EPIDEMIOLOGICAL CHARACTERISTICS OF Q FEVER IN THE SREM DISTRICT OF VOJVODINA, FROM 2011 TO 2020

Q fever is a globally widespread zoonosis, which has a large number of animal reservoirs. In humans, Q fever is asymptomatic in 60% of cases but may appear in the form of acute or chronic disease. Th e disease is mostly asymptomatic in animals but may be a cause of abortions or stillbirths. Th e aim of this study was to present and analyze the epidemiological characteristics of Q fever in Srem district, in the Autonomous Province of Vojvodina, Republic of Serbia. We conducted a descriptive study using data from the registries of infectious diseases of the Institutes of Public Health of Vojvo-dina and Institute of Public Health Sremska Mitrovica in the period 2011-2020


INTRODUCTION
Q fever is a globally widespread zoonosis caused by the small obligate intracellular gram-negative bacterium Coxiella burnetii (C. burnetii) (Angelakis and Raoult, 2010). Th e large number of reservoirs, the possibility of transmission by diff erent routes, the resistance of the microorganism and the small infectious dose place C. burnetii in category B on the list of agents suitable for biological weapons (Parker et al., 2006). Among the many reservoirs, for humans the most important are ruminants: goats, sheep and cattle. Infected animals shed pathogens via body fl uids. Q fever is mostly asymptomatic in animals, but it causes abortions and stillbirths (Fournier et al., 1998).
Th e most common route of transmission of C. burnetii to humans is by inhalation of contaminated dust or aerosols from infected feces, urine and birth products (Reimer, 1993; CDC, 2019), thus Q fever is an occupational hazard. Q fever is less commonly transmitted through the consumption of raw milk and unpasteurized dairy products, blood transfusions, from a pregnant woman to her fetus, or through sexual contact (Fournier et al., 1998;Maurin and Raoult, 1999;Milazzo et al., 2001). In humans, Q fever is asymptomatic in 60% of cases and in the rest of symptomatic cases, it may occur in the form of acute or chronic disease (Ristić et al., 2018). Due to asymptomatic course, Q fever is oft en an unrecognized and under-registered disease.
In aim to better understand the reasons for the endemic maintenance of Q fever, we analyzed the epidemiological characteristics of Q fever in Srem district, in the Autonomous Province of Vojvodina (APV), Republic of Serbia, for ten consecutive years (from 2011 to 2020).

Study area and data collection
Th e study was conducted in Srem district, located in its southwestern part of the APV, with a population of 312,278 inhabitants according to the latest population census (SORS, 2011). Th e source of data was the database from the registries of infectious diseases of the Institute of Public Health of Vojvodina and the Institute of Public Health of Sremska Mitrovica responsible for the surveillance on communicable diseases. Th e data of Q fever in humans included individual case reports, epidemiological studies, epidemiological questionnaires, reports on laboratory tests, reports of outbreaks and annual reports for the observed ten-year period (2011-2020). According to the European Center for Disease Prevention and Control (ECDC) case defi nition, we classifi ed Q fever cases as probable and confi rmed cases (ECDC, 2018).
antibody response to C. burnetti (IgM and/or IgG to phase II antigen) was used as a laboratory criteria for the confi rmation of Q fever. Paired serum samples tested at least two weeks apart were taken for patients whose fi rst serum test results were equivocal or negative.

Veterinary data and laboratory confi rmation of animal Q fever
Data on Q fever laboratory confi rmation of animals in Srem district were obtained from the Scientifi c Institute of Veterinary Medicine, Novi Sad. Samples of goats, sheep and cattle were tested under the auspices of mandatory annual monitoring. By order of the Republic Veterinary Inspection, the Veterinary Scientifi c Institute, Novi Sad conducted epizootiological investigation in the households with livestock in the case of Q fever in humans. In the case of reports of abortions in domestic animals, serological analysis was performed and in case of positive fi ndings, again, an epizootiological investigation in the household was conducted. Th e ELISA method was used for the analysis of blood samples (detection of IgM/IgG antibodies against C. burnetti) of goats, cattle and sheep (OIE, 2021). Data about results of molecular analysis and data about seroprevalence of Q fever in specifi c years (2011, 2012 and 2013) are lacking and are not presented.

Data analysis and statistics
Descriptive method was applied. Th e data were analyzed chronologically, demographically, and topographically for the observed period. We used the basic statistical indicators, general and specifi c incidence rates. Statistical data processing included descriptive analysis (mean, minimum and maximum, as well as standard deviation). Th e Mann-Whitney U test was applied to test the diff erences of Q fever incidence between two of each age group (0 -19, 20 -59, ≥ 60 years). For all data analyses the SPSS soft ware, version 21.0 was used and p-value < 0.05 was considered statistically signifi cant. For mapping, we used the program Quantum GIS (QGIS) version 3.4.

RESULTS
A total of 76 cases of Q fever were registered in the period 2011 -2020, in Srem district. Of the total number of cases, 92.11% (N = 70) were laboratory confi rmed and 7.89% (N = 6) were classifi ed as probable cases. Males were 3.2 times more aff ected compared to females (M : F = 3.22 : 1). Th e highest incidence rates of Q fever were reached in 2012 (14.41/100,000; N = 45) and in 2017 (7.04/100,000; N = 22) ( Figure 1). Th ere were no registered cases of Q fever during the four years (2016, 2018, 2019, and 2020) of this observed period. Cases of Q fever were registered in all age categories, except in children aged < 10 years. Th e mean age (± standard deviation) of patients was 39.51 ± 14.17 years. Th e youngest patient was 14, and the oldest was 75 years old. Agespecifi c incidence of Q fever during the observed ten-year period is shown in Figure 2. Th e highest age-specifi c incidence rates were recorded in 2012, in the age groups 20 -29, 30 -39 and 40 -49 (32.17/100,000; 28.76/100,000; 18.38/100,000 respectively) and in 2017, in the age groups 30-39, 40-49 and 50-59 (16.78/100,000; 16.08/100,000; 7.82/100,000; respectively). Th e majority of patients (82.89%; N = 63) belonged to the working-age population (aged 20 -59 years). Age specifi c incidence (mean, minimum and maximum, as well as standard deviation) are shown in Table 1. Based on the Q fever data in Srem district during the ten-year period (2011 -2020), a statistically signifi cantly higher incidence of Q fever was observed in the age category of the working-age population (aged 20 -59 years) compared to the age category from 0 -19 years (U = 291.5; p = 0.022) ( Table 2). Regarding the seasonal distribution, the majority of confi rmed cases were registered in the period from January to May (96.05%; N = 73). Th e highest percentage of cases (60.53%; N = 46) was detected in February, and then equally in the months of April and May (13.16%; N = 10) (Figure 3). Most patients (89.47%; N = 68) were registered in the western part of Srem district, in the municipalities of Šid and Sremska Mitrovica (incidence rates 64.35/100,000 and 57.54/100,000; respectively) ( Figure 4). Th e lowest incidence rate (1.52/100,000) was registered in the municipality of Stara Pazova. Cases of Q fever were registered in six out of seven municipalities of the district with the exception of the municipality Inđija. Most (92.11%; N = 70) Q fever cases were detected in outbreaks and the rest (7.89%; N = 6) occurred as sporadic cases. One third of patients (32.89%; N = 25) were hospitalized due to the severity of the disease or complications. Th ere were no registered deaths due to acute Q fever. Direct contact with animals was probably route of Q fever transmission in Srem district, registered in 50.00% (N = 38) of cases (Table 3). Airborne spread of C. burnetti was probably route of transmission for the other half of patients (47.37%; N = 36). In only two cases the route of transmission of C. burnetti was not established. Overall, three outbreaks of Q fever were notifi ed in Srem district in the observed period, in municipalities of Ruma (2011), Sremska Mitrovica (2012) and Šid (2017) with registered 5, 43 and 22 patients, respectively. In the outbreak in the municipality of Ruma, four patients probably contracted Q fever through direct contact with animals (due to occupational exposure), and one patient through airborne transmission. Epizootiological tests were also carried out in six households in the municipality of Ruma (the blood of the sheep and goats was sampled and tested with an ELISA test), and the results for the presence of specifi c antibodies against the causative agent of Q fever were negative. In the 2017 outbreak, which originated from a goat farm in the vil- During the period from 2014-2020, total of 3,962 animals (821 goats, 3,076 sheep and 65 cattle) were tested for C. burnetii in sera samples in Srem district (Table 4). Of all tested animals, the highest seroprevalence was reported in goats (44.58%). Th e highest seroprevalence was recorded in the municipality of Šid in goats (55.54 %), and also in the municipality of Šid in sheep (11.88%) and in Stara Pazova in cattle (80.00%).

DISCUSSION
Our study shows that Q fever is still endemic disease in Srem district with majority of human cases registered in outbreaks. Th e highest incidence of Q fever in the observed ten-year period was registered in the western part of the Srem district. Men were more aff ected as well as the working-age population. One third of patients were hospitalized, none with death outcome. Looking at the season, the largest number of infected people was recorded in February while probable routes of transmission were direct contact with animals and airborne spread of infection.
Th e results of our study showed that there were more cases among men in Srem district, which may be linked to their higher likelihood of occupational exposure. Due to the protective role of sex hormones, women develop symptomatic Q fever less frequently (Leone et al., 2004) Th e average annual incidence rate of Q fever in Srem district in the observed period was 1.7 times higher than the incidence rate notifi ed in APV and fi ve times higher than the average annual incidence rate recorded in the Serbia in the same period (IPHS, 2021). Th e average annual incidence of Q fever in Srem district (2011 -2020) was approximately two and 1.2 times lower compared to average incidences registered in the endemic areas of Central Banat and South Banat (IPHS, 2018; IPHS, 2021, Ristić et al., 2018). Compared to the average annual incidence rate of Q fever (incidence of 0.17/100,000) in the countries of the European Union (EU), the average annual incidence rate in Srem district was 14.3 times higher in the period 2011 -2020 (ECDC, 2017; ECDC, 2021).
Th e lowest incidence of Q fever in EU, in the observed ten-year period was reported in 2020 (ECDC, 2021), when, interestingly, not a single case of the disease was registered in Srem district, nor in the territory of APV. Th is is explained by the fact that the Coronavirus disease 2019 (COVID-19) emerged in 2020, which is why all communicable diseases, including Q fever, were underreported. Besides, it is possible that pandemic mitigation measures such as wearing masks and washing hands also reduced the chance of contracting Q fever, since it is an airborne infection.
In our study, there were no registered cases among children < 10 years which is similar to the situation in EU, where the small number of patients with positive samples was recorded in children under 10 years of age (ECDC, 2017). In accordance with the above, we found a statistically signifi cantly higher incidence of Q fever in the working age population (20 -59 years) compared to the age category of 0 -19 years old. Th e reason for this fi nding may be that children and adolescents infected with C. burnetti are more oft en asymptomatic than adults and may have milder symptoms of the disease, so cases of Q fever are less oft en registered among them, which is in agreement with other studies (Maltezou and Raoult, 2002;Terheggen and Leggat, 2007).
Seasonal distribution of Q fever, with the majority of patients in Srem district registered in the period January -May, may be related to the fact that during these months the lambing and kidding of sheep and goats are done and other activities in which humans are in contact with potentially infected animals. During the huge outbreak in the Netherlands, it was observed that the majority of patients appeared, at the end of winter and at the beginning of spring (Schneeberger et al., 2014). Several studies have pointed out the role of high wind speed and low rainfall on the easier transmission of microorganisms and the higher incidence of Q fever (Tissot-Dupont  Outbreaks in the village of Noćaj (municipality of Sremska Mitrovica -2012) and in the village of Kukujevci (municipality of Šid -2017) contributed to the peaks of incidence registered in 2012 and 2017 as well as to the highest incidence rates recorded in these municipalities. Th e reason for this may be the fact that these two municipalities are close to the border of Croatia, where the highest incidence of Q fever in Europe was registered in 2012 and 2017 (ECDC, 2019). Not a single case of Q fever was registered in Inđija, even though there was an outbreak in a goat farm with a high number of positive animals. Th e most possible reason may be the under-registration of human Q fever cases, as well as the asymptomatic clinical course of the disease.
An extensive epidemiological investigation of Q fever outbreak in village Noćaj (Sremska Mitrovica) established that the most likely route of transmission for more than 2/3 of the patients was airborne, while the rest reported direct contact with animals (Medić et al., 2012). Our results also showed that half of patients had no contact with animals and that they most likely contracted Q fever which was transmitted through the air. Th is claim is in agreement with previous studies that C. burnetii can be transmitted by wind through aerosols and dust over long distances (within 5 km of the source of infection) (Arricau-Bouvery and Rodolakis, 2005; Debeljak et al., 2018). Furthermore, the other half of the cases of Q fever were registered in people who had direct contact with animals, similar to other studies conducted in APV (Ristić et al., 2018;Popov et al., 2019).
Some limitations of our study should be listed. First, documentation of Q fever is limited by the availability of laboratory testing, and it is certain that a majority of asymptomatic infections were not recorded. Th is suggests that the incidence rates of Q fever are likely to be higher than registered. Earlier studies on Q fever found a positive correlation between the number of infected people and animals in Srem district (Ristić et al., 2018), but more extensive studies are needed to address the low incidence or absence of cases in municipalities where positive animals have been registered. No data for the outcomes of chronic forms of Q fever were obtained during this study.

CONCLUSION
Our results show that Q fever is still an endemic disease in Srem district and represents a signifi cant problem of veterinary and human medicine. More effi cient disease control requires strengthening surveillance with continued monitoring the disease in animals, and more eff ective cooperation between sectors of human and veterinary medicine.

АCKNOWLEDGEMENT
Th e authors gratefully acknowledge the contribution of all physicians and epidemiologists who participated in the monitoring of Q fever in Srem district during 2011 -2020 for their invaluable contribution to this study.

Author's Contribution:
SP and SM made contributions to conception and design of the article; SP wrote the manuscript. TP contributed in results analysis and the way of results presenting. ND performed the statistical analysis. SS, DB, ML and JJK were involved in the data collection. SM and TP revised the manuscript critically. SP prepared the fi nal draft of the manuscript. All authors read and approved the fi nal manuscript.