Community case management approach to SAM treatment in Angola

Location: Angola

What we know already:  CMAM is a successful treatment approach for acute malnutrition. The conventional CMAM model of outpatient treatment centres (OTPs) may not always be feasible in contexts where there is particularly weak health infrastructure and low staffing capacity. 

What this article adds: A community case management (CCM) approach using volunteers achieved high cure rates and coverage. Further operational research is recommended to refine this approach and identify solutions to implementation challenges, such as access to medical treatment in low capacity rural settings, effective MOH RUTF supply chains, and consistent, appropriate volunteer incentives.

Introduction

The 2012 drought in Angola was described as the worst agricultural season since 1978. An estimated half million children under-five were affected by acute malnutrition in 10 provinces. The provinces of Bié, Huambo, Kwanza Sul and Zaire had the highest global acute malnutrition (GAM) rates, initially estimated at 18-25%¹, with 14-20% moderate acute malnutrition (MAM) and 4-5% severe acute malnutrition (SAM). In partnership with the Ministry of Health (MOH), UNICEF, Africare and People In Need, World Vision launched CMAM programming in these four provinces.   

At the onset of this nutritional crisis, capacity to respond was low with limited reach of the rural health system due to a lack of physical infrastructure and health staff.  In June 2012, there were only 24 inpatient centres across the country. As admission numbers to these centres doubled, mortality rates ranged from two to four times SPHERE standards², an indication of the sub-optimal quality of services. It was clear at design phase that it would be challenging to achieve high coverage and significantly impact the high GAM rates using the standard CMAM model. It was proposed therefore to train and supervise Community Health Activists (CHAs)³ to deliver Ready to Use Therapeutic Food (RUTF) based on a community case management (CCM) model. From November 2012 to January 2013, under CERF funding, activities began with recruitment and training of trainers for Municipal Supervisors, cascading to training for Communa (commune) Supervisors and CHAs.  In January 2013, mass screenings were undertaken across all four provinces. In February 2013, ECHO funding enabled the response to continue with further training, screening, treatment and nutrition education, until December 2013.

Programme summary

Although CHAs are used by non-governmental organisations (NGOs) in Angola to deliver health and nutrition messaging and by the MOH for campaigns such as polio immunisation, they are not a recognised cadre of health workers for delivering routine health and nutrition services. Despite this, given the urgent nutritional situation, the MOH approved using CHAs in a CCM approach for this CMAM project. Across the four provinces, UNICEF provided support to the health system, training staff and opening inpatient therapeutic centres (Unidade Especial de Nutricao (UENs)) to provide care for SAM cases with medical complications and outpatient centres (Programa Terapeutico para Pacientes em Ambulatorio (PTPAs))⁴ treating MAM and SAM cases without medical complications, within a 3 km radius. Other projects have found 3 km to be the maximum distance families can reasonably travel to access services. Beyond 3km, CHAs were used to deliver CMAM.  See Box 1 for a summary of the health system and programme setup.

The project mobilised a network of over 2,000 CHAs covering 21 municipalities and 76 communes. Each CHA served two to five villages (around 100 households) with responsibility for screening, treatment, referral, follow up and nutrition education.  CHAs were recruited by traditional leaders and local administrators. They were required to be literate, respected by the community, and have knowledge of health; many had past experience in community health programmes. Training in detection and treatment of SAM and MAM was provided by World Vision and partners using UNICEF materials. A strong on-the-job training and supervision component was implemented to ensure quality of implementation. Each group of CHAs was assigned to a Communa Supervisor who ensured coverage of all villages in the commune and was responsible for collating nutrition data at the end of each month. These Communa Supervisors were MOH staff, often trained nurses. At village level, delivery of services varied, with some CHAs going house-to-house while others set up a temporary site for treatment. The CHAs workload was around 2-3 days a week. 

The use of CHAs required some adjustments to standard CMAM protocols, which were agreed in advance with UNICEF and the MOH:

• Admission and discharge criteria were based on MUAC (mid upper arm circumference) measurements and oedema detection. The discharge criteria used were MUAC >12.5 cm, no oedema, and no complications. Children who met these criteria were kept in the programme for two extra weeks to prevent relapse. 
• The quantity of ready to use therapeutic food (RUTF) is normally determined based on a child’s weight on admission. As CHAs used only MUAC, individualised RUTF calculations were not possible. All SAM patients were therefore supplied with two packages of RUTF per day, and MAM patients with one packet of Ready-to-Use Supplementary Food (RUSF) per day. This protocol was based on the estimation that the majority of children with SAM were likely to be in the 3.5 to 5.9 kg range. However, the accuracy of this estimation and the impacts of the distribution protocol were not monitored.
• All SAM cases not requiring in-patient treatment should receive routine medical care. However, CHAs were not permitted to administer these, and the alternative plan was to refer children with SAM to the government health clinics in each municipality. However, this proved unfeasible and no other solution was found, despite extensive discussions. The project relied on either MOH-run campaigns or health centre administration of vitamin A and albendazole. It is unclear to what extent antibiotics were provided through these mechanisms, but it is likely many SAM patients did not receive them. Only cases with complications and severe generalised oedema (grade +++) were directly referred to the health facilities. This was due to geographic characteristics of the project areas, and the low coverage and capacity of health services.

In the second half of the project, as CHAs' workload decreased, delivery of nutrition education sessions and counselling on infant and young child feeding (IYCF) became more systematic. These activities were addressed not only to the mothers of children enrolled in the project, but to the community in general, as a means of preventing malnutrition.

Half way through the programme, it was identified that even within the 3 km radius of a health facility, children with malnutrition were not attending the health facility. Hence World Vision and partners also utilised CHAs within the 3 km radius to identify children and refer them to the PTPA.

Extensive community mobilisation took place through village traditional leaders (Sobas) to promote better feeding practices and CMAM activities, using the project’s advocacy materials. Screening activities were supported by the sobas (who conducted weekly meetings), traditional practitioners, teachers and the church leaders. This good relationship built with the community fostered strong participation and a corresponding low default rate.

Due to the small sample size in the initial survey, the high GAM rate in Zaire Province turned out to be significantly lower at 5.6% compared to the initial estimation of 19.8%. Therefore implementation in Zaire Province closed in August 2013 due to low caseload.

Programme outcomes

A total of 705,058 children were screened by CHAs between February and November 2013, an estimated 85.2% of the under-five population. During the project lifetime, 23,865 children were admitted for SAM treatment and 53,229 for MAM. This represents 9.3% of all under-fives in the intervention areas. In addition 176,144 mothers of children enrolled in the programme received training and counselling on nutrition and improved IYCF practices.

The programme had high cure rates, low defaulter and death rates, all exceeding SPHERE standards (Figure 1). An independent end of project evaluation⁵ in late 2013 demonstrated SAM cure rates of 93.8%.  There were some problems with CHA reporting, limiting the accuracy of these results, but even with a reduction of 10% as a margin of error, they remain good.

The programme maintained an average of 2,062 CHAs working out of 2,225 trained. Coverage surveys were undertaken in all three provinces, although due to the patchiness of coverage it was not possible to identify an overall coverage figure. In those areas where CHAs were particularly active, coverage was estimated to be 82.1%. However factors including limited number of CHAs available for implementation, the lack of RUTF and poor transport solutions (see challenges) meant the programme was unable consistently to achieve high coverage and reach its full potential.

Another indicator of effective implementation, median MUAC on admission for SAM cases, was 114 mm (Huambo), 113 mm (Bie), and 112 mm (Kwanza Sul), demonstrating excellent screening procedures. The low death rates also suggest that the CHAs were effective in early case finding and timely referrals to in-patient facilities. 

Key successes and challenges

Successes

The project successfully mobilised a vast network of community volunteers who were highly motivated and involved in the programme. This was recognised as an important booster to coverage in the SQUEAC survey⁶. The mobilisation of CHAs and utilisation of a CCM approach extended CMAM services to rural areas where the health system does not function and health care seeking behaviour is poor. Independent evaluators recognised that the CHAs generally demonstrated a sound ability to identify and treat SAM and MAM cases using the simplified CMAM protocol, as well as referring children with complications to health facilities. The project also fostered strong links between the Commune Supervisors and the CHAs, ensuring support for quality implementation.

This project's services were adaptable to the needs of the communities.  The close proximity of CHAs to communities enabled frequent support and monitoring. Nutrition education sessions and treatment could be provided at a time and place which suited the community and increased participation. The project achieved good integration with the existing health system when Communa Supervisors worked for the MOH, which also built MOH capacity.  

The project also raised the profile of malnutrition within the government, church and traditional leaders. Using local administrative and church leaders for social mobilisation proved a very effective strategy for engaging the communities and increasing understanding of malnutrition. Improving nutrition has not been a priority issue on the government agenda in Angola. Previous initiatives to provide treatment services for SAM have not been sustained. The CMAM project partners therefore worked hard to advocate for greater recognition of nutrition issues within national policy. A significant success was the government’s commitment to provide RUTF to MOH-run UENs and PTPAs in 2014, assuming this responsibility from UNICEF. Furthermore, nutrition was named as one of three priority health areas for the government in 2014. 

Challenges

Stock outs of RUF from the MOH were a challenge throughout the project despite advocacy efforts of World Vision and partners.  This was one of the main weaknesses of the intervention. This challenge arose because of an unrealistic expectation that the MOH could manage the RUTF supply. UNICEF received donor funding for the initial supply, but the MOH did not have the capacity to take over after this. Apart from lack of supplies at central level, distribution from the municipal to the communa level was also an issue with a lack of adequate budget for transport. CHAs also faced transportation issues to deliver RUTF in the communities. 

There was a gender imbalance in the CHA workforce with the majority of CHAs being men. The mid-term evaluation highlighted this and attempts were made to recruit more women. Nonetheless, this was a challenge for several reasons including insecurity for women travelling long distances, women's commitments in the home, and higher levels of illiteracy amongst women. The service did not appear to suffer on account of this imbalance. Male CHAs reported being comfortable counselling women on IYCF issues and there was no specific finding in the evaluations that women objected to male CHAs doing this. However, as a principle, World Vision would like to empower women to undertake these roles. Thus in future programmes, more efforts will be necessary at the design stage to ensure a better gender balance.

CHAs’ literacy, education levels and ability to report was a challenge throughout the project. What were perceived as simple tasks (e.g. recording names and numbers) posed great constraints, especially during the early stages of implementation. The programme carried out extensive capacity building to build CHA skills in assessment, referral and monitoring.   Following the mid-term evaluation, even more emphasis was placed on training CHAs and strengthening mentoring and supervision by municipal and Communa Supervisors. Improvements in CHA performance were recognised by the final evaluator. 

Adherence to international treatment protocols was not attempted or achieved, as described earlier, although overall cure rates were high. Identified children were meant to attend a health facility for the medications which form part of the standard protocol as CHA were not, due to policy constraints, allowed to dispense medications. A stronger connection between the UNE/PTPA coverage area and the community treatment provided by CHAs would have been desirable to track attendees (data on who attended are not available) and to ensure attendance. 

Learning and recommendations

Despite mobilising over 2,000 CHAs, this proved insufficient. The number of CHAs required to cover the intervention area was underestimated during the design phase due to inaccurate municipal level population figures. This resulted in patchy or low coverage. Where population planning figures are questionable, projects should plan for a review and readjustment of CHA numbers early in the implementation phase. However, this requires adequate budget flexibility to allow for additional recruitment, training and supervision if necessary. 

CHA incentives were a challenging aspect of the intervention.  Inappropriate incentives were initially introduced (mobile phone credit was proposed but due to poor mobile reception in programme areas, this proved inappropriate), then adjusted following the mid-term evaluation with different partners providing incentives (each were unique to the project partners for their respective implementation areas and resourced using private funds).  Although this led to an increase in retention and motivation of the volunteers, it is preferable for a contextually appropriate incentive or compensation scheme to be determined at the design stage and implemented consistently. It should be compatible with other initiatives, including other NGO programmes and MOH protocol for frontline workers. 

Ensuring access to a stable and regular supply of RUTF, RUSF, therapeutic milk and medication is vital.  Implementing partners need to assure buffer stocks are in place before proceeding with a CMAM project.  In addition, advocacy should take place to address supply chain blocks, ensure lines of responsibility are clearly delineated and that capacity to deliver is adequate. Ensuring routine medications are provided in this case would require policy and practice change regarding CHA permitted tasks, a workable system of mobile medical teams or a feasible and sustainable system of referral and transport to clinics for an initial check.

Finally, the challenge remains of the inability of the health system in Angola to reach 3km beyond the facility, leaving large rural areas underserved. The project has shown CHAs can be utilised to extend the reach of the health system.  Advocacy should continue for CHAs to be included in the health service to ensure better outreach. One of the desired outcomes was to integrate CMAM into municipality plans.  Despite the increased awareness of addressing malnutrition, municipalities do not have the funds to include nutrition activities in their budget. In addition, there are policy limitations regarding the remit of CHAs. The MOH appears now to be open to reviewing these issues, thanks to advocacy activities building on this project by UNICEF and World Vision. The Minister of Health commissioned a study to assess the feasibility of using a network of CHAs in the health system, including for nutrition activities and the policy has now been submitted to parliament for approval. 

Conclusions

Very few projects to date have implemented CMAM using a CCM model but this experience demonstrates the potential of this approach, particularly where health system capacity is very low. The use of a large and far-reaching network of CHAs with a CCM approach enabled remote communities to be reached with CMAM services, which appeared to be of good quality in terms of cure rates for those covered. The proximity of the CHAs to the communities also enabled close tracking of children and follow-up of treatment. Considering the limited health system capacity, a community-based approach is clearly the most effective method to reach children in Angola. Going forward, further work is needed to strengthen this CCM approach, improving the ability to adhere to the full CMAM protocol. Beyond Angola, this project raises the potential for CHWs to deliver CMAM in other contexts. Adapting and innovating the accepted CMAM model, within the boundaries of a framework of operational research and evaluation, may help close the gap in coverage of prevention and treatment services for acute malnutrition.  

For more information, contact: Mariacristina Armellin, Grants Portfolio Manager, World Vision email: mariacristina.armellin@worldvision.org.uk and Hasha Hayden Smith (Acting Senior Nutrition and Child Health Adviser), email: hasha.haydensmith@worldvision.org.uk