Challenges to effective malaria control in refugee settings: experiences from Chad and Tanzania, 2004

Malaria is an important cause of illness and death among refugees. The majority of today’s refugees live in malaria-endemic areas: of the 19 million people of concern to the United Nations High Commissioner for Refugees (UNHCR), two-thirds live in malaria-endemic areas. Many factors may promote vulnerability to malaria illness and death among refugees. As other articles in this issue describe, pregnant women and young children are particularly at risk of severe illness and death: in many refugee situations, women of child-bearing age and children make up the majority of the population. Refugee camps are often sited on marginal lands that promote breeding sites for malaria vectors. Refugees may be malnourished, and access to adequate food may be limited, particularly in the phase immediately following flight. Travel may take refugees through or to areas of higher malaria endemicity than their place of origin. Control programmes may have broken down due to the conflict that caused population flight, or may never have been implemented. This article uses two case studies from 2004 to highlight the challenges of implementing effective and up-to-date malaria control in refugee situations: an emergency situation in Chad, and a more stable situation in Tanzania.

Case study 1: Chad

More than 200,000 refugees fled into Chad from the Western Darfur region of Sudan following the eruption of fighting and systematic violence against the population in 2003. Refugees are currently settled in 12 camps along the border with Sudan. Women and children make up the majority of the refugee population.

Eastern Chad lies in a zone of unstable seasonal malaria with a short transmission season, similar to that of West Darfur, the refugees’ place of origin. The wet season extends from June to October, with up to 600mm of rain per year; rainfall decreases as one moves from south to north. The majority of the host and refugee population therefore has little or no immunity, and is at risk of severe malaria and death during the transmission season.

Health services for refugees are delivered by a number of different bodies. Local capacity in this region of Chad is very limited. Health coordination is shared between the World Health Organization (WHO), UNHCR and the Chadian Ministry of Health. Following an assessment in mid-2004, these agencies reached a consensus on malaria control interventions, including:

• standardised case definition and confirmatory early diagnosis using rapid diagnostic tests (RDTs);
• prompt treatment with artemisinin-based combination therapy (ACT); and
• prevention, using both indoor residual spraying (IRS) and insecticide treated nets (ITNs).

Amodiaquine plus artesunate was selected as first-line therapy for uncomplicated malaria in refugee settlements on the assumption that chloroquine resistance was widespread in the area. At the time of the emergency, chloroquine was still being used as first-line treatment in Chad, although a pilot ACT programme in one southern district had begun. Choroquine resistance had been demonstrated in Sudan, and the treatment protocol had recently been changed to ACT. Nevertheless, delays in procuring supplies for some agencies held up the implementation of the new protocol, and training of clinic staff was not undertaken in all camps. Not all clinicians were convinced by the new protocol or by RDTs.

Shelters in the camps and host communities were targeted for IRS. However, the approach was not standardised. In the north, one agency sent a malaria specialist to coordinate malaria control activities in three camps, and to assist in training in two additional camps. IRS was completed in northern camps ahead of the rainy season. Effective implementation was delayed beyond the start of the rainy season in the south due to delays in the procurement of supplies, a lack of trained personnel, the early onset of the rains and logistical limitations.

ITNs were distributed through a number of different delivery mechanisms and partners, and to different groups of beneficiaries. Deficiencies in monitoring and coordination resulted in duplication or, at times, lack of coverage. Failure to implement effective community information, education and promotion, compounded by food shortages, resulted in nets being resold. The exception to this was the distribution of insecticide treated shelters (dumerias) in one camp where shelters were too flimsy to support either IRS or ITNs; reportedly, the dumerias enjoyed good community acceptance and retention.

Regular health coordination meetings were established. Fully effective coordination was limited by large distances, poor infrastructure and telecommunications, insufficient professional staff to allow some to be released to attend meetings, and high turnover of personnel. However, malaria was included in outbreak preparedness plans. Monitoring and evaluation was limited by the lack of an effective common reporting system. Basic health and mortality data were not reported reliably from all camps.

By the end of 2004, no increase in malaria incidence during the rainy season was observed, and case fatality rates remained low. These observations may have perhaps been due in part to more effective malaria control measures. Strengths in the response include the consensus among partners on an intervention strategy, based on a situation analysis, and the early adoption of ACT as first-line treatment. IRS was implemented effectively in some camps.

At the same time, however, no common organisational framework and division of responsibilities was agreed. This resulted in the fragmented implementation of interventions, which weakened the overall response. Improved coordination, in the form of regular sectoral meetings, would have identified priority areas for action, as well as the weaknesses of the various agencies. A data collection system with agreed indicators would have improved monitoring capacity. An overall focal point for malaria could also be appointed to ensure the coordination of activities in all camps and communication between partners, and to facilitate the pooling of logistics, expertise and other resources.

A more coherent operational framework between partners for IRS would have identified priority areas such as the south, and allowed for the pooling and redirection of resources. Pre-positioning of equipment and supplies before the onset of the rains would have ensured a more timely response. In retrospect, the focus in some camps on prevention campaigns of probably limited effectiveness was at the expense of adequate supplies, training and supervision to ensure the provision of life-saving treatment.

A thorough assessment of the preconditions necessary for an ITN programme in an emergency may have led to the distribution being postponed until the next rainy season, except for inpatient beds. Long lead times for net procurement meant that there was insufficient time for community education and participation. In a community that is not familiar with nets, and which faces competing survival priorities, net resale is likely. In many settings, ITNs and IRS were provided to the same household, which was an unnecessary drain on resources. For the future, a common ITN distribution plan could be drawn up specifying target groups, along with distribution mechanisms, community participation strategies, logistics plans and reporting duties.

Case study 2: Tanzania

At the end of 2004, some 400,000 refugees from Burundi and the Democratic Republic of Congo were living in 14 long-standing camps in western Tanzania. Several agencies provide health services to these populations.

Malaria is the number-one cause of illness and death among children under five years of age in these camps. Transmission is perennial, with seasonal peaks associated with the rainy season (October to May). The first formal malaria assessment for refugee camps was conducted in 1998, when reports of very high numbers of malaria cases and malaria deaths prompted development of a comprehensive and effective response to malaria control in Tanzania. At the time, children and women of childbearing age, at high risk of severe malaria and death, accounted for 75% of the population.

In view of chloroquine resistance, the first-line treatment protocol for uncomplicated malaria in refugee camps was changed from chloroquine to sulfadoxine-pyrimethamine (SP). This change was ahead of national protocol. Health workers were trained on the presumptive treatment of malaria, first-line treatment was decentralised to health posts, and active case-finding and follow-up at the community level were added. At the same time, intermittent preventive treatment (IPT) in pregnancy with SP (two doses) was implemented, with high levels of antenatal coverage. These changes were incompletely implemented, there was great variation between partners and confirmatory diagnosis was carried out on an ad hoc basis.

Approaches to prevention include environmental control (draining of standing water in camps, larviciding in some camps), IRS and ITNs. The approach to prevention is not consistent: prevention has alternated between IRS and ITN distribution, or both have been implemented simultaneously. Widespread IRS implementation began in most camps in 1998. Implementation varies by agency and by camp. A variety of insecticides is used, and spraying is consistently late, after the commencement of the rains, at considerable annual cost. As for ITN, although large-scale distribution began in late 1998, by August 1999 20% of the nets were missing. Resale was common, often to provide money to add variety to or supplement inadequate food rations, and poorer households were more likely to sell nets to buy food. In settings where individual households have competing survival priorities, community IRS might be more effective in preventing malaria than ITNs. Community information, education and engagement were lacking, and this contributed to low retention rates.

Coordination was improved with the establishment of camp- and district-level malaria task forces. Standard data collection (morbidity, mortality) was implemented in 2000 in all camps, supplemented by entomological data and periodic net-retention surveys in some camps.

The Tanzania experience demonstrates that elements of an effective malaria control programme can be implemented in refugee settings: assessment, improved access to efficacious treatment, implementation of appropriate preventive measures, establishment of coordination mechanisms and standardised data collection. But it also highlights some of the practical difficulties of ensuring that control programmes are effective in reducing malaria illness and death. Late spraying of IRS is common in malaria-endemic areas. Poor ITN retention is also typical where community engagement is lacking, and where there are competing survival priorities. The treatment protocol again needs updating in the face of increasing SP resistance.

As in many stable settings, there is a need for formal periodic reassessment and mid-term evaluation. The challenge today is to implement an effective and coordinated prevention campaign across all camps, up-to-date treatment protocols with effective antimalarial drugs (ACT) and improved diagnostic coverage. Implementation plans should also include community engagement, the training and supervision of health workers, and monitoring and evaluation.

Conclusion

The Chad and Tanzania examples described here demonstrate the feasibility of implementing malaria control in emergency and refugee settings. The prompt implementation of an interim treatment protocol using ACT and RDTs in Chad should be used as a model for other emergencies. The approval of the change in the malaria treatment protocol in Tanzanian refugee camps, ahead of any changes to the national protocol, is an example that can be followed in other stable settings. Successful conduct of IRS in many of the camps in Chad, despite logistical hurdles, is testimony to a capacity among agencies for good planning and implementation, and for sharing resources and expertise. In Tanzania, a common disease surveillance system and regular surveys strengthened programme monitoring and evaluation.

Some things could have been done better. The lack of community participation, for example, had a negative effect on ITN programmes. In Tanzania, evaluation and reassessment are required before planning changes to treatment and prevention strategies. In Chad, the response was sometimes not prioritised towards life-saving needs. In retrospect, the preconditions for IRS and ITN programming were not met in all settings, and the focus should have been on implementing effective treatment. More effective coordination could have improved the response in both Chad and Tanzania. Strengthened training and supervision in implementing new treatment protocols was also required. Common indicators for programme monitoring would have improved the response in Chad.

These examples show that malaria control in refugee settings is possible, both in emergency situations and in stable environments. Adherence to common principles for up-to-date and effective malaria treatment and prevention – for Africa this includes ACT, confirmatory diagnosis, treated nets or indoor residual spraying – is essential to ensure that priority areas are addressed adequately. Ongoing documentation of field experience by practitioners will also be important to further develop and refine these common principles.

Nadine Ezard is Senior Public Health Officer, UNHCR Geneva (ezard@unhcr.ch). John Haskew is former Health and Nutrition Officer, UNHCR Chad. Lucas Machibya is Associate Public Health Officer, UNHCR Tanzania. Raoufou Makouis Medical Coordinator, UNHCR Tanzania.

References and further reading

P. Bloland and H. Williams, Malaria Control During Mass Population Movements and Natural Disasters (Washington DC: National Academies Press, 2003).

M. A. Connolly (ed.), Communicable Disease Control in Emergencies: A Field Manual (Geneva: WHO, 2005), http://www.who.int/infectious-disease-news/IDdocs/whocds200527/whocds200527chapters/index.htm.

v D. Warrel and H. Gilles (eds), Bruce Chwatt’s Essential Malariology, Fourth Edition (London: Arnold, 2002).

WHO, Malaria Control in Complex Emergencies: An Inter-agency Field Handbook (Geneva: WHO, forthcoming, 2005).

WHO/Roll Back Malaria consultation on ‘Best Practices and Lessons Learnt: Implementing Malaria Control in Complex Emergencies in Africa 2000–2004’, November 2004.