Menu ENN Search
Language: English Français

% of Stunting prevelence or Mean WHZ score

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


Normal user

5 May 2015, 12:24

We are developing one programme one nutrition sensitive social protection programme on maternity cash transfer with intensive behaviour change communication. Below were the some points for project information and I have some question around the outcome indicators on our research programme.

Activities: maternity cash transfer (3 mth gestation to child 2 year of age) + BCC
Enrollment criteria for cash transfer: only pregnant mother (within 3 mth gestation to 9 mth gestation)
Graduation criteria: when their child reach 2 yrs age
project period: 3 years
Estimate beneficiaries: approximately 5000 mother and child pair (per arm) will be enrolled in 3 yrs project period

Regarding research, we have tow arms to test our hypothesis - there was a different between the impact of Cash + BCC and BCC only on stunting rate?
Arm 1: 191 villages - received cash transfer and BCC
Arm 2: 191 villages from different township but has same pop characteristic as arm 1 - only received BCC

But, we have limited budget to cover the robust M&E. So, in that case, what will be the best case of research methodology to apply and what will be the best outcome indicators to test the difference in both arms (stunting prevalence or mean WHZ score) or how about the power on detection of difference outcome indicators.


Mark Myatt

Frequent user

6 May 2015, 08:52

OK. Here is some "hand-waving" to get this started ...

I assume:

(1) You will compare your outcomes in "Arm 1" with your outcomes in "Arm 2" at an end point.

(2) Your outcome is some measure of stuntedness and when you say "mean WHZ score" you mean "mean HAZ score".

(3) The end-point is "graduation" at two years of age. You will collect data (e.g. height and age) at or near "graduation".

You can use either prevalence HAZ < -2 of mean HAZ for your outcome.

Using means may be more efficient. The difficulty with using means is that we often want to work with prevalences as they can be easier to understand.

It is not too difficult convert between means and prevalences when the quantitative indicator (e.g. HAZ) is quasi-normal. If we assume that the mean HAZ is -1 with SD = 1.1 we would expected about:

  > pnorm(q = -2, mean = -1, sd = 1.1) * 100
  [1] 18.16511

That is 18.2% with HAZ < -2.

If we manage to shift the mean HAZ upwards to -0.7 with SD = 1.1 we would expect about:

  > pnorm(q = -2, mean = -0.7, sd = 1.1) * 100
  [1] 11.86389

That is 11.9% with HAZ < -2. That is equivalent to a reduction of about 6.3% in the prevalence of stunting.

I used R above but the calculation can be done in (e.g.) Excel using:

  =100 * NORMDIST(-2, -1, 1.1, 1)       <- yields 18.165107
  =100 * NORMDIST(-2, -0.7, 1.1, 1)     <- yields 11.863892

You can play with these calculations (using available survey data to get expected means and SDs) to finds the difference between means that corresponds to a desired difference in prevalence. I will work continue by using what I have done above.

Ignoring design effects (intra-cluster correlations), a simple sample size calculation to detect this difference with 90% power at the 5% significance level would be:

  n = ((1.28 + 1.96)^2 * (1.1^2 + 1.1^2)) / ((-1) - (-0.7))^2
  n = 282

children in each arm.

You have a complex design. This means that you will likely need a larger sample size (e.g. double what you calculate using standard formulae would probably be OK) and you will need to analyse the data using appropriate methods (these are available in most statistical packages). You will need to weight within-community results by population size.

The sample size will also need to be increased to account for losses to follow-up (LTFU). About 20% LTFU is probably OK. The sample size would then be:

  n = 2 * (282 * 1.2)
  n = 677

in each arm.

What you have above suggests that you will have no problems reaching this sample size. In that case you could collect data from about:

  m = (677 / 5000) * 191
  m = 26

communities in each arm which could be selected at random. I would increase this to at least m = 30. That will reduce costs (i.e. you will only need to sample is m = 60 communities rather than m = 382 communities).

There are other designs. A longitudinal design (e.g) will need fewer children but you will need to measure them more often (e.g. quarterly over two years (that is 8 measurements) rather than once at graduation. This may be more expensive and data management and analysis will be complicated. Sequential trials are robust and very efficient and will typically cut the required sample sizes in half. Such trials are not commonly used in the emergency nutrition field. If you decide to go for a sequential design then you should consult an experienced statistician.

What I have here is "rough and ready" and is intended only to illustrate how you could proceed. These types of trial can be very difficult to get right. You may want to consider consulting a statistician with experience with clustered trials.

I hope this is of some use.


Normal user

23 Jan 2016, 19:18

Thanks Mark.

Another question, which also related to above question.

In this project scenario, I'd like to see the difference of stunting status between sibling from each beneficiaries (both control and intervention). Because, I assume, we might got certain amount of Pregnant mother who have under five children from our program beneficiaries. Those children were not received any of our program intervention, but now the new baby will start received benefit (almost cover his window of opportunities for proper development). So, I assume the stunting status of new baby who bot our program benefits and his/her elder brothers/sisters were not.

So, in this case, how can we estimate how much # of pregnant mother who have under 5 children and what will be the best estimation of sample size need to compare these two community for study. In more detail, when the program endline perform children of pregnant mother who are registered to program from initial enrollment (enrollment will start at April 2016 and cash transfer end at Oct 2018) were age interval between 24 months - 30 months. So, in this case what will be the meaningful age group to compare the stunting status of their older sibling and that group.

thanks in advance.

best regards,

Mark Myatt

Frequent user

25 Jan 2016, 11:33

This is a "quasi-experiment" in the sense that there is no randomisation between control and intervention groups with the controls just "found". Such studies need to be interpreted with care as bias and confounding can easily creep in. In the case (e.g.) of a secular trend of reducing stunting you may see a difference in favour of the intervention group even when an intervention effect was absent.

You have identified a potential bias with age. Ideally you would want to compare children at the same ages. If (e.g.) you had a pair with the control ages 33 months at at recruitment then the comparison would be with the intervention child at (or very close to) 33 months. In many cases stunting occurs before about 30 months and the child remains stunting for some time after that (i.e. until the pre-adolescent growth spurt). You may, therefore, want to compare measurements at 30 months (intervention) and >= 30 months (control). Note that age can be subject to considerable measurement error and this will likely effect the control group more than the intervention group.

Data-analysis should be straightforward. Something like a paired t-test on HAZ should do the job.

I'd be tempted to choose the simplest design. You may find the quasi-experimental design a bit complicated to run.

I hope this is of some help.

Back to top

» Post a reply