In September 2000, Rift Valley fever (RVF) was detected in Saudi Arabia and Yemen, causing human deaths and major losses in the livestock population. It was considered the first documented RVF outbreak outside the African continent, although the virus could have been endemic in the wadi zones for some years in cryptic foci.
The first reports of widespred abortion in sheep and goats in Saudi Arabia were in August and early September 2000. Human cases of RVF were first recognized on about 11 September, and the virus was isolated and identified by the Center for Disease Control and Prevention (CDC), Atlanta, on 19 September and at the Ministry of Agriculture Virology Laboratory in Jeddah.
Epidemiological features
As usual, the onset of Rift Valley fever was characterized by many abortions in sheep, goats, cattle and camels. The first cases were encountered at Al Humayrah, 70 km east of Jizan (see map), where over 90 percent of the pregnant sheep and goats aborted.
The clinical disease was not usually observed in the adult sheep or goats before the abortions. In a two-week period, 2·699 abortions and 943 deaths were recorded, mostly in sheep and goats, and the total number of abortions has been estimated to be in the order of 8-10·000.
The appearance of RVF in some flocks has been dramatic, with 60-90 percent of the pregnant females aborting within a period of 10-14 days. In the drier wadi zones to the north, the number of abortions reported was fewer, with only 5 to 20 percent of the flock affected.
A constant finding has been the random nature of the occurrence of the disease, in a village or community, where some flocks are affected and others not at all. This is reflected in the serological results. Indeed, in one area of 20 flocks, only seven were shown to have been infected with RVF. This situation has been encountered in all RVF epizootic situations in Africa and Egypt.
Sequencing shows that the virus strain is closely related to that isolated in the Horn of Africa in 1997-98.
Small ruminants in infected areas
Geographical distribution
Clinical observations suggestive of RVF in animals were made at many widely distributed foci throughout the Jizan region. The outbreak appeared to be multifocal in origin and was coincident with similar RVF outbreaks in Yemen.
The cases in humans and animals were invariably associated with the wadi systems (the alluvial floodplains of the rivers as they emerge from the mountains) and were particularly serious at the upper watershed of the Jizan dam. Most of the problems were recognized in the regions close to the mountains, where terraced agriculture utilizes the rainfall catchment in the narrow wadi clefts and valleys.
Cases were also seen in the more isolated wadi systems, higher up into the mountains. No RVF cases were seen in sheep or goats kept on the mountains themselves nor in the dry sandy Tihama regions. The wellhead irrigation systems have not been associated with RVF outbreaks; the surface water generated by these systems persists for only short periods of time and does not allow sufficient generation time for mosquitoes.
The northern limit for the disease was originally thought to be at latitude 17°75' N, but it has become apparent that RVF cases have been occurring much further north, up to and possibly beyond latitude 19° N. There is some evidence to show that abortions started in August in some areas. A decline in the number of new cases was noticed in the second and third weeks of October.
Typical ecosystem close to the mountains where RVF is prevalent
Incidence in humans
There have been 70 deaths in Jizan Province and approximately 400 cases have been confirmed by ELISA (IgM·+ve), many with severe clinical signs including visual disturbances. Cases in humans appeared to be more numerous in the Al Aridah township area, which was one of the worst affected. Most of the men sleep outside at night, which was said to be because of the absence of electricity, whereas women sleep inside; the majority of the cases were seen in males. Although cases are occurring in village people associated with livestock in their daily work, no slaughterhouse staff nor veterinarians appear to have been infected. Here, the Jizan dam and lake, holding up the waters of the Jizan wadi, have created a higher water table over a wide area upstream. This in turn has created more mosquito breeding sites and greater vector populations.
In September 2000, suspected RVF involving human deaths and abortion storms in animals was reported from Hodeidah Governorate, El Zuhrah district in the area of Wadi Mawr.
This outbreak was reported at the same time as the one in Saudi Arabia. More than 90 percent of the confirmed cases have been within the area encompassed by the north and south canal systems in this wadi where there is an abundance of standing water pools in all irrigated areas ready for cultivation and numerous small rainfall pools. The water pools are associated with the river flow in the wadi bed, with irrigation channels which flood fields and with habitation.
Epidemiological features
An ephemeral fever-like disease was reported in Wadi Mawr on 5 September, and abortions in sheep and goats on 9 September. Many more abortions with some mortality in young animals were again reported from 15 to 25 September. In addition, towards the end of August and in early September, there were abortion storms in sheep and goats, when up to 90 percent of the pregnant animals in a single flock were affected. Although no adult animals died, some deaths were seen in very young lambs and kids and others up to six months of age. Few cattle aborted, but some young calves died after showing signs of RVF (some 30 percent of the total livestock population of Yemen are to be found in Hodeidah Governorate).
Animal sera have been tested for IgM to RVF virus, including those collected by the investigation teams in their work in the immediate environment of El Zohrah, sera sent in from many parts of the country where RVF was suspected, and some collected as a part of the wadi sero survey. These results have shown that recent RVF virus activity has occurred in Sada'd, Hajah, Dahrar and Hodeidah Governorates.
Incidence in humans (WHO Weekly Epidemiological Record, No. 48. 1/12/2000)
Between 7 August and 7 November 2000, 1·087 suspected cases were identified, including 121 (11 percent) people who subsequently died. These figures do not reflect all ongoing transmission owing to the inability of teams to visit all the affected areas. The clinical spectrum of disease is typical of that associated with RVF and includes patients with haemorrhagic disease, encephalitis, retinitis and uncomplicated RVF. Most patients (75 percent) reported exposure to sick animals, handling an aborted foetus or slaughtering animals in the week prior to illness. Of the 490 patients who underwent serological testing, 136 (26 percent) had IgM class antibody to RVF virus and 17 (3 percent) had weak reactions to the serological tests. Serologically confirmed disease transmission has been detected in 16 districts throughout the coastal plain and adjacent mountains.
RVF virus has not previously been recognized in Yemen and Saudi Arabia. Whether this current outbreak of disease is the result of a recent introduction or is an extremely rare epizootic occurrence in an existing enzootic area is not known at present. The simultaneous appearance of foci of RVF activity in Jizan Province of Saudi Arabia and in many of the wadis in Yemen, from the Saudi border southwards, does suggest the simultaneous emergence of RVF virus at widely distant foci. This is characteristic of RVF virus emergence in epizootic areas in Africa in response to regional climatic conditions.
Although it is well known that favourable climatic conditions characterized by persistent and above average rainfall are essential for the disease to reach epizootic proportions, the 2000 rainy season does not appear to have created such dramatic conditions. Interpolated estimated rainfall data generated by the Climate Prediction Center (CPC) of the United States National Oceanic and Atmospheric Administration (NOAA) were extracted for Al Humayrha region (a 100 km2 study area centred on Al Humayrha - average rainfall estimate by month from 1995 to 2000), which was the first recognized focus and one of the most affected areas in Saudi Arabia. It shows above average rainfall in 1996 and 1997 which reverted to normal during the next years. The study of Vegetation Index images, derived from data from the Advanced Very High Resolution Radiometer (AVHRR) instrument on board the NOAA polar orbiting satellites, reflects the same kind of pattern.
However, it is also believed that the area of most interest in terms of rainfall patterns could be the catchment area, further east in the mountains, which might reveal different conditions during the 2000 rainy season.
Although the above-mentioned satellite images did not show climatic conditions as dramatic as expected for this period with regard to such an epizootic, swamps and flooded areas present in different parts of the country definitely provided good conditions for mosquito breeding.
Typical swamp area susceptible to mosquito breeding
|
The Tihama describes the whole of the coastal plain in the Arabian peninsula in the west and southwest of Saudi Arabia and Yemen. It consists of low hillocks bordering the north-south mountain chain. Wadis or river valleys occur in the Tihama and constitute the alluvial floodplains of the rivers as they emerge from the mountains. These soils are sands, loamy silts and clays, of very low salt content, with little humus or nitrogen. The first ecozone is of Panicum and Cyperus grasslands, which are seasonally flooded. In many areas, terraced catchments have improved the utilization of water from the rivers at these points. The second and larger zone borders the rivers as they proceed to the sea, and consists of Acacia zizyphispina and Dobera spp., with some grasses and bare earth. The riverine zones are characterized by the presence of Dactyloctenium grasses and are altered in some wadis by lateral canal systems extending north and south of the rivers. The rivers are seasonal in flow, largely fed by the mountain catchment areas, and are dry for much of the year. Towards the sea is a belt of land with a high salt content, planted with Salsola spp. The Tihama biotope has been the focus of extensive agricultural development over the last 20 to 30 years as greater use is made of the available water resources for cultivation. These changes have a direct impact upon the environment, creating a more extensive habitat for RVF mosquito vectors. In Yemen, for example, the largest wadi (Wadi Mawr) covers some 18·000 hectares watered by the canal systems. The methods used for the utilization of the spate flow in the wadis are very similar in Saudi Arabia and Yemen. Agriculture is practised in the riverine alluvial deposits and surrounding sandy soils. Water flow is directed by channel systems into field units, and new areas are flooded sequentially. Together with the rainfall, this results in many large and small water pools suitable as breeding sites for certain mosquito species. The changes in the wadi systems, which have been made to ensure more effective use of the available water, are also those that favour the development of more extensive breeding sites for the mosquito species that are believed to be of greatest importance in amplifying and transmitting the RVF virus. Additional ecozones where primary RVF virus amplification may take place following the emergence ofAedes mosquitoes are in the wet highland plateau grasslands planted with Acacia combretum and allied species. These are found in Thaiz and Ibb Governorates and possibly also in Sa'dah. The virus may also be expected to occur along the wadi river beds far up into the mountain zones, especially where these broaden out into alluvial plains with pockets of clay soil. |
Control measures were based on vector-control (insecticide spraying) associated with restrictions on animal movement and sensitization campaigns to limit the spread of the disease in the human population.
In order to improve the early detection of Rift Valley fever in West Africa and the control of potential epizootics of the disease in the future, a regional disease surveillance system has been implemented in Mali, Mauritania and Senegal through a FAO Technical Cooperation Programme project (TCP/RAF/8931). The project started in April 2000, before the beginning of the rainy season, with the following objectives:
The table on p. 10 summarizes the approach adopted, including the information collected, the tools employed and the main expected outcomes of the project.
In the context of the National Animal Disease Surveillance Systems, a network of sentinel herds (small ruminants) was established in the three countries at the beginning of the project. Sentinel herd locations were chosen in potential high risk areas, based on ecological considerations and the suitability of these places to harbour the virus (proximity to rivers, swamps, dams, etc.). An average of 30 animals per herd were sampled each time and all the animals sampled were clinically examined by field agents. Collected serum samples were analysed at national level by central veterinary laboratories (namely CNERV-Mauritania, LCV-Mali and LNERV-Senegal). These samples were tested for IgM and IgG antibodies to RVF in order to reveal recent and past viral circulation. A total number of 31 herds were visited five times during the period and around 5·000 samples were analysed.
The principle was to discard the animals in the herd which were IgG positives and keep sampling the susceptible animals only (IgG free), to reveal a recent viral circulation. A regional database was set up at the coordination unit in Dakar to record and analyse all the data related to the disease (serosurveillance surveys, suspicions and outbreak notification) generated by the project. Historical data and results of past serological survey campaigns which started ten years ago in Senegal will also be computerized before the end of the project, in order to highlight the long-term disease trends.
The necessary feedback to the project's stakeholders, and more particularly to decision-makers, was given through a newsletter which was distributed after the first results were obtained. The three bulletins (a fourth one is pending) have been key elements in information sharing at regional level and risk management. A geographical representation of the results obtained from serological surveys was presented in each bulletin in order to give a better understanding of the situation in terms of viral circulation.
Communication and training materials (a booklet, video and poster) were produced by the project to raise local awareness of Rift Valley fever consequences on livestock and human health as well as to train field agents in disease recognition.
|
Rift Valley fever surveillance system in Mali, Mauritania and Senegal (TCP/RAF/8931) |
|||
|
OBJECTIVES |
INFORMATION REQUESTED AND COLLECTED |
TOOLS |
OUTCOME |
|
EARLY DETECTION |
Viral circulation Clinical manifestations |
Sentinel herds (IgM antibodies) Active and passive disease search |
Disease analysis RVF alerts in Mali -> Joint veterinarians/physicians field investigation mission and risk assessment No recent viral circulation in Senegal and Mauritania |
|
RISK ANALYSIS |
Climate indicators Epidemiological data Current and historical data (past outbreaks) - results of serological surveys |
Climate forecast models available on the Web Satellite images provided by FAO-Artemis (NDVI, rainfall estimate) National Veterinary Services - TCP Mauritania |
Risk assessment in Bulletins Nos 1, 2 and 3 Risk rating |
|
INFORMATION SHARING at regional level |
Above-mentioned information |
Bulletins Nos 1, 2 and 3 |
Increased regional awareness Early reaction Risk management Emergency preparedness |
|
DATA RECORDING AND ANALYSIS |
Above-mentioned information |
Regional database linked to a geographic information system (GIS) |
Past and present occurrence of the disease recorded Disease trend - predicting models |
|
COMMUNICATION AND TRAINING |
Posters, video, booklet, radio messages |
Increase in the number of suspected cases |
|
Sentinel herd monitoring in Mauritania by a field agent
Rift Valley fever epizootics generally appear following an association of favourable conditions, including the combination of extreme climatic events such as El Niño, above average rainfall, changes of hydrological conditions (dams, changes in irrigation scheme, etc.), overall immunity of the animal population (referring to its susceptibility to the disease) and the general conditions of the animals (especially after a period of drought). The disease is frequently detected at the end of an epizootic, when the disease in humans is reported and cannot be related to another pathology. When reaching this advanced stage of an epizootic, control measures are very often useless and the disease dies out by itself. Retrospective studies are often carried out but major difficulties are encountered in assessing the real impact of Rift Valley fever on the livestock sector in comparison with other pathologies which developed in the same context.
The RVF regional surveillance system currently implemented through the FAO project aims at strengthening the capacity of national veterinary services to detect the early signs of the disease and react accordingly should the disease occur. The decision for action stems from objective information generated by the project (laboratory results and reported suspicions) and the perception of a risk given by environmental parameters such as the rainfall pattern, climate forecast and NDVI data.
Viral circulation in 2000
The situation in 2000 was particularly quiet, especially in Mauritania and Senegal where the virus does not seem to have circulated during the whole rainy season. After two years of viral activity, especially in Mauritania, no IgM antibodies were detected in sentinel herds of both countries from June to October. The epizootic cycle in Mauritania in fact began in 1998, with some recurrent activity in 1999 (due to persistent favourable climatic conditions). This period of activity ended in 2000 with conditions back to normal and a good general immunity coverage in the livestock population.
However, some very low level viral circulation has been encountered in Mali in July 2000, although the disease has not been detected. Four animals out of 30 were found to be IgM positive, and this seropositivity was confirmed after the animals were retested.
IgG antibodies were detected at different levels in the three countries. Moderate to high levels of seropositivity were found in Mauritania, where in one region up to 57 percent of the animals (20 animals out of 35) were found to be IgG positive in one herd (Hodh Ech Chargui).
The detection of IgG antibodies in 2000 is likely to be related to the above average rainfall in the previous rainy season, which reverted to normal in 2000. Interpolated estimated rainfall generated by the Climate Prediction Center (CPC) for the region of Hodh El Gharbi in Mauritania shows a peak in 1999 (see graph below left).
Estimated rainfall images can also be derived from the dekadal cold cloud duration (CCD) images, using a technique developed by the TAMSAT group of the University of Reading in the United Kingdom. The estimations are based on a linear regression between CCD and rainfall using historical data, as opposed to the "real time" interpolation approach adopted by the United States NOAA Climate Prediction Center, and can be displayed as maps. With images available back to 1988, average estimated rainfall has been calculated over the period 1988 to 2000. Using Windisp software developed by FAO, a map expressing the difference between August 1999 (3rd dekad) and the average data obtained over the past 12 years was created (see map below). This model shows that above average rainfall was expected (>27 mm above average) in the eastern part of Mauritania in August 1999.
Although at a lower level than in Mauritania, IgG antibodies were detected in Mali in most of the sentinel herds. Only one herd did not show any antibodies.
Increased awareness and improvement in emergency management practices
The National Animal Disease Surveillance Systems already in place in each country were supported in their task by the project activities, which resulted in a significant increase in disease awareness. Through the passive surveillance system, several suspected RVF outbreaks in humans were reported from Mauritania (Trarza region and Hodh El Gharbi) and suspicions of RVF in animals were reported from Senegal. All the suspicions investigated were found negative.
Emergency management practices were improved and actions were taken on the basis of laboratory results obtained by sentinel herd monitoring. After detection of viral circulation in Mali (IgM antibodies detected in July), a joint mission involving the veterinary services and the Ministry of Public Health went immediately into the field to investigate the suspicions. No clinical signs of Rift Valley fever were found during the investigation and field agents closely monitored the situation throughout the rainy season. Appropriate advice was given to the farmers to report to the veterinary services any abnormal event, such as stillbirth and abortions, and to avoid contact with infectious material. Follow-up activities showed, several months later, that the infection within the herd did not evolve and that the situation was under control.
The prompt reaction of the veterinary services showed that there has been a positive impact on the decision-making process through a better assessment of the risk associated with RVF viral circulation. As observed during the investigation mission in July and the surveillance period covering the rest of the rainy season, it is important to understand that RVF viral circulation is not always associated with the clinical signs of the disease, showing that the expression of the disease is the complex result of risk factors present at the same time. Consequently, a multifactorial approach must be adopted to understand the epidemiology of the disease and to set up a RVF surveillance system.
Rift Valley fever geographical distribution
The disease was observed in Senegal and Mauritania in the past and there is serological evidence of low level viral circulation in the three countries.
Since 1987, the most severe form of the disease has been seen in Mauritania where several human deaths were experienced in 1987 and 1998. Most of the outbreaks were reported from the dry Sahelian Acacia savannah, as far north as the Tagant region. Although the outbreak in 1987 was related to the installation of the Diama dam on the Senegal river basin, it is still unclear what are the determining factors of emergence of Rift Valley fever epizootics in the region. More recent oubtreaks show that above average rainfall is likely to contribute also to the expression of the disease in this part of Africa.
Now that climatic conditions are back to normal, weather patterns and major changes in hydrological conditions should be closely monitored during the next two to three years in order to detect any changes which could foster another epidemic. Vector-borne disease monitoring and surveillance of environmental parameters are also fundamental at this time of global warming and extreme weather events.