Enable low bandwidth mode Disable low bandwidth mode
FEX 28 Banner

Locally produced RUTF in a hospital setting in Uganda


By Tina Krumbein, Veronika Scherbaum, and Hans Konrad Biesalski

Tina Krumbein is a graduate nutritionist. This article forms part of her diploma thesis submitted to the Department of Biological Chemisty and Nutrition, University of Hohenheim, Germany.

Veronika Scherbaum holds a MSc degree in Mother and Child Health Nutrition. Since 1998, she has been a lecturer in Nutrition in Developing Countries at University of Hohenheim, Germany. In the 1980s she worked for several years in Ethiopia as a nutritionist. More recently, she has been involved in consultancies (mainly for evaluation of SFPs, TFPs and breastfeeding promotion) in Afghanistan, Iraq, Zimbabwe and Darfur.

Hans K Biesalski is director of the Department of Biological Chemistry and Nutrition. He is actively involved in research on vitamin A metabolism and is supervisor of studies in International Nutrition. He is a member of IVACG and chair of a couple of studies dealing with Vitamin A Deficiency Disorders (VAD).

A mother spoonfeeds her baby in the NU

This pilot study was financially supported by the Eiselen Foundation, Ulm and the German Medical Missionary Team.

This article describes some of the preliminary findings of a recent study that introduced locally produced Ready to Use Therapeutic Food (RUTF)1 into a hospital setting in Uganda, comparing its use to F100 during the rehabilitation phase. This article focuses particularly on the experiences producing local RUTF in a hospital setting, including cost comparisons with the routinely used F100.

Kumi Hospital (formerly Kumi Leprosy Centre) is located in Kumi district, about 8 km east of Kumi Town in Eastern Uganda. Established as a general hospital in1997, most of the services and programmes of Kumi Hospital are donor funded. In 1996/97, a poor regional harvest led the Medical Superintendent of Kumi Hospital to request a nutritionist from the German Medical Missionary Team (GMMT) to help with the construction of a Nutrition Unit (NU). The NU was subsequently established in September/October 1998.

Development of the Nutrition Unit (NU)

Initially the NU treated severely malnourished children in the rehabilitation phase, with stabilisation and initial care offered in the paediatric ward of the hospital. The NU also provided food (NU diet, see box 1) for malnourished children who were admitted to other wards or could stay near the hospital. At this stage, other ward staff provided medical care to the children. To improve the care and monitoring of the children, especially at night, and to avoid absence of the mothers/patients on some days, a 16-bed extension to the NU was opened in March 2001, where the children could be accommodated with their caretaker. In May 2004, the NU became a ward with its own employed nursing staff and a dedicated doctor for daily supervision. Severe clinical cases (e.g. those who required naso-gastric feeding) continued to be first admitted to the paediatric ward, with transfer to the NU once stabilised. Malnourished children without obvious medical complications were directly admitted to the NU ward.

Food storage room in the Nutrition Unit

All children admitted to the NU are treated according to WHO's 10-step-guidelines (1999). F75 (during the initial phase) and F100 (for the rehabilitation phase) are produced from fresh cow's milk, oil, sugar, and vitamin-mineral complex (CMV therapeutic). The majority of children are weighed daily and weight gain/kg/day was calculated.

As well as treating severe malnutrition, the NU is also actively involved in:

  • Identification of malnourished children in the area through home visits or immunisation outreach work.
  • Teaching caretakers about basics of nutrition, hygiene and primary health care. Caretakers also worked in the garden of the NU growing some of the food for the programme like groundnuts, vegetables and fruits.
  • Counselling of the caretakers in the NU and during home visits, by a trained social worker.
  • On the job training and supervision of NU staff.

Since the NU began in 1998, almost 1000 patients have been admitted. Additionally the NU has taken care of outpatients enrolled in the feeding programme but not admitted to the NU.

Study objectives

The main objectives of the study were:

  • To compare weight gain, duration of stay and other parameters of two groups of children, one group receiving locally produced F-100 and the other locally produced RUTF, during the rehabilitation phase.
  • To develop a suitable procedure for small scale local production of RUTF.
  • To develop recommendations for use in the NU on how to produce and administer RUTF.
  • To assess and compare prices of F-100 and RUTF (both locally produced)

The study was carried out in the NU between September and December 2005.

Box 1

Preparation of the NU-diet

The NU provides locally available food for the patients (porridge, lunch, fruits, porridge and supper) several times daily and one meal per day for the caretakers. The caretaker meal consists of posho (a stiff porridge made from maize flour) and red kidney beans. Every 3-4 weeks, two staff members go to Kumi town or to Mbale (25 km southern of Kumi) and buy food. It is necessary to hire a vehicle from the hospital for this.

Preparation of therapeutic diets

The therapeutic diets (F75, F100, ReSoMal) for all malnourished patients, including those on other wards, are prepared in the 'staff-kitchen' of the NU by one member of the staff, usually the nutritionist. When the nutritionist is not on duty, other staff prepare the diets. Therapeutic milks are prepared in the morning as soon as fresh cow's milk4 and water are boiled. Oil, sugar and CMV therapeutic are used to prepare therapeutic milk according to recipes hung up in the kitchen. The therapeutic milks are given out to the caretakers once a day in the morning and they are told to boil the milk again before giving it to the child. Each mother has to bring a plastic jug (they are commonly used and easy to obtain in the area). The name of the child is written on the jug and the diet is given out according to the calculated amount for the child. ReSoMal is also prepared in a jug, once in the morning and again during the day if necessary.

Every day, the staff give out food for the caretakers and the children who receive the NU-diet. The caretakers prepare their own lunch and the meals for the children on firewood or charcoal stoves in a kitchen outside the NU-building.

Food storage

One room in the NU is used to store food. This room is also the kitchen for the staff. Posho, beans, groundnuts, rice and sugar are kept in bags on the floor. Fruit and vegetables are covered and kept either on shelves or in the fridge. Flour (soya, rice, millet), fresh peas and green grams are kept in buckets on shelves. Vegetable oil is bought and kept in 20 litre cans.

Study outline and preliminary findings

During the first weeks of the study, children entering the rehabilitation phase (after their appetite returned) were randomly divided into two groups, one group receiving F-100 and the other local RUTF. However, some children did not like the taste or could not manage2 RUTF during this early stage of rehabilitation and so random allocation was abandoned. Instead, children whose appetite had improved and who liked the taste of the RUTF were managed using RUTF with small amounts of the NU diet3. These children (n=10) were compared to those who had selected to receive either F-100 exclusively for several days first or immediately a combination of F-100 with small amounts of the NU diet (n=5). All three children who were suffering from oedematous forms of malnutrition on admission received F-75 during the initial phase. At the beginning of the rehabilitation phase, two kwashiorkor cases chose the F- 100 group and one opted into the RUTF group. Thus self-selection determined the majority of assignments to the F100 or RUTF group, based on the appetite and food preference of the malnourished child.

Weight gain

Weight gain averaged 9.7g/kg/d in the F-100 group and 7.3 g/kg/d in the RUTF group. The mean duration of stay was 32.6 days in the F- 100 group and 28.5 days in the RUTF group. On admission, all children in the F-100 group were severely wasted (W/H z-score <-3), compared to only 22% in the RUTF group. The fact that more severely wasted children self-selected for F-100 than RUTF needs further investigation.

Components of local RUTF

Local production of the RUTF took place in the kitchen of the NU. A procedure for production was developed during August/September 2005 based on an Instruction Manual for local Production of Plumpy'nut written by Nutriset.

As there were only a few5 children receiving RUTF, only small amounts had to be produced each time - 250 or 500 g. A 500g portion was based on peanut paste (125 g), full cream milk powder (150 g), vegetable oil (75 g), sugar (140g), and vitamin-mineral mix (CMV Therapeutic; 8g). All ingredients for RUTF (See box 2) are available in Kumi or Mbale throughout the year. Although seasonal prices vary (especially sugar), ingredients remain available and their prices within the means of the NU.

Initial production trials

Preparation area for therapeutic feeds in the NU kitchen.

The method for local RUTF production is outlined in box 3. In the first trials in the NU, the oil was not heated however this unfavourably altered the taste of the product. Furthermore, local staff feared that uncooked oil might cause diarrhoea in some children. Therefore the oil was subsequently warmed.

Initially the sugar was not ground which made the product very crumbly and difficult to eat. The sugar was subsequently ground before mixing with the milk powder, which made the product smoother and more palatable. It was feasible to grind all the sugar because there were only small amounts for each production cycle (140 g sugar for 500g RUTF).

In the beginning, an electric mixer, which was brought from Germany, was used to stir the ingredients. Subsequently, mixing was achieved through using a spoon and mashing and shaking ingredients in a closed container until the mixture was a paste.

This proved to be as effective as the electric mixer and was independent of electricity - an important consideration in an area where power cuts are common.

Training staff

After learning about RUTF production, the NU staff members had a chance to ask questions about the product and the production. A sheet with basic information about the product and quantities required for children was included with other recipes in the NU kitchen.


Table 1 Cost-comparison of 1000kcal equivalents of F-100 (using fresh cow milk or full cream milk powder) and local RUTF
Fresh cows milk Full cream milk powder (FCM) Local RUTF
Amount Cost Amount Cost Amount Cost
880 ml fresh cows milk 440 Ush 110g FCM 1320 Ush 57g FCM 784 Ush
75g sugar 113 Ush 50g sugar 75 Ush 53g sugar 80 Ush
20g veg oil 35 Ush 30g veg oil 53 Ush 29 ml veg oil 29 ml veg oil 46 USh
Half a msp* CMV 110 Ush Half a msp CMV 110 Ush 3g CMV 106 Ush
        48 g peanut paste 83 Ush **
Cost per litre: 698 (32 cent) Cost per litre: 1558 (71 cent) Cost per 190g:*** 1004 Ush (46 cent)

Costings calculated in Ugandan Shillings (Ush) and euro (cents).
*Measuring spoon included with CMV therapeutic
** When bought at the local market
*** Costing calculated from weights and prices to produce 500g RUTF (2625 kcal): where cost/500g divided by 2.63 to produce cost/190g RUTF (1000 kcal)


Box 2

Peanuts (or groundnuts) are very common in Kumi and are grown in the NU-garden. For the study, they were taken from this year's harvest. After harvesting, the mothers dried, peeled6, roasted without salt and milled them using a wooden mortar. The paste was stored in a tin. Before making the RUTF, the quality of the paste was checked by checking the colour, smell and taste.

For the study, 7.5kg of full cream milk powder (enriched with Vitamins A and D) was bought in in Kampala and Mbale supermarkets at a cost of 30,000-35,000 USh (13.50-16 euro) per 2.5kg tin. Before using the powder, the use-by date, appearance, colour (white to yellow, creamy), taste and the smell (milky, bit sweet) were all checked.

Sugar and oil were already regularly bought by staff of the NU for the preparation of F-75 and F- 100. Sugar was purchased in Kumi town in 50 kgbags, which cost 75,000 Ush (34 euro) per bag (the price of 1 kg of sugar is 1,500 USh). It was impossible to stop ants getting into the bags so, before using the sugar, impurities were removed and quality was checked.

Vegetable oil was regularly bought in cans of 20 litres. The price of one can was 35,000-40,000 Ush (16-18 euro). The cost of 1 litre was 1,750- 2,000 USh. Quality was checked before use.

Vitamin-mineral-premix (CMV therapeutic) was already available in the NU since used in the preparation of F-75 and F-100. It is regularly ordered from Nutriset in France. The costs were as follows:

1 kg of CMV: 15.69 euro
1 carton of 6 tins of 800 g each (4.8 kg): 75.31 euro
Transport to Kumi from France: 62 euro/carton
Total price per carton (6 tins; 4.8 kg):302,080 Ush (137.31 euro)/carton (3 cents/g) CMV was stored in 800 g tins in the kitchen. Before use, the useby date and quality were checked.

Box 3

  1. First the vegetable oil is warmed through. Then the amount of peanut paste and the amount of (cooled) oil are weighed with a digital weighing scale to the nearest 0.1 g and both are mixed with a spoon until the mixture becomes homogenous (this takes about 5 minutes).
  2. The sugar is then weighed and ground with a pestle and mortar to a finer powder. The required amount of milk powder and CMV are weighed and mixed with the sugar in a container. Mixing takes about 4-5 minutes.
  3. The peanut-oil-mixture is then poured into the powder-mixture and again mixed with a spoon until it becomes a kind of paste. It is not only stirred but also pressed/mashed against the container.

Advising mothers

The paste was given out to the children in cups with a lid. The mothers were told to feed the child with a spoon out of the cup and to offer sufficient amounts of drinking water.

Cost comparison

The cost of producing local F-100, using both fresh cows milk and full cream milk powder (FCM), and for locally produced RUTF were calculated and compared, based on quantities of each that contain 1000 kcal (see table 1). One litre (1000ml) of F-100 based on fresh cow's milk costed 698 USh (32 cent)7 which is cheaper than the local RUTF produced in this study ((920 USh (42 cent)/1000kcal where groundnuts did not need to be purchased, and remains cheaper if groundnuts had to be bought (1004 USh (46 cent)/1000kcal). The high cost of milk powder is the main reason for the higher price of RUTF, which also accounts for the high cost of FCM based F100 - over double the price of fresh milk based F100.


Different degrees of malnutrition require different feeding options. For severely malnourished children with/or without complications, during the initial stage of rehabilitation it is appropriate to offer small amounts of RUTF in addition to F-100 and to observe which type of therapeutic food a child prefers.

Appetite does not always remain consistent during the early stages of recovery. Some children who re-developed signs of fever or diarrhoea temporarily lost their appetite again. In such situations, it is important to have access to alternative types of therapeutic food.

Locally produced RUTF was well accepted by the majority of less severely malnourished children in the NU. The duration of stay was also shorter in these children. This suggests that earlier discharge combined with a weekly check-up and distribution of RUTF (i.e. homebased therapeutic feeding until full recovery) is a realistic option for certain children and their caretakers, especially for those who live close by.

Local production of RUTF in the NU in Kumi hospital is feasible. The means for production (spoon, cups, boxes, a fridge) were already available in the NU, and practicalities, like grinding sugar or manual mixing, were possible since only small amounts were being prepared. However, local production did rely on imported vitamin-mineral mix, which in this instance was already being supplied. This may be a constraint where a supply and a budget are not in place. Only small amounts were produced, therefore the caster sugar, which was used instead of icing sugar, could be ground easily before use. This could become more difficult when larger amounts of RUTF have to be produced locally (e.g. during the rainy season when there are typically higher admission rates).


A recovering child feeds himself porridge in the NU

Different types of locally prepared RUTF (e.g. with milk powder in supervised feeding settings and without milk powder for locally prepared take home rations) should be explored to increase the variety and safety of therapeutic food and to support the local economy.

In an institutional setting (hospital/NU) the right time to offer RUTF to (severely) malnourished children with complications needs to be further investigated. There needs to be access to different types of therapeutic food for children at various stages of malnutrition.

Children who are discharged early (before full recovery) should receive a take-home ration of RUTF and be followed-up on a weekly basis. The right time for discharge needs to be negotiated between the caretaker and hospital/NU staff members.

For further information, contact: Dr. Veronika Scherbaum, email: scherbau@uni-hohenheim.de

1Based on Plumpy'nut®, a peanut paste semi-solid RUTF produced by Nutriset, France

2It is not only the taste, some children were unable to swallow RUTF - perhaps this is due to the consistency of peanutbased RUTF that may be difficult to manage during early stages of rehabilitation.

3The reason for early introduction of small amounts of the NU-diet was to increase the variability of food offered and to familiarise the child again with the taste of family food (e.g. a lunch/supper consisted of cereals, beans, dried fish flour, oil and vegetables which was offered ad libitum)

4Fresh cow milk is daily bought at 500 USh per litre from a farm near the hospital.

5In general, fewer malnourished children are admitted during the dry season as compared to the rainy season. In 2005, a total of 190 children were admitted in the Kumi hospital/ NU with a case fatality of 11.6%. It is interesting to note that no child died during the study period, which may be attributed to increased attention and care given to these children by all staff members.

6To prevent aflatoxin contamination mothers were told to exclude all peanuts that showed any discoloration or other irregular appearance.

7F-100 prepared with full cream milk would cost 1558 USh (71 cent) (see table 1)

Imported from FEX website


About This Article

Article type: 
Original articles

Download & Citation

Recommended Citation
Citation Tools