Menu ENN Search
Language: English Français

Any robust experiences using LQAS for assessing and monitoring nutrition programs using Valadez's common health survey method?

This question was posted the Assessment forum area and has 5 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Wendy Dyment

Sr. Health and Nutrition Advisor, Medair

Normal user

23 Jun 2011, 15:32

I have searched the ENN forum and followed past LQAS survey discussions on this site; however, as a health person, I am wondering why I have not been able to find widespread use and recommendation of the method of LQAS frequenlty used for baseline surveys in monitoring in health programs. Specifically the selection of 19 households in at least 5 supervision areas as outlined by Joseph Valadez in 2001 in "Using LQAS for Assessing Field Programs in Community Health in Developing Countries" and also widely publicized by the CORE group. I've seen similarities in SQUEAC and saw reviews such as FANTA using much larger sample sizes such as 67 x 3 and 33 x6. I've also seen it's use with IYCF indicators. However the simplicity and benefit of a baseline survey and ongoing monitoring of coverage, prevalence, and other useful indicators integrated with regular supervision of decentralized OTPs or SFPs making supervision areas, I have not seen discussed within the nutrition sector or this forum. (I realize it cannot be used for mortality.) So any robust experiences? Did I miss something in the nutrition sector? Or any good reasons why it should not be adapted here as well? We've used it with some of our integrated health and nutrition programs and found it useful, simpler and felt as confident with the results as some of our classic 30 x 30 surveys.

Mark Myatt

Consultant Epideomiologist

Frequent user

15 Aug 2011, 12:59

I frequently use LQAS in other fields (e.g. trachoma prevalence, program coverage, HIV drug resistance, TB drug resistance, &c.). I have no clear idea why it has not taken off in for prevalence of wasting. I have some ideas ...

(1) The method offers little savings over a SMART survey when you spend a lot of time travelling to cluster locations.

(2) The method is not ratified by SMART.

(3) The method offers less than SMART. I think that LQAS is most useful when you want to do things with small samples and, for me, that usually means in small areas or rare populations. I can imagine using LQAS to map prevalence or risk but using it to get a wide-are classification seems (to me) to not be playing to the strengths of the method.

Just my tuppence.

Tariq Khan

Normal user

4 Oct 2011, 13:27

Dear all,
Can anyone tell me some differences between SMART methodology and LQAS. I know and implement SMART methodology in Afghanistan but still want to know the highlighted differences between SMART and LQAS....

Thanks.

Mark Myatt

Consultant Epideomiologist

Frequent user

5 Oct 2011, 14:28

First ... SMART is a survey method and LQAS is a data-analysis technique. LQAS is a classification technique that can tell you if (e.g.) prevalence is above or below a critical value using a small sample size.

In the emergency nutrition field "LQAS" usually means a way of analysing data from a cluster-sampled survey (like SMART) with a small sample size (e.g. 33 clusters of 6 kids). This is just like a SMART survey but with fewer kids sampled from each cluster.

Both the SQUEAC and SLEAC coverage survey methods use LQAS techniques to classify coverage levels (SLEAC) and test hypotheses about coverage (SQUEAC).

Tariq Khan

Normal user

28 Jan 2013, 20:46

Dear Mark,
What is the difference between SMART methodology and the one used for National Nutrition Survey?
And, can we use SMART methodology for a National Nutrition Survey?
Explaination will be appreciated..........

Thanks.

Mark Myatt

Consultant Epideomiologist

Frequent user

29 Jan 2013, 03:40

Different countries will use different methods for national nutrition surveys. One approach would be for national survey to be made of of a patchwork of SMART surveys. The constituent SMART surveys could cover single districts or single livelihood zones. National estimates could be made using a population weighted average of the SMART survey results. A national survey would usually not have any spatial gaps. Indicators could be mapped at the level of individual SMART surveys. Depending on the indicators you need ... it may be possible to replace SMART surveys with smaller, cheaper, and faster surveys.

Another approach would be to use a spatial sampling method such as S3M. The advantage of this approach is that it allows finer scale mapping of indicators than would be practicable with the SMART approach.

I hope this helps.

Back to top

» Post a reply