26 Crimean-Congo haemorrhagic fever

Crimean-Congo haemorrhagic fever (CCHF) is an acute disease of humans, caused by a tick-borne virus which is widely distributed in eastern Europe, Asia and Africa. Cattle, sheep and small mammals such as hares undergo inapparent or mild infection with transient viraemia, and serve as hosts from which the tick vectors of the virus can acquire infection. Despite serological evidence that there is widespread infection of livestock in nature, infection of humans is relatively uncommon. Humans acquire infection from tick bite, or from contact with infected blood or other tissues of livestock or human patients, and the disease is characterized by febrile illness with headache, malaise, myalgia, and a petechial rash, frequently followed by a haemorrhagic state with necrotic hepatitis. The mortality rate is variable but averages about approximately 30%. Inactivated vaccine prepared from infected mouse brain was used for the protection of humans in eastern Europe and the former Soviet Union in the past, but the development of a modern vaccine is inhibited by limited potential demand. The voluminous literature on the disease has been the subject of several reviews from which the information presented here is drawn, except where indicated otherwise (Chumakov 1974; Hoogstraal 1979, 1981; Watts et al. 1989; Swanepoel 1994, 1995; Swanepoel and Burt 2004; Burt and Swanepoel 2005; Whitehouse 2004; Ergunol and Whitehouse 2007; Ergunol 2008).

Descriptions of a disease in eastern Europe and Asia resembling CCHF can be traced back to antiquity, but a condition given the name Crimean haemorrhagic fever was first recognized in an outbreak affecting about 200 soldiers and peasants who were exposed to ticks while harvesting crops and sleeping outdoors on the Crimean Peninsula in 1944. In the following year it was demonstrated through the inoculation of human subjects with filtered suspensions of ticks and tissues of patients, that the disease was caused by a tick-transmitted virus. However, the virus itself was only isolated in laboratory hosts, namely mice, in 1967. In 1969, it was shown that the agent of Crimean haemorrhagic fever was identical to a virus named Congo which had been isolated in 1956 from the blood of a febrile child in Stanleyville (now Kisangani) in what was then the Belgian Congo, and since that time the two names have been used in combination.

During the three decades which followed the initial description of the disease in the Crimea, the presence of the virus came to be recognized in many east European and Asian countries, in some instances as a result of the conducting of deliberate surveys, but often as a consequence of the occurrence of nosocomial outbreaks or large epidemics, many of which were precipitated by circumstances which involved the exposure of large numbers of humans to ticks, such as the implementation of major land reclamation or resettlement schemes in Bulgaria and the Soviet Asian republics. Although an outbreak involving 90 cases of the disease occurred in Khazakstan in 1989 (Lvov 1994), large epidemics are now apparently less frequent in countries of the former Soviet Union, with the decrease being ascribed to the adoption of more intensive agricultural practices and the reduction of populations of the wild hosts of the tick vectors by hunting. More recently, outbreaks of the disease in Eurasia have resulted from the exposure of people to blood and ticks from slaughter stock imported from Africa and Asia to countries in the Near East such as Saudi Arabia, the United Arab Emirates and Oman, plus large scale exposure of war refugees to outdoor conditions in Kosovo, Albania, Macedonia and the Afghanistan-Pakistan border area (El Azazy et al. 1997; Khan et al. 1997; Papa et al. 2002a, 2002b; Scrimgeour et al. 1996; Williams et al. 2000; Avšicˇ-Županc 2007). Occurrence of the disease has also been confirmed in Turkey and in Iran, where the presence of the virus had long been suspected on serological grounds (Karti et al. 2004; Chinikar 2007). More than 1,100 cases of CCHF have been recorded in Turkey since 2002, and it is believed that an increase in tick populations was triggered by climate change, altered grazing practices and prohibition of the hunting of wild hosts of ticks (Vatansever et al. 2007).

In Africa, only 15 cases of the disease were reported prior to 1981, eight of them laboratory infections, and only one patient had developed haemorrhagic manifestations and died. Since then, sporadic cases of haemorrhagic disease and deaths have been diagnosed regularly each year in southern Africa, probably as a result of increased awareness among clinicians, and a few cases of severe disease have also been recorded West and East Africa (Dunster et al. 2002). Contrary to earlier speculation, therefore, it is now evident that the disease which occurs in Africa is no less severe than that in Eurasia.

It is evident from surveys conducted in Africa and Eurasia that there is extensive circulation of the virus in livestock and wild vertebrates, with very high antibody prevalence rates occurring in adult livestock in some areas. In contrast, the prevalence of antibody in rural human populations is generally low, (<1–2%), but there are notable exceptions, as in northern Senegal where up to 20% of people had antibody in locations where nomadic shepherds had regular contact with sheep and slept outdoors where they were exposed to ticks. The evidence suggests that the disease of humans is probably under diagnosed in many countries due to lack of awareness and/or non-availability of appropriate medical and laboratory services, but also that there is generally a low rate of transmission of infection to humans as discussed below.

The causative agent of CCHF is a member of the Nairovirus genus of the family Bunyaviridae, which at present contains 32 viruses arranged in seven serogroups on the basis of antigenic affinities, with CCHF virus, Hazara from Pakistan, and Khasan from the former USSR constituting one of the serogroups. All members of the genus are believed to be transmitted by either ixodid or argasid ticks, and only three are known to be pathogens of humans, namely, CCHF, Dugbe and Nairobi sheep disease viruses. Dugbe commonly causes mild infection of sheep and cattle in West Africa and is infrequently associated with benign febrile illness of humans. Nairobi sheep disease virus, which is believed to be identical to Ganjam virus of India, is a pathogen of sheep and goats in East Africa and India which occasionally causes benign illness in humans.

Nairoviruses are spherical, 90–120 nm in diameter, and have a bilipid-layer envelope from which glycoprotein spikes project. The virions contain three major structural proteins: two envelope glycoproteins G1 and G2 with molecular weights 72–84x10³ and 30–40x10³ Da respectively, a nucleocapsid protein N (48–54x10³ Da), and minor quantities of a large protein L (>200x10³ Da), believed to be the viral transcriptase. Hazara virus is unique in having three glycoproteins. The viruses have a three-segmented, single-stranded RNA genome which is in the negative-sense (complementary to mRNA). Each RNA segment, L (large), M (medium) and S (small), is contained in a separate nucleocapsid within the virion. The L RNA segment (molecular weight 4.1–4.9x10⁶ Da) codes for the viral transcriptase, the M segment (1.5–2.3x10⁶ Da) for the G proteins, and the S segment (0.6–0.7x10⁶ Da) for the N protein. Precursors of the glycoproteins have been found in infected cells, but non-structural proteins found during the replication of viruses of other genera of the Bunyaviridae have not as yet been demonstrated in association with nairoviruses.

Nairoviruses attach to receptors on susceptible cells, are internalized by endocytosis, and replicate in the cytoplasm. The virions mature by budding through endoplasmic reticulum into cytoplasmic vesicles in the Golgi region, which are presumed to fuse with the plasma membrane to release virus.

Phylogenetic studies indicate that there are regional differences in virus strains, and confirm that reassortment and recombination occur in nature (Hewson et al. 2004; Deyde at al. 2006; Lukashev 2005). The epidemiological implications of these observations are not yet clear, but lower death rates occur in Turkey than elsewhere. The differences should be borne in mind in developing molecular diagnostic techniques (Duh et al. 2006).

The mechanisms of pathogenesis by CCHF virus are incompletely understood, but it can be surmised that there may be some initial replication in tissues at the site of inoculation, with subsequent replication of virus to high titres in macrophages/monocytes resulting in haematogenous and lymph-borne spread of infection to regional lymph nodes and certain target organs such as the liver which is a major site of virus replication. Infected cells release cytokines, chemokines and other pro-inflammatory mediators which lead to coagulation defects and fall in blood pressure with development of intractable shock and multi-organ failure (Karti et al. 2004; Papa et al. 2006; Ergunol et al. 2006b; Bray 2007; Doganci 2007; Fisgin et al. 2008; Ergunol 2008). It is postulated that release of pro-inflammatory cytokines by T-helper lymphocytes and macrophages triggers the occurrence of haemophagocytosis in bone marrow, and that endothelial damage associated with the occurrence of disseminated intravascular coagulopathy is caused by cytokine storms rather than directly by viral infection (Karti et al. 2004; Doganci 2007; Fisgin et al. 2008). Moreover, there is evidence of the formation of circulating immune complexes with activation of complement, and this would contribute to damage of the capillary bed and hence to the genesis of the skin rash and renal and pulmonary failure. Endothelial damage leads to platelet aggregation and degranulation, with activation of the intrinsic coagulation cascade. Tissue damage in organs such as the liver would result in further release of procoagulants into the bloodstream, and the impairment of the circulation through the occurrence of disseminated intravascular coagulopathy would in turn contribute to further tissue damage. Damage to the liver would limit clearance of fibrin degradation products and impair synthesis of coagulation factors to replace those consumed. Abnormalities in clinical pathology values observed in patients indicate that the occurrence of disseminated intravascular coagulopathy is probably an early and central event in the pathogenesis of the disease.

In the past, CCHF virus has been propagated and titrated most commonly by intracerebral inoculation of suckling mice. The virus is non-pathogenic for other laboratory animals, including rabbits, guinea pigs and monkeys. It can be grown in a wide variety of primary and line cell cultures, including Vero, CER, BHK₂₁, and SW13 cells, but it is poorly cytopathic and hence infectivity is titrated by plaque production or demonstration of immunofluoresence in infected cells.

Little information is available on the stability of CCHF virus, but infectivity is destroyed by low concentrations of formalin or beta-propriolactone. Being enveloped, the virus is sensitive to lipid solvents. It is labile in infected tissues after death, presumably due to a fall in pH, but infectivity is retained for a few days at ambient temperature in separated serum, and for up to three weeks at 4°C. Infectivity is stable at temperatures below −60°C, but is rapidly destroyed by boiling or autoclaving.

Experimentally inoculated domestic ruminants and small mammals, such as little susliks, hedgehogs, hares and myomorph rodents, were found to undergo inapparent infection or mild fever and viraemia, with maximum recorded titres of infectivity ranging from 10^2.7 to 10^4.2 mouse intracerebral 50% lethal doses/ml (MICLD₅₀/ml), and with a demonstrable immune response. The virus was not abortigenic in heifers and ewes inoculated late in pregnancy (Swanepoel and Shepherd 1983–8, unpublished observations). However, when ticks of a laboratory strain of Hyalomma truncatum capable of causing sweating sickness, a toxicosis, were inadvertently placed on CCHF-infected sheep and cattle in the course of tick infection experiments, some of the animals became severely ill (Shepherd et al. 1991). Thus, animals which undergo simultaneous infection with CCHF virus and specific tick-borne pathogens of livestock in nature, constitute a source of infection for humans who treat or butcher sick animals; an observation which would explain the circumstances under which some patients have been observed to acquire infection in the former USSR and in South Africa, as discussed below.

Passerine birds and chickens were found to be refractory to infection, but they may be capable of infecting ticks despite failing to circulate detectable levels of virus, while ostriches have been shown to develop viraemic infection, as discussed under Epidemiology below.

The incubation period is generally short, ranging from one to three days (maximum nine) following infection by tick bite, and is usually five or six days (maximum 13) in persons exposed to infected blood or other tissues of livestock or human patients. There is usually very sudden onset of illness with fever, rigors, chills, severe headache, dizziness, neck pain and stiffness, sore eyes, photophobia, malaise, and myalgia with intense backache or leg pains. Nausea, sore throat and vomiting are common manifestations early in the disease and some patients experience non-localized abdominal pain and diarrhoea at this stage. Fever may be intermittent and patients may undergo sharp changes of mood over the first two days, with feelings of confusion and aggression. By the second to fourth day of illness they may exhibit lassitude, depression and somnolence, and have a flushed appearance with injected conjunctivae or chemosis. Tenderness of the abdomen localizes in the right upper quadrant, and hepatomegaly may be discernible. Tachycardia is common and patients may be slightly hypotensive. There may be lymphadenopathy, and enanthem and petechiae of the throat, tonsils, and buccal mucosa.

Patients develop a petechial rash on the trunk and limbs on day three to six of illness, and this may be followed rapidly by the appearance of large bruises and ecchymoses, especially in the anticubital fossae, upper arms, axillae and groin. Development of a haemorrhagic tendency may be evident only from the oozing of blood from injection or venipuncture sites, but epistaxis, haematemesis, haematuria, melaena, gingival bleeding and bleeding from the vagina or other orifices may commence on day four to five of illness, or even earlier. There may also be internal bleeding, including retroperitoneal and intracranial haemorrhage. Severely ill patients enter a state of hepato-renal and pulmonary failure from about day five onwards and progressively become drowsy, stuporous and comatose. Jaundice becomes apparent during the second week of illness. Deaths generally occur on days five to 14 of illness. Patients who recover usually begin to improve subjectively on day nine or 10 of illness, but asthenia, conjunctivitis, slight confusion and amnesia may continue for a month or longer (Swanepoel et al. 1987; Ergunol 2008).

Viraemia has been detected from the time of onset up to day 13 of illness, with highest titres occurring during the first five days. The viraemia is of greater intensity and longer duration in humans than in lower animals, with a maximum recorded titre of 10^6.2 MICLD₅₀/ml, but is less intense than the viraemias commonly recorded in the other so-called formidable viral haemorrhagic fevers, such as Marburg, Ebola and Lassa fevers.

Changes in clinical pathology values are more marked in fatal than in non-fatal infections, and abnormalities recorded during the first few days of illness include leucocytosis or leucopenia, elevated serum aspartate transaminase (AST), alanine transaminase (ALT), gamma-glutamyl transferase, lactic dehydrogenase, alkaline phosphatase and creatine kinase levels, thrombocytopenia, prolonged activated partial thromboplastin (APTT) and thrombin times, elevated prothrombin ratio and fibrin degradation product levels, and depression of fibrinogen and haemoglobin values. Bilirubin, creatinine and urea levels increase and serum protein levels decline during the second week of illness.

Specimens to be submitted for laboratory confirmation of a diagnosis of CCHF include blood from live patients and, in order to avoid performing full autopsies, heart blood and liver samples taken with a biopsy needle from deceased patients. On account of the propensity of the virus to cause laboratory infections, and the severity of the human disease, investigation of CCHF is generally undertaken in maximum security laboratories in countries which have appropriate biosafety regulations and facilities.

Virus can be isolated from blood and organ suspensions in a wide variety of primary and line cell cultures, including Vero, CER and BHK₂₁ cells, and identified by immunofluorescence. Isolation and identification of virus can be achieved in 1–5 days, but cell cultures lack sensitivity and usually only detect high concentrations of virus present in the blood of severely ill patients during the first five days or so of illness. Suckling mice inoculated intracerebrally are more sensitive than cell cultures for the isolation of virus present in blood in low concentrations for up to 13 days after the onset of illness, but they take 6–9 days to succumb to the infection. Virus antigen can sometimes be demonstrated in the blood of severely ill patients with intense viraemia, or in liver suspensions, by enzyme-linked immunoassay. Viral nucleic acid can be demonstrated in serum and liver homogenates of patients by the reverse transcription-polymerase chain reaction (RT-PCR) (Burt et al. 1998;  Duh et al. 2006). Observation of necrotic lesions compatible with CCHF infection in sections of liver, provides presumptive evidence in support of the diagnosis.

Antibodies, both IgG and IgM, become demonstrable by indirect immunofluorescence in a few patients from day four of illness, but most commonly become detectable from day seven onwards, and are present in the sera of all survivors of the disease by day nine at the latest. The IgM antibody activity declines to undetectable levels by the fourth month after infection, and IgG titres may begin to decline gradually at this stage, but remain demonstrable for at least five years. Recent or current infection is confirmed by demonstrating seroconversion, or a four-fold or greater increase in antibody titre in paired serum samples, or IgM antibody activity in a single sample. The antibody responses may also be demonstrated by enzyme-linked immunoassay. Patients who succumb rarely develop a demonstrable antibody response and the diagnosis is confirmed by isolation of virus from serum, or from liver specimens (Burt et al. 1994).

The disease must be distinguished from the other viral haemorrhagic fevers which partially overlap in distribution with CCHF: Lassa fever, Marburg disease, Ebola fever, Omsk haemorrhagic fever, Kyasanur Forest disease, and the haemorrhagic fever with renal syndrome (HFRS) group of diseases associated with hantavirus infections. Other febrile illnesses which can be acquired from contact with animal tissues within the same geographic range as CCHF include Rift Valley fever, Q fever, brucellosis and systemic anthrax, while diseases which can be acquired from ticks include Q fever and tick-borne typhus (Rickettsia conorii infection commonly known as tickbite fever). However, severe forms of many other common infections may resemble CCHF, including the various types of viral hepatitis, malaria and bacterial septicaemias.

Macroscopic and microscopic lesions seen in CCHF are suggestive, but not pathognomonic of the disease. Lesions in the liver vary from disseminated foci of coagulative necrosis, mainly mid-zonal in distribution, to massive necrosis involving over 75% of hepatocytes, and a variable degree of haemorrhage, with little or no inflammatory cell response. Lesions in other organs include congestion, haemorrhage and focal necrosis in the central nervous system, kidneys and adrenals, and general depletion of lymphoid tissues. Fibrin deposits may be seen in small blood vessels in parenchymatous organs including the liver.

Patients should be treated under conditions of barrier-nursing for the protection of medical personnel. Theoretically, therapy appropriate for disseminated intravascular coagulopathy, such as the use of heparin, could be applied early in the course of the disease, but patients rarely come to medical attention at a sufficiently early stage, and the procedure is considered to be risky so that it should only be contemplated by clinicians well versed in the treatment of haemostatic failure. Standard treatment consists of replacement of red blood cells, platelets and other coagulation factors, plus protein (albumin) and intravenous feeding as indicated by clinical pathology findings (Ergunol et al. 2007; Ergunol 2008).

Neutralizing antibody responses to nairovirus infections are inherently weak, and although immune plasma from recovered patients has been used in therapy, there has been no controlled trial with a uniform product of proven virus-neutralizing ability. Moreover, treatments have been initiated at various stages of illness up to and including terminal coma, so that no firm conclusions can be drawn on the efficacy of the treatment. Ribavirin inhibits virus replication in cell cultures and suckling mice (Peters and Shelokov 1990), and promising results have been obtained in limited trials on human patients with the intravenous formulation of the drug, but there have been no definitive studies. The rapid course and gastrointestinal complications to the disease render oral treatment less effective. Oral treatment for prophylaxis should only be applied very selectively to persons with severe exposure to infection, such as needle stick with confirmed infected blood, since the side effects of the treatment can cause highly inconvenient confusion if the drug is used indiscriminately during outbreaks.

The results of in vitro studies suggest that interferons and other immunomodulators may have a role in treatment of CCHF infection (Ergunol et al. 2007; Ergunol 2008).

The mortality rate is approximately 30% (range 20–50%), but this can be reduced considerably by careful monitoring of patients and the application of appropriate blood product replacement therapy. In southern Africa it was found that the occurrence during the first five days of illness of any of the following clinical pathology values is highly predictive of fatal outcome: leucocyte counts ≥10x10⁹/L; platelet counts ≤20x10⁹/L; AST ≥200U/L; ALT ≥150U/L; APTT ≥60 seconds; and fibrinogen ≤110mg/dL. Curiously, leucopenia early in the disease does not have the same poor prognostic connotation as leucocytosis, and all clinical pathology values may be grossly abnormal after day five of illness without necessarily being indicative of a poor prognosis (Swanepoel et al. 1989). Modifications to these predictive values have been proposed in Turkey, but the mortality rate associated with CCHF virus circulating in that country appears to be inherently lower (5%) than elsewhere (Ergunol et al. 2007; 2008). Elevated serum levels of pro-inflammatory cytokines, tumour necrosis factor-α and interleukin-6, are also significantly higher in patients with fatal outcome (Ergunol et al. 2006b; Papa et al. 2006). Viraemia is most intense in severe disease, and a viral load of ≥1x10⁹ genome copies/ml as determined by quantitative RT-PCR was found to be indicative of fatal outcome (Cevik et al. 2007). Since an antibody response is rarely demonstrable in fatal illness, the occurrence of a detectable immune response is generally a favourable sign (Burt et al. 1994).

The causative agent of CCHF is widely distributed in eastern Europe, Asia and Africa: the presence of the virus or antibody to it has been demonstrated in the former USSR, Albania, Serbia, Kosovo, Macedonia, Bulgaria, Greece, Turkey, Hungary, France, Portugal, Saudi Arabia, Kuwait, Dubai, Sharjah, Iraq, Iran, Afghanistan, Pakistan, India, China, Egypt, Ethiopia, Mauritania, Senegal, Burkina Faso, Benin, Nigeria, Central African Republic, Zaire, Kenya, Uganda, Tanzania, Zimbabwe, Namibia, South Africa, and Madagascar. However, the evidence for France and Portugal is based on limited serological observations and needs to be confirmed.

Although CCHF virus has been isolated from at least 31 species of ticks, including two argasids and 29 ixodids, there is no definitive evidence for most species that they are capable of serving as vectors, and in many instances virus recovered from engorged ticks may merely have been present in the bloodmeal imbibed from a viraemic host. Argasid ticks are unlikely to be vectors since the virus failed to replicate in three species inoculated intracoelomically. Members of three genera of ixodid ticks, Hyalomma, Dermacentor and Rhipicephalus, have been shown to be capable of transmitting infection transstadially and transovarially, but the bulk of the evidence suggests that Hyalommas are the principal vectors in nature, and in broad terms the known distribution of CCHF virus coincides with the world distribution of members of this genus of ticks. The prevalence of antibody to CCHF virus in the sera of wild vertebrates in southern Africa was found to be highest in large herbivores (the size of kudu antelope and greater), which are known to be the preferred hosts of adult Hyalomma ticks, and in small mammals up to the size of hares which are the preferred hosts of immature Hyalommas; wild mammals of intermediate size, which are parasitized by other genera of ticks, generally lacked evidence of infection. Virus or antibody has also been demonstrated elsewhere in the sera of small mammals of Eurasia and Africa, such as little susliks, hedgehogs, hares and certain myomorph rodents, and in some instances it has been shown that these hosts develop viraemia of sufficient intensity to infect ticks. Furthermore, it has been demonstrated that CCHF virus can be passed from infected to non-infected ticks which feed together on non-inoculated or immune mammals which fail to develop demonstrable viraemia. The phenomenon of ‘non-viraemic’ transmission of infection between ticks, which had been demonstrated earlier with other viruses, is believed to be mediated by factors present in tick saliva. Transovarial transmission of virus in ticks occurs with low frequency, but appears to be facilitated when virus is transmitted venereally from infected males to females.

Certain passerine birds and domestic chickens were found to be refractory to CCHF virus, while guinea fowl developed transient viraemia of low intensity following experimental inoculation, and an antibody response which was demonstrable for a few weeks only. Antibody to CCHF virus was detected in the sera of certain ground-frequenting birds in Senegal, notably the red-billed hornbill (Tockus erythrorhyncus), and it was shown that infection could be transferred between infected and noninfected ticks feeding on these birds through non-viraemic transmission (Zeller et al. 1994a; 1994b). Immature ticks of some species of Hyalomma, notably H. marginatum rufipes in Africa, utilize ground-frequenting birds as hosts, and it has long been accepted that the millions of birds which migrate annually on a north-south axis between Africa and Eurasia can serve to disseminate CCHF virus through the carriage of transovarially-infected immature ticks. The implication of the findings in Senegal is that birds may well play a significant role in infecting ticks, and that even those with a limited flight range, such as hornbills, can disseminate infected ticks locally. Antibody to CCHF virus was found in farmed ostriches (hosts to adult Hyalomma ticks) in South Africa, and following the occurrence of two outbreaks of the disease among workers at ostrich abattoirs it was shown that these birds develop viraemic infection similar to that in domestic ruminants (Swanepoel et al. 1998).

High prevalences of antibody occur in domestic ruminants in areas infested by Hyalommas and the virus causes inapparent infection or mild fever in cattle, sheep and goats, with viraemia of sufficient intensity to infect adult ticks. However, since transovarial transmission of infection in ticks occurs with low frequency, the role of livestock in the circulation of the virus is probably limited: infection of adult ticks followed by transovarial transmission is unlikely to sustain the virus. Hence, it is believed that the infection of immature ticks on small mammals and possibly ground-frequenting birds, constitutes an important amplifying mechanism.

Sheep, goats and cattle generally acquire natural infection with CCHF virus early in life in areas with high challenge rates, and are viraemic for about a week. Hence, it is found that humans become infected when they come into contact with the viraemic blood of overtly healthy young animals in the course of performing procedures such as castrations, vaccinations, inserting ear tags or slaughtering the animals. Young ruminants are innately resistant to specific tick-borne diseases of livestock, such as anaplasmosis, babesiosis and cowdriosis, but animals which are raised under tick-free conditions and moved to infested locations later in life may acquire the tick-borne diseases at the same time that they become infected with CCHF virus; consequently humans also become infected from contact with viraemic blood in the course of treating sick animals or butchering those that die. Common source outbreaks involving more than one case of the disease can occur when several people are exposed to infected tissues. The available evidence suggests that the infection in humans is acquired through contact of viraemic blood with broken skin, and this accords with the fact that nosocomial infection in medical personnel usually results from accidental pricks with needles contaminated with the blood of patients, or similar mishaps. Infection appears to be limited to those who have contact with fresh blood or other tissues, probably because infectivity is destroyed by the fall in pH which occurs in tissues after death, and there has been no indication that CCHF virus constitutes a public health hazard in meat processed and matured according to normal health regulations. Many human infections result directly from tick bite, and it has been observed that people can also become infected from merely squashing ticks between the fingers. Some patients are unable to recall contact with blood or other tissues of livestock, or having been bitten by ticks, but live in or have visited a rural environment where such exposure to infection is possible. Town dwellers sometimes acquire infection from contact with animal tissues or tick bite while on hunting or hiking trips.

The majority of patients tend to be adult males engaged in the livestock industry, such as farmers, herdsmen, slaughtermen, and veterinarians, but this changes where women and children participate in tending livestock, or where refugee populations are exposed to the outdoors. The observation that infection of humans is relatively uncommon despite serological evidence of widespread infection occurring in livestock, may be explained by the facts that viraemia in livestock is short-lived, and of low intensity compared to that in other zoonotic diseases such as Rift Valley fever, and that humans are not the preferred hosts of Hyalomma ticks. The low prevalences of antibody generally found in populations at risk, and the relative paucity of evidence of inapparent infection encountered among the cohorts of cases of the disease, suggests that infection is frequently symptomatic.

The control of the vectors of CCHF virus through the use of acaricides is impractical, particularly under the extensive or nomadic farming conditions which prevail in the arid areas where the disease is most prevalent. In South Africa, a single acaricide treatment of farmed ostriches followed by quarantine for at least 14 days is used to ensure that the birds are non-viraemic on arrival at abattoirs, and the same principle could be applied to other slaughter animals.

Stockmen, veterinarians, slaughtermen, and others involved with the livestock industry should be made aware of the disease and take practical steps to limit or evade exposure of naked skin to fresh blood and other tissues of animals, and to avoid handling and being bitten by ticks. Precautions should include the use of gloves and other protective clothing in slaughtering and treating animals, or in performing autopsies. Pyrethroid acaricides, such as permethrin, can be used at low concentration (0.05%) to kill ticks which come into contact with human clothing (Screck et al. 1980), and in some countries liquid or aerosol formulations are commercially available for this purpose: clothing is either dipped in the liquid and dried, or sprayed with the aerosol. Inactivated mouse brain vaccine for the prevention of human infection has been used on a limited scale in Eastern Europe and the former USSR, but the sporadic and unpredictable occurrence of the disease renders it difficult to identify target populations. A corollary to this problem is that the development of a safe and effective modern vaccine is inhibited by limited potential demand.