Cost effectiveness of community-based and inpatient therapeutic feeding programmes to treat SAM in Ethiopia

Health Extension Worker testing the appetite of a malnourished child, Menkere health post, Tigray region

By Asayehegn Tekeste, Kebede Deribe, Dr Mekitie Wondafrash and Dr Girma Azene

Asayehegn Tekeste is a Public Health with a Masters in Public Health (MPH). He has worked in management of emergency nutrition with Save the Children USA and UNICEF Ethiopia. Currently he is working as Senior Capacity Building Advisor in Save the Children USA-Ethiopia.

Kebede Deribe is a Public Health Specialist with a Masters in Public Health (MPH). He has been working in complex emergencies in Darfur for the last three years. He has worked with World Relief and American Refugee Committee International. Currently he is working as Health Coordinator with Merlin in South Darfur, Sudan.

Dr. Mekitie Wondafrash is a Medical Doctor and Assistant Professor of Nutrition in Jimma University, where he teaches nutrition for undergraduate and postgraduate students. He has conducted nationwide surveys and consultancies in Ethiopia.

Dr Girma Azene is a health economist and has been working as Head of the Health Planning Department of the Ministry of Health in Ethiopia.

Dr Girma Azene is a health economist and has been working as Head of the Health Planning Department of the Ministry of Health in Ethiopia.

A retrospective study was recently conducted in SNNPR¹, Ethiopia to determine the average cost of treatment of a severely malnourished child in a therapeutic feeding centre (TFC) and in a community based therapeutic care (CTC) programme, to determine the effectiveness of TFC and CTC programmes as measured by the clinical outcomes, and to compare the cost effectiveness of the two approaches. CTC programming comprises community screening, an outpatient therapeutic programme (OTP) for uncomplicated cases and a stablisation centre (SC) for complicated cases².

Data were collected in Shebedino woreda of the Sidama zone in the SNNPR from February to April 2007. The woreda is located south of the capital of the regional state Awassa. One of nineteen Woredas of Sidama Zone, it has 50 rural and three urban kebeles with an estimated population of 315,354 and population density of 630 people per km2. Farming, which combines both crop cultivation and livestock rearing, is the major economic activity for 98% of households living in rural areas of the district. An average farmland shared by a household is less than or equal to 0.5 hectare.

In recent years (2000 - 2005), there have been unfavourable rainfall patterns, hampering crop production resulting in serious food shortages in lowland and mid highland areas of the district. In 2004, the information obtained from the Woreda Disaster Prevention and Preparedness Commission (DPPC) office revealed that 32 kebeles were in need of emergency food aid. Of those 32 kebeles, 22 of them had been receiving general ration from the regional DPPC. Save the Children US started running CTC programmes in the Woreda in July 2004.

Morocho TFC was opened on July 24, 2003, located within a health post compound. Initially the centre's capacity was planned to accommodate 100 children. However, as there were 122 children admitted within the first week, the capacity was revised to accommodate 200 children.

The Shebedino Woreda CTC programme was opened on 25th of July 2005 in response to a global acute malnutrition (GAM) prevalence of 16% with aggravating factors, identified in a May 2005 nutrition survey. The programme established one SC at Leku health centre and eight OTP sites.

Sample size and study method

The sample size for the study was determined using the sample size determination formula for two population means. Using a 15 % allowance for non-response, missed cases and out migration, the final sample size was determined at 328.

Stratified random sampling was employed to select the study subjects (patient cards and caregivers). First, the list of kebeles where the children served by the respective programmes came from were stratified into 'near', 'medium' and 'far' to the Morocho TFC based on information from the qualitative study. The calculated sample size was allocated proportionally to the number of children treated in each stratum. For these selected study subjects, patient card reviews were undertaken. The caregivers of these selected children were interviewed for the opportunity costs they incurred.

Health extension workers and community health assistants (CHAs) who spoke the local language, and lived in the study kebeles were recruited and collected the data for the household survey.

A semi-structured guide was used to collect data on the costs to caregivers from interviews and from Focus Group Discussions (FGD). Ten people were interviewed as key informants. Two FGDs each for TFC and CTC caregivers from Telamo and Remeda kebeles were conducted. The main purpose of the interviews and FGDs was to estimate the age of employees in the area, determine the daily wage rate and productivity during specific seasons and then calculate and value productivity loss or wage loss, time spent to reach the programme site, transportation costs and waiting time to obtain services.

Administrative records and reports were reviewed to gather data on costs and outcomes of the programmes. Patient cards were reviewed to identify direct costs of treatment for each beneficiary. The financial records and statements from Save the Children US's Ethiopia Country Office in Addis Ababa, Awassa, (EHNP) office and Project Units were also reviewed.

The 'societal perspective analysis', which considers costs to all sectors of the society, was used. Collection of cost data included both the direct costs of programme implementation and opportunity cost (economic cost) of the two alternative models under consideration. Since the cost estimation exercise in this study included use of cost data at different time periods, adjustments were made to account for inflation using appropriate indices, gross domestic product (GDP) deflators and exchange rates.

The effectiveness of the two programmes was measured in terms of cure rates, or number of children cured from severe acute malnutrition (SAM) as identified from their follow up care records. In this analysis, a child was consider to be cured if discharged on fulfilling the criteria of weight for height ?85 % for two consecutive weighing and no oedema for ten consecutive days.

Results

A total of 328 patient cards/records of children cured in the programmes were reviewed. Out of these, 306 children (157 CTC and 149 TFC) were traced back to their households to interview their caregivers.

Cost for routine medicines and medicines for treatment of complications

The two programmes were run by a humanitarian organisation and therefore no user fees were incurred by the caregivers and families of the children.

Among the 164 children treated in the CTC programme, only 19 children (11.6%) had to seek care in the SC. The average length of stay of a child admitted to the SC was 13.3 days. The rate of infection in the SC was 12.8% (21 children). The cost for additional treatments for these medical problems/complications in the CTC programme was estimated at $0.17 per child.

In the TFC group, 71 children (43.3%) were treated for complications. A significantly greater number of children in the TFC group had medical complications compared to CTC children (P<0.001). The cost for additional treatments for these medical problems/complications in TFCs was estimated at $0.38 per child.

The average cost for medicines in CTC was $1.92 per child compared to $2.51 in the TFC.

Cost of therapeutic food The cost of therapeutic food provided per child was found to be $42.94 for TFC and $55.53 for the CTC programme.

Cost of therapeutic food

The cost of therapeutic food provided per child was found to be $42.94 for TFC and $55.53 for the CTC programme.

The total cost of food for caregivers in the TFC was $11.64 per child compared to $0.15 per child treated in the CTC programme.

In the TFC, soap, a jerry can, a blanket, a pair of bed sheets and insecticide treated net (ITN) were given to the beneficiaries while in the centre and at discharge. The cost was $23.25 per child. Some of these items were also given to CTC beneficiaries, as well as a bucket to carry the bimonthly dry ration. The average unit cost of non-food items provided in the CTC programme was $13.77.

Personnel cost

The average unit cost of staff per child for the TFC ($122.36) was more than three times that of the CTC programme ($37.1).

Capital depreciation and utilities

Based on the financial record reviews and reports for the period under study, capital depreciation and utilities costs of the programmes were estimated. The space used for treatment, capital items, utility at the sites and SC, vehicle operation and supplies were included in this cost category. The medical equipment used in the programmes were considered as recurrent expenditures, assuming that within the setting of emergency therapeutic feeding, their life is not expected to exceed one year.

The capital depreciation and utilities cost of the TFC was $50.47 per child and the CTC programme cost was $17.92 per child. A major portion of the overhead costs in both programmes went to vehicle rental. Here also, TFC costs were about three times more than CTC programming costs.

Total institutional costs

The total institutional cost incurred to treat a child was calculated by adding all cost categories discussed above. The institutional cost in the TFC was $262.62 per child which was more than twice that of the CTC programme. (Table 1).

*Here utilities include vehicle fuel and operation, electricity, maintenance, etc. RUTF: Ready to Use Therapeutic Food

Costs to caregivers

Caregivers and families spend money while seeking treatment for the child. These include costs of transportation, food and lodging. Generally, the average costs to caregivers for the TFC was $1.45 per child compared to $0.92 per child for the CTC programme.

Opportunity cost to caregivers

The cost of lost productive time for caregivers while in programme was calculated based on their occupational status and the total time spent on caring for the child during treatment. Assuming caregivers were productive all the time, the opportunity cost in the TFC was $20.92 per child and in the CTC $5.88 per child.

Combining direct expenditure with the opportunity costs, the economic costs to caregivers was about $21.93 per child for the TFC and $6.29 per child treated in the CTC programme.

When caregiver and institutional costs are combined, the TFC costs $284.56 per child while the CTC programme costs $134.88.

Outcomes of the interventions

Out of a total of 693 children admitted to Morocho TFC, 616 children were cured with a cure rate of 88.9 %. There were no reported deaths in the TFC.

Out of the 660 children discharged from Shebedino CTC during the period under review, 612 were cured. The cure rate was 92.7 % for CTC. The death rate in CTC (1.2%) was higher than the TFC but very much lower than the Sphere standard.

The average cost per cured child in the TFC is 320 USD compared to 145.5 USD for the CTC programme (Table 2).

Discussion

Caregivers' direct costs were higher in the TFC than in the CTC group. Since TFCs are far from home, caregivers tend to spend more money to reach the centres compared to the CTC distribution sites. The average length of stay of children in the TFC (?27 days) and CTC (?42 days) was similar to findings reported elsewhere.

The direct cost of supplies for treatment and cost of medicines used to treat complications was three times higher for the TFC. The cost of therapeutic food for children was the only cost element that was higher in the CTC programme than in the TFC. The greater length of stay and lower rate of weight gain compared to TFC children might explain this.

The significantly higher average cost of caregivers' food in the TFC might be due to a smaller proportion (11%) of children in CTC who required inpatient treatment.

Three times more cost was incurred for professionals in TFCs, as these are large intensive care centres where the skill of professionals required is higher. Seventy-one full time support staff were required to operate the centre for three shifts round the clock. The staff that ran the CTC programme comprised only two staff, in addition to volunteers and parttime workers.

Vehicle rental took a significant portion of the overhead cost in CTC as teams needed to be transported to the sites every day.

Caregivers' costs (both direct and opportunity cost of lost productive time) were 3.5 times higher in the TFC as caregivers spent many days away from home with their child. The lower expenditure on drugs in the CTC group may also partly be due to community mobilisation and outreach activities helping families or community volunteers to detect malnutrition and seek treatment earlier.

With all the costs considered, the CTC model was more cost effective than the TFC model One of the main limitations of the study is the fact that it is based upon retrospective information and recall bias may occur especially in the estimation of opportunity cost to caregivers. Another limitation was that the estimation of costs to caregivers in terms of productivity losses assumes similar patterns of work and does not take account of the seasonal variation in work availability.

Conclusions

The findings suggest that CTC is more cost effective than inpatient therapeutic care. The cure rate in the CTC was above 92% during the period under study, which was well above that of the TFC. Since therapeutic food costs were a significant part of the costs of CTC, local production of ready to use therapeutic food (RUTF) should be encouraged to cut the costs of care. Further comprehensive and prospective studies in drought prone pastoralist areas or amongst socio-culturally different populations are recommended.

For more information and a full report of this study, contact: Kebede Deribe, P.O.Box 2082 code 1250, Addis Ababa Ethiopia.
Tel: + 251 911937708,
email: kebededeka@yahoo.com

¹Southern Nations, Nationalities, and Peoples Region

²See Field Exchange 40 for a comprehensive series of articles on community based management of acute malnutrition in Ethiopia.