Decentralisation and scale up of outpatient management of SAM in Ethiopia (2008-2010)

By Sylvie Chamois

Sylvie Chamois has been a Nutrition specialist with UNICEF Ethiopia and Burundi for the past 9 years. Before joining UNICEF, she spent 6 years working as a nutritionist for Action Contre la Faim, mostly in emergency settings.

Disclaimer: the findings, interpretations, and conclusions in this article are those of the authors. They do not necessarily represent the views of UNICEF, its Executive Directors, or the countries that they represent and should not be attributed to them.

This article describes experiences and observations around the successful decentralisation and scale up of the outpatient management of severe acute malnutrition in Ethiopia in the period 2008 to 2010. It includes achievements, the challenges around monitoring and reporting and ongoing steps needed to strengthen and assure service quality.

Following up on the experience of decentralising outpatient therapeutic care in 100 districts of Ethiopia in 2008 (Field Exchange, No 36¹), this article describes how this was further rolledout in 606 districts of Amhara, Oromia, SNNP and Tigray regions between 2009 and September 2010.

Background

Ethiopia has one of the highest children under-five mortality rates, with malnutrition contributing to 57% of all children deaths². The 2005 Demographic and Health Survey revealed that 2.2% of children under-five are severely wasted³, a condition associated with a mortality rate of 13% (7.3 to 18.7%)⁴.

When adopting the Millennium Development Goals (MDGs), the Government of Ethiopia (GoE) committed to halve by 2015 children under-five malnutrition (MDG 1) and mortality (MDG 4). This has been reflected in its Plan for Accelerated and Sustained Development to End Poverty (PASDEP, 2005- 2010⁵). The strategy under the third Health Sector Development Programmes (HSDP III, 2005-20105) addresses the major causes of child mortality that account for 90% of under five deaths, i.e. pneumonia, neonatal conditions, malaria, diarrhoea, measles, HIV/AIDS and malnutrition.

A series of nutrition articles published in the medical journal, The Lancet, in January 2008 defined the magnitude and consequences of undernutrition and demonstrated the availability and potential benefits of proven interventions, including the management of severe acute malnutrition. Therefore, increasing coverage of and access to Outpatient Therapeutic Programme (OTP) is one of the key elements that contribute to the achievements of the MDGs 1 and 4, among other key food security and nutrition interventions currently being implemented in Ethiopia.

How the OTP decentralisation started in 2008

UNICEF has been advocating for the integration of the management of severe acute malnutrition into the Ethiopian government health system since 2004. With Government leadership, in- and out-patient care had been successfully integrated into 165 hospitals and health centres by January 2008. However, the idea of integrating outpatient management of severe malnutrition into the Health Extension Programme⁶ (i.e. allowing health extension workers to provide curative services) was still being discussed at the time of the March/April 2008 rain-failure in the southern part of the country.

In May 2008, dramatic and rapid increases of severe acute malnutrition levels were reported in Oromia and Southern Nations, Nationalities and People's (SNNP) regions. This was the result of the poor performance of the March/April rains in the southern part of the country combined with the prevailing high market prices (food prices in rural Ethiopia had risen by 250% between 2006 and 2008⁷). In these two regions alone, 193 districts were affected where over 23 million people lived. The Ministry of Health (MOH) and international non-governmental organisations (NGOs), with UNICEF support, began implementing emergency feeding programmes under the coordination of the Emergency Nutrition Coordination Unit (ENCU).

The MOH concluded that the best option to prevent high mortality due to malnutrition was to decentralise the outpatient management of severe malnutrition to the health post/sub-district level. In July 2008, UNICEF was requested to support the Family Health Department of the Ministry to rollout OTPs in 100 drought affected districts of Oromia and SNNP regions (See Map 1). This involved 1,239 health posts and 2,478 health extension workers.

This programme has been described in a previous issue of Field Exchange referred to earlier (No 36). Key elements of the programme were as follows. In July and August 2008, all the 2,478 health extension workers were trained in the identification of severe acute malnutrition, referral of the complicated cases to inpatient facilities and management of the uncomplicated cases of severe acute malnutrition. As of November 2008, 51 districts (50% of the initial plan) were managing OTPs in 455 health posts (36% of the total number of health posts in the 100 districts), raising the service coverage from 38 to 65% in the two regions' affected areas. A total of 27,739 children were reported to have been admitted in the 455 therapeutic feeding sites with overall positive performance indicators: 77.6% recovery, 0.7% mortality and 4.2% defaulter rates.

The MOH has since endeavoured to strengthen the existing OTPs in the 455 health posts and effectively continue the rolling-out of the plan to the remaining 784 health posts in the two regions and more widely nationally. As of September 2010, there are over 6,400 health posts delivering OTP services and 280 inpatient Therapeutic Feeding Units (TFUs) in 691 districts of Ethiopia out of which 89 are supported by NGOs. This scale up by GoE is a remarkable achievement. The article below details some of the key developments in achieving this.

How the OTP rollout was implemented

Advocacy and coordination

Post- 2008 experience of decentralising outpatient management of severe acute malnutrition in 100 districts, the GoE was keen to implement the rollout through their system and to ensure maximum integration and capacity development within the existing decentralised Health Extension Programme. UNICEF's role was to support GoE in this process.

It was also necessary to find the right fit for a tripartite partnership between GoE, NGOs and UNICEF, each playing their role for the rollout to be successful. NGOs had difficulties engaging in the process as the GoE preferred to implement the programme itself. There were concerns expressed by the NGO community about the proposed speed of scale up for this type of programme and fears that quality of service may suffer. There may also have been some concerns about the possible lack of NGO role in this scale up.

UNICEF advocacy role was to find the right link between the GoE and the support NGOs could provide. This was achieved through three types of coordination meetings:

Type 1: Involving all parties: nutrition cluster meetings were held at the ENCU and attended by NGOs, UN agencies, relevant government counterpart agencies and donors.

Type 2: Between NGOs, donors and UN agencies: regular meetings were organised by UNICEF as a means of encouraging donor support and to ensure good coordination with the NGO sector, and

Type 3: Between UNICEF and the Government.

This resulted in agreement that there should be joint monitoring of the OTPs (and TFUs) by all stakeholders, e.g. GoE, NGOs and UN agencies. NGO collaboration was to be requested by the GoE in those geographical areas where capacity for OTP (and TFU) rollout was limited. The predominant NGO role in these areas should and would be to support and build up Government capacity. These meetings were held for approximately one year from May 2009 and proved useful during a difficult transition period for various stakeholders.

Coordination at regional level was led by the Regional Health Bureaus through the Health and Nutrition Task Forces.

Donor support

UNICEF developed an emergency nutrition response plan for 2009 and 2010 and donors were invited to contribute funding. Donor support for the rollout of OTP in the four target regions was secured from the Humanitarian Response Fund (HRF), OFDA, Governments of Japan and Spain, CIDA, DFID, SIDA and ECHO¹¹.

Development of regional action plans

UNICEF helped to instigate the development of action plans in the four target regions (Amhara, Oromia, SNNPR and Tigray) in conjunction with the Regional Health Bureaus. These four regions were selected on the basis that they were the most severely affected by drought in 2008, their high population density and existence of a functional Health Extension Programme. Regional planning took place within regional task force meetings and involved NGOs where they were able to pledge commitment to training and monitoring.

Adaptation of the strategy to regional contexts (Afar and Somali regions)

Afar and Somali regions (arguably regions most prone to food insecurity) were not initially included in this scale up due to the mobility of their populations and consequent difficulties of integrating OTPs into existing health structures. Furthermore, Somali and Afar are not population dense so that health posts would serve only a few children, also security and access are continuously problematic in the regions. However, GoE with UNICEF and NGO support, fielded mobile health and nutrition clinics based upon guidelines developed in March 2008. Mobile teams visit communities every one to two weeks to implement OTP management as well as Integrated Management of Childhood Illnesses, promotion of hand washing and safe water. As of September 2010, there are seven mobile clinics run by the GoE in Afar, including two using camels. In Somali region, the GoE runs twenty mobile clinics while NGOs support another seven.

Technical assistance for training and follow-up

During the scale up, Regional Health Bureaus organised trainings with UNICEF and NGO support mainly targeted to health extension workers and HEP supervisors, but also including local zone and district officials. Training of health workers and doctors on the inpatient management of severe acute malnutrition was also provided with the aim of setting up a minimum of one inpatient unit per district for the management of complicated cases. UNICEF recruited six nutrition specialists to support the organisation and conduct training, as well as post-training visits and supportive supervision. Training was similar to that carried out during the 2008 decentralisation programme in Oromia and SNNPR.

Provision of supply and logistic support

In 2009, an amount of Ready to Use Therapeutic Food (RUTF) was imported, although the 'Hilina Enriched Food Processing Centre', the Nutriset franchised company, provided most of the supply. From 2010 onwards, UNICEF is planning to only procure RUTF locally¹². It had been anticipated that NGOs would procure their own RUTF where they were supporting districts, however it appears that some preferred or expected UNICEF to procure the commodity on their behalf. Furthermore, a number of NGOs advocated for provision of contingency stocks in the event of shortages.

UNICEF also became involved in some of the logistics of the programme where GoE needed additional support. UNICEF would therefore, in some cases, distribute RUTF to zonal or district level. For the remote districts, provision for supply transportation down to the health post level was included into the regional action plans.

Enhanced programme monitoring and quality insurance

Ensuring good programme monitoring and quality proved to be challenging with the rapid multiplication of OTP sites. UNICEF supported the GoE working on the three following aspects of programme monitoring and quality insurance:

a) Improved report quality, timeliness and completion

One priority was to increase and maintain high monthly statistics report quality, timeliness and good completion rates.

Figure 1 shows that the report completion rate was 69% in the four target regions before the OTP rollout (Jan - Aug 2008) and dropped to 42% with the OTP expansion (Sep 2008 to Dec 2009).

In September 2009, the Minister of Health sent a letter to Regional Health Bureaus to emphasise the importance of getting timely and accurate reports for enhanced programme monitoring. In addition, UNICEF recruited in November 2009 four technical assistants, one attached to each regional ENCU, to look at reporting systems, identify bottlenecks for quality and timely reports and take appropriate actions in conjunctions with the GoE. This resulted in the achievement of 77% report completion rate between January and August 2010. Reports usually reach the federal level within a maximum of one month after the end of each month of activity.

Other longer-term actions initiated by GoE and UNICEF on reporting are:

1. Development of a user-friendly software to be decentralised to the District Health Offices (currently under development)¹⁴. It is expected that this tool will improve reporting as well as district and regional capacity to analyse/use data for programme management. The software will also include growth monitoring, vitamin A supplementation and deworming indicators.
2. Piloting of health extension workers sending OTP data using SMS for improved timeliness and completion rate (ongoing and results expected in 2011).
3. Integration of nutrition indicators into the Health Management Information System (not yet achieved)¹⁵.

b) Development of joint monitoring tool and system

Following on the recommendation to undertake joint monitoring of the programme (GoE, NGOs and UNICEF), a sub-working group chaired by UNICEF was formed with ACF, Concern, GOAL, MSF-Greece and MSF-Holland. They met weekly between May and June 2009 to develop a monitoring checklist for field assessments. The tool was then discussed with the regional authorities in Oromia and SNNPR, tested and modified between July and August 2009. It was agreed where NGOs operate, that they were to be the partner of the MOH and UNICEF for all joint monitoring work. They were also invited to participate as much as possible outside of their traditional support area.

In SNNPR, by the first week of August 2009, a team composed of Regional Health Bureau, NGOs¹⁶, regional ENCU and UNICEF, made the final modifications to the checklist for OTP and TFU assessment, grading individual components of the practices observed. It was agreed that a three tier grade banding could be useful to identify the level of programme support needed at district level, where:

• Grade over 70%: working very well with minor support needs only
• Grade 50-70%: working well but with some technical and logistic support needs requiring attention
• Grade under 50%: has major support needs to refresh skills of staff and where operational systems need to be established

In districts where NGOs provide minimal support to MOH, the NGO in collaboration with MOH would address the agreed support needs. In all other districts, it was suggested that the Regional Health Bureau request additional support from UNICEF and/or selected NGOs.

The testing in Oromia and SNNPR also allowed for Regional Health Bureaus and partners to learn more about the resource needs to manage regular joint monitoring, especially in terms of key personnel time and transportation needs. It was initially thought that the six additional staff recruited by UNICEF to support the OTP rollout in the four regions would also be involved in the joint supervision. However, this additional capacity coupled with the GoE and NGO capacity was not sufficient to establish adequate monitoring of the programme (it takes a full week for a monitoring team to assess the practice of health extension workers and health workers in 40-50% of all OTP sites and all TFUs in one district).

Table 1: Summary of results from joint monitoring assessments in the four target regions (Amhara, Oromia, SNNPR and Tigray), September 2009 to September 201017
		Number and % of districts scoring:
Region*	Number of districts assessed	< 50%	50-70%	> 70%
Amhara	59	20 (34%)	27 (46%)	12 (20%)
Oromia	53	9 (17%)	30 (57%)	14 (26%)
SNNPR	3	0	3 (100%)	0
Tigray	16	6 (38%)	9 (56%)	1 (6%)
Total	131	35 (27%)	69 (53%)	27 (20%)

*Joint monitoring started in September 2009 in Amhara and Oromia and in April and June 2010 in Tigray and SNNPR respectively.

 

Table 2: Number of TFUs per district and hotspot priority in the four target regions (Amhara, Oromia, SNNPR and Tigray), September 201018
Hotspot priority number*	Total number of districts	Number of districts with no TFU	Number of districts with at least 1 TFU	% of districts with TFU
1	176	70	106	60%
2	96	52	44	46%
3	99	75	24	24%
Total	371	197	174	47%

* Districts affected by food insecurity and in need of humanitarian assistance are classified into hotspots priority 1 (high), 2 (medium), 3 (low) and 4 (not affected).

 

Table 3: Summary results from TFP coverage surveys conducted in each of the four target regions, 201019
District assessed and date*	OTP period coverage	OTP point coverage
Wadela, Amhara region, March 2010	19%	10.4%
Arsi Negele, Oromia region, January 2010	48.8%	21.2%
Mareko, SNNP region, January 2010	60.9%	37.8%
Tahtay Machew, Tigray region, March 2010	56.2%	27.3%

 

Consequently, in September 2009 UNICEF developed a Project Cooperation Agreement with the NGO Population Service International (PSI) to deploy twenty field monitors in an effort to boost programme quality in all hotspot districts in the four largest regions. A total of 131 districts were monitored from September 2009 onwards. As of September 2010, 20% of districts were working very well with minor support needs only (ranked over 70%), 53% were working well but with some technical and logistic support needs requiring attention (50- 70%) and 27% have major support needs to refresh skills of staff and where operational systems need to be established (under 50%), (Table 1).

In spite of the impressive increase in service access that has been achieved in districts that have rolled out OTP to health posts at subdistrict level, the monitoring has helped to identify key areas where the programme needs strengthening.

The availability of one functional TFUs per district as planned is still limited. Sixty per cent of 'priority 1' hotspot districts have at least one TFU, compared with 46% of districts in hotspot priority 2 and 24% in hotspot priority 3 (see Table 2). Also see Map 2.

The low proportion of referral from OTP to TFU in some districts is also a concern and should be studied further. It is not known how much of this is due to very early detection reducing the number of complicated cases, or as highlighted in the four TFP coverage surveys conducted recently (Table 3), for other reasons affecting access. These include low level of active community mobilisation and lack of skills among some health extension workers to identify and refer the complicated cases, caregiver refusal to go to the TFU because it is too far, lack of capacity to pay for transport, lack of food for caregivers, opportunity cost for caregivers of staying away from their home, etc.

Health Extension Programme supervision by MOH is not always working adequately due to lack of trained staff skilled to mentor/supervise OTP activities. Some supervisors are sanitarians, not health staff, and are not familiar with the programme. Some districts lack the full quota of supervisors (one per five health post). Lack of capacity and effective management from District Health Offices managers and lack of transport to facilitate visits to OTP sites may also contribute to shortcomings. When assessing the programme, the monitors (GoE, PSI/ UNICEF and NGO) also build the capacity of the HEP supervisors and District Health Offices managers using the checklist and providing on-the-job support.

Health extension workers in most districts observed so far urgently need follow up training to refresh/strengthen their skills. OTP protocols are not always being maintained, and errors in anthropometry are still observed. This can result in moderate acute malnutrition cases included in the programme, and unduly long length of stay. Recording and reporting also remains a challenge. A system for regular supply management and distribution is yet to be established in many districts.

Where Therapeutic Feeding Programmes and Community-Based Nutrition (CBN) programmes co-exist, the District Health Office has sometimes failed to capitalise on the opportunities provided by the CBN programme enhanced training and support for health extension workers and community health workers. Community mobilisation needs to be strengthened in districts that have as yet not included it in the training or programme set up.

Another issue is that some NGOs are still in implementation mode, providing nurses to work alongside health extension workers for each OTP (involved in recording, reporting, supply management and child treatment). It is intended that NGOs work to develop capacities by mentoring, supporting quarterly reviews and developing District Health Offices skills to manage monthly reporting and supply management systems, rather than providing staff to help service delivery. On the other hand, some NGOs are still facing difficulties working on capacity building and dialogue is still required to achieve the smooth partnership that UNICEF is promoting.

In addition, there is insufficient access to safe water at health post level. It was recently suggested in Amhara that whether OTP sites have sanitary latrines and safe water should be an additional indicator for inclusion during assessments and monitoring.

Routine drugs are not always being administered by the health extension workers, although the State Minister of Health authorised the provision in a letter dated January 2010. This has also been included in the Community Case Management of Childhood Illnesses together with the community treatment of malaria, diarrhoea and pneumonia (January 2010).

c) Partnership for programme coverage surveys

UNICEF developed a partnership with MOH and Concern to build the capacity of the Ethiopian Health and Nutrition Research Institute (EHNRI) to assess TFP programme coverage using the Centric Systematic Area Sampling (CSAS) methodology. The approach has been used previously by NGOs in small programme areas (individual districts). For exposure to the methodology and capacity building, one coverage assessment was completed by the EHNRI with Concern technical support in each of the four regions in January and March 2010 (Table 3). Findings and recommendations were discussed at regional level and action plans were developed to address the issues (see Box 1).

In September 2010, a consultant was brought in by Concern and UNICEF to develop and propose to the MOH a coverage survey methodology that would allow assessment of programme coverage over wider geographic areas (at the moment, the methodology does not allow for assessment of more than one district at a time). A subsequent step will be to pilot and validate the new methodology before building the capacity of regional authorities to plan for and conduct regular surveys as part of their programme monitoring.

Progress and results

The Government of Ethiopia has, within a short space of time, managed to provide access to services for the majority of families affected by severe acute malnutrition in four regions of the country. In less than two years, service coverage for severe acute malnutrition has reached 49% and 48% of health posts and health centres respectively running OTPs and 17% and 92% of health centres and hospitals respectively running TFUs (Table 4). Consequently, there is now earlier detection of severe acute malnutrition cases reducing the number of complicated cases needing specialised inpatient care.

* Districts affected by food insecurity and in need of humanitarian assistance are classified into hotspots priority 1 (high), 2 (medium), 3 (low) and 4 (not affected). Source: UNICEF Ethiopia.

Furthermore, key monitoring results in terms of SPHERE standards²⁰ are impressive. Between January 2008 and August 2010, a total of 370,559 children were reported to have been admitted to in and out-patient therapeutic feeding sites in the four regions with overall positive performance indicators: 82% recovery, 0.7% mortality and 5% defaulter rates (Table 5). Also see Figure 2.

Source: TFP database, ENCU/ DRMFSS/ MOH, Ethiopia

Lastly, the GoE is now in a much better position in terms of national capacity and preparedness to respond to any increases in severe acute malnutrition levels.

Table 5: OTP & TFU per formance in the four target regions (Amhara, Oromia, SNNPR and Tigray), Jan. 2008 to Aug. 2010
	TFP performance	SPHERE standards
% of reports completed	63.4%	-
Number of admissions	370,559	-
Cured	272,261 (82.5%)	> 75%
Died	2,481 (0.7%)	< 10%
Defaulter	15,392 (4.7%)	< 15%
Medical transfer	3,627 (1.1%)	-
Transfer from OTP to TFU and from TFU to OTP	28,357 (8.6%)	-
Non respondents	8,054 (2.4%)	-

Source: TFP database, ENCU/ DRMFSS/ MOH, Ethiopia.

Key contributing factors to success

There have been a number of important contributory factors to the success of this programme. Key amongst these are:

• Central and regional government commitment to develop policies and guidelines on decentralised treatment of severe acute malnutrition and to integrate services into the wider decentralised health programme. The outpatient management of severe acute malnutrition is now fully part of the Community Case Management of Common Childhood Illnesses implemented through the Ethiopian Health Extension Programme.
• Advocacy to promote this approach and to provide technical assistance to all relevant field level actors.
• High educational levels, technical skills and commitment amongst programme staff working within the health sector, e.g. health extension workers, nurses and doctors.
• Enhanced coordination between GoE, UNICEF, NGOs and donors creating an environment in which all stakeholders could contribute skills and resources to best effect.
• Simplicity and efficacy of OTP approach both for the service providers (health extension workers) and service users (patients and caregivers).

Ways forward

Key next steps are to continue the OTP roll-out at health post level and expand TFU coverage to reach a minimum level of one TFU per district. It will also be important to include therapeutic feeding items in the Essential Drug/Commodity List to resolve some of the supply and logistic issues. Similarly, the inclusion of basic TFP indicators into the Health Management Information System (new admissions and performance) will be an important step towards fully integrating and sustaining the programme, together with advocacy for an increased allocation of funds to the Health sector to absorb the programme cost. Continued joint supervision visits are also essential to address service quality and build the capacity of the supervisors in charge of programme implementation. Linkages with other programmes must also to be developed. These should include improved access to safe water and health services by working with Ministry of Water Resources and linking with community health workers of the Health Extension Programme, and improved and expanded management of moderate acute malnutrition and prevention of severe acute malnutrition (this will become an even greater priority as Ethiopia adopts the new WHO Growth Standards). Finally, after two years of implementation, it is becoming increasingly urgent to conduct a comprehensive programme evaluation and to study a number of issues, including the reasons for low TFU admission rate in some districts.

For more information, contact: Sylvie Chamois, email: schamois@unicef.org

---

¹Sylvie Chamois (2009). Decentralisation of out-patient management of severe malnutrition in Ethiopia. Field Exchange, Issue No 36, July 2009. p12. http://fex.ennonline.net/36/decentralisation.aspx.

²National Strategy for Child Survival in Ethiopia, Federal Ministry of Health, Ethiopia, July 2005.

³Defined as weight-for-height < -3 zscore.

⁴Steve Collins, Nicky Dent, Paul Binns, Paluku Bahwere, Kate Sadler and Halister Hallam. Management of Severe Acute Malnutrition in Children. The Lancet, Vol. 368, December 2006.

⁵The PASDEP and HSDP III cover the period Sept 2005-Aug 2010, which corresponds to the Ethiopian calendar years 1998-2002.

⁶The Health Extension Programme delivers primary health and nutrition services at sub-district level by two female Health Extension Workers based in health posts; there is one health post per sub-district covering an average of 5,000 people.

⁷FEWS NET Ethiopia food security updates, USAID and WFP, 2008.

⁸Vitamin A is supplemented to children 6-59 months old and deworming tablets are administered to children 2-5 years old.

⁹Are screened for malnutrition: children 6-9 months old, visibly pregnant women and women breastfeeding a less than 6 months old infant.

¹⁰Eligibility for TSF: Children with MUAC < 12 cm and/or bilateral oedema and women with MUAC < 21 cm. One TSF ration is composed of 25 kg of CSB and 3 litres of oil for 3 months.

¹¹See acronyms listing on page 92.

¹²As of September 2010, there is only one local RUTF factory in Ethiopia and a second one is planning to start production in October 2010.

¹³Source: TFP database, ENCU/DRMFSS/MOH, Ethiopia.

¹⁴At the moment, the database is kept by the ENCU at federal level.

¹⁵Health extension workers salary is conditioned to the submission of the HMIS reports to the District Health Office.

¹⁶Concern, GOAL, Save the Children-US, Samaritan's Purse and International Medical Corps (IMC).

¹⁷Source: UNICEF Ethiopia.

¹⁸Source: UNICEF Ethiopia.

¹⁹Source: EHNRI.

²⁰The SPHERE project, Humanitarian Charter and Minimum Standards in Disaster Response.