Impact of community mobilisation activities in Uganda

By David Doledec, ACF-USA

A qualified nurse, David Doledec joined Action Contre la Faim in 2002, working in Democratic Republic of Congo, Sudan, Liberia and Uganda before joining the ACF-USA headquarters team as a Nutrition Advisor in 2007.

The author would like to acknowledge the ACF-USA teams in both Kampala and Lira offices, the village leaders, community health workers, traditional birth attendants and other community members who assisted the teams in their home communities, and the individual families who participated in the survey. The author also acknowledges the work of the consultant, Dr Mark Myatt, who adapted the methodology used in this survey.

Measuring height during the ACF survey

In July 2006, ACF-USA implemented an out-patient programme for the treatment of severely malnourished children in the main internally displaced people (IDP) camps of the districts of Lira and Apac, Northern Uganda. ACF-USA nutritional centres were already present in the IDP camps where they also served the resident communities. The beneficiaries admitted at that time were 90% IDPs and 10% from the resident population.

Assessing coverage

In January 2007, ACF-USA enlisted the assistance of a consultant to assess the programme and in particular, to measure its coverage and identify weaknesses. An adaptation of the Centric Systematic Area Sampling (CSAS) methodology was developed to suit the context of the IDP camps and a coverage survey was carried out¹. The survey teams were comprised of ACF-USA local staff, who were mostly IDPs residing in the camps and who had had previously been recruited and trained by ACF USA. The surveys, which took place during a two week period in January 2007, were conducted in the main camps where the nutritional centres were located, followed by further surveys in nearby camps (one centre area was defined as a quadrat).

The results showed that the programme coverage was low^2,3:

• Point coverage: 25.8% (95% CI = 15.0%, 39.7%)
• Period coverage: 50.5% (95% CI = 39.3%, 61.9%)

Several of the main contributing factors and programme weaknesses were identified through the survey including:

• Significant interface problems between supplementary and therapeutic programmes - Many severe cases were being treated in the Supplementary Feeding Centres (SFC).
• Poor and inadequate communication with caretakers at time of referral and in the centres - On several occasions, the caretakers reported that the centre staff were eitherrude or set out to shame them. Others reported that they had been referred to a centre but were not given a clear date for attending the centres or told where to go for treatment.
• Discrepancies between the referral and admission criteria - This led to high rates of rejection. The admission criteria in the SFCs were MUAC<120 mm and WHM<80%⁴, while criteria for referral used by home visitors was MUAC<125 mm. Several referred children were rejected from centres due to these inconsistencies.

Identification of these problems resulted in a complete re-organisation of the programme. Supplementary feeding and outpatient therapeutic programme (OTPs) centres were integrated into a single unit with teams combined into one. Trainings were provided for all staff members and a community mobilisation campaign was started in March 2007, following Valid International guidelines⁵.

Monitoring progress

In order to monitor programme performance as well as the progress and impact of the community mobilisation activities, a second coverage survey, using the traditional CSAS methodology, was conducted in January 2008. However, the situation in the main IDP camps of the districts had changed since the previous coverage survey leading to substantial population changes. In early 2007, a massive return process began in the districts of Lira, Apac and Oyam. Between March and November 2007, 99.5% of the 466,103 IDP's were reported to have left the camps to return to their areas of origin⁶. Such massive population movements were likely to result in a significant impact on the programme and its coverage.

Following the 2008 survey, the coverage rates were found to be:

• Point Coverage: 37.8% (95% CI = 23.8% - 53.5%)
• Period Coverage: 54.1% (95% CI = 40.8% - 66.9%)

The survey highlighted three main reasons for severe cases of malnutrition not being enrolled in the programme. These was a lack of awareness about malnutrition and the programme, a lack of time for caretakers to attend the distribution centres, and too lengthy a distance to reach the closest centre. Problems that were identified during the 2007 survey appeared to be less of an issue. In addition, the fact that coverage rates had improved in a context where they could easily have worsened (as the population was much more spread out), was seen as encouraging with regard to the efforts that had been made to improve the program. Community mobilisation was still ongoing and seemed to show a positive impact. However, the second coverage survey threw up more questions than answers. As the surveyed communities had faced such a dramatic change with the returnee process, it was not possible to consider the measured coverage rate as an indicator of the effectiveness of the community mobilisation activities.

Evaluating effectiveness of mobilisation

Community mobilisation activities were conducted with the aim of developing a comprehensive detection and referral network at the community level. In addition, through regular meetings in all communities, it aimed to improve people's awareness of malnutrition, its prevention and the treatment programme. In order to evaluate the effectiveness of the community mobilisation activities, the team collected several indicators:

1. The referral system Figure 1 highlights clearly that the level of referral from the community increased significantly throughout the year. Community mobilisation activities had an impact on both referrals from the community volunteers trained by the nutrition team and on self-referrals. 'Other' data mainly represents referrals through nutrition surveys and screenings conducted by the team in non-catchment areas.
2. Attendance Admission trends is another set of information to consider in conjunction with data on referrals. As shown in Figure 2, monthly attendance in the programme increased significantly in 2007 compared with 2006. As a result of this increase, the team increased the number of centres from 5 to 10, as well as opening a therapeutic feeding centre (stabilisation centre) in the area.
3. Prevailing malnutrition rates An analysis of attendance in the TFP requires knowledge of the changing rates of acute malnutrition in the population. As can be seen in Tables 1 and 2, the rates of acute malnutrition, though steady in the Apac and Oyam districts, had increased slightly in Lira district since 2005⁷.
4. Other factors It is also necessary to consider important factors or events that may impact attendance and referal indicators, as well as community mobilisation itself. If massive population movements occur in areas where community mobilisation activities have already been implemented, it is likely that the same mobilisation process will need to be conducted again. Several communities in Northern Uganda increased dramatically in size during the return process. The team therefore had to go back to these communities to train more community volunteers in community mobilisation.

Conclusions

There is now a wide range of evidence showing the positive impact of community mobilisation on the treatment of acute malnutrition. It improves programme coverage, acceptance and sustainability. However, measuring the actual outcomes of community mobilisation requires an analysis of the whole programme, as coverage surveys alone cannot fulfil this objective, especially in complex and rapidly evolving contexts. A comprehensive understanding of a programme is therefore only possible through analysis of a wide range of available information, including malnutrition rates, coverage rates, referrals to nutritional centres, attendance, and context analysis.

For further information, contact: David Doledec, email: dd@aah-usa.org

¹All survey reports are available on ACF-USA website, http://www.actionagainsthunger.org under Resources

²SPHERE standards recommend 70% coverage in IDP / refugee camps settings

³CI: Confidence Interval

⁴MUAC: Middle Upper Arm Circumference, WHM: Weight for Height % of the median

⁵Community-based Therapeutic Care Manual. Valid International. Available at http://www.validinternational.org

⁶IASC working Group report, November 2007

⁷Results expressed using the NCHS reference tables (1977).