Journal of Infection (2006) 53, 308e314

www.elsevierhealth.com/journals/jinf

Prevalence of Brucella antibodies in rural and

suburban communities in three provinces of
Turkey: Need for improved diagnosis and
prevention
Sukran Kose a, Henk L. Smits b,*, Theresia H. Abdoel b, Yusuf Ozbel c

a
Department of Clinical Microbiology and Infectious Diseases, Tepecik Educational Hospital,
Yenisehir, Izmir, Turkey
b
KIT Biomedical Research, Royal Tropical Institute/Koninklijk Instituut voor de Tropen (KIT),
Meibergdreef 39, 1105 AZ Amsterdam, The Netherlands
c
Department of Parasitology, Ege Medical University, Izmir, Turkey

Accepted 13 December 2005

Available online 7 February 2006

KEYWORDS Summary Objective: To determine the seroprevalence of Brucella-specific anti-

Brucellosis; bodies in rural and suburban communities in different provinces of Anatolia.
Seroprevalence; Method: Cross-sectional seroepidemiological study on serum samples collected in
Point-of-care test; communities in two relatively developed provinces in west Anatolia with an official
Turkey; low prevalence of brucellosis and in one province in southeast Anatolia with a high
Risk factor prevalence.
Result: The seroprevalence of brucellosis in the two provinces in the west Anatolia
appears to be high and ranged from 2.9 to 8.5% in Rose Bengal test and from 0 to
5.6% in Wright serum agglutination test at a titer equal or higher than 1:100. The
seroprevalence in communities in the province in southeast Anatolia was lower
and this might well be attributed to vaccination of livestock in the year preceding
the survey.
Conclusion: Adherence to traditional farming practices and lifestyle, and a prefer-
ence for fresh dairy contribute to the high seroprevalence of brucellosis. Vaccina-
tion of livestock is of utmost importance and the consumption of fresh milk and
dairy products prepared from unpasteurised milk should be halted. Better access
to laboratory testing is needed for the confirmation and management of brucellosis.
ª 2005 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: þ31 20 5665470; fax: þ31 20 6971841.

E-mail address: [email protected] (H.L. Smits).

0163-4453/$30 ª 2005 The British Infection Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jinf.2005.12.022
Seroprevalence of brucellosis 309

Introduction respectively.8 Not all health care facilities in

Turkey have access to laboratory testing and labora-
Brucellosis is a zoonotic disease that may have tory testing is not always used to confirm the
a major veterinarian, public health and economic clinical diagnosis of brucellosis. The Brucella
impact. Brucellosis in human presents as an acute IgM/IgG flow assay is a simple and rapid field test
febrile illness with undulant fever, which might for the confirmation of brucellosis that may be
progress to a more persistent form. The pathogen used outside the laboratory.9 The clinical utility
may affect almost any organ notably the musculo- of this new test for the diagnosis of patients with
skeletal, cardiovascular and central nervous an active infection may depend on the prevalence
systems with sometimes serious complications.1e3 of Brucella-specific antibodies in a given popula-
Animals are the almost-exclusive source of infec- tion. Therefore in this study we applied the flow
tion for human. Brucella melitensis has host prev- assay on samples collected during the serosurvey
alence for goats and sheep, and Brucella abortus to confirm its specificity.
for cattle. Although, both species infect man, Bru-
cella melitensis is more infectious than B. abortus.
The pathogen is transmitted to man through Methods
close contacts with infected animals or animal
products and through the consumptions of unpas- Study area and blood collection
teurised milk and of dairy products such as fresh
cheeses and yogurts that are prepared from non- Cross-sectional serosurveys with random inclusion
pasteurized milk.4 Farmers, animal caretakers, of inhabitants were performed in communities in
veterinarians, butchers and slaughterhouse workers (i) Odemis located in west Anatolia 114 km south-
have an increased risk of infection because of east of Izmir in Izmir province, (ii) Burhaniye
their professional activities.5 located 160 km north of Izmir in Balikesir province
In Turkey, B. melitensis is more common than and (iii) Sanliurfa province located in Southeast
B. abortus and is the main cause of infection in Anatolia (Fig. 1).
humans.6 However, a recent study from Turkey did The survey in Odemis was conducted in March
not show differences in clinical characteristics of 2003 in three villages, named Kofundere, Orhan-
bacteremic brucellosis caused by B. melitensis gazi and Uzundere, and a nearby urban district
and B. abortus.7 Brucellosis is widespread with an named Buyuk with blood samples collected from
official annual incidence rate of 22.9 per 100,000 249 individuals belonging to 66 households. The
and prevalence rates vary widely for different prov- survey was repeated in the same three villages in
inces.8 The prevalence of brucellosis is highest in April 2004 with samples collected from 237 in-
the provinces in the southeast of Turkey where rates dividuals of a group of 56 other households. The
above 100 per 100,000 are reported for Diyarbakir, survey in Burhaniye was performed in August 2004
Siirt, Sirnak, Hakkari, Batman and Aksaray. The with samples collected from 126 individuals from
prevalence rates for these six provinces are 109.3, 22 households in the village of Dutluca. The survey
130.3, 203.4, 166.6, 313.8 and 410.1, respectively. in Sanliurfa was performed in June 2004 with
However, the disease is still fairly common in other samples collected from 45 villagers in Kurudere
more developed parts of the country. village, 135 individuals from an urban district
At primary health care centers in the relatively named Harrankapi, and 40 individuals from the
developed province of Izmir in west Anatolia the town of Siverek of 51 households in total. House-
diagnosis of brucellosis is fairly common and many holds of which the head of the household was
patients are referred to the larger hospitals in the present were selected and after his permission was
cities for laboratory testing and treatment. The obtained all household members that were present
seroprevalence of Brucella-specific antibodies in were asked to participate.
a community provides an indication of exposure During the 2003 village survey in Odemis one of the
to the pathogen. In this study we determined the investigators completed a structured questionnaire
seroprevalence in rural and suburban communities inquiring into demographic, health and risk factor
in Odemis in the province of Izmir and in Burhaniye data for all participants entered in the study by oral
in the nearby province of Balikesir. For comparison interview. The median age was 27.3 years (range: 4e
a serosurvey was performed in communities in San- 83 years; 25e75 percentile: 15e37 years).
liurfa province in southeast Anatolia. In 2001, the Ethical permission for the study was obtained
reported prevalence rates of brucellosis in Izmir, from the medical ethical committee of the Ege
Balikesir and Sanliurfa were 3.8, 10.0 and 85.7, University Medical School. The purpose of the
310 S. Kose et al.

Figure 1 Geographical location of study sites (C) in Izmir, Balikesir and Diyarbakir provinces in west and southeast
Anatolia.

study was explained to all individuals and oral The Brucella IgM/IgG flow assay was performed for
informed consent was obtained before blood was all samples collected at the Odemis study sites in
collected. In case of children permission was 2003 and the Sanliurfa study sites in 2004. For
obtained from one of the parents. the samples collected at the Odemis and Burha-
niye study sites in 2004 the flow assay was only
Serology performed for the RB test positive samples.

The Rose Bengal (RB) test with antigen purchased Results

from Chromatest, Linear Chemicals, Spain, and
Wright, 2-mercaptoethanol (2-ME) and Coombs Seroprevalence in RB test was 4.0% (95% Confi-
tests with antigens obtained from the Veterinary dence Interval, CI: 2.2e9.5) for samples collected
Control and Research Institute in Pendik, Istanbul, in Odemis in 2003 and ranged from 2.0% for the
Turkey were performed and interpreted according village of Uzundere to 5.6% for the village of
to routine methods and criteria. Briefly, RB test Kofundere (Table 1). None of the samples col-
was performed by mixing a drop of serum spotted lected in the urban district in Odemis tested posi-
on a glass plate with an equal volume of antigen. tive in RB test. In 2004 the seroprevalence in RB
The result was read after 2 min. Two-fold serial test in Odemis was slightly higher (5.1%, 95 CI:
dilutions starting at 1:50 were tested by Wright’s, 3.1e9.4) and ranged from 2.9 to 8.5% for the dif-
2-ME and Coombs tests to determine Brucella ferent communities. The mean seroprevalence in
antibody levels. The tests were performed in test Wright test for samples collected in Odemis in
tubes and the following day agglutination was 2003 and 2004 was 5.2 and 2.1%, respectively, at
looked for under a binocular. a titer of 1:100 and above. At a one-titer step
The Brucella IgM and the Brucella IgG flow higher cut-off value the seroprevalence was 2.4
assays were performed exactly as described previ- and 0.8%, respectively. In the village of Dutluca
ously.9 Briefly, the flow assay is performed by the in the Burhaniye study area, the seroprevalence
addition of 5 ml serum followed by 130 ml running by both RB and Wright test in 2004 was 4.0% (95%
fluid to the sample pad in the sample well of a plas- CI: 1.5e9.5), while Coombs test confirmed the
tic assay device. The assay is read after 10e15 min presence of specific antibodies in three out of
by visual inspection for staining of the antigen line five of these samples. None of the samples col-
in the test zone of the assay device. Tests are lected in the village of Kurudere or in the town
scored negative when no staining is observed and of Siverek in the Sanliurfa study area were sero-
positive when a reddish staining is observed. Stain- positive. In the Harrankapi urban district, the sero-
ing intensity is subjectively rated 1þ for weak prevalence in RB and Wright tests was 2.2%, while
staining, 2þ for moderately strong staining, 3þ seroprevalence for the three communities in the
for strong staining, and 4þ for very strong staining. Sanlurfia study area together was 1.4% (95% CI:
Rose Bengal, Wright and 2-ME tests were per- 0.4e4.3).
formed for all samples. Coombs test was per- Eleven of the 17 samples that had at
formed for the samples collected in 2004 only. a titer  1:200 in Wright test were sensitive to
Seroprevalence of brucellosis 311

Table 1 Prevalence of antibodies to Brucella spp. in communities in Izmir, Balikesir and Diyarbakir in west and
southeast Anatolia
Study site, year (no. tested) Percent positive in the following assays
RB Wright test 2-ME Coombs Flow assay
1:100 1:200 IgM IgG
Odemis, Izmir, 2003 (264) 4.0 5.2 2.4 0.4 nt 3.6 1.2
Kofundere (124) 5.6 5.6 1.6 0.8 nt 1.6 0
Orhangazi (75) 2.7 5.3 4.0 0 nt 5.3 0
Uzundere (50) 2.0 4.0 2.0 0 nt 2.0 0
Buyuk (15) 0 0 0 0 nt 0 0
Odemis, Izmir, 2004 (237) 5.1 2.1 0.8 0 0 nt nt
Kofundere (70) 8.5 0 0 0 0 nt nt
Orhangazi (72) 2.9 2.8 1.4 0 0 nt nt
Uzundere (91) 3.3 3.3 1.1 0 0 nt nt
Burhaniye, Balikesir, 2004
Dutlaca (126) 4.0 4.0 4.0 4.0 2.4 nt1 nt
Sanliurfa, Diyarbakir, 2004 (220) 1.4 1.4 1.4 0 0 0 0
Kurudere (45) 0 0 0 0 0 0 0
Harankapi (135) 2.2 2.2 2.2 0 0 0 0
Siverek (40) 0 0 0 0 0 0 0
nt ¼ not tested.
1
One of the RB positive samples tested in the flow assays tested 2þ positive in the IgM and 1þ positive in the IgG flow assay.

the addition of reducing agent in 2-ME test indicat- of these eight individuals had Brucella-specific
ing that agglutinating was due to reactivity of IgM antibodies at the time of the survey. Members of
antibodies. A total of 13 samples, 12 from Odemis eight other households mentioned illness diag-
and one from Burhaniye, reacted in the Brucella nosed as brucellosis in family members who did
IgM/IgG flow assay. Of the 13 flow test positive not participate in the survey.
samples nine tested positive for IgM, two tested
positive for IgG and one tested positive for both
IgM and IgG antibodies. The staining intensity was Discussion
weak (1þ) in 11 samples and only two samples,
one IgM positive sample and one IgG positive sam- Many recent clinical studies have shown that
ple gave a moderately strong (2þ) staining. None brucellosis is still a common health problem in
of the participants presented with clinical illness Turkey often causing severe clinical illness with
at the time of the surveys. complications.2,10e17 In this country adults as well
Virtually all (93.9%, 95% CI: 84e98) households as children are affected.18e20 The disease is pres-
of the Odemis area keep livestock. Most (68%) ent in rural and urban areas and is mostly associ-
households keep sheep, 41% keep cattle and 19% ated with the consumption of contaminated food
have goats. Mixed farming, e.g., keeping cattle products.21,22 The number of reported cases varied
and sheep and/or goats was practiced in 36.4% between the different provinces. Most cases are
(95%CI: 25e49). Abortions in livestock (particularly reported from southeast Anatolia.8 For instance,
in sheep) were very common and reported by 62% 1231 cases were reported in 2001 from the prov-
of the households. More than half (61.4%, 95% CI: ince of Sanliurfa in southeast Anatolia, while only
55e68) of the individuals that were interviewed 130 cases were reported from the province of Izmir
had activities that brought them in close contacts in the most western part of the country. For com-
with their animals and the majority (85.1%, 95% CI: parison, in the same year, 1490 cases were re-
80e89) consumed fresh cheeses and other dairy ported from the province of Diyarbakir, one of
products prepared from unpasteurised milk. In- the provinces with the highest incidence of brucel-
dividuals who did not have contact with animals or losis. A countrywide surveillance study performed in
consume fresh dairy products were only 8% (95% CI: 1990 showed an average seroprevalence of 1.8% in
5e12). Eight participants reported to have been Wright test at a cut-off titer of 1:40 by testing over
treated for brucellosis during the last 12 months. 70,000 serum samples.23 A highest seroprevalence
The local doctor had diagnosed these cases. None of 3.6% was observed in Diyarbakir. The relatively
312 S. Kose et al.

high seroprevalence with a predominance of the fact that the living conditions, farming prac-
Brucella-specific IgM antibodies observed in the tices and food habits of farmers in this area are not
present study for the communities in Izmir (4.0 unlike those in communities in Izmir and Balikesir.
and 5.1% in RB test in 2003 and 2004, respectively) However, during the collection of the serum
and Balikesir (4.0% in RB test in 2004) indicates samples in the Sanliurfa area we learned that
that the population is exposed to the pathogen cattle in the area had been vaccinated in 2003,
regularly. Most of the villagers in the Odemis and the year preceding the survey. Mass vaccination of
Burhaniye study areas in these relatively devel- livestock is essential to control brucellosis and to
oped provinces are farmers who have retained reduce the number of human cases.27 If vaccina-
their traditional habits and farming practices and tion indeed has been effective and has reduced
who raise a small number of animals for their transmission, a lower number of reported cases
nutritional needs and their lives depend almost might be expected for subsequent years in these
entirely on their livestock. Abortion of livestock communities. In a seroprevalence study that was
is a problem at most farms and is suggestive of in- conducted in 2002, and included about 15,000 pa-
fection with Brucella. Animals are not vaccinated, tients attending a university hospital in Diyarbakir,
kept closely together in small stables during winter in a province neighboring Sanliurfa, a seropreva-
and animals from different households use the lence of 2.3% in RB test and of 1.3% in SAT test
same pastures. Most villagers have a preference at titers of 1:160 and above was found.28 These
for unpasteurised milk and traditionally prepared values are similar to those obtained during the
soft cheeses and even in suburban areas where present survey for the samples collected in Harran-
dairy products prepared from pasteurized milk kapi district.
are available, cheeses prepared from unpasteur- Several other seroprevalence studies performed
ised milk are preferred because it is tastier. Hence in other provinces have been published. Colak and
infections may spread among the farm animals and coworkers found a seroprevalence of 15.7% in RB
are easily transmitted to humans. The eight cases test and of 0.4% in Wright test at titers of 1:160
of brucellosis recalled by the participants inter- and above for samples collected in the Afyon
viewed in the 2003 Odemis survey indicate province in 1991, a province with a prevalence of
a need for laboratory testing. The local doctor brucellosis of 38.9 per 100,000 in 2001.8,29 In
had diagnosed these patients based on clinical a more recent study performed in Malatya in
signs and symptoms. Patients with brucellosis 1996, RB test seropositivity was 2.9% among 245
remain seropositive for a long period even if the samples collected from butchers and was 1.6%
patient has been treated. As none of these possi- for 2852 others, the majority of whom had been
ble cases tested positive during the survey the in contact with animals.30 The seroprevalence in
reported cases are questionable and might be mis- SAT for these two groups was 0.8 and 1.0%, respec-
diagnosed. However, if such cases could be con- tively. In another recent study a seroprevalence of
firmed by laboratory testing this should alert 6.5% in RB test was reported for the Denizli prov-
health officials and physicians in the area to deal ince.31 A seroprevalence of 3.4% reported by RB
with the disease appropriately. Although the sur- test in 2000 and 2001 among 1850 individuals living
veys presented here were performed in spring in rural areas around Kayseri.32 In Bolu province in
and summer during which seasons most cases of the northwest of Turkey the seroprevalence among
brucellosis can be expected, no brucellosis cases 2204 people was 1.3% in RB test and 0.8% in SAT at
were observed. Also during earlier surveys per- a cut-off value of 1:160.33 A much higher seropre-
formed in Turkey none of the participants had valence of 26.7% in RB test and of 7.7% in Wright
active brucellosis.24,25 A study performed in Iran test was found in a recent study performed in vil-
showed that active surveillance for brucellosis in- lages in the province of Van in the eastern part
creases the detection of patients with active bru- of Turkey.34 In 2001, prevalence rates for brucello-
cellosis when screening slaughterhouse workers.26 sis for Malatya, Denizli, Kayseri, Bolu and Van were
Slaughterhouse workers are at risk of contracting 8.6, 8.8, 23.8, 47.7 and 37.9 per 100,000, respec-
brucellosis because of their occupational activi- tively.8 The seroprevalence rates obtained in these
ties. Possibly active screening is useful for specific studies thus do not correlate well with the preva-
risk groups. lence of brucellosis in these provinces based on
The relatively low seroprevalence in the San- the reported number of cases. This might be
liurfa area with no seropositive individuals found in explained by differences in epidemiological condi-
the village of Kurudere and in the town of Siverek tions or could be due to differences in serological
was somewhat surprising given the high number of methods. For instance the very high seropreva-
cases reported from this province and because of lence reported for Van could be due to active
Seroprevalence of brucellosis 313

transmission at the time of sampling.24,34e36 Also, 7. Dokuzoguz B, Ergonul O, Baykam N, Esener H, Kilic S,
in the absence of laboratory support brucellosis Celikbas A, et al. Characteristics of B. melitensis versus
B. abortus bacteraemias. J Infect 2005;50:41e5.
could be easily misdiagnosed and standardization 8. Turkish Ministry of Health Statistics. <http://www.saglik.
of diagnosis and reporting is essential for the col- gov.tr/extras/istatistikler/ger_bil_2001/t_32.htm>.
lection of accurate data on disease prevalence. 9. Smits HL, Abdoel TH, Solera J, Clavijo E, Diaz R.
In Turkey brucellosis is likely under-reported.33 Immunochromatographic Brucella-specific immunoglobulin
The official reported number for all provinces com- M and G lateral flow assays for rapid serodiagnosis
of human brucellosis. Clin Diagn Lab Immunol 2003;10:
bined has shown a yearly increase from 37 in 1970 1141e6.
to 10,742 in 2000 and 151,510 in 2001.8 10. Bodur H, Erbay A, Akinci E, Colpan A, Cevik MA, Balaban N.
Earlier it was demonstrated that the specificity Neurobrucellosis in an endemic area of brucellosis. Scand
of the Brucella IgM/IgG flow assay is 96% for pa- J Infect Dis 2003;35:94e7.
tients hospitalized at a hospital in east Turkey.24 11. Namiduru M, Gungor K, Dikensoy O, Baydar I, Ekinci E,
Karaoglan I, et al. Epidemiological, clinical and laboratory
The sensitivity of the flow assay was determined features of brucellosis: a prospective evaluation of 120
to be 96% in a study performed in Spain, and 91% adult patients. Int J Clin Pract 2003;57:20e4.
for adult patients and 97% for pediatric patients 12. Kadikoylu G, Tuncer G, Bolaman Z, Sina M. Brucellar orchitis
in the study performed in east Turkey.9,24 Our pres- in Innerwest Anatolia region of Turkey. A report of 12 cases.
ent study confirms the high specificity of the flow Urol Int 2002;69:33e5.
13. Gungur K, Bekir NA, Namiduru M. Ocular complications asso-
assay and supports its use as a simple and rapid ciated with brucellosis in an endemic area. Eur J Ophthal-
field test for the confirmation of brucellosis. mol 2002;12:232e7.
In conclusion, seroprevalence rates show that in 14. Geyik MF, Gur A, Nas K, Cevik R, Sarac J, Dikici B, et al. Mus-
rural and suburban communities of Turkey expo- culoskeletal involvement of brucellosis in different age
sure to Brucella is still very common also in the groups: a study of 195 cases. Swiss Med Wkly 2002;132:
98e105.
relatively well-developed western provinces. How- 15. Metin A, Akdeniz H, Buzgan T, Delice I. Cutaneous findings
ever, no cases with active brucellosis were found encountered in brucellosis and review of the literature.
during the surveillance studies. Vaccination and Int J Dermatol 2001;40:434e8.
other control measures to prevent exposure and 16. Tasova Y, Saltoglu N, Sahin G, Aksu HS. Osteoarthricular
disease are very important. In Turkey many cases involvement of brucellosis in Turkey. Clin Rheumatol
1999;18:214e9.
of brucellosis are diagnosed based on clinical signs 17. Akdeniz H, Irmak H, Anlar O, Demiroz AP. Central nervous
and symptoms and misdiagnosis could be common. system brucellosis: presentation, diagnosis and treatment.
Therefore such patients either should be referred J Infect 1998;36:297e301.
for laboratory testing to one of the major hospitals 18. Ertekin V, Altinkaynak S, Selimoglu A, Alp H. Childhood
that offer laboratory testing or the introduction of brucellosis in eastern Turkey. Trop Doct 2004;34:57e8.
19. Ciftci E, Ince E, Dogru U. Pyrexia of unknown origin in
laboratory testing at local health care centers children: a review of 102 patients from Turkey. Ann Trop
should be considered. The Brucella IgM/IgG flow Paediatr 2003;23:259e63.
assay may well be used for this. The assay is simple 20. Caksen H, Arslan S, Oner AF, Cesur Y, Ceylan A, Atas B, et al.
to perform, provides a rapid result and is both Childhood brucellosis is still a severe problem in the eastern
sensitive and specific. region of Turkey. Trop Doct 2002;32:91e2.
21. Tasbakan MI, Yamazhan T, Gokengin D, Arda B, Sertpolat M,
Ulusoy S, et al. Brucellosis: a retrospective evaluation.
Trop Doct 2003;33:151e3.
References 22. Tasova Y, Saltoglu N, Yilmaz G, Inal S, Aksu HSZ. Brucellosis:
an evaluation of clinical, laboratory and treatment fea-
1. Young EJ. An overview of human brucellosis. Clin Infect Dis tures. Turk J Infect 1998;9:263e5.
1995;21:283e9. 23. Cetin ET, Coral B, Bilgic A, Bilgehan H, Sipahloglu U,
2. Gur A, Geyik MF, Dikici B, Nas K, Cevik R, Sarac J, et al. Gurel M, et al. Determination of brucellosis in humans in
Complications of brucellosis in different age groups: a study Turkey. Turk J Med Sci 1990;14:324e34.
of 283 cases in southeastern Anatolia of Turkey. Yonsei Med 24. Irmak H, Buzgan T, Evirgen O, Akdeniz H, Pekcan
J 2003;44:33e44. Demiroz A, Abdoel TH, et al. Use of the Brucella-IgM and
3. Memish Z, Mah MW, Al Mahmoud S, Al Shaalan M, Khan MY. IgG flow assays in the serodiagnosis of human brucellosis
Brucella bacteraemia: clinical and laboratory observations in an area endemic for brucellosis. Am J Trop Med Hyg
in 160 patients. J Infect 2000;40:59e63. 2004;70:688e94.
4. Bikas C, Jelastopulu E, Leotsinidis M, Kondakis X. Epidemi- 25. Sumer H, Sumer Z, Alim A, Nur N, Ozdemir L. Seropreva-
ology of human brucellosis in a rural area of north-western lence of Brucella in an elderly population in mid-Anatolia,
Peloponnese in Greece. Eur J Epidemiol 2003;18:267e74. Turkey. J Health Popul Nutr 2003;21:158e61.
5. Omer MK, Assefaw T, Skjerve E, Tekleghiorghis T, 26. Karimi A, Alborzi A, Rasooli M, Kadivar MR, Nateghian AR.
Woldehiwet Z. Prevalence of antibodies to Brucella spp. Prevalence of antibody to Brucella species in butchers,
and risk factors related to high-risk occupational groups in slaughterers and others. East Mediterr Health J 2003;9:
Eritrea. Epidemiol Infect 2002;129:85e91. 178e84.
6. Refai M. Incidence and control of brucellosis in the near 27. Roth F, Zinsstag J, Orkhon D, Chimed-Ochir G, Hutton G,
east region. Vet Microbiol 2002;20:81e110. Cosivi O, et al. Human health benefits from livestock
314 S. Kose et al.

vaccination for brucellosis: case study. Bull World Health 33. Karabay O, Serin E, Tamer A, Gokdogan F, Alpteker H,
Organ 2003;81:867e76. Ozcan A, et al. Hepatitis B carriage and Brucella seropreva-
28. Atmaca S, Ozeniki T, Akpolat N, Elci S, Suay A, Arika E. lence in urban and rural areas of Bolu province of Turkey:
Brucellosis seroprevalence in Southeast Turkey (Diyarbakir). a prospective epidemiologic study. Turk J Gastroenterol
Turk J Med Sci 2004;34:251e5. 2004;15:11e3.
29. Colak H, Usluer G, Karaguven B, Kose S, Ozgunes I. Kirsal 34. Ceylan E, Irmak H, Buzgan T, Karahocagil MK, Evirgen O,
alanda seroepidemiyolojik bruselloz arastirmasi. [Seroepi- Sakarya N, et al. Van iline bagli bazi koylerde insan ve
demiological screening for brucellosis in rural areas]. Turk hayvan populasyonunda bruselloz seroprevalansi. Van Tip
J Infect 1991;5:83e6. Dergisi 2003;10:1e5.
30. Durmaz R, Durmaz B, Rafiq M, Sonmez E. Seropositivity of 35. Schelling E, Diguimbaye C, Daoud S, Nicolet J, Boerlin P,
brucellosis in Malatya, Turkey. Tr J Med Sci 1997;27: Tanner M, et al. Brucellosis and Q-fever seroprevalences
125e8. of nomadic pastoralists and their livestock in Chad. Prev
31. Kaleli I, Kocoglu T, Ozen N, Aksit F. Denizli yoresinde brusel- Vet Med 2003;61:279e93.
loz prevalansi. Turk J Infect 1999;13:231e3. 36. Omer MK, Skjerve E, MacMillan AP, Woldehiwet Z. Compar-
32. Cetinkaya F, Nacar M, Koc AN, Gokahmetodkus S, Aydin T. ison of three serological tests in the diagnosis of Brucella
Prevalence of brucellozis in the rural area of Kayseri, infection in unvaccinated cattle in Eritrea. Prev Vet Med
Central Anatolia, Turkey. Tr J Med Sci 2005;35:121e6. 2001;48:215e22.