The intersectoral response to the malaria epidemic in Ethiopia in 2003: an assessment

This article reviews the epidemic transmission of malaria in Ethiopia, and the indicators that warned the Ethiopian government about the likelihood of an epidemic from April to December 2003. It examines the responses undertaken by different actors to the epidemic, and places the timing of the response in the context of the epidemic season. Finally, it indicates lessons learnt, with a view to increasing the effectiveness of future responses in similar situations.

Malaria transmission in Ethiopia

According to the Ethiopian Federal Ministry of Health, malaria in Ethiopia ranks first in communicable diseases, with 65% of the population living in malarious areas. The problem is compounded by more frequent epidemics, combined P. vivax and P. falciparum infections, and increasing drug and insecticide resistance. Depending on location, malaria transmission is perennial and endemic (with a transmission period of 7–12 months), seasonal and endemic (4–6 months) or strongly seasonal or epidemic (1–3 months). ‘Endemic areas’ are defined as ‘areas with significant annual transmission, either seasonal or perennial’. An epidemic, by comparison, is the occurrence of disease cases in a population or region that clearly is in excess of normal expectancy. The term ‘epidemic’ may acquire different meanings according to the timeframe in which it is considered. In Ethiopia, in the ‘epidemic areas’, high numbers of malaria cases are expected during particular months, and very few or none during the rest of the year. This is considered strongly seasonal or epidemic transmission. However, some epidemic-prone areas may not experience any seasonal epidemic in some years if environmental conditions are not conducive. Aside from this seasonal variation, there is also an inter-annual variation in the number of malaria cases. In Ethiopia, it seems that there is a major malaria epidemic every eight to ten years, and concerns that 2003 would be a major epidemic year prompted the intersectoral response this paper reviews. The epidemic occurred at different times depending on the region, but overall it is believed to have started in April, and lasted until the end of December.

Warning indicators

In April 2003, many areas of Ethiopia began reporting an abnormal number of malaria cases. In July, the Health Ministry indicated that a major epidemic (‘an inter-annual epidemic’) could occur in the coming months, and prepared an emergency malaria project proposal for submission to donors.

The Ministry’s rationale was based on meteorological indicators, ten-year caseload data from the three largest regions indicating an emerging epidemic, malaria case trends reported by regional health bureaus and vulnerability indicators. These indicators included factors such as drought and malnutrition, low background immunity between inter-epidemic years, population movement and a recent water harvesting development policy that had increased the number of mosquito breeding sites.

Stakeholders’ roles

Most donors opted to respond to the epidemics through the existing health system, with UNICEF as the lead agency. The donors most involved in the response were, by decreasing order of financial contribution, the US, the UK, the Netherlands, Norway, the World Health Organisation (WHO) and Ireland. From the approximately $7.8 million provided by donors to respond to the malaria emergency, 88% (almost $7 million) was channelled through UNICEF. UNICEF played a key role in coordinating the response, procuring the bulk of malaria drugs and supplies and providing the majority of operational funds requested by the regions for training, indoor residual spraying, outbreak investigation and response, social mobilisation and supervision. In addition, UNICEF appointed an expert in malaria epidemic forecasting and detection to determine the existence, likely duration and extent of the epidemic. According to the consultancy report, malaria epidemics were observed in all drought-affected areas in which health facility data for 2003 were obtained, and extensive supporting primary evidence from a range of organisations suggests that the observed malaria epidemics were not a local phenomenon.

From the government side, the lack of an official declaration of a malaria emergency or epidemic by the Ministry of Health created an accountability vacuum at all administrative levels: regional, provincial and district. Government regional health bureaus (RHBs) were decisive in accessing UNICEF funds for operational costs. The only conditions on accessing these funds were the liquidation of previous UNICEF funds and the preparation of a plan for the activities to be undertaken at provincial level. Delays in satisfying these two requirements were the main causes of late access to operational funds by regional authorities.

NGOs also played an active role: Médecins Sans Frontières, Merlin, Care, World Vision and Goal supported the provincial and district health desks and offices in their areas of operation. This support mainly consisted of coordination assistance, logistics, outbreak response and the provision of drugs and supplies. Most NGOs met intervention costs from their own budgets, or without prior donor commitment.

Strategy

The strategy of the response was agreed by the Ministry of Health, WHO and UNICEF at the beginning of October. It consisted of rapid anti-malaria drug distribution, outbreak response, indoor residual spraying (IRS), insecticide treated net (ITN) distribution and community mobilisation and education.

Rapid anti-malaria drug distribution

During the malaria epidemics, health authorities, NGOs and community representatives raised concerns about the efficacy of sulphadoxine-pyrimethamine (SP), the first-line anti-malarial drug in Ethiopia for uncomplicated P. falciparum. In addition, the Drug Administration and Control Authority in Ethiopia did not allow the import of non-registered anti-malarial drugs during the epidemic, and the Health Ministry discouraged attempts to introduce effective drugs (ACT, artesunate suppositories and artemether injections) on an emergency basis. Therefore, SP accounted for the bulk of the drugs ordered by UNICEF for the epidemic.

Generally, the expected impact of rapid anti-malaria drug distribution on morbidity and mortality was greatly diminished by delays in delivering drugs to villages, the late availability of operational funds for the regions (which would have eased drug distribution problems at lower levels), the inappropriateness of the malaria diagnosis and treatment guidelines for Ethiopia in an epidemic context, and the reluctance of the government to introduce effective drugs.

Donors provided substantial funds to ensure adequate drug supplies. Overall, 16.5 million anti-malaria treatments were ordered between July and November 2003, and 13.8 million treatments had been distributed to the provinces by the end of December (see Figure 1). Most drugs did not reach community levels until the end of the epidemic season; 2.8 million treatments (17% of the total requested) did not arrive in the country until after the end of the epidemic season.

Outbreak investigation and response

Outbreak investigation and response was complicated by the considerable distances between health centres and affected communities, inadequate protocols for early response, many districts’ lack of knowledge or adherence to the guidelines on malaria control and the general problems of drug availability. In addition, the outbreak investigation and response was dependent on funds to cover health workers’ costs and vehicle rentals. As most operational funds did not arrive at the district level before mid-November, the outbreak response was weak, and dependent on the ability of local authorities to divert funds and mobilise NGOs in affected areas.

Indoor residual spraying

Indoor residual spraying (IRS) in Ethiopia has long been one of the most popular interventions for malaria control. IRS requires high coverage if any protection is to be achieved, and the required coverage must be achieved before and throughout the transmission season. Timeliness is particularly important in the control of epidemics. Therefore, the requirement that effective coverage should be maintained during the entire transmission season implies that spraying of the whole area to be protected should be completed before the beginning of that season.

For a variety of reasons, it is very unlikely that effective spraying operations were undertaken during this malaria epidemic season. Malaria transmission in Ethiopia’s seasonal epidemiological settings usually ends in December, IRS coverage must be achieved beforehand and maintained during the transmission season, operational funds were not available to the regions until November, financial flows to districts were slow, donors were reluctant to provide funding because IRS guidelines in Ethiopia involved the use of dichlorodiphenilchloroethane (DDT), an insecticide banned in the US in 1972, and organising spraying teams is time-consuming.

Insecticide treated net (ITN) distribution

While ITNs are effective once they are in the field and used correctly, they are not recommended as a first-line strategy for acute emergencies because: (i) they take a long time to arrive in-country and take up a large proportion of the budget; (ii) achieving a high geographic coverage requires complicated logistics; and (iii) achieving effective usage and avoiding resale and inappropriate use all require ongoing community mobilisation.

There is no tradition of ITN use in Ethiopia, and UNICEF started the implementation of a countrywide campaign to increase insecticide treated net coverage through a revolving fund stock with pre-treated nets in 2000. To respond to the emergency, UNICEF had ordered 290,000 pre-treated nets, but the agency decided to mobilise its ITN stock to rapidly obtain coverage, and integrate it with the ordered ITNs after arrival.

An estimated 57,000 ITNs were distributed during the malaria epidemics. Most of these were old pre-treated nets with uncertain insecticide content. ITN leakage was reported in Somali Region, and beneficiary targeting and effective use were not monitored. Even in ideal circumstances, this intervention could only have provided protection to 57,000 households, accounting for about 20% of the ITNs envisaged for the malaria emergency. This supports the view that only immediately available stocks of long-lasting ITNs in emergencies will be effective in stopping an ongoing epidemic.

Community mobilisation and education

Although community mobilisation was planned for district levels, operational funds did not arrive at these levels before mid-November. Thus, the degree of timely mobilisation achieved through donor financing was low. Here again, the degree of community mobilisation achieved during the malaria epidemic season depended mainly on the initiative of the provincial and district health desks to divert existing funds and to mobilise local stakeholders.

Chronology of the response

An analysis of the response chronology indicates that all the key steps in the malaria response were significantly delayed. This translated into late regional access to drugs, ITNs and operational funds. By the end of November 2003, only 50% of the funds made available for the malaria response had been used. Figure 2 illustrates the funds used versus each donor contribution by the end of November.

Lessons learned

Many constraints of the response to the 2003 malaria epidemic season relate to chronic structural deficits within the existing health system, which require long-term investment and development. The main challenges specific to this epidemic response were:

• A lack of consistency between the Ministry of Health’s request for donor help and its own commitment to the malaria response, as shown in the failure to declare an epidemic, and a reluctance to accept measures such as ACT to deal with the emergency.
• Problems of coordination, leadership and clarification of roles in the response.
• Limitations in regional capacity to absorb and utilise large amounts of cash rapidly and effectively, and slow accounting of funds advanced for other activities, such as polio and measles campaigns.
• Lack of a fast-track system for customs clearance of emergency life-saving supplies.

The main lessons learned during the 2003 malaria epidemic response in Ethiopia are:

1. A comprehensive and effective response cannot be implemented without sufficient government commitment.
2. A clear management and leadership framework should already be in place, in which roles and responsibilities are clearly set out.
3. Disbursement of emergency funds to main emergency implementers should not be contingent on the prior liquidation of other programme funds.
4. A contingency stock of emergency items should be pre-positioned.
5. An assessment of possible funding constraints should be made ahead of time.
6. A fast-track customs system should be developed before the next epidemic.
7. The Federal Ministry of Health should re-evaluate its outbreak response guidelines and examine its drug policy before the next epidemic.
8. Systems for ensuring effective and continuous surveillance should be in place.
9. NGOs should receive adequate financial support to respond to epidemics.

Yolanda Barbera Lainezwas malaria emergency technical advisor for the Mentor Initiative from September to December 2003, seconded to the Disaster Assistance Response Team (DART) of the US Agency for International Development’s Office of Foreign Disaster Assistance (OFDA). Her email address is: yolabarb@tiscali.it.

References and further reading

World Health Organisation, Malaria Early Warning Systems: Concepts, Indicators and Partners. A Framework for Field Research in Africa, 2001.

S. Hay, Malaria in Ethiopia 2003/2004: Report of an Individual Consultancy to UNICEF Addis Ababa Country Office, 2003.

J. A. Najera and M. Zaim, Malaria Vector Control: Decision Making Criteria and Procedures for Judicious Use of Insecticides, WHO, 2002.