Reviews in Medical Virology REVIEW Rev. Med. Virol. 2016; 26: 183–196. Published online 29 February 2016 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/rmv.1877 Repetitive dengue outbreaks in East Africa: A proposed phased mitigation approach may reduce its impact Marycelin Baba1,2*, Jandouwe Villinger1 and Daniel K. Masiga1 1 Martin Lüscher Emerging Infectious Diseases Laboratory (ML-EID), International Centre of Insect Physiology and Ecology (icipe), Nairobi, Kenya 2 Department of Medical Laboratory Science, P.M.B. 1069, University of Maiduguri, Maiduguri, Nigeria S U M M A RY Dengue outbreaks have persistently occurred in eastern African countries for several decades. We assessed each outbreak to identify risk factors and propose a framework for prevention and impact mitigation. Seven out of ten countries in eastern Africa and three islands in the Indian Ocean have experienced dengue outbreaks between 1823 and 2014. Major risk factors associated with past dengue outbreaks include climate, virus and vector genetics and human practices. Appropriate use of dengue diagnostic tools and their interpretation are necessary for both outbreak investigations and sero-epidemiological studies. Serosurvey findings during inter-epidemic periods have not been adequately utilised to prevent re-occurrence of dengue outbreaks. Local weather variables may be used to predict dengue outbreaks, while entomological surveillance can complement other disease-mitigation efforts during outbreaks and identify risk-prone areas during inter-epidemic periods. The limitations of past dengue outbreak responses and the enormous socioeconomic impacts of the disease on human health are highlighted. Its repeated occurrence in East Africa refutes previous observations that susceptibility may depend on race. Alternate hypotheses on heterotypic protection among flaviviruses may not be applied to all ecologies. Prevention and mitigation of severe dengue outbreaks should necessarily consider the diverse factors associated with their occurrence. Implementation of phased dengue mitigation activities can enforce timely and judicious use of scarce resources, promote environmental sanitation, and drive behavioural change, hygienic practices and community-based vector control. Understanding dengue epidemiology and clinical symptoms, as determined by its evolution, are significant to preventing future dengue epidemics. Copyright © 2016 John Wiley & Sons, Ltd. Received: 22 September 2015; Revised: 3 February 2016; Accepted: 4 February 2016 INTRODUCTION Dengue infection is underreported and often misidentified, but its global incidence in 2012 was estimated at almost 400 million in 128 countries in contrast to only nine countries that experienced severe dengue epidemics before the 1970s [1,2]. Possibly, many countries (especially in Africa) that *Correspondence to: M. Baba, Martin Lüscher Emerging Infectious Diseases Laboratory (ML-EID), International Centre of Insect Physiology and Ecology (icipe), P. O. Box 30772, Nairobi, Kenya. E-mail: marycelinbaba@gmail.com Abbreviations used DF, dengue fever; DHF, dengue haemorrhagic fever; DSS, dengue shock syndrome; DENV, dengue virus; CFR, case fatality rate; YF, Yellow fever; IgM, Immunoglobulin M; IgG, Immunoglobulin G; PRNT, plaque reduction neutralisation test; MNT, mouse neutralisation test; MCNT, micro neutralisation test; POS, post onset of symptoms; NS1, non-structural protein 1; IEP, inter-epidemic period. Copyright © 2016 John Wiley & Sons, Ltd. experienced dengue outbreaks did not report because of diagnostic limitations. Differential diagnosis of febrile illnesses for dengue is difficult to attain amidst several endemic diseases (malaria, chikungunya, etc.) with atypical symptoms. Although global estimates of dengue infections vary by year, nearly 500 000 episodes of dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) occur annually, with over 20 000 dengue-related deaths [2]. The World Health Day campaign focused on dengue in 2014 to emphasise its public health importance [3]. Dengue viruses (DENV) belong to the genus Flavivirus (family: Flaviviridae) and comprise four related serotypes (DENV1-4) with antigenic crossreactivity, but no cross protection. Sylvatic DENV transmission (between mosquito and monkeys) or extrinsic virus stock being exchanged vertically 184 across mosquito generations within tree holes or domestic homes may cause DF/DHF/DSS among residents of towns situated adjacent to these rural or peri-urban niches [4]. Generally, DENVs rely on transmission by mosquito vectors with Aedes aegypti serving as the principal vector in most locations, while Aedes albopictus serves as a secondary vector in other areas [5]. In islands across the Indian Ocean, the primary vector of all dengue outbreaks has been Ae. albopictus, whereas Ae. aegypti served as the secondary vector [6,7]. These mosquito vectors live in close association with human populations and prefer to breed in domestic water containers [5,6]. This review highlights the public health importance of dengue, its enormous burden in eastern Africa, risk factors associated with past outbreaks, its predictability and investigations strategies adopted. Because of the enormous socio-economic impact of dengue on human health, we examine limitations in responses to past outbreaks in eastern Africa between 1823 and 2015 and propose an adjusted framework for mitigating against future dengue outbreaks. METHODOLOGY All published, peer-reviewed literature, published country reports and the World Health Organisation library database were reviewed using the search terms ‘dengue outbreaks’, ‘dengue and ecology’, ‘vectors of dengue’, ‘epidemiology of dengue’, ‘dengue and climate’, ‘dengue in East Africa’, ‘Prediction of dengue outbreaks, ‘genetics of dengue virus’, ‘kinetics of dengue’, ‘seroprevalence of dengue’ and ‘economic estimates of dengue’. We searched for publications available in English as of 10 January 2016, in MEDLINE, EMBASE, Agora and World Health Organisation Hinari electronic databases as well as ProMED-mail posts. We also examined abstracts presented at international forums for information on dengue in eastern Africa from the 1800s to 2016. About 200 published English articles were reviewed, assuming that reports in other languages would not change the conclusions of this article. History of dengue outbreaks in East Africa and Indian Ocean islands Repetitive dengue outbreaks have occurred in seven out of ten countries in eastern Africa and Copyright © 2016 John Wiley & Sons, Ltd. M. Baba, J. Villinger and D. K. Masiga three islands across the Indian Ocean (Figure 1) as far back as 1823. Four dengue outbreaks occurred in Tanzania between 1823 and 1926 [8], during which the Swahili word for dengue (believed to be caused by an evil spirit) was ‘Ki-dinga pepo’ (which means cramp-like seizure), where ‘Pepo’ is ‘to sway, reel, stagger or totter’ [9] and ‘Ki’ is diminutive [9]. The modern usage of ‘dinga’ or ‘denga’ in Swahili does not explain the term ‘ki-dinga’. After the 1823 dengue outbreak in Cuba, the Spanish word ‘dengue’ came into general use in medical literature [10]. Dengue outbreaks were reported in Reunion and Mauritius in 1851, while in 1870, a dengue pandemic started in Tanzania, spread towards Egypt, Saudi Arabia, Yemen (Aden), India, China, Indonesia, Indochina (Vietnam, Laos, Cambodia), and back to Mauritius and Reunion in 1873 [11]. In Tanzania, the 1826 and 1870 dengue outbreaks were considered to be more likely chikungunya instead of dengue [9–14]. It is, however, unclear how the outbreak that spread to other parts of the world was known and identified as dengue in the affected countries but was considered as Chikungunya in Tanzania where the outbreak was thought to have originated. In Kenya, the first outbreak of dengue occurred (Malindi and Kilifi) from 1860 to 1868 [8], and in Somalia and Eritrea, they occurred simultaneously from 1897 to 1899 [8,11,12]. Mozambique and Seychelles were dengue-endemic countries in the early 1900s [15], but between 1975 and 1996, the list was extended to include Comoros, Ethiopia, Somalia, Tanzania, Réunion and Mauritius [13]. In 1926, a ‘probable’ dengue outbreak was reported in Seychelles [11,13] and confirmed in Comoros between 1943 and 1948 [6]. Four years later, in Tanzania (Makonde Plateau), an outbreak re-occurred in 1952–1953 [16,17]. Subsequently, simultaneous dengue outbreaks reoccurred in Seychelles and Reunion in 1977–1979 and in Pemba (Mozambique) between 1982 and 1983 [18,19]. In 1982, dengue outbreaks re-occurred in Malindi and Kilifi (located 68 km north of Mombasa (Kenya) in the same locations where the first outbreak occurred in 1860 [20]. That outbreak was speculated to have spread from Seychelles, which experienced dengue outbreaks between 1977 and 1979 [21]. The basis of that speculation is not clear because dengue outbreaks took place in Kenya 3 years after Seychelles. In Somalia, four major dengue outbreaks associated with DEN-2/3 occurred in Jubbaada Hoose (Kismayu), and in Shabeellaha Hoose (Afgoi), Banaadir (Mogadishu) in 1982 and Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa 185 Figure 1. Map of eastern African countries indicating frequencies of major dengue outbreaks over the several decades (1823–2014). Number of outbreaks in specific countries is indicated in the red spheres 1993 [22–25]. These outbreaks mostly affected the US military troops engaged in the mission ‘Operation Restore Hope’ in Somalia [26,27]. In the local Somalian language, dengue was then described as ‘Jejeebiye’, which means ‘bone breaking sickness’ [25]. Although no death was recorded during these epidemics, they caused high rates of morbidity and hospitalisation [25]. Mozambique experienced Copyright © 2016 John Wiley & Sons, Ltd. DENV-3 outbreaks in 1984–1985, which resulted in two deaths [18], while Comoros experienced DENV-1 in 1948, DENV-2 in 1984 and a DENV-1 epidemic in 1993 affecting 56 000–75 000 people [28]. Port Sudan City had DENV-1 and 2 outbreaks in 1985–1986 involving 17 cases [29]. Dengue outbreaks re-occurred in Mozambique (1983) [18,19,27,30] and in Djibouti (1991–1992), resulting in 12 000 cases [22]. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 186 In 1998 and 2000, dengue outbreaks hit Djibouti again [31], and in 2004–2005, DENV-3 re-emerged in Port Sudan with 312 cases of DHF and 12 deaths (CFR = 3.8%) and spread to northern Kenya [31]. Dengue outbreaks re-occurred in Eritrea in 2005 [27,33] and Djibouti in 2008 [30]. In 2009, dengue outbreaks hit Seychelles and Reunion [8], but the details of these outbreaks with respect to the type of clinical manifestations and serotypes involved were not reported. Dengue outbreaks re-occurred in Tanzania [33] and Port Sudan city in Sudan [34] in 2010, resulting in 100 and 3765 cases, respectively. In Port Sudan city, the outbreak was caused by DENV-1/2 but the serotype involved in Tanzania is unknown. In 2011, South Kordofan experienced a dengue outbreak involving 299 cases and 71 deaths [35]. These outbreaks further strained the war-ravaged nation’s tattered health system [36]. More dengue outbreaks re-occurred in 2011 in Mogadishu (Somalia) and Mandera (Kenya) (borders with Somalia and Ethiopia), involving 143 (three deaths) [37] and 2070 cases (seven deaths) [38–40], respectively. The dengue outbreak in Mandera highlighted the cross border nature with eight and five cases imported from Ethiopia and Somalia, respectively [39]. In 2013, more dengue outbreaks re-emerged in Somalia (23 cases, no deaths) [41] and Tanzania (20 laboratory-confirmed cases, 3 deaths) [42]. Similarly in Mandera/Wajir (Kenya), 300 estimated dengue cases with three deaths occurred in January 2013 [43–45]. This outbreak later spread to Mombasa in April 2013, resulting in 153 confirmed cases (one death) [45–47]. The serotypes of DENV associated with the 2013 outbreaks in Mombasa include DENV-1 (69%), DENV-2 (28%), and DENV-3 (3%) and 1 co-infection (DENV-1/2) [47]. Interestingly, dengue re-emerged with 239 cases in Mombasa in early 2014 [48]. As a major shipping port and international tourist destination, Mombasa may serve as an entry point of mosquito vectors and DENVs through human commerce. The introduction of new serotypes/strains across the Indian Ocean, where the four DENV serotypes are endemic, may contribute to sporadic and persistent dengue outbreaks in Kenya. Mozambique and Tanzania had dengue outbreaks in 2014 involving 243 (no deaths) and 400 cases (three deaths), respectively [49]. In 2015, dengue outbreaks hit Mozambique with 110 cases (no death) [50] and Sudan (254 cases, 83 deaths) [51]. Copyright © 2016 John Wiley & Sons, Ltd. M. Baba, J. Villinger and D. K. Masiga Identified risk and epidemiological factors associated with past dengue outbreaks Climate Different climate conditions could result in temporal and spatial changes in temperature, precipitation and humidity in ways that affect the biology and ecology of disease vectors [52]. In Kenya, higher temperature (29–31°C) [53] in Kilifi, Mandera and Mombasa counties (where repeated dengue outbreaks have occurred) [11,20,38,39,42–48] enhanced vector competence of Ae. aegypti populations for DENV transmission in contrast to the same mosquito vector from Nairobi with average annual temperature of 26–28°C [54]. Alternatively, this difference in transmission competence may be due to other variables such as genetic differences in the vector [55]. Climate change may facilitate dengue outbreaks in endemic areas and emergence of the virus in new regions. Human practices Numerous commercial vessels transiting across the Indian Ocean may be the main vehicles conveying the four DENV serotype and its mosquito vectors from India to Kenya [56]. Additionally, socioeconomic activities (hunting, farming, woodcutting, clearing the rainforest, washing and bathing in the river, collecting water from the river and storing it in containers indoors) are common practices in epicentres of past dengue outbreaks and could predispose victims to accidental transmission from sylvatic cycle. Generally, accelerated urbanisation in Africa, resulting in increasing numbers of overcrowded, informal settlements or ‘shanty towns’ characterised by low-grade housing, inadequate water supply and storage, and poor roads, sanitation and waste management services, could provide favourable mosquito vector breeding sites. In 2011, a dengue epidemic in Mandera (Kenya) spread due to poor sanitation, resulting in up to 5000 cases within weeks and overwhelmed the limited health facilities with few medical personnel [40]. Untimely implementation of outbreak response activities and delay in outbreak confirmation due to atypical symptoms akin to malaria contributed to the spread of the 2011–2014 dengue outbreaks in Mandera [38–40,42–44] and Mombasa (Kenya) [44–48]. Lack of accessible, affordable and appropriate diagnostic reagents in Africa is an impediment to systematic surveillance and contributes to underreporting and underestimation of dengue infections in Africa. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa The 2011 and 2013, dengue outbreaks in Kenya were signalled by an increased number of febrile illnesses [40,45] without haemorrhagic symptoms. Therefore, case definitions based on non-specific and haemorrhagic symptoms led to missed cases. Unlike in 2011, adequate community awareness conducted during the 2013 dengue outbreak in Mandera (Kenya) may have contributed to changes in risky behaviours that led to a reduction of cases from 2070 (seven deaths) in 2011 [39] to 300 (three deaths) in 2013 [43–45]. Different categories of water storage containers that provide favourable breeding sites for DENV vectors were found in households affected by the 2010 dengue outbreaks in Sudan [34] and Kenya in 2014 [57]. Patients’ reluctance to travel long distances for medical care and others opting for treatment from traditional healers contributed to the spread of dengue. Human settlements and irrigation canals enhanced mosquito vector diversity in Sudan [58]. The quality of the breeding water is important to the survival of mosquito eggs [59]. Genetics The variability in susceptibility of DENV infections and disease expression among different races was associated with sustained transmission in Africa [32]. Lower rates of DHF/DSS were obtained among blacks compared with whites during the 1981 and 1997 dengue epidemics in Cuba [32]. In East Africa, cases of DHF/DSS with fatal outcomes clearly suggest that race may not be an important factor in the severity of DENV infections. In 2004–2005, 312 cases of DHF were reported and 37 of 312 (11.9%) were DSS with a mortality rate of 3.8% (n = 12) in Port Sudan [31]. In 2010, 58% and 11.1% of 113 dengue cases were DHF and DSS, respectively, in Kassala (Sudan). [60]. Similarly, ten deaths due to outbreaks of DHF/DSS occurred in Kenya [61], four in Somalia [62], two in Mozambique [18] and three in Tanzania [42]. Although details of incidence of DHF/DSS by age, risk factors and fatality rate are lacking, the occurrence of deaths due to these conditions contradicts claims that outbreaks of DHF/DSS have not been reported in Africa [32,63,64]. The hypothesis that the low rate of DENV infection is caused by cross protection from other endemic flaviviruses in Africa [65] cannot holistically explain several outbreaks of both dengue and Copyright © 2016 John Wiley & Sons, Ltd. 187 yellow fever (YF) in South Kordofon. In 2005, YF and dengue outbreaks involving 615 cases (183 deaths) [66] and 312 cases (12 deaths) [31], respectively, occurred in South Kordofon. Additionally, YF outbreaks with 44 cases (14 deaths) in 2013 [66] and a dengue outbreak resulting in 299 cases (71 deaths) in 2011 [35] reoccurred in South Kordofon. Although the level of immunity against either dengue or YF viruses in individuals affected in these outbreaks was not determined, we speculate that environmental or ecological factors may influence the initiation of outbreaks of flaviviruses in different geographical zones. Our speculation is supported by a report that different strains of Aedes mosquitoes vary in their competence in transmitting DENV in different geographical locations [32,67]. Ae. aegypti populations in Kilifi, Kenya (where repetitive dengue outbreaks have occurred) are more competent in disseminating dengue infection compared with Nairobi, Kenya (with no history of dengue outbreak) [54]. In addition, susceptibilities of the vector to different DENV genotypes also differ [32]. Ae. aegypti mosquitoes are more susceptible to infection with DENV-2 of the Southeast Asian genotype, which is also more virulent than the American genotype [68]. High virulence of DENV genotypes correlate with incidence and epidemics of DHF/DSS [64]. American DENV-2 and DENV-3 genotypes are comparatively less virulent than Asian genotypes of the same serotype [69]. The 2013 dengue outbreak in Mandera only affected individuals who were not protected after the 2011 epidemic [45]. This implies that the serotypes/ genotypes implicated in 2011 (seven deaths) and 2013 (three deaths) were most likely the same. Genotyping of dengue isolates from outbreaks is needed in East Africa for better understanding of evolutionary trends, virulence, transmissibility and molecular epidemiology of the disease. Kinetics of immune response to dengue viruses and serological outbreak investigations Dengue virus infections can be diagnosed serologically (EIA, plaque reduction neutralisation, immunofluorescence antibody, haemagglutination inhibition), virus isolation and RT-PCR [70,71]. Non-structural protein 1 antigen assay is sensitive, specific, has similar detection rates of acute dengue as RT-PCR and antibody assays [72,73], is Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 188 detectable from day 1 up to day 18 post onset of symptoms and differentiates DENV from other flaviviruses [74]. Notably, sero-epidemiological studies for DENV require more than one type of serological test because of high cross-reactivity among flaviviruses [75]. However, only EIA was commonly used in sero-epidemiological studies to determine dengue burden in some eastern African countries [76–79]. Although EIA on acute serum may be sensitive, it cannot accurately differentiate virus species nor DENV serotypes because of the broad cross-reactivity of IgG and IgM antibodies against flaviviruses [72]. However, the ability of plaque reduction neutralisation [80–83] or mouse neutralisation test [83] to neutralise specific virus serotypes makes it a useful tool for assessing the immune status of a given population and identifying risk-prone areas for future DENV outbreaks [82]. Acute dengue infections can be confirmed by sero-conversion from negative to positive IgM antibody or demonstration of a fourfold or greater increase in antibody titres in paired (acute and convalescent) sera [74] or detection of non-structural protein 1 and antibody from a single serum [73]. In primary DENV infections, IgM antibody is detectable within 4–5 days after onset of fever for up to 90 days. But IgG antibodies appear about a week after onset of fever and peak several weeks before it declines to detectable level for decades and longer [75]. Therefore, DENV IgM from the acute phase serum indicates infection that occurred 2–3 months before sample collection, while DENV IgG antibody denotes previous exposure to the virus. Secondary DENV infections induce detectable IgG antibodies on the first day of symptoms before IgM or both IgG and IgM rise quickly simultaneously and peak within 2 weeks after onset of symptoms [84]. Thereafter, IgM wanes but remains detectable in 30% of patients within 2 months after onset of symptoms while IgG declines slowly over 3–6 months. Positive RT-PCR and virus isolation may be obtained if the specimen is collected within 0–7 days post onset of symptoms. Because of delayed health-seeking behaviour in Africa, RT-PCR results should be interpreted with caution. Complementing RT-PCR with one or more serological tests in East Africa [85] will provide better diagnostic efficiency especially when the date of onset of symptoms is uncertain during specimen collection. Copyright © 2016 John Wiley & Sons, Ltd. M. Baba, J. Villinger and D. K. Masiga Underutilization of seroprevalence findings to prevent future dengue outbreaks Routine and active surveillance of animals, humans and mosquitoes for DENV with appropriate laboratory confirmation of circulating serotypes can facilitate outbreak prediction [85]. However, inadequate interpretation of serosurveys can limit their utility. Although neutralising homotypic DENV IgG antibodies can provide lifetime immunity against the infecting serotype [86], primary DENV infections can also induce either non-neutralising or sub-optimal heterotypic antibodies that may lead to greater disease severity in subsequent infections with a different serotype by ‘antibodydependant enhancement’ [87,88]. In a previous study, the presence of dengue IgG in 27.7% of a study population in Sudan in 2012 was considered indicative of disease burden [83]. However, the remaining 72.3% that are at risk of Flavivirus infections were not considered as a concern. Within the same year of dengue outbreak in Sudan [89], a YF outbreak ensued, resulting in 849 cases and 171 deaths, and in 2013, 44 cases with 14 deaths [90]. We speculate that the occurrence of DF and YF outbreaks indicated that the majority (72.3%) of the residents were susceptible to Flavivirus infections. Similarly, a study among personnel in the Djibouti army revealed that only 8.5% of the population had Flavivirus antibodies, and 5 months later, dengue outbreak ensued [21]. Additionally, a report that 1.2% of the study population had a neutralising antibody against DENV-2 in Western Kenya [78] indicates that 99% of the population would be at risk if DENV-2 epidemics recur. Therefore, proper utilisation and careful interpretation of sero-epidemiological findings could inform policy decision towards preventing re-occurrence of dengue outbreaks. SOCIO-ECONOMIC IMPACTS Despite the repetitive emergence of a severe and fatal form of dengue epidemic, the disease is not considered a major public health problem by policy makers in Africa. In Puerto Rico, the total annual economic cost of dengue between 2002 and 2010 was $46.45m with 48% borne by individual households, 24% by the government and 22% by insurance providers [91]. In the Americas, the cost was estimated at $2.1bn per year on average with a range of $1–4bn in sensitivity analyses and Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa substantial year to year variation, excluding vector control [92]. However, the total cost on the health system could be deduced from a report on a meningitis epidemic (2007) in Burkina Faso, which attracted a cost of $7.1m ($0.69 per capita) on the public health system and $ 2.3m on households of cases [93]. Like with the meningitis epidemic, dengue epidemics can strain health services in affected areas with a large increase in medical consultations. There is a need to use standardised methods to correctly estimate the economic burden of dengue in endemic countries in Africa. Individuals suffering from dengue-related morbidity cannot perform their normal economic activities and are attended to by their relatives and/or the community, during which, families incur losses that may be difficult to quantify in terms of time, cost of hospitalisation, diagnostic testing and supportive treatments [94]. Overall, the number of deaths and disease adjusted life years (heavily driven by mortality) calculations for dengue remain low compared with RVF and YF [92]. Nevertheless, the impact of the death of a person is difficult to quantify in terms of monetary value. The loss of a productive member of the family because of infectious disease morbidity [94,95] exerts serious socio-economic consequences on not only the immediate family of the diseased, but the village, community, district, province, country and region at large [96,97]. Deaths may result in orphans or parents and relatives that can no longer afford school fees or sustain themselves. The deceased’s families could be further stressed by spending additional funds for the traditional practices in respect of mourning/funeral ceremonies. Predictability of outbreaks Weather variables impact on the magnitude of dengue distribution [98,99] and the changes in El Niño Southern Oscillation affect the incubation period, life cycle, egg development, biting rates, infectivity and survival rates of both vectors and the virus [100,101]. Additionally, El Niño Southern Oscillation has been reported as a good predictor of dengue cases in Mexico [102,103], and weather variables were used to develop a dengue outbreaks forecasting model in Singapore [98]. The model correctly predicted 5/5 dengue epidemics with a lead period of 16 weeks in 2011 [98]. In agreement with Hii et al. [98], we hypothesise that integrating local weather variables with risk and epidemiological Copyright © 2016 John Wiley & Sons, Ltd. 189 (entomological data) and information on circulating serotypes in different vulnerable ecologies can be used to develop local forecasts. We assume that, since 2006–2007, RVF outbreaks were successfully predicted with a lead period of 2–6 months in East and South Africa [104]; dengue, YF and Chikungunya outbreaks can be similarly predicted. Such a prediction can enhance decision-making on the timing to upscale vector control operations, vaccination (where applicable) and adequate utilisation of limited resources to prevent future outbreaks. Limitations of past outbreak response activities The confirmation of the 2013 dengue outbreak in Mombasa was delayed by 2 months due to failure to recognise initial symptoms [44]. Although DENV serotypes implicated in Mandera and Mombasa were identified, serotypes were not genotyped. However, since 2013 and 2014, all East African countries that experienced dengue outbreaks responded meticulously to curb and mitigate its spread. Outbreak response activities include sensitization of health workers, training of health workers, and dissemination of health education to drive risky behavioural change, vector control and active case surveillance [31,44,49]. Although dengue outbreaks were successfully contained, their persistence (Table 1) with significant impact on human health remains a public health concern. In Sudan, malaria vector control methods were unsuccessfully adopted for dengue epidemics without understanding the bionomics and ecology of dengue vectors. Consequently, Sudan experienced repetitive DF and DHF for a decade. Although this error was corrected by adopting community-based integrated vector control programmes during the 2010 outbreak, 738 cases and six deaths due to DHF/DSS still occurred [1]. A new systematic approach that may drive timely and coordinated dengue mitigation activities with resultant positive impact in all facets of life is needed. Strategies to prevent outbreaks and reduce their socio-economic impacts Strategizing disease-mitigation efforts during interepidemic, prediction and outbreak periods may prevent re-occurrence of dengue outbreaks. During IEPs, activities such as assessment of attitudes, beliefs and perception will generate baseline information for Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 190 M. Baba, J. Villinger and D. K. Masiga Table 1. Reported dengue outbreaks in eastern African countries and Islands across the Indian ocean Dengue serotype Year Country Unknown Unknown 1823 1851 Unknown 1870–1873 Unknown Unknown 1860–1868 1897–1899 Unknown Unknown DENV-1 Unknown Unknown 1926 1943 1948 1952–1953 1977–1979 DENV-2 1982 Tanzania Reunion and Mauritius Tanzania, La Reunion and Mauritius Kenya Somalia and Eritrea Seychelles Mayotte Comoros Tanzania Seychelles and La Reunion Kenya Unknown DENV-2 DENV-2 DENV-1, -2 and -3 1982 1984 1985–1987 1992–1993 Somalia Comoro Somalia Somalia Unknown DENV-1 and -2 1984–1985 1985–1986 Mozambique Sudan DENV-2 DENV-1 1991–1992 1993 Djibouti Comoros Unknown DENV-1 DENV-3 1996 1998, 2004–2005 Eritrea Djibouti Sudan Unknown Unknown 2005 2005 Djibouti Eritrea State or Province Makonde plateau Malindi and Kilifi Hargeysa Banaadir (Mogadishu), Shabeellaha Hoose (Afgoi), Jubbaada Hoose (Kismayu) Pemba Port Sudan city Type of diagnostic tests used Number of cases (deaths) References Unknown Unknown Unknown Unknown [12] [15] Unknown Unknown [11,15,16] Unknown Unknown Unknown Unknown [11] [11,12] Unknown Unknown Serology Unknown Serology Unknown Unknown Unknown Unknown Unknown [11,13] [6] [28] [16,17] [18,19,21] Serology, virus isolation Unknown Serology Serology Unknown Unknown [20] Unknown Unknown Unknown Unknown [24] [28] [25] [22–25] Serology Virus isolation, RT-PCR RT-PCR Virus isolation, Serology Unknown Serology, RTSerology, RT-PCR Unknown Unknown Unknown (2) 17 (0) [18,19] [29] 12 000 (0) 56 000–75 000 [22] [28] 237–240 (0) Unknown 312 (12) [30] [31] [30] Unknown Unknown [31] [32] Continues Copyright © 2016 John Wiley & Sons, Ltd. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa 191 Table 1. (Continued) Dengue serotype Year Country Unknown 2009 DENV-1 and -2 DENV-3 2010 Seychelles and La Reunion Sudan Port Sudan city 2010 Tanzania Dar es Salaam DENV-1 and -2 DENV-1, -2 and -3 DENV-3 2011 Sudan South Kordofan 2011 Somalia Mogadishu 2011 Kenya Mandera 2013 2013 Kenya Kenya 2013 Somalia DENV-3 DENV-1, -2, -3, and DENV-1/2 coinfection DENV-2 and -3 State or Province Type of diagnostic tests used Number of cases (deaths) References Unknown Unknown [8] Serology, RTVirus isolation, RT-PCR, ELISA, Serology, RTRT-PCR 3765 (12) [32,34] 17 (0) [33] 299 (71) [35] 143 (3) [37] 5000 (4) [38–40] Mandera/Wajir Mombasa (Coast) RT-PCR, ELISA RT-PCR RT-PCR 190 (3) 210 (0) [42–44] [45] Mogadishu Unknown 23 (0) [41] health education campaigns [58]. Cascaded training of trainer’s workshops for health workers with contributions from geospatial analysis, database management and entomology experts is a necessity. The East Africa Public Health Laboratory Networking Project [105] could be used to organise training workshops at the regional, national, district and constituency levels. Each trainee can further train the staff of his/her hospital/clinic and provide feedback at the national level. Such a network of trained health workers will facilitate timely dissemination of health related information [106]. Limiting the training of health workers to outbreak periods can be described as performing ‘a noble act at the wrong time’ with little or no impact. Mobile phone-based infectious disease surveillance systems could be adopted for systematic [107] and active surveillance. Internet access may be limited in many parts of East Africa, but the cellular phone network is extensive and can be used as the data collection platforms during human health Copyright © 2016 John Wiley & Sons, Ltd. surveys in private and public health care settings as developed by using EpiSurveyor (www.datadyne. org). Simple close-ended questionnaires could be filled out in remote areas without cellular service and transmitted in an area of network reception. The location of each questionnaire could be collected with global positioning system software. The use of mosquito nets, vector repellents, protective clothing, regular environmental sanitation and proper coverage of water containers should be encouraged in risk-prone areas and its environs during rainy seasons. Surveillance for disease vectors is an important tool for the identification of risk-prone ecologies, and point-of-care diagnosis for DENV may improve early detection and reporting of a new outbreaks. Activities during risk prediction lead periods should include refresher training workshops for health workers to update knowledge and skills on appropriate diagnostic testing and disease reporting mechanisms. A well-defined sensitive case definition for suspected cases should be Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 192 M. Baba, J. Villinger and D. K. Masiga developed, and systematic disease surveillance may be intensified in risk-prone areas. Public awareness campaigns could be intensified through mass media and meetings with village heads and elders in affected areas. Affected communities could be mobilised as active participants and not spectators. Widely publicised health education will drive behavioural change and promote personal and community-driven protection against mosquito bites. Use of personal protective equipment in healthcare settings may be enforced to prevent possible nosocomial transmission and disease spread. During the outbreak phase, co-ordinated implementation of existing outbreak response machineries should be enforced immediately. Intensified dissemination of health promotion and disease prevention information throughout the affected communities is crucial. Definite and sensitive case definition should be distributed to all health centres in affected areas. Visits to affected households and collection of relevant information may aid the identification of epidemiological factors that can facilitate more cost-effective outbreak prevention/control [103]. During visits to affected households, symptomatic persons should be referred to the nearest local health care facilities for proper management [103] and further testing. The local health facility in the affected districts and environs should be capable of recognising and diagnosing the disease accurately. Aliquots of clinical specimens (acute and convalescent) should be shipped to the reference laboratories in the region for detailed and robust diagnostic procedures. Screening for broader panels of arboviruses during outbreak investigations may contribute to effective disease mitigation. There may be a need to extend disease-mitigation strategies beyond affected areas, collecting data retrospectively by reviewing hospital/laboratory/ clinic records [93,103] and temporarily prohibiting human movements to and from the affected areas to contain the spread of the disease. CONCLUSION Dengue outbreaks have persistently occurred in eastern African countries for several decades resulting in significant impacts on human health. Dengue outbreaks are multifactorial and prevention and mitigation against their effects should necessarily take these factors into consideration. Understanding the kinetics of immune responses to DENV infections will facilitate the choice of serological tests and proper interpretation of diagnosis for management of cases. Adequate utilisation of seroprevalence findings could prevent future dengue outbreaks. The limitations of past dengue outbreak responses and enormous socio-economic impacts demand for more stringent and effective measures to curb future outbreaks. Understanding dengue epidemiology and clinical symptoms as determined by its evolution may be significant to preventing future dengue epidemics. CONFLICT OF INTEREST The authors have no competing interest. ACKNOWLEDGEMENTS The technical support of David Makori is highly appreciated. We are also grateful to Damaris Matoke, Maamun Jeneby and Geoffrey Jagero for their moral and technical support. Funding for this review was provided by the Institute of International Education for Scholar Rescue Fund (IIE-SRF) (to MM Baba) and the German Academic Exchange Service (DAAD) (to M M Baba). REFERENCES 1. World Health Organization. Dengue and o http://www.who.int/campaigns/world- severe dengue. Fact Sheet N 117, 2015. health-day/2014/en/. Media centre. Available: http://www. 4/01/2016. who.int/mediacentre/factsheets/fs117/ on cycle 6. Julvez J, Ragavoodoo C, Gopaul AR, Mouchet J. Human diseases transmitted by Culicidae in southwest Indian Ocean 4. Joshi V, Singhi M. Study of role of zoonotic en/. Accessed on 27/02/2015. Accessed of dengue virus in islands. Bulletin de la Société de Pathologie Exotique 1998; 91(1): 99–103. 2. Murray NEA, Quam MB, Wilder-Smith A. maintaining/amplifying endemic DF and 7. Ramchurn SK, Moheeput K, Goorah SS. Epidemiology of dengue: past, present causing DHF in different settings of Rajas- An analysis of a short-lived outbreak of and future prospects. Clinical Epidemiology than, India. India; 2006; 16–19. dengue fever in Mauritius. Euro Surveil- 5. Weaver SC, Reisen WK. Present and future 2013; 5: 299–309. 3. World Health Day campaign: World Health Organization. Available: arboviral threats. Antiviral Research 2010; 85(2): 328–345. Copyright © 2016 John Wiley & Sons, Ltd. lance 2009; 14(34): 1–3. 8. Kleef VE, Bambrick H, Hales S. The geographic distribution of dengue fever and Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa 193 the potential influence of global climate albopictus (Skuse). Bulletin of the World change. TropIKA. 2010; 1–18. Health Organization 1980; 58(6): 937–943. alence of dengue infection: a cross-sectional 9. Valdés K, Alvarez M, Pupo M, Vázquez S, 22. Kodier GR, Gubler DJ, Cope SE, et al. Epi- survey in mainland Tanzania and on Pemba Rodríguez R, Guzmán MG. Human Den- demic dengue 2 in the city of Djibouti island, Zanzibar. International Journal of gue antibodies against structural and non- 1991–1992. Transactions of the Royal Society structural proteins. Clinical and Diagnostic of Tropical Medicine and Hygiene 1996; Laboratory Immunology 2000; 7(5): 856–857. 33. Vairo F, Nicastri E, Meschi S, et al. Seroprev- Infectious Diseases 2012; 16: e44–e46. 34. Seidahmed OME, Siam HAM, Soghaier MA, et al. Dengue vector control and 90(3): 237–240. 10. Halstead SB. Dengue: Tropical Medicine 23. Kanesa-thasan N, Chang GJ, Smoak BL, surveillance during a major outbreak in a and Practice. Pasvol G and Hoffman SL. Magill A, Burrous MJ, Hoke CH. Molecu- coastal Red Sea area in Sudan. Eastern Imperial College Press: London, UK, lar and epidemiologic analysis of dengue Mediterranean Health Journal 2012; 18(12): 2008. virus isolates from Somalia. Emerging In- 1217–1224. 11. Hirsch A. Geographical and Historical Pathology. Acute Infectious Diseases. The fectious Diseases 1998; 4(2): 299–303. 35. Kuznetsova T. Dengue fever kills 71 in Su- 24. Saleh AS, Hassan A, Scott RM, Mellick New Sydenham Society, London, 1883. PW, Oldfield EC, Podgore JK. Dengue in 12. Gubler DJ. Dengue and Dengue Hemor- north-east Africa. Lancet 1985; 2(8448): rhagic fever, Gubler DJ, Kuno G (eds). CABI, Oxfordshire, 1997. 13. Gratz N, Knudsen A. The Rise and Spread work, 2011. 36. Schelle S. Dengue fever kills 71 in Sudan epidemic. ArmMed Media 2005. Available 211–212. 25. Botros BA, Watts DM, Soliman AK, et al. at: http://www.health.am/fever/more/ Serological evidence of dengue fever dengue_fever_kills_71_in_sudan_epidem- of Dengue, Dengue Haemorrhagic Fever among and its Vectors: A Historical Review. Ge- Journal of Medical Virology 1989; 29(2): neva World Health Organ, 1996. dan epidemic. Armanian Medical Net- refugees, Hargeysa, Somalia. ic/. Accessed on 01/06/2015. 37. Bosa HK, Montgomery JM, Kimuli I, Lutwama JJ. Dengue fever outbreak in 79–81. 14. Carey DE. Chikungunya and dengue: a 26. Sharp TW, Wallace MR, Hayes CG, et al. case of mistaken identity? Journal of the Dengue fever in U.S. troops during Opera- of three dengue virus serotypes. 16 History of Medicine and Allied Sciences tion Restore Hope, Somalia, 1992–1993. International 1971; 26(3): 243–262. American Journal of Tropical Medicine and Diseases. Cape Town, South Africa. April 15. Van Rooyen C, Rhodes A, 2nd edn. Virus Diseases of Ma. Thomas Nelson and Sons: New York, N. Y., 1948. Hygiene 1995; 53(1): 89–94. Mogadishu, Somalia 2011: co-circulation Congress on th Infectious 2–5, 2014. 27. Sang RC. Dengue in africa. In: report of 38. Standard Digital News-Kenya. State the scientific working group on dengue. confirms outbreak of dengue fever in 16. Ulmann E. The Geographical Distribution WHO. pp 50–52: Working Paper 3.3., 2006 Mandera. Kenya; 30 September 2011. of Dengue up to 1957. World Atlas of Epi- 28. Boisier P, Morvan JM, Laventure S, et al. Available at: http://www.standardmed demic Diseases, Part III. Falk-Verlag, 1957. Dengue 1 epidemic in the Grand Comoro ia.co.ke/article/2000043805/state-con- 17. Robinson MC. An epidemic of a dengue- Island (Federal Islamic Republic of the firms-outbreak-of-dengue-fever-in-mand like fever in the southern province of Comores). March-May 1993. Annales de la Tanganyika. The Central African Journal of Société Belge de Médecine Tropicale 1994; Medicine 1956; 2(11): 394–396. 74(3): 217–229. health action: horn of Africa crisis health up- 18. Gubler DJ, Sather GE, Kuno G, Cabral JR. 29. Hyams KC, Oldfield EC, Scott RM, et al. Dengue 3 virus transmission in Africa. Evaluation of febrile patients in Port American Journal of Tropical Medicine and Sudan, Sudan: isolation of dengue virus. Hygiene 1986; 35(6): 1280–1284. American Journal of Tropical Medicine and 19. Messer WB, Gubler DJ, Harris E, era. Accessed on 10/01/2016. 39. World Health Organization. Humanitarian Hygiene 1986; 35(4): 860–865. date, 2011. Available at http://www.who. int/hac/crises/horn_of_africa/update_ nov2011/en/. Accessed on. 02/07/2014. 40. Integrated Regional works (IRIN). Information Net- Kenya: Medics over- Sivananthan K, de Silva AM. Emergence 30. de Laval F, Plumet S, Simon F, Deparis X, and global spread of a dengue serotype 3, Leparc-Goffart I. Dengue surveillance Humanitarian News and Analysis, 2011. subtype III virus. Emerging Infectious Dis- among Accessed: eases 2003; 9(7): 800–809. Emerging Infectious Diseases 2012; 18(2): port/93848/kenya-medics-overwhelmed- 342–343. as-dengue-fever-spreads. 20. Johnson BK, Ochieng D, Gichogo A, et al. French military in Africa. whelmed as dengue fever spreads. http://www.irinnews.org/reAccessed on Epidemic dengue fever caused by dengue 31. Malik A, Earhart K, Mohareb E, et al. type 2 virus in Kenya: preliminary results Dengue hemorrhagic fever outbreak in of human virological and serological stud- children in Port Sudan. Journal of Infection Emergency Weekly Health Update 2013. and Public Health 2011; 4(1): 1–6. Highlights: Reporting January 2013 - ies. East African Medical Journal 1982; 59(12): 781–784. 04/01/2016. 41. World Health Organization Somalia 32. Amarasinghe A, Kuritsk JN, Letson GW, Epidemiological weeks 1–3. Available 21. Metselaar D, Grainger CR, Oei KG, et al. An Margolis HS. Dengue virus infection in at http://www.who.int/hac/crises/som/ outbreak of type 2 dengue fever in the Sey- Africa. Emerging Infectious Diseases 2011; sitreps/somalia_eha_january2013.pdf. chelles, probably transmitted by Aedes 17(8): 1349–1354. Accessed on 02/07/2014. Copyright © 2016 John Wiley & Sons, Ltd. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 194 M. Baba, J. Villinger and D. K. Masiga 42. African Field Epidemiology Network reliefweb.int/report/mozambique/out- (AFENET). Dengue fever spreading along break-dengue-fever-nampula. the East African coastal region: Tanzania on 03/01/2016 and Kenya FELTP residents investigate Regional Information Net- 62. Mail and Guardian. Suspected dengue fever outbreak hits Somalia, April 2005. Available 51. Reliefweb. Sudan: Humanitarian Bulletin Issue 44 | 26 October – 1 November 2015 outbreak in Dar es Salaam, 2013. 43. Integrated Accessed [EN/AR]. Available at: http://reliefweb. at: http://mg.co.za/article/ 2005-04-20-suspected-dengue-fever-outbreak-hits-somalia. Accessed on 02/0/ 2014. works (IRIN). New cases of dengue int/report/sudan/sudan-humanitarian- 63. Gubler DJ, Clark GG. Dengue/dengue fever, bulletin-issue-44-26-october-1-november- hemorrhagic fever: the emergence of a 2015-enar. Accessed on 03/01/2016 global health problem. Emerging Infectious kala-azar reported in Kenya’s north. UN Office for the Coordination of Diseases 1995; 1(2): 55–57. Humanitarian Affairs, 2013. Available 52. Patz JA, Githeko AK, McCarty JP, Hussain at: http://www.irinnews.org/printreport. S, Confalonieri U, De Wet N. Climate 64. Nimmannitya S. Dengue haemorrhagic fe- aspx?reportid=97506. change and infectious diseases. In Climate ver: current issues and future research. Change and Human Health: Risks and Re- Asian-Oceanian Journal of Pediatrics and Accessed on 15/01/2016. 44. Akhwale W. Dengue fever outbreak re- sponses, McMichael A, Campbell- Child Health 2002; 1: 1–21. Lendrum AD, Corvalan C, et al (eds). 65. Gubler DJ. The changing epidemiology of sites. East Afr Public Health Lab Netw WHO: Geneva, Switzerland, 2003; 103–132. yellow fever and dengue, 1900 to 2003: full Newsletter, Issue 07. 2013. Available at: 53. Kilifi: annual weather averages. Available circle? Comparative Immunology, Microbiol- sponse: Kenya sets up sentinel surveillance http://www.eaphln-ecsahc.org/kenya/? at: http://www.holiday-weather.com/ wpfb_dl=39. Accessed on 10/01/2016. kilifi/averages/. Accessed on 04/01/2016. ogy and Infectious Diseases 2004; 27(5): 319–330. 54. Chepkorir E, Lutomiah J, Mutisya J, et al. 66. World Health Organization. Yellow fever response in Mandera east district from Vector competence of Aedes aegypti popu- in Sudan – update. Global Alert and Re- 10th to 24th February 2013. East Africa lations from Kilifi and Nairobi for dengue Public Health Laboratory Networking 2 virus and the influence of temperature. 45. Matini W and Badel M. The dengue fever Project Report. Available at: http:// www.eaphln-ecsahc.org/kenya/?wpfb_ dl=31. Accessed on 28/01/2016. 46. Njeru I. Dengue/DHF UPDATE (33): Asia, Africa, Pacific: Africa-Kenya (Mombasa). Parasit Vectors 2014; 7: 435. sponse, 2013. 67. Cox J, Brown HE, Rico-Hesse R. Variation in vector competence for dengue viruses 55. McBride CS, Baier F, Omondi AB, et al. does not depend on mosquito midgut Evolution of mosquito preference for binding affinity. PLoS Neglected Tropical humans linked to an odorant receptor. Nature 2014; 515(7526): 222–227. Diseases 2011; 5(5): e1172. 68. Armstrong PM, Rico-Hesse R. Efficiency Division of Disease Surveillance and Re- 56. Sang RC, Dunster LM. The growing threat of dengue serotype 2 virus strains to infect sponse, Kenya Ministry of Public Health of arbovirus transmission and outbreaks and disseminate in Aedes aegypti. American and Sanitation, 2013. Available at: http:// in Kenya: A review. East African Medical Journal of Tropical Medicine and Hygiene promedmail.org/direct.php?id=20130428. Journal 2001; 78(12): 655–661. 1676860. Accessed on 10/01/2016. 2003; 68(5): 539–544. 57. Sang R. Dengue re-emergence in Kenya 69. Rico-Hesse R. Microevolution and viru- 47. Ellis EM, Neatherlin JC, Delorey M, et al. A and the approaches for vector control. lence of dengue viruses. Advances in Virus household serosurvey to estimate the The International Congress of Tropical magnitude of a dengue outbreak in Mom- Paediatrics, 2014. basa, Kenya, 2013. PLoS Neglected Tropical Diseases 2015; 9(4): 1–10. 48. UNICEF. Kenya Country Office monthly Research 2003; 59: 315–341. 70. World Health Organization (WHO). Den- 58. Ahmed O. Rift Valley fever, a resurgent gue: Guidelines for Diagnosis, Treatment, threat: Case studies from Sudan and the Prevention and Control: New Edition. Kingdom of Saudi Arabia, 2010. WHO: Geneva, 2009. humanitarian situation report, June 2014. 59. Mwangangi JM, Mbogo CM, Muturi EJ, 71. Simmons CP, Farrar JJ, Nguyen VV, Wills Available at: http://www.unicef.org/ap- et al. Spatial distribution and habitat char- B. Dengue. New England Journal of Medicine peals/files/UNICEF_Kenya.pdf. Accessed acterisation of Anopheles larvae along the on 30/01/2016. Kenyan coast. Journal of Vector Borne Dis- 49. WHO epidemic and pandemic alert and eases 2007; 44(1): 44–51. 2012; 366: 1423–1432. 72. Kassim FM, Izati MN, TgRogayah TA, Apandi YM, Saat Z. Use of dengue NS1 response. Dengue fever outbreak in Mo- 60. Abdallah TM, Ali AAA, Karsany MS, antigen for early diagnosis of dengue virus zambique and Tanzania (Situation as of Adam I. Epidemiology of dengue infec- infection. Southeast Asian Journal of Tropical 14 May 2014), 2014. Available at : http:// tions in Kassala, Eastern Sudan. Journal of Medicine and Public Health 2011; 42(3): 562– www.afro.who.int/pt/grupos-organicos- Medical Virology 2012; 84(3): 500–503. e-programas/dd. Accessed on 10/01/ 61. Mghenyi C. Beware of Dengue fever, 2016. 50. Reliefweb: outbreak of dengue fever in 569. 73. Datta S, Wattal C. Dengue NS1 antigen Kemri warns. Kenya: KEMRI, 2013. Avail- detection: able at: http://www.the-star.co.ke/news/ diagnosis of dengue virus infection. Indian Nampula. REPORT from Mozambique artic…er-kemri-warns. News Agency 2015. Available at: http:// 05/06/2014 Copyright © 2016 John Wiley & Sons, Ltd. Accessed on A useful tool in early Journal of Medical Microbiology 2010; 28(2): 7–10. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv Repetitive dengue outbreaks in East Africa 195 74. Centres for Disease Control and Preven- 83. Halstead SB, Sukhavachana P, Nisalak A. impact of meningitis epidemics for the tion. Dengue homepage-Laboratory guid- Assay of mouse adapted dengue viruses public health system in Burkina Faso. ance 2012. in mammalian cell cultures by an interfer- Available at: http://www.cdc.gov/den- ence method. Proceedings of the Society for 94. Luqman M, Sattar T, Farid S, Warraich IA, gue/clinicalLab/laboratory.html. Experimental Biology and Medicine 1964; Khan WA. Effects of dengue incidence on Accessed on 04/01/2016. 115: 1062–1068. socio-economic status of patient’s family: and diagnostic testing, Vaccine 2011; 29(33): 5474–5480. 75. Wahala WMPB, de Silva AM. The human 84. World Health Organization, 2nd edn. a comparative analysis of Multan and antibody response to dengue virus infec- Chapter 4: Laboratory Diagnosis. Dengue Lahore city (Pakistan). Journal of Economics tion. Viruses 2011; 3(12): 2374–2395. Haemorrhagic Fever: Diagnosis, Treat- and Sustainable Development 2013; 4(13): 76. Mease LE, Coldren RL, Musila LA, et al. Seroprevalence and distribution of ment, Prevention and Control. World 28–39. 95. Sanders EJ, Marfin AA, Tukei PM, et al. Health Organization: Geneva, 1997. arboviral infections among rural Kenyan 85. Toan NT, Ross S, Prisco G, Nante N, adults: a cross-sectional study. Virology Viviani S. Dengue epidemiology in se- Kenya, Journal 2011; 8: 371. lected endemic countries: factors influenc- investigations. ing the expansion factors as estimate of Tropical Medicine and Hygiene 1998; 59(4): 77. Labeaud A, Banda T, Teng C et al. Dengue Tropical Medicine and and chikungunya seroprevalence in rural underreporting. coastal Kenya. Week Adv Sci Improv Care International Health 2015; 20(7): 840–863. [Internet]. 2013 Oct 2; Available at: First recorded outbreak of yellow fever in 1992–1993. I. Epidemiologic American Journal of 644–649. 96. Boutayeb A. The impact of infectious dis- 86. Guzman MG, Alvarez M, Rodriguez- eases on the development of Africa. In https://idsa.confex.com/idsa/2013/web Roche R, et al. Neutralizing antibodies af- Handbookof Disease Burdens and Quality program/Paper41647.html. Accessed on ter infection with dengue 1 virus. Emerging of Life Measures, Preedy VR, Watson RR 27/02/2015. Infectious Diseases 2007; 13(2): 282–286. (eds). Springer: New York, NY, 2010; 78. Awando J, Ongus J, Ouma C, Mwau M. 87. Goncalvez AP, Engle RE, St. Claire M, Seroprevalence of anti-dengue virus 2 Purcell RH, Lai CJ. Monoclonal antibody- 97. World Health Organization Guide to iden- serocomplex antibodies in out-patients mediated enhancement of dengue virus in- tifying the economic consequences of dis- with fever visiting selected hospitals in ru- fection in vitro and in vivo and strategies ease ral parts of western Kenya in 2010–2011: a for prevention. Proceedings of the National Organization: Geneva; 2009. Available at: Academy of Science 2007; 105(22): 9422–9427. http://www.who.int/choice/publications/ cross sectional study. The Pan African Medical Journal 2013; 16: 73. 88. OhAinle M, Balmaseda A, Macalalad AR, 1171–1188. and injury. World d_economic_impact_guide.pdf. Health Accessed 79. Ochieng C, Ahenda P, Vittor AY et al. Sero- et al. Dynamics of dengue disease severity prevalence of infections with dengue, Rift determined by the interplay between viral 98. Hii YL, Zhu H, Ng N, Ng LC, Rocklöv J. Valley fever and chikungunya viruses in genetics and serotype-specific immunity. Forecast of dengue incidence using tem- Kenya, 2007. PLoS ONE 2015; 10(7): Science Translational Medicine 2011; 3(114): perature and rainfall. PLoS Neglected Tropi- e0132645. 114–128. 80. Russell PK, Nisalak A. Dengue virus iden- 89. Soghaier MA, Mahmood SF, Pasha O, et al. on 10/01/2016. cal Diseases 2012; 6(11): e1908. 99. Heng B, Goh K, Neo K. Environmental tification by the plaque reduction neutrali- Factors associated with dengue fever IgG Temperature, Aedes aegypti House Index zation test. Journal of Immunology 1967; 99: sero-prevalence in South Kordofan State, and Rainfall as Predictors of Annual Epi- 291–296. Sudan, in 2012: reporting prevalence ra- demics of Dengue Fever and Dengue tios. Journal of Infection and Public Health Haemorrhagic Fever in Singapore. Minis- 81. Timiryasova TM, Bonaparte MI, Luo P, Zedar R, Hu BT, Hildreth SW. Optimization and validation of a plaque reduction 2014; 7(1): 54–61. try of Environment: Singapore, 1998. 90. World Health Organization. Global alert 100. Fouque F, Carinci R, Gaborit P, Issaly J, neutralization test for the detection of neu- and response. Yellow fever in Sudan – up- Bicout DJ, Sabatier P. Aedes aegypti sur- tralizing antibodies to four serotypes of date. 2013. vival and dengue transmission patterns dengue virus used in support of dengue 91. Halasa YA, Shepard DS, Zeng W. Eco- vaccine development. American Journal of nomic cost of dengue in Puerto Rico. Tropical Medicine and Hygiene 2013; 88(5): American Journal of Tropical Medicine and 962–970. Hygiene 2012; 86(5): 745–752. in French Guiana. Journal of Vector Ecology 2006; 31(2): 390–399. 101. Yang HM, Macoris MLG, Galvani KC, Andrighetti MTM, Wanderley DMV. 82. Thomas SJ, Nisalak A, Anderson KB, et al. 92. Shepard DS, Coudeville L, Halasa YA, Assessing the effects of temperature on Dengue plaque reduction neutralization Zambrano B, Dayan GH. Economic im- dengue transmission. Epidemiology and In- test (PRNT) and secondary dengue virus pact of dengue illness in the Americas. infections: How alterations in assay condi- American Journal of Tropical Medicine and tions impact performance. American Journal of Tropical Medicine and Hygiene 2009; 81(5): 825–833. Hygiene 2011; 84(2): 200–207. fection 2009; 137(8): 1179–1187. 102. Hurtado-Díaz M, Riojas-Rodríguez H, Rothenberg SJ, Gomez-Dantés H, 93. Colombini A, Badolo O, Gessner BD, Cifuentes E. Short communication: impact Jaillard P, Seini E, Da Silva A. Costs and of climate variability on the incidence of Copyright © 2016 John Wiley & Sons, Ltd. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv 196 M. Baba, J. Villinger and D. K. Masiga dengue in Mexico. Tropical Medicine and Sciences of the United States of America International Health 2007; 12: 1327–1337. 2009; 106(3): 955–959. 106. Archer B, Thomas J, Weyer J, et al. Epidemiologic investigations into outbreaks of Rift Valley fever in humans, South Africa, 103. Colón-González FJ, Lake IR, Bentham G. Cli- 105. East Africa Public Health Laboratory Net- mate variability and dengue fever in warm working Project. Available at: http:// 2008–2011. Emerging Infectious Diseases and humid Mexico. American Journal of Tropi- eaphln-ecsahc.org/index.php/compo- 2013; 19(12): 1918–1925. cal Medicine and Hygiene 2011; 84: 757–763. nent/content/article/113-eaphln-twgs/lab- 104. Anyamba A, Chretien J-P, Small J, et al. 107. Robertson C, Sawford K, Daniel SLA, Nelson networking-and-accreditation/10-lab- TA, Stephen C. Mobile phone-based infectious Prediction of a Rift Valley fever outbreak. networking-and-accreditation, Accessed on disease surveillance system, Sri Lanka. Emerg- Proceedings of the National Academy of 02/07/201. ing Infectious Diseases 2010; 16(10): 1524–1531. Copyright © 2016 John Wiley & Sons, Ltd. Rev. Med. Virol. 2016; 26: 183–196. DOI: 10.1002/rmv