Hospital-based studies/ surveys can tell you little about incidence / prevalence in the community. This is because a selection bias (often a set of selection biases) occur.

You should also check for bilateral pitting oedema.

Dear Iyke:

In answer to your questions:

1. Absolutely not! Hospitalized children are hospitalized because they are ill. They do not represent in any way children in the community. In order to assess the nutrition status of children in the community, you must select a RANDOM sample of children from the community.

2. Regarding other parameters, what do you want to measure? The anthropometric measurements you list are the most commonly used to measure child nutritional status; however, there are a zillion other measurements which have been proposed, most of which are not feasible in field data collections. But if you are thinking of measuring things other than nutritional status, of course there are other indicators.

I hope this helps.

Dear Sir
From a personal experience in two different counties in my country the nutrition status clients you see per specified period does reflect what is at community level. The higher the numbers the more the probability that its a community problem.Because referrals are done from a community.

This is for both inpatient and outpatient.

The severity of the problem is indicated by the number of inpatient. This can be seasonal or continuous.

The other parameter that we use and it helps explain the nutrition status is the Dietary diversity score(DDS) that looks at the number of food groups a child has has consumed in a specified period. Is assesses if a child is getting all the required nutrients.

Dear Iyke,

I truly wish it was as easy as that. Hospital based childhood nutritional assessments could theoretically reflect the situation in the community, but there so many factors that can bias the information that it cannot be relied upon. Here are a few of the problems I've seen with this data.

1) Even if you have people in the communities trained to detect children who are malnourished, not all of these people will be as active in referring the children. You may end up with more referrals from some communities simply because those trained screeners are more active.
2) Being farther away from a hospital increases the demands on the family taking their child to the hospital, so you may get a higher percentage of the cases from nearby communities and a lower percentage from communities farther away. The nutritional situation may be different in the more distant communities.
3) Even if the care for malnourished children provided by the hospital is free, there is an opportunity cost for the caregivers in time and travel. The poorest families, who are often the most likely to have malnourished children, are sometimes the least likely to take their children for treatment because they cannot stop working long enough or don't have money for the transportation. This is especially important in rural areas were distances may be longer and transportation relatively more expensive.
4) If the supply pipeline is irregular, changes in admissions are often more reflective of the breaks in supply than from changes in actual prevalence.
5) Depending on where you are, one type of malnutrition (kwashiorkor or marasmus) may be considered serious and the other not, which may change the likelihood of whether or not parents bring children to the hospital.
6) Some malnutrition may be linked to illness. The parents may bring to hospital children with a fever, but not think to bring a child that is malnourished but appears to be otherwise healthy. This could give you the idea that malnutrition is mostly due to illness, when perhaps there are more cases caused by local childcare practices.

I am sure some of our colleagues could come up with other factors that could affect the hospital-based data.

For the survey parameters, as Mark suggested, testing for bi-pedal edema should be included. Sometimes people include head circumference for small children, but I don't see this as often.

Good luck

Good morning!
What data tools will be best to collect nutritional status data and analyse same with relative ease amongst a patient population in the hospital or community?
Thanks.

I agree with all the above comments, children at hospitals are generally nothing like the background community. As well as oedema, remember to include infants under 6 months.

An easy tool is Epidata: https://www.epidata.dk/

The most important thing is that collecting a small amount of data that is complete and measured accurately is far better than having many variables or subjects but missingness or data quality problems. It is easy to be over-optimistic. First, write down the exact questions you want to answer, exactly which population you are including and excluding, then decide who is going to measure and ask questions, train them fully and ensure they are measuring according to standard procedures from survey guidelines, then try a pilot.

Jay