates

Coverage estimation was done at the state level. Table 8 presents the results. Bauchi has the highest coverage estimate at 56.9% while Sokoto has the lowest coverage estimate at 5.3%.


Table 8: Coverage estimates per state

State SAM prevalence4 Coverage estimation 95% CI
Kebbi 1.2% 12.6% 11.7%-13.4%
Sokoto 1.3% 5.3% 4.0%-6.6%
Zamfara 1.3% 11.4% 9.9%-12.3%
Katsina 5.4% 37.9% 32.2%-41.9%
Kano 3.9% 41.8% 35.0%-48.6%
Gombe 0.07% 14.5% 10.2%-18.9%
Jigawa 3.7% 30.9% 27.6%-34.3%
Bauchi 2.5% 56.9% 49.9%-64.4%
Adamawa 0.04% 48.0% 41.5%-54.5%
Borno 2.2% 31.4% 23.2%-39.6%
Yobe 1.5% 26.6% 24.5%-28.8%

Chi-square test performed per state indicate that only Adamawa and Kebbi state had homogeneous coverage across the LGA surveyed5. This means that the overall estimates for these two states (12.6% and 48% for Kebbi and Adamawa respectively) most likely is the coverage across all the LGAs providing CMAM services in the two states. For the rest of the states, however, the overall estimates should be taken into context given within-state variability of coverage as shown by the per LGA classification in the previous section and the chi-square testing performed.

The overall point coverage for all states is 36.6 % (32.3 – 40.9%). Again, this result should be taken in context of high variability of coverage across the LGAs.


Barriers to service uptake and access

A questionnaire was administrated to mothers of SAM cases children who were not in program, in order to identify the barriers of the program. To have a good answer to question, the terminologies of malnutrition in local language were used by surveyors to facilitate understanding of the questions to the mothers.

Overall barriers to service and access for the 11 states
The Pareto chart in Figure 6 shows the overall barriers for the program of all states.


Figure 6: Overall barriers

Discussion on overall barriers found
The results of the survey showed that the two main barriers of the program in 11 states are maternal ignorance of malnutrition and lack of knowledge of CMAM program. These two causes remain the two main barriers to coverage of the program as shown in the other programs in other countries6.

Regarding ignorance of malnutrition, the majority of mothers did not know that their children were having severe acute malnutrition despite the use of local terminology in asking questions, they were thinking that it was their normal stature. Thus if a mother does not know that her child was malnourished it was not possible for them to bring her to the program for treatment. This situation calls for an awareness of malnutrition should be carried out in all states.

Ignorance of CMAM program and ignorance of how the program works come respectively in second and third position. Regarding knowledge about the program, the CMAM program in Nigeria is integrated into the activities of the primary health center and not as an emergency activity. Hence according to the fact that we found ignorance of the program like a barriers, raising awareness about the program should also be strengthen. Concerning the barriers about ignorance how the program works, the major barriers were misconceptions concerning the program and the fear of being rejected by the program.

In misconceptions of program, belief in traditional medicine comes first and the fear of being hospitalized in second. These two major barriers on misconceptions of the program and the fear of being rejected calls an important part of sensitization about the program in community, should focus on how the program works. The following barriers are related to constraints by mother. In the community, the mother is often the key person who is responsible for bringing the child into the program and all barriers in general are focused on her. Thus, these barriers regarding constraints and responsibility of the mother in society include the following impediments due to maternal health and also impediments related to the cultural position. In this section also, the refusal of Husband was a major barrier to the coverage due to the cultural position of the mother.

Regarding health service, stock out of RUTF was a major cause of barrier to coverage because without input, a CMAM program cannot operate. Other barriers related to health service as shown in the table are inappropriate treatment (lack of RUTF), attitudes and negative attitudes or inappropriate advice of health workers. Regarding geographical accessibility, distance and transportation problems were significant barriers and just four people have mentioned the problem of insecurity, this was because surveyors went only to places which were safe.

Coverage barriers for each state
The summaries of barriers to the program of different states are represented in the Figure 6 to 16. SQUEAC investigations which are going to be carried out in these state will provide detailed explanation of these barriers.



The coverage in the northern states of Nigeria is moderate with an overall estimate of 36.6% (95% CI: 32.3% – 40.9% ).

Coverage classification was performed for each LGA, in which we have found 27 LGAs having low coverage, 40 LGAs with moderate coverage and 4 LGAs with high coverage. Barriers to coverage have been reported for each state. The survey coverage has allowed under capacity building, to train agents of MOH and NBS on the SLEAC methodology.



According to overall barriers found by SLEAC coverage survey, some preliminary recommendations have been formulated which can be applied to all programs of each state:


Emphasis on community mobilisation
  • Sensitization
  • Awareness through voluntary on malnutrition should be focused on the knowledge of the early signs of malnutrition, consequences and also prevention. This awareness can be performed in several ways, through posters, radio broadcasts etc.

    Awareness about malnutrition should also include awareness about the program, how it works, admission criteria, explaining the phenomenon of rejection, the advantage of management of SAM cases in the program compared to traditional treatment etc.

  • Include community leaders on sensitization of CMAM program activities
  • The community leaders as village chief, religious leaders should be involved in sensitization of the program to facilitate the acceptance of program by community. It is also important to educate traditional healer in CMAM activities so they can participate in referring cases. Husbands, chiefs and families are also key persons to be involved in this awareness to facilitate greater ownership of the program by households

  • Active case finding by volunteers (monthly)


To ensure that each village has a volunteers with a MUAC tape

For each LGA with CMAM activities, it is important to ensure that each village has a volunteer who performs screening activities at least once by month and each volunteer should have a MUAC tape

  • Perform regular refresher training for volunteers, consider incentives.
  • For the volunteers, it is important to conduct regular refresher training as an incentive


Service delivery
  • Strength integration CMAM program in routine activities of Health center
  • In other activities of the Health Centre (consultations, immunization activities, etc.) it would be important to integrate the screening cases MAS

  • Harmonize visit of mothers with many constraints with schedule of CMAM program
  • For mothers who have many constraints it is important to discuss with them to harmonize a specific program of visits to the health center.

  • Avoid stock out of RUTF of each program


Access issues

Organize mobile treatment to far villages and give RUTF for two weeks for people who live far from the health center


Implement the program treatment of moderate acute malnutrition

In the active case finding of the SLEAC survey, several cases of moderate acute malnutrition have been identified and it is essential and important to set up a program of support for moderate acute malnutrition to prevent relapse of severe acute malnutrition after being discharged


Conduct SQUEAC investigations

Perform SQUEAC investigations at least one by state in order to understand different barriers and boosters and provide strong and evidence-based recommendations for the program


1 As cited in ‘Commission Decision on the financing of humanitarian actions in West Africa from the 10th European Development Fund’. European Commission, 2010

2 ACF, Save the children, Valid, MOH Nigeria, report on Assessing Coverage of CMAM Services in Nigeria a& Building Government Monitoring Capacity, 2013.

3 State-level and overall coverage estimate aggregates are only meaningful if coverage across the LGAs within a state are not significantly different from each other.

4 Prevalence of MUAC SAM of SMART survey 2013

5 Although Sokoto had low coverage classification across all its LGAs, one LGA stood out as almost being classified as moderate and has a significantly higher coverage than all other LGAs in Sokoto. This is most likely the reason why based on chi-square testing, Sokoto was assessed to have heterogeneous coverage.

6 Guerrero, S., Myatt, M. and Collins, S., Determinants of coverage in community-based therapeutic care programmes: towards a joint quantitative and qualitative analysis, August 2009, Disasters. 34 (2), pp 571-585.