Outpatient therapeutic programme (OTP): an evaluation of a new SC UK venture in North Darfur, Sudan (2001)

Summary of internal evaluation¹ by Anna Taylor (headquarters nutrition advisor for SC UK)

North Darfur experienced a severe drought in 1999 and 2000. This caused widespread crop failure, disruption to markets, a decrease in earning power of all families and a generalised decline in food security.

In October 2000, a SC UK and Development & Rehabilitation Committee (DRC) food economy assessment of North Darfur concluded that during 2001, between 17,192 and 26,057 MT of relief grain would be required in order to prevent loss of life and destitution. The report recommended that this food be distributed from March 2001, to preempt the start of an extended hunger gap and to ensure that food was available before the planting season in July.

From October 2000 onwards, repeated efforts by SC UK to mobilise donor support for food distribution proved unsuccessful. Nutrition surveys subsequently carried out in April 2001 revealed the following prevalence of acute malnutrition:

* Z-Score < -2 or Oedema
** Z-Score < -3 or Oedema

Following this survey, funding was successfully secured by SC UK to respond. Funds available were inadequate to do blanket supplementary feeding or further general ration distribution. Subsequently in August 2000, an outpatient therapeutic programme (OTP), using ready to use therapeutic food (RUTF), and a supplementary feeding programme (SFP) were established. Overall, 836 severely malnourished children were treated in the OTP between the period August 11- December 12 2001. In addition, the targeted SFP enrolled approximately 24,000 children and 23,000 pregnant and lactating women, in 10 rural councils of North Darfur, during the same five-month period. General ration distribution of 15,000 MT of grain was also made between May and October 2001.

It was hoped that outpatient therapeutic care would offer the following programmatic advantages:

• Avoid setting up numerous therapeutic feeding centres (TFCs) that would have been beyond the capacity of any operating agencies. The large geographical area and scattered population would have required the construction, staffing, and equipping of dozens of TFCs in order to achieve acceptable levels of coverage and equity of provision.
• Avoid having children concentrated in centres risking cross infection and being treated by medical staff with limited training and competency.
• Allow the community greater participation in the programme and reduce the cost to families often associated with participation in an inpatient programme.
• May prove cheaper than inpatient TFCs since the costs of technical staff and physical infrastructure would be reduced.

Outpatient Therapeutic Programme design and content

Children were screened in the first instance using MUAC and by checking for nutritional oedema.

• Infants aged 6-11 months with a MUAC <110mm and children aged 12-59 months with a MUAC <115mm were immediately assessed using weight for height.
• Infants aged 6-11 months with a MUAC 110- 115mm and children aged 12-59 months with a MUAC 115-119mm immediately received supplementary food and were assessed one week later using weight for height.

Inpatient treatment

Children with <70% weight for height or with nutritional oedema were immediately seen by the medical team comprising a medial assistant and team nurse.

Children who presented with

• oedema
• severe dehydration,
• no appetite and visibly not eating RUTF when offered
• temperature exceeding 39 oC
• increased respiratory rate (> 35 if over 2 years, > 40 if 1 - 2 years or > 50 if less than 1 year) or
• any sign of illness

were referred in the first instance to a hospital (four hospitals were being supported by GOAL to give therapeutic care).

Where a child could not be taken to the hospital due to the distance, they were referred to a stabilisation centre (three stabilisation centres were set up in health centres located in areas least accessible to the hospitals).

Treatment in the stabilisation centre included:

• rehydration where necessary
• treatment with a systematic antibiotic (amoxycillin)
• routine fansidar
• folic acid and vitamin A
• measles vaccination
• feeding with F75 by naso-gastric tube, and
• daily examination and monitoring.

Children were discharged from the stabilisation centre to the OTP when appetite was shown to have returned (successful eating of RUTF for more than one day). Children were transferred to the hospital if they vomited more than 50% of the feed after 6 consecutive meals, had a body temperature greater than 39C, had a high respiratory rate or failed to regain appetite after 5 days in the stabilisation centre.

Outpatient management

Children admitted into the OTP received:

• systematic antibiotic treatment (amoxycillin)
• chloroquine (according to the Ministry of Health treatment guidelines)
• folic acid and vitamin A
• measles vaccination, and
• 14 packets of RUTF per week (500 kcals per packet).

Children in the OTP were visited daily by the Community Nutrition Worker (CNW) who checked the child for appetite (sachets of RUTF eaten the previous day), diarrhoeal history, thirst, and dehydration. They confirmed the presence of a carer and watched the child consume RUTF. In addition, children had a full medical examination and weight for height review weekly, carried out by the team nurse or medical assistant (health professional category just below a doctor in Sudan).

The CNW was also responsible for

• Identification of severely malnourished infants through community screening
• Instruction in the use of therapeutic foods through home visits
• Identification (alongside the medical assistant and nurse) and inpatient referral of severely undernourished children who failed to thrive in the outpatient programme (usually due to anorexia or co-morbidity)
• Hygiene promotion activities and the distribution of soap to mothers of severely undernourished children with skin diseases
• Promotion of the use of ORS through home visits
• Acting as a 'contact point' between the community and all aspects of the outpatient therapeutic programme.

Children were discharged from the OTP to the supplementary feeding programme when they had reached 75% weight for height for 4 consecutive weeks.

In total the programme had 100 distribution points and employed 290 field staff including medical assistants and nurses, CNWs, and Team leaders.

A note about data quality

Since this was a new type of intervention for SC UK, an especially designed monitoring system was put in place to measure programme outcomes. However, it proved to be inadequate and would have to be substantially revised for any future programme. Specific problems encountered were inconsistencies between daily and weekly reporting forms; some children's outcomes not reported at the end of the programme due to the rush to close; and failure of the system to track children through OTP and SFP.

Outpatient Therapeutic Programme outcomes

Mortality

Traditional house in North Darfur.

A mortality rate of 2.9% is very low and well within Sphere standards. However, the mortality rate is difficult to interpret since children who may have died after discharge to the SFP are not reflected in the figures. The expected number of deaths (using the Prudhon index) was compared to the observed deaths in the OTP and the discharge criteria were taken into account in this analysis. Half (51%) of expected deaths of children without oedema (n=744) were actually observed while almost all (92%) those expected occurred for children with oedema (n=62). Rates did vary according to location (see ranges in Table 1).

Other indicators

The average rate of defaulting was skewed by the high rate of defaulting in one location (El Fasher town) which had the largest number of children and where up to 34% of children defaulted from the OTP. This was mainly because the children were from pastoralist families who only stayed in the town for a few days at a time. Just over one-third of children (36%) defaulted from another location (Tina), a pastoral area where only 24 children were admitted.

Discharge rates were therefore also low in these two centres. With the exception of these two locations, defaulting rates in the other nine locations remained below 14%. Readmission rates were approximately 1.0% of total admissions.

Child recovering from malnutrition at home.

Mean length of stay in the OTP was estimated at 25 days for wasted children and 35 days for oedematous children. Once again this data must be interpreted in the context of the OTPs' discharge criteria (75% weight for height), which is lower than the 85% weight for height threshold typically used in TFCs.

Mean weight gain was 6.6g/kg/day for wasted children and 1.8g/kg/day for oedematous children.

Children in the stabilisation centres

Only 17 children were admitted to the stabilisation centres, of whom three children died in the centre and two after discharge to the OTP (included in the death rate in table 1).

Coverage

OTP Outreach workers.

Table 2 shows the range of estimated coverage rates achieved by rural council. The rates are comparable to rates achieved in well-run TFCs operating in high population densities. More children were admitted in locations where prevalence of malnutrition was higher. Taking into account the children admitted into the hospital, the presence of TFCs did not increase coverage substantially (35% (17-79%)) (See Table 2).

Malnourished infant with infant. (SCUK 2002)

There are two possible reasons for the lower than expected coverage rates. First, part of the programme period overlapped with the planting season, which influenced the rate of admissions into the programme and secondly, a miscalculation was made in setting the registration targets, which may have affected the rate of case finding.

The estimates in Table 2 are problematic for several reasons.

1. The numerator is the total number of children admitted (minus re-admissions and transfers) over the 3-month project period. Coverage should be estimated at a single point in time and could therefore be as little as a third of the estimated coverage in Table 2.
2. These figures are based on 20% of the population being under five years, whereas the Bureau of Statistics recommends this figure to be 16.6%. The data for the pastoralist areas could underestimate the real coverage due to overestimation of population in these areas. Both of these factors could mean that the coverage figures are underestimated.
3. The estimates of malnutrition were based on a survey conducted in April 2001 and compared to children admitted into the programme August - December 2001. Malnutrition could have increased in the run up to the harvest in October / November or, as in West Darfur, could have declined.

Conclusions

The programme admitted 836 children, which is probably several times more than a therapeutic feeding programme could have achieved within a 5- month period (from funding to closure) in North Darfur.

Mortality rates were very low, probably due to a combination of not being presented with the risk of poor care and cross infection in a TFC, and because rates of oedema and complicated malnutrition were low. The reduced costs to the community through a decentralised programme have not been evaluated but the overall programme costs (approximately £260 per child) do not vary substantially per beneficiary from costs per child in a TFC. Coverage rates were not as high as hoped though they were of the order achieved in well run TFCs in areas of much higher population density. Defaulting rates were generally very low except in the pastoral areas.

The future replicability of this programme needs to be considered in the light of the fact that in North Darfur, SC UK has

1. a good knowledge of both the macro-economic and micro-economic (i.e. household food economies) context of the intervention setting
2. a good knowledge of the geography of the intervention setting
3. ready availability of qualified staff and other resources such as offices and vehicles
4. good relations with the intervention population
5. good relations with local government officials at all levels
6. good relations with the government health systems allowing local health staff (i.e. medical assistants and nurses) to be seconded to the programme for both service delivery and training of community nutrition workers as well as the use of primary and secondary level health facilities
7. strong leadership and good support from SC UK national and international offices
8. the existence of a strong community ethic in the intervention population and
9. the availability of a ready to use therapeutic food (RUTF) acceptable to the intervention population.

Future recommendations

The results of the evaluation from the programme in North Darfur give grounds for cautious optimism. The programme was implemented with strong technical support in the phases of design and early implementation, was thoroughly and independently evaluated and the results have been disseminated. We consider these to be essential components of any future work using RUTF or aiming to treat children with severe malnutrition outside of the TFC or hospital setting. There are also some priority areas for research and analysis in any future pilots of outpatient care:

1. More experience on the potential complementary functions of traditional TFCs, hospitals, stabilisation centres (as conceived in North Darfur) and outpatient care, appropriate protocols for each level need to be developed and models for systems of referral need to be piloted.
2. Examine how the needs of infants aged less than 6 months old weighing under 4kg or with oedematous malnutrition can be addressed in a programme with an outpatient component.
3. More careful analysis of the varying weight gains which children achieve on RUTF is required to understand its suitability for different conditions.
4. Evaluate the long term prognosis of children receiving outpatient care compared to those who are discharged from a TFC.
5. Examine the extent of the reduced cost to communities of a decentralised outpatient programme compared to an inpatient programme.
6. Review whether there is greater opportunity for improved psychosocial stimulation during therapy through an outpatient programme than through an inpatient programme.
7. Investigate the degree of compliance with systematic antibiotic / micronutrient regimens that can be sustained in an outpatient programme.
8. Priority should be given to the use of other suitable RUTFs, such as BP100, and to the development of more scientifically evaluated products with a view to reducing prices. Nutriset Plumpy Nut is currently prohibitively expensive for routine use.
9. Monitoring frameworks for home based care need to be piloted and best practice established. Experience can be drawn from monitoring for inpatient care.

Acknowledgments Many people contributed their expertise and commitment to this project and its evaluation and they deserve our thanks. The consultants who have been involved in the project cycle are Steve Collins (surveys and project design), Yvonne Grellety (evaluation) and Mark Myatt (surveys and evaluation). This report draws on project documents, survey reports and the consultants' reports.

¹For more information please contact Anna Taylor, Nutrition Adviser, Save the Children UK. A.Taylor@scfuk.org.uk