MALARIA CONTROL IN THE
LUBOMBO SPATIAL DEVELOPMENT AREA

Contents

1. Project Overview
2. Funding
3. Impact of Intervention to date
3.1 Malaria Infection Rates
3.2 Hut Exiting Malaria Vector Mosquitoes
3.3 Insecticides
3.4 Malaria Information Systems
3.5 Tourism
4. Management
5. Training
6. Developing a GIS Based Decision Support System
7. Regional Population Movement
8. Future Sustainability
9. Regional Malaria Control Commission

1. Project Overview

The Lubombo Spatial Development Initiative (LSDI) is a programme by the governments of Mozambique, Swaziland and South Africa to develop the Lubombo region into a globally competitive economic zone. The geographic region targeted by this initiative is broadly defined as eastern Swaziland, southern Mozambique and north-eastern KwaZulu Natal, an area linked by the Lubombo mountains. It also aims to create sustainable employment and equity in access to economic opportunity in the region.

In July 1999 President Mbeki, President Chissano and His Majesty, King Mswati III signed the General Protocol which put in place a platform for regional cooperation and delivery. In October 1999 the Lubombo Malaria Protocol and tri-national malaria programme was launched. In December 1999 the World Heritage Convention Act was promulgated and the Greater St Lucia Wetlands Park inscribed on the World Heritage Convention list. In June 2000 the three countries signed the Lubombo Transfrontier Conservation and Resource Area Protocols (TFCA). The malaria component of the LSDI project is managed by the Regional Malaria Control Commission (RMCC), comprised of malaria control programme managers, public health specialists and scientists from the three countries.

Although the malaria control project addresses a number of aspects central to increasing the effectiveness of malaria control in the two highest risk malaria provinces in South Africa and Swaziland, the primary emphasis was to extend malaria control to southern Mozambique. There is increasing consensus that even if malaria control measures are optimal in South Africa and Swaziland (i.e. effective drugs and insecticides in place), disease incidence could only be further reduced by a regional approach to control. There is also increasing evidence that malaria control is a positive precursor to development with the situation prior to malaria control in South Africa supporting this view, given the well documented negative effects of malaria on tourism and agricultural development in the 1930’s. The LSDI malaria programme is targeted at creating a platform for development, the beneficiaries being communities in some of the areas with the lowest socio-economic development in the region as well as tourism, business and governments.

The effectiveness of the malaria control programme in the long-term will be assessed by the incidence of malaria over time in Mozambique as well as in the neighbouring malarious areas of South Africa and Swaziland. The success will not only be measured using process (e.g. spraying coverage) and biological markers (e.g. parasite prevalence rates, health facility patient numbers and mosquito vector reductions), but also by the effects on tourism e.g. bed occupancy, job creation and risk perceptions, in all three countries over the course of the 5 year period (2000 – 2004). The malaria control component in Mozambique has been implemented in phases (Figure 1) starting with Zone 1 which is the area extending from KwaZulu-Natal to Maputo. Zone 1A is the areas surrounding the MOZAL Plant that has been sprayed as part of their social responsibility campaign. Zone 2A comprises part of the Boane District, and Zone 2 extends north to approximately ¼ way up the Kruger National Park. The regional malaria situation prior to and post intervention in Mozambique is broadly evident in Figures 1 and 2.

Figure 1. Regional distribution of malaria: pre-implementation

Figure 2. Regional distribution of malaria : post implementation

2. Funding

Funding for the project for the first two years (2000/2001) was by The Business Trust, MOZAL, The Department of Health in South Africa and the Ministry of Health in Mozambique (Table 1). Funding for 2003 is secured from these sources. A proposal to the Global Fund requesting five years of funding was submitted in 2002 and signed by the three participating countries. To date, 70% of the funding for the project has come from the private sector.

Table 1: Funding for the LSDI project for 2002

3. Impact of Intervention to Date

3.1 Malaria Infection Rates

Base line surveys have been carried out at a number of localities in the LSDI zones of Mozambique, in three localities in South Africa and four in Swaziland. This was done to document the extent of the malaria problem prior to the implementation of malaria vector interventions in the area, and to allow comparison post intervention. These included surveys on parasite prevalence, indoor mosquito numbers, and Knowledge, Attitudes and Perceptions (KAP) regarding malaria. Baseline malaria surveys will continue to be undertaken each year in June to evaluate the effectiveness of the intervention.

The baseline malaria survey conducted in southern Mozambique in December 1999 showed average malaria infection rates based on the HRP-2 antigen (children aged 2 to <15 years of age) of 64%, with infection rates of 90% at Catuane, just across the border from Ndumu Game Reserve, the highest risk area in South Africa. Infection rates in the area surrounding the MOZAL Industrial Park (Zone 1A) were also high (>85%) prior to vector control. By June 2002 these combined control efforts had resulted in a 70% reduction in malaria prevalence in children in Zone 1 and a 50% in Zone 1A (Table 2, Figure 3). The reductions in prevalence in Zones 1 and 1A are statistically significant as outlined in Table 3.

Table 2. Malaria control Zones in Mozambique indicating size in Km2, estimated population (2002), number of sentinel sites and parasite prevalence pre and post house spraying (bold).

Figure 3: Average Parasite Prevalence in children aged 2<15 years of age in Zones 1 and 1A of Mozambique, 1999 – 2002

Table 3. Statistical significance of results from Zones 1 and 1A

At 4 of the 7 sentinel sites in Zone 1 in Mozambique, the prevalence of the disease has been reduced to less than 20%, attaining the 5 year objective after just two years. Hospital admissions in Zone 1 have also shown a reduction since the implementation of vector control (Figure 4).

Figure 4: Inpatient data for malaria hospital admissions in southern Mozambique by malaria season (July to June) : 1997 –2002

Following three years of sustained effort of implementing malaria vector control in the LSDI area, the overall prevalence of the disease had dramatically decreased in Zone 1, Mozambique. By June 2001 there was a 40% reduction in malaria prevalence in Zone 1. Malaria incidence had decreased by 64% in Swaziland despite no changes in insecticide or drug policy during this period and by 76% in KwaZulu-Natal where both drug and insecticide policy had changed. By June 2002 there was a 70% decrease in prevalence in Zone 1, an 80% decrease in malaria incidence in Swaziland and a 91% decrease in KwaZulu Natal against the 2000 baseline (Figure 5).

Figure 5. Regional reductions in malaria in the 2001/2002 season in comparison to baseline indicators in 1999/2000

3.2 Hut exiting malaria vector mosquitoes

Traps to catch mosquitoes exiting from houses were fitted to homes at each of the sentinel sites. These are cleared daily by the homeowners, and the mosquitoes preserved in labeled and dated containers for analysis. Table 4 shows the number of window traps in place in each zone in 2002. Figures 6 and 7 show the average number of mosquitoes caught per trapping day in Zones 1 and 1A. Two vector species have been identified i.e. An. arabiensis and An. funestus. The vector populations in both Zones 1 and 1A showed seasonal peaks in summer prior to house spraying. Numbers were starting to increase in both areas in late 2000, early 2001, but decreased dramatically after the first spraying round, further decreasing after the second spray round and remained low thereafter.

Table 4: Window traps in place in the respective Zones in Mozambique, 2002.

Figure 6: Average number of Mosquitoes captured per trapping day in Zone 1:
1999-2002

Figure 7: Average number of Mosquitoes captured per trapping day in Zone 1A: 2000- 2002

3.3. Insecticides

The initial technical proposals identified pyrethroids as the insecticide to be used in the spraying component of the LSDI. However, with the discovery of high levels of pyrethroid resistance in An. funestus, meetings were held with the RMCC as well as national and international experts to recommend an alternative to the use of this insecticide. Based on data, it was unanimously agreed the best course of action would be to use DDT. The alternative recommendation was that a carbamate such as Bendiocarb be used. The latter was used due to Mozambique declining to allow DDT use. More recently, carbamate resistance (Propoxur) has been reported from an area in close proximity to the control area and studies are currently in place to assess the extent of the problem.

Resistance by An. funestus to synthetic pyrethroids was first detected in December 1999 in the southern area of Maputo Province within the LSDI programme, Zone 1. Since then, extensive collections have been made at different points of the country to ascertain the distribution of resistance of malaria vectors to different insecticides used in malaria control activities. F1 generation An.funestus were subjected to standard WHO susceptibility tests using all major insecticide families, e.g., pyrethroids, organophosphates, carbamates and DDT. To date, 20 locations have been sampled, and the susceptibility tests on these F1 generations showed that resistance of An.funestus to synthetic pyrethroids is concentrated in the southern region of Mozambique, in particular, the localities of Moamba, Catuane, Catembe, Bela Vista and Boane in Maputo Province. In Gaza Province, resistance of adult An. funestus was also found to synthetic pyrethroids.

North of Gaza Province, susceptibility of An. funestus (F1) was found to be greater than 96 percent in all areas. Data collections are still in progress with a focus on areas of agricultural importance. The above data has been entered into a GIS map and has provided an informed basis for insecticide choice for the LSDI malaria vector house spraying program.

Increasing levels of insecticide resistance and a limited number of available insecticides restricts the options in respect to the residual house spraying programme in southern Mozambique. Discussions emanating from the discovery of pyrethroid and carbamate resistance have emphasized the need to consider rotational insecticide use as the only way forward and to avoid fixing resistant genes in the vector population. These findings have implications for the future of malaria control in the region.

House Spraying:
Table 5 outlines the house spraying activities that have been carried out in the different Zones in Mozambique. Both Zones 1 and 1A have had 3 spray rounds since 2000 and are current being sprayed for the fourth time. Zone 2A was first sprayed in 2001 and is currently being sprayed for the second time while the first spray round is taking place in Zone 2.

Table 5 : Completed house spraying activities in the respective Zones in southern Mozambique.

Bioassays :
Bioassays were conducted in four localities to investigate the residual effect of Bendiocarb in order to determine when re-spraying is required. Ten houses in each of the four localities, Bela Vista, Catuane, Namaacha and Ponto du Ouro, were bioassayed at monthly intervals. Following one hour of exposure to the insecicide, the 24 hour mortality decreased from 100% one month after spraying to 80% at 5 months and 19.26% at 7 months post spraying. These results indicate that re-spraying should optimally take place five months after the first round of spraying

3.4 Malaria Information Systems

Malaria Information Systems (MIS) were developed and implemented for each of the partner-sectors with modifications being made on an ongoing basis. This computerised system allows the input, management and output of malaria case data which is used for both management and research purposes. It includes a spatial component using a geographic information system (GIS) which is being customised to minimise end-user skill requirements and optimise access to the different data sets. The data collected during routine operations and entered into the MIS consists of both in- and out-patient data of confirmed and presumptive cases. The input screen mirror the data collection forms and the automatic-linking and drop-down list minimising data entry errors. Pre-designed outputs are provided in the form of maps, graphs or tables (i.e. number of can refills per week per person). This allows problems to be identified and addressed on an ongoing basis.

The data is transferred to the MRC on a monthly basis where it is used for research which is fed back to the partner-sectors. Spatial data has been collected for the region and includes administrative boundaries, population, health facility locations, towns and other relevant information. Figure 7 indicates the 2002 estimated population based on the 1997 census using boundaries delineated by members of the Ministry of Health in Maputo for Zone 1. Figure 4 indicates the number of admissions recorded at two health facilities in Mozambique on a seasonal basis (July to June). The graph clearly indicates a decrease in the number of confirmed admissions following the implementation of the spraying programme in 2000.

3.5 Tourism

A tourism survey (88 tourist facilities) conducted in the LSDI area during 2000/2001, found that malaria was perceived as the principle negative determinant on bed occupancy. Cancellations were recorded from tourist facilities in all the districts of the LSDI during the 2000 malaria season, with an average cancellation figure of 44% being recorded in southern Mozambique. This was largely due to the floods early in 2000 resulting in tourists being concerned about a possible malaria epidemic which was widely reported on in the press.

A similar study to determine the influence of malaria on tourism and to assess the perception of risk amongst local and international tourists that visit the LSDI area will commence in January 2003 in Mpumalanga Province. All tourist facilities visited in the 2000/2001 study will be revisited in January/February 2003 to assess any change that might have taken place.

A tourist information booklet providing appropriate information regarding prophylaxis in risk areas has been produced (2001) and was made available to tourism facilities. Small-scale malaria risk maps are being produced in order for appropriate malaria risk information to be distributed to tourist facility owners, tourists, the media and business.

Figure 7. Estimated population by locality in Zone 1 based on 1997 Census data.

The reduction in malaria cases in KwaZulu-Natal since 2000 (Figure 8) are attributed to a number of new interventions such as:
1. the introduction of an effective drug (Atemether-lumefatrine) in KwaZulu-Natal.
2. insecticide policy changes in Mpumalanga and KwaZulu-Natal Provinces to DDT.
3. the regional approach to malaria control between South Africa, Swaziland and Mozambique and the extension of vector control to southern Mozambique is believed to have had a major influence on malaria incidence in the Lubombo corridor.

Malaria cases in KwaZulu-Natal decreased dramatically from 42 395 in 2000, to 7811 in 2001. It is anticipated that the malaria case reductions in Swaziland, South Africa and southern Mozambique will have a positive influence on tourism in the Lubombo corridor.

Figure 8: Total KwaZulu-Natal malaria cases from 1998-2002.

Figures 9 and 10 indicate the location of the tourist facilities with regard to small scale variations in malaria incidence in KwaZulu-Natal Province, and show the decrease in case incidence from 1999/2000 to 2001/2002. In the 1999/2000 malaria season, 57% of tourist facilities were in areas of > 50 malaria cases per 1000 people, and 42% were in areas where 5-25 malaria cases per 1000 people were recorded (Figure 9).

A major reduction in malaria cases in the 2001/2002 malaria season was achieved. Only 3% of tourist facilities were in areas where 5-25 malaria incidences per 1000 people were recorded and 96% where in areas were the malaria incidence was very low, with 0.001-5 cases per 1000 people being recorded (Figure 10). The reductions in the 2001/2002 malaria season indicate the positive effect of the regional approach to malaria control in the Lubombo corridor. It is essential that this information is made available through all mediums to encourage tourists to visit the area.

Figure 9: Malaria incidence per 1000 people: July 1999 to June 2002

Figure 10: Malaria incidence per 1000 people: July 2001 to June 2000

4. Management

Management of the programme consists of five tiers:
i. Tri-Lateral Ministers meeting
ii. LSDI management
iii. RMCC management
iv. Management structures in Mozambique
v. Research and control management

Management structures were set up at a Provincial and District level in Zone 1 in Mozambique that permitted the implementation of the programme with the help of external experts (RMCC and scientists) while at the same time, building capacity at both the National and Provincial level in Mozambique.

5. Training

The foundation of a successful, efficient and effective spraying programme is optimally trained staff at every level. Experience in this regard was lacking in Mozambique, and training was therefore a key priority before a spraying programme could be introduced. It was also conducted on an ongoing process once spraying had started. Table 6 indicates the number of spray operators who have been trained during the course of the initiative to undertake the indoor residual house spraying for vector control.

Table 6. Number of spray operators trained from January 2000 to December 2002 per LSDI Zone.

Training of field staff, whether spray operators or supervisors, followed a similar pattern i.e. 85% practical and 15% theory. However, supervisors received more in-depth training on environmental hazards, toxicity, first aid and safe handling/disposal of insecticides. Training of supervisors and spray persons has taken place each year. The Mozambican programme mangers assisted Mpumalanga in training their spray operators in 2002.

Training was extended to include intervention assessment and in this regard, window-trap caught mosquitoes were morphologically identified in Mozambique, and residual efficacy bio-assays carried out. The latter required the maintenance of an insectary and the ability to undertake both susceptibility and biochemical resistance testing which are increasingly being done in country and will lead to a postgraduate degree. Training has been undertaken to staff the insectory in Maputo, to equip field entomologists with the necessary research techniques, field staff in the use of global positioning system (GPS) receiver hand-held units, and office staff in the use of the MIS An important factor identified prior to the implementation of the spraying programme was the necessity to adequately supervise the spray operations. Due to the vast area to be sprayed, supervision of spray operators’ activities on a daily basis was virtually impossible. A fourth generation relational database (Microsoft access) was therefore designed as an information repository for all spraying activities and the data generated from computerized reports made it possible to evaluate productivity and spraying performance on an ongoing basis. Quality control was undertaken by the malaria control programme managers of Swaziland, KwaZulu-Natal and Mpumalanga during each spraying round.

6. Developing a GIS based Decision Support Systems

The Department of Arts Culture Science and Technology (DACST) have recently funded a project entitled “Developing a GIS based Decision Support Systems (DSS) for the Lubombo Spatial Development Initiative (LSDI)”. The project takes cognisance of the fact that malaria control, and development in general, have a spatial component which is ideally suited to be supported by a GIS based Decision Support System (DSS).

The project aims to 1) develop and implement a GIS based DSS and 2) develop malarial prediction models for the region to be used as health management tools within the LSDI.

Establishing a DSS which will consist of further development of the Malaria Information System (MIS), a repository for spatially referenced data, and a web-based spatial and statistical query and analysis tool for information dissemination. This will support;
- the extension of malaria control to southern Mozambique,
- the assessment of the intervention effects on tourism,
- the incorporation of a malaria control component into water resource management

7. Regional Population Movement

Population movement has long been recognized as a possible factor in the spread of malaria in the LSDI area. Cross-border movements between Mozambique (where malaria was not controlled for many years) and South Africa and Swaziland has been seen as a reason for the persistence of malaria in the border areas of Kwazulu-Natal, Mpumalanga and Swaziland. However, the relationship between population movement and malaria transmission has never been formally investigated in order to establish the implications of such movements for malaria control. A study has therefore been conducted on the movement of people within the region to better understand its role on the dynamics of malaria transmission from a regional perspective and is under final analysis.

8. Future Sustainability

From an operational perspective, starting a malaria control programme in a largely underdeveloped rural area, and in an area designated for industrial development, was successful, and the necessary skills to run and evaluate the control programme are in place. The future sustainability of the programme, the first regional project of this nature in Africa that aims to create a platform for development, is reliant on appropriately skilled personnel, funding, and access to effective insecticides and anti-malarial drugs. As outlined, training has been ongoing, and an appropriate skills base exists in the region to effectively implement a vector control programme based on house spraying. An application has been made to the Global Fund towards financially sustaining the programme which has funds for 2003. Implementation of effective treatment in all the LSDI areas is being phased in through the SEACAT project. In regard to continued viability of effective insecticides, it is envisaged that a rotational spraying programme will be evaluated.

9. REGIONAL MALARIA CONTROL COMMISSION

At a recent meeting held by the RMCC it was decided that absent members no longer employed in their RMCC capacity would be replaced.

MOZAMBIQUE
Ministry of Health
Avertino Barreto
Abdul Mussa
Samuel Mabunda
Elizabeth Streat
Sonia Casimiro
Manual Dinis

SWAZILAND
Ministry of Health
Simon Kunene
Quinton Dlamini

SOUTH AFRICA
South African Medical Research Council
Brian Sharp
Rajendra Maharaj
Carrin Martin
Francois Maartens
Mano Konar

University of Capetown, Department of Pharmacology
Karen Barnes
Department of Health, KwaZulu-Natal Province:
Joatham Mthembu
Isaac Hatting
Department of Health, Mpumalanga Province:
Kobus la Grange
Marlize Booman
Department of Health, Communicable Disease Control Directorate:
Devanand Moonasar