Menu ENN Search
Language: English Français

Calculating beneficiary numbers of SAM/MAM using prevalence

This question was posted the Assessment and Surveillance forum area and has 3 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Ali Maclaine

Normal user

19 Jan 2011, 18:41

Hi, I am trying to work out estimated beneficiary numbers using prevalence and incidence for SAM and MAM.
For SAM as I understand it you take the total population of your area e.g. 450,000, then take estimate of under five population 20%. We have the prevalence 3.1% that gives 2790 (450000*20%*3.1 = 2790). But we know that the actual number of children to treat every year is two to three times higher than prevalence. Is it 2 times higher or 3 times higher?
Assuming it is 3 then I have been told different methods of calculating it, is it
Using example above 2790+ (2790*3) = 11160 OR again using example above 2790*3 = 8370 ?

Then how do you do this for MAM?
Confused of London!

Rogers Wanyama

Emergency Nutrition Specialist

Normal user

20 Jan 2011, 05:17

Hi Annonymous 169
A similar question was posted in 2009. Log to the link below.
Hope it helps

Anonymous 81

Public Health Nutritionist

Normal user

20 Jan 2011, 07:53

Dear Annonymous 169

To add on Rogers. you can also find similar question with reply from the following link.


Mark Myatt

Frequent user

20 Jan 2011, 09:54

Others have pointed out:


One approach is to use the formula:



POP = Total population
U5% = Percentage of total population age 6-59 months
EP% = Estimated prevalence
CFI = Correction factor to estimate incidence from prevalence

The value of CFI is uncertain (see the links above). A CFI of 2.0 (in this formula) is broadly in-line with published estimates. This level of CFI is used for SAM. I do not think we have a good idea of the value of CFI for MAM (so I would use 2.0).

The problem with:


is that it takes a "fairy tale" view of programming in the sense that it assumes a coverage proportion of 100%. I have looked at coverage of TFC, OTP (in both CTC and CMAM guises) and SFP and have never seen coverage above 89%. Here are some rules of thumb for different program types:

TFC : Typical range 0.5 - 5% (maximum seen is c. 30%)
OTP : Typical range 20% - 80% (minimum 8%, maximum 89%)
SFP : Typical range 5% - 20% (limited data available)

We have to face it ... most programs achieve coverage below SPHERE minimum standards.

The point is that we need to account for coverage in the formula:


Using your data and assuming the program will hit the SPHERE minimum of 50% we have:

EXPECTED = 450000 * 20% * (3.1 / 100) * 2.0 * (50 / 100)

Another approach is to use the formula:

EXPECTED = POP * U5% * EP% + (POP * U5% * EP% * CFI)

That is prevalent cases + incidence cases.

A value of 1.6 is used for CFI (this is a published estimate). This level of CFI is used for SAM. I do not think we have a good idea of the value of CFI for MAM (so I would use 1.6).

This formula also fails to account for coverage. A better formula is:

EXPECTED = POP * U5% * EP% * IC + (POP * U5% * EP% * CFI * AC)


IC : Initial phase coverage (often low)
AC : Achieved coverage (i.e. after the first few months)

It is sensible to use:

IC = AC / 2

as the average between starting at zero and achieving 50% some time later.

Using your data and:

AC = 50%


IC = 50% / 2 = 25%

we get:

EXPECTED = 450000 * 0.2 * 0.031 * 0.25 + 450000 * 0.2 * 0.031 * 1.6 * 0.5

The two methods give similar answers.

You have to be aware that there are big sources of error in both of these approaches:

POP : Subject to secular change but also displacement and migration
U5% : Subject to secular change &c. and public health shocks
CFI : An informed guess base on limited data
EP% : For SAM this will be very imprecise (e.g. 1.15%, 95% CI = 0.38%; 2.57%).

And COVERAGE is not known. We have to be realistic about what we will achieve. I'm sure that the agencies that "achieve" 8% coverage started out thinking they would get 80% coverage.

BEWARE : You need to use EP% for your program admission criteria. If you use MUAC then EP% is for the MUAC case-definition not the W/H case-definition.

So ... a short answer ... there is no really correct way. There are different ways of getting an informed guess which, given the same assumptions, give roughly the same answers.

I hope this helps.

Back to top

» Post a reply