Transactions of the Royal Society of Tropical Medicine and Hygiene (2007) 101, 707—713
available at www.sciencedirect.com
journal homepage: www.elsevierhealth.com/journals/trst
Brucellosis as a cause of acute febrile illness in Egypt Gregory J. Jennings a, Rana A. Hajjeh a,d, Fouad Y. Girgis a, Moustafa A. Fadeel a, Mohamed A. Maksoud a, Momtaz O. Wasfy a, Nasr El Sayed b, Padmini Srikantiah c,d, Stephen P. Luby d, Kenneth Earhart a, Francis J. Mahoney a,d,e,∗ a
US Naval Medical Research Unit-3 (NAMRU-3), Cairo, Egypt Egyptian Ministry of Health and Population, Cairo, Egypt c San Francisco General Hospital, San Francisco, CA, USA d Centers for Disease Control and Prevention, Atlanta, GA, USA e Eastern Mediterranean Regional Office, WHO, Cairo, Egypt b
Received 12 July 2006; received in revised form 26 February 2007; accepted 26 February 2007 Available online 17 April 2007
KEYWORDS Brucellosis; Brucella melitensis; Brucella abortus; Surveillance; Incidence; Egypt
Summary To develop better estimates of brucellosis incidence, we conducted populationbased surveillance for acute febrile illness (AFI) in Fayoum governorate (population 2 347 249), Egypt during two summer periods (2002 and 2003). All hospitals and a representative sample of community healthcare providers were included. AFI patients without obvious etiology were tested for brucellosis by culture and serology. Incidence estimates were calculated adjusting for sampling methodology and study period. Of 4490 AFI patients enrolled, 321 (7%) met the brucellosis case definition. The estimated annual incidence of brucellosis per 100 000 population was 64 and 70 in 2002 and 2003, respectively. The median age of brucellosis patients was 26 years and 70% were male; 53% were initially diagnosed as typhoid fever. Close contact with animals and consumption of unpasteurized milk products were associated with brucellosis. The high incidence of brucellosis in Fayoum highlights its public health importance, and the need to implement prevention strategies in humans and animals. © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.
1. Introduction ∗ Corresponding author. Present address: NAMRU-3, PSC 452 Box 5000 (code 304), FPO AE 09835, Egypt. Tel.: +11 202 276 5287; fax: +11 202 276 5414. E-mail address:
[email protected] (F.J. Mahoney).
Brucellosis is a zoonotic disease that is widely distributed throughout the developing world. In 1992, WHO reported that 86 (49%) out of 175 countries were affected, with an estimated population at risk of 2.4 billion persons (WHO,
0035-9203/$ — see front matter © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.trstmh.2007.02.027
708 1996). Surveillance data suggest that brucellosis incidence is increasing in many countries of the Mediterranean Region and the Middle East (WHO, 1997). However, it is not clear whether the increase reflects improved surveillance and better recognition of disease or a true increase in disease incidence. Estimates of brucellosis incidence in Egypt are based primarily on surveillance data reported from infectious disease hospitals (n = 108) throughout the country. Data from these hospitals indicate considerable variations in disease incidence by region, with rates ranging from 0.2 to 26/100 000 (data from National Information Center for Health and Population). The diagnosis of brucellosis in this surveillance system is based almost entirely on clinical criteria with little laboratory confirmation of disease. Laboratory-based surveillance among patients with acute febrile illness (AFI) admitted to infectious disease hospitals throughout Egypt reported that 11% of patients had evidence of acute brucellosis (Afifi et al., 2005). While these data indicate that brucellosis is widespread in Egypt, there is little reliable information on disease incidence. To better characterize the epidemiology of dis-
G.J. Jennings et al. ease and better define disease burden, population-based surveillance for patients with AFI was established in Fayoum Governorate (population 2.3 million). This paper summarizes the results of surveillance activities over a 2 year period and discusses the implications of these studies for public health control measures.
2. Materials and methods Fayoum Governorate, located ∼100 km southwest of Cairo (Figure 1), is geographically isolated by desert, which limits population mobility, making it a suitable site for population-based surveillance. The governorate is divided into six administrative districts and is largely agrarian, with a total estimated population of 2 347 249. Population-based surveillance for patients with AFI was conducted in two time periods, including June to October 2002 and June to October 2003. Case-finding methods involved recruitment of a representative sample of primary care providers from all tiers of health service in the study area. These tiers included the infectious disease hospital, all six district general hospitals, all infectious disease or ‘fever’ specialists and a random sample of rural health units and primary care providers. We defined primary care providers as general practitioners, internal medicine physicians or pediatricians. In 2002, approximately 10% of rural health units (13/138) and primary care providers (18/186) participated in the surveillance. All providers were randomly selected from a comprehensive roster of practicing physicians in the catchment area. Provider enrollment was expanded to 20% in 2003, to obtain better estimates of disease incidence at the district level. None of the selected primary care providers refused to participate in this study.
2.1. Case definitions AFI was defined as any individual ≥1 year of age with a temperature greater than 38 ◦ C at the time of visit, or history of fever of more than 2 d duration, and no identified cause of fever such as diarrhea, hepatitis and respiratory tract infections, or alternatively any patient with a clinical diagnosis of typhoid fever or brucellosis. A case of brucellosis was defined as an AFI patient with laboratory-confirmed brucellosis, established by the isolation of Brucella spp. from blood culture or a serologic titer ≥1:320 by the tube agglutination test.
2.2. Data collection
Figure 1 (A) Location of Fayoum Governorate, ∼100 km southwest of Cairo, Egypt. (B) Brucellosis incidence for 2003 for Fayoum Governorate and its six administrative districts.
All patients meeting the case definition of AFI were invited to participate in the study. Providers were trained to conduct a standardized clinical and laboratory evaluation, including collection of blood cultures and serum samples at the time of clinical evaluation. Data from consenting patients were collected using a brief, standardized questionnaire that included information on demographic and clinical characteristics. In 2003, questions were added to the questionnaire regarding antibiotic use, occupation and animal exposures.
Brucellosis as a cause of acute febrile illness in Egypt
2.3. Incidence calculations We calculated the incidence of brucellosis after considering the provider-sampling scheme, diagnostic test sensitivity and specificity and duration of sampling (Crump et al., 2003). To account for provider sampling, we derived a representative arithmetic multiplier for each of the five tiers of healthcare providers. In 2002, 89% of fever specialists were enrolled for a multiplier of 1.1. In 2003, 93% of fever specialists were enrolled for a multiplier of 1.1. In 2002, 9.4% of rural health units and 9.7% of primary care providers were enrolled for multipliers of 10.6 and 10.3, respectively. In 2003, 19.4% of rural health units participated for a multiplier of 5.2, and 18.5% of primary care providers were enrolled for a multiplier of 5.4. Because the majority of brucellosis cases can be detected with a combination of blood culture and serology, we assumed that test sensitivity approached 100%; consequently we used a multiplier of one. To adjust the 20 week study period to represent an entire year, we multiplied the number of patients with brucellosis identified during surveillance by 2.6 (Table 1) using estimates of seasonality based on the monthly distribution of hospitalized patients with brucellosis in Egypt from 1999 to 2003 (Afifi et al., 2005).
2.4. Statistical analysis Data were double-entered into a Microsoft Access database. Population incidence was calculated using annualized census estimates. To characterize risk factors for disease, we compared brucellosis patients with other AFI patients who were culture negative for typhoid fever. Basic descriptive analysis of demographic data and age-adjusted prevalence ratios were calculated using Epi Info 2000 (CDC, Atlanta, GA, USA). A P-value 1:160, WHO/CDS/CSR/ISR/99.2), thus the incidence may actually under-represent overall burden of disease. Although this study focused on Fayoum Governorate, studies of AFI patients have found that brucellosis is a common cause of disease in all parts of Egypt. In a study of hospitalized patients with AFI in diverse regions of the country, 12 to 40% had laboratory-confirmed brucellosis (Afifi et al., 2005). These data also suggest that disease is prevalent in both urban and rural settings in all parts of the country. As brucellosis is usually transmitted through exposure to infected animals or consumption of contaminated products, these studies imply that Brucella infection in domestic animals is widespread in Egypt (Refai, 2002). The incidence rates observed in Fayoum Governorate are consistent with those reported throughout the Middle East, which range from 8/100 000 in the Gaza Strip (Awad, 1998; Memish, 2001) to 40/100 000 in Saudi Arabia (Memish, 2001), 46/100 000 in Jordan (Dajani et al., 1989) and 85/100 000 in Kuwait (Mousa et al., 1987). A recent population-based serosurvey in southern Saudi Arabia demonstrated serological evidence of past exposure to Brucella among 20% of the population, with more than 2% having evidence of active disease (Alballa, 1995). Although brucellosis is recognized as a common cause of AFI in various parts of Egypt, this study revealed that it is often misdiagnosed and mistreated. More than half of all brucellosis cases were misdiagnosed as typhoid, and only a quarter of them received treatment with more than one antibiotic. We could not assess duration of therapy in this study. However, well-established clinical guidelines for management of brucellosis require treatment with two antibiotics for a minimum of 6 weeks in order to minimize the risk of relapse (Hall, 1990; Solera et al., 1997). It is therefore important to educate physicians in Egypt about proper treatment of brucellosis, and to develop national treatment guidelines. Furthermore, it is important to improve laboratory capacities available to physicians in Egypt in order to better diagnose brucellosis, especially as
712 the clinical presentation of this disease is relatively nonspecific. In the current study, we found that the majority of brucellosis patients were adult males, with approximately one-third citing their principle occupation as farmer. Exposures associated with infection included close contact with sheep, cattle or buffalo, exposure to an animal abortus (aborted fetus) and consumption of dairy products such as soft cheeses and yogurt. As these dairy products are often unpasteurized and produced locally, they represent a plausible source of infection. Consumption of raw milk was not identified as a risk factor for disease. However, this exposure was quite common among other AFI patients as well, indicating that a much larger sample would be needed to demonstrate a significant association. The current study was not designed to evaluate the proportion of brucellosis attributable to food versus animal exposures, and it is difficult to evaluate the potential impact of different prevention strategies. Better risk factor studies are needed to assess attributable risk due to these different exposures. Despite this limitation, the high frequency of exposures associated with well-defined risk factors highlights the need for extensive efforts to educate the population regarding potential food- and animal husbandry-related exposures. In addition to the hazard it represents to human health, Brucella infection has a considerable economic impact on animal husbandry, including significant loss in agricultural productivity due to abortions, sterility, decreased milk production and costs associated with veterinary care (Roth et al., 2003). In addition, the disease can be an impediment to animal movement and exportation. To better define the economic impact of disease, quality data are needed on the prevalence of disease in livestock populations, and estimates of control costs to both producers and consumers. However, surveillance and control of brucellosis in Egypt and other countries in the Eastern Mediterranean Region are complex, due to the diversity of domestic animal species that are infected, different modes of transmission from animals to humans, and the challenges of disease control. Animal husbandry is the primary occupation for a large sector of the rural population in Egypt, where the livestock population includes 2.8 million cattle, 3.2 million buffalo, 3.7 million sheep, 2.9 million goats and 220 000 camels (Refai, 2002). Seroprevalence studies of domestic animals indicate that Brucella infection rates vary widely among different species and in different parts of the country (A. El-Taweel, unpublished data; A.M. Montasser, unpublished data) (Refai, 2002). Vaccination of livestock populations is a well-recognized and agreed-upon strategy for control of brucellosis. However, delivery strategies are challenging, particularly in the Middle East, with diversity of species distributed over wide geographic regions. In Egypt, animal husbandry practices include close contact among different species; thus control programs targeting one species are likely to have a limited impact. In this project, B. melitensis was, with a single exception, the only species isolated. Brucella melitensis is typically associated with sheep and goats; however, the species is being increasingly recognized as a source of infection for both cattle and buffalo (Corbel, 1997; Hamdy and Amin, 2002; Stevens et al., 1994). The predominance of B. meliten-
G.J. Jennings et al. sis as a cause of disease in humans suggests that it has adapted to cattle and buffalo in Egypt, as exposure to these animals and their dairy products has been recognized as a source of infection. This cross-species adaptation creates a challenge to vaccination strategies, because vaccines such as Rev1 were developed to combat B. melitensis infection in sheep and goats (Zygmunt et al., 1994), while RB51 and Strain 19 were developed to reduce infection by B. abortus in cattle (Stevens et al., 1995). The use and efficacy of these vaccines in alternate hosts requires additional investigation (Denes, 1997). Coupled with the uncertainty regarding vaccine effectiveness is the uncertainty of program effectiveness. Considerable resources and program experience are necessary to implement and sustain animal vaccination programs with substantial investment from both the public and private sector. Experience in many countries suggests that it can take decades to eradicate brucellosis with a combined strategy of vaccination followed by test and slaughter activities. To establish an economic framework for such an investment, it is recommended that veterinary and national health authorities work together to evaluate available data on disease burden and cost-effectiveness of control strategies. Given the challenges of implementing control programs in animals, it is recommended that national health authorities and veterinary services work together to ensure pasteurization of commercial dairy products, increase public awareness of brucellosis and minimize high-risk occupational and domestic exposures such as handling an abortus. In addition, efforts are needed to educate clinicians on recognition of disease, to establish laboratory capacity to confirm the diagnosis and to ensure availability of appropriate antibiotics for treatment. Disclaimer: The opinions and assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the Navy Department, Department of Defense, the U.S. Government, or the Egyptian Ministry of Health and Population. Authors’ contributions: GJJ, RAH, KE and FJM conceptualized and designed the study; FYG, PS, NES and SPL carried out the clinical assessments; MAF, MAM and MOW carried out the immunoassays and microbiological examination; GJJ, RAH, KE, MOW and FJM drafted the manuscript. All authors read and approved the final manuscript. FJM is guarantor of the paper. Acknowledgements: We would like to acknowledge the Fayoum Governorate Health Directorate and the physicians and laboratory personnel at hospitals and clinics in Fayoum Governorate for their valuable contributions to this surveillance study. We would also like to recognize Mr Muhammad El Mofty and Mohammed Adel for their valuable assistance in managing the Fayoum data and Dr Guillermo Pimentel for assistance during manuscript preparation. Funding: Global Emerging Infections Surveillance System, United States Department of Defense, and the US Agency for International Development, work unit #847705.82000.25GB.E0018.
Brucellosis as a cause of acute febrile illness in Egypt Conflict of interest: None declared. Ethical approval: Institutional review boards of NAMRU-3 and the US Centers for Disease Control and Prevention.
References Afifi, S., Earhart, K., Azab, M.A., Youssef, F.G., El-Sakka, H., Wasfy, M., Mansour, H., El-Oun, S., Rakha, M., Mahoney, F., 2005. Hospital-based surveillance for acute febrile illness in Egypt: a focus on community-acquired bloodstream infections. Am. J. Trop. Med. Hyg. 73, 392—399. Alballa, S.R., 1995. Epidemiology of human brucellosis in southern Saudi Arabia. Am. J. Trop. Med. Hyg. 98, 185—189. Awad, R., 1998. Human brucellosis in the Gaza Strip, Palestine. Eastern Mediterranean Health J. 4, 225—233. Bricker, B.J., Halling, S.M., 1994. Differentiation of Brucella abortus bv. 1, 2, and 4, Brucella melitensis, Brucella ovis, and Brucella suis bv 1 by PCR. J. Clin. Microbiol. 32, 2660—2666. Corbel, M.J., 1997. Brucellosis: an overview. Emerg. Infect. Dis. 3, 213—221. Crump, J.A., Youssef, F.G., Luby, S.P., Wasfy, M.O., Rangel, J.M., 2003. Estimating the incidence of typhoid fever and other febrile illnesses in developing countries. Emerg. Infect. Dis. 9, 539— 544. Dajani, Y.F., Masoud, A.A., Barakat, H.F., 1989. Epidemiology and diagnosis of human brucellosis in Jordan. Am. J. Trop. Med. Hyg. 92, 209—214. Denes, B., 1997. Serological findings obtained in cattle herds immunized with the Brucella melitensis Rev.1 and the B. abortus B19 vaccine in Mongolia. Acta Vet. Hung. 45, 33—43. Hall, W.H., 1990. Modern chemotherapy for brucellosis in humans. Rev. Infect. Dis. 12, 1060—1099.
713 Hamdy, M.E., Amin, A.S., 2002. Detection of Brucella species in the milk of infected cattle, sheep, goats and camels by PCR. Vet. J. 163, 299—305. Memish, Z., 2001. Brucellosis control in Saudi Arabia: prospects and challenges. J. Chemother. 13 (Suppl. 1), 11—17. Mousa, A.M., Elhag, K.M., Khogali, M., Sugathan, T.N., 1987. Brucellosis in Kuwait: a clinico-epidemiological study. Trans. R. Soc. Trop. Med. Hyg. 81, 1020—1021. Refai, M., 2002. Incidence and control of brucellosis in the Near East region. Vet. Microbiol. 90, 81—110. Roth, F., Zinsstag, J., Orkhon, D., Chimed-Ochir, G., Hutton, G., Cosivi, O., Carrin, G., Otte, J., 2003. Human health benefits from livestock vaccination for brucellosis: case study. Bull. World Health Organ. 81, 867—876. Solera, J., Espinosa, A., Martinez-Alfaro, E., Sanchez, L., Geijo, P., Navarro, E., Escribano, J., Fernandez, J.A., 1997. Treatment of human brucellosis with doxycycline and gentamicin. Antimicrob. Agents Chemother. 41, 80—84. Stevens, M.G., Hennager, S.G., Olsen, S.C., Cheville, N.F., 1994. Serologic responses in diagnostic tests for brucellosis in cattle vaccinated with Brucella abortus 19 or RB51. J. Clin. Microbiol. 32, 1065—1066. Stevens, M.G., Olsen, S.C., Cheville, N.F., 1995. Comparative analysis of immune responses in cattle vaccinated with Brucella abortus strain 19 or strain RB51. Vet. Immunol. Immunopathol. 44, 223—235. WHO, 1996. The World Health Report, 1996, Executive Summary. World Health Organization, Geneva. WHO, 1997. Brucellosis fact sheet N173. World Health Organization, Geneva. Young, E.J., 1995. An overview of human brucellosis. Clin. Infect. Dis. 21, 283—289. Zygmunt, M.S., Debbarh, H.S., Cloeckaert, A., Dubray, G., 1994. Antibody response to Brucella melitensis outer membrane antigens in naturally infected and Rev1 vaccinated sheep. Vet. Microbiol. 39, 33—46.
