Prevalence, Risk Factors, and Antimicrobial Resistance Profiles of Thermophilic Campylobacter Species in Humans and Animals in Sub-Saharan Africa: A Systematic Review

Thermophilic Campylobacter species are clinically important aetiologies of gastroenteritis in humans throughout the world. The colonization of different animal reservoirs by Campylobacter poses an important risk for humans through shedding of the pathogen in livestock waste and contamination of water sources, environment, and food. A review of published articles was conducted to obtain information on the prevalence and antimicrobial resistance (AMR) profiles of thermophilic Campylobacter species in humans and animals in sub-Saharan Africa (SSA). Electronic databases, namely, PubMed, Google Scholar, Research4life-HINARI Health, and Researchgate.net, were searched using the following search terms “thermophilic Campylobacter,” “Campylobacter jejuni,” “Campylobacter coli,” “diarrhea/diarrhoea,” “antimicrobial resistance,” “antibiotic resistance,” “humans,” “animals,” “Sub-Saharan Africa,” and “a specific country name.” Initially, a total of 614 articles were identified, and the lists of references were screened in which 22 more articles were identified. After screening, 33 articles on humans and 34 on animals and animal products were included in this review. In humans, Nigeria reported the highest prevalence (62.7%), followed by Malawi (21%) and South Africa (20.3%). For Campylobacter infections in under-five children, Kenya reported 16.4%, followed by Rwanda (15.5%) and Ethiopia (14.5%). The country-level mean prevalence in all ages and under-five children was 18.6% and 9.4%, respectively. The prevalence ranged from 1.7%–62.7% in humans and 1.2%–80% in animals. The most reported species were C. jejuni and C. coli. The AMR to commonly used antimicrobials ranged from 0–100% in both humans and animals. Poultry consumption and drinking surface water were the main risk factors for campylobacteriosis. The present review provides evidence of thermophilic Campylobacter occurrence in humans and animals and high levels of AMR in SSA, emphasizing the need for strengthening both national and regional multisectoral antimicrobial resistance standard surveillance protocols to curb both the campylobacteriosis burden and increase of antimicrobial resistance in the region.


Introduction
Diarrhoea remains the main cause of morbidity and mortality in low-and middle-income countries (LMICs) [1][2][3]. Worldwide, under-five children experience approximately 1.4 billion episodes of diarrhoea each year, with several medical checks, hospitalizations, and around two million deaths. Over 78% of diarrhoea cases are found in the LMICs [4]. e burden of diarrhoeal diseases is complicated by the lack of appropriate case management [5], limited ability to detect the aetiologies [6], and antimicrobial resistance [7]. e most common aetiologies of diarrhoea include bacteria such as Escherichia coli, Vibrio cholerae, Campylobacter jejuni, Salmonella spp., Aeromonas spp., and Yersinia enterocolitica; viruses mainly rotavirus, norovirus, sapovirus, and adenovirus; and protozoa largely Entamoeba histolytica, Giardia spp., and Cryptosporidium spp. [8,9]. Of the bacterial aetiologies, Campylobacter is a leading cause of gastroenteritis in both high-, middle-, and low-income countries, responsible for 400-500 million cases of diarrhoea each year [10]. e clinically important Campylobacter species are C. jejuni and C. coli, which are responsible for about 98% of all human Campylobacter gastroenteritis cases [11,12].
In most cases, campylobacteriosis does not require any antimicrobial therapy except in severe cases, especially in immune-deficient or immune-suppressed individuals [13,14]. e recommended drugs are macrolides (mostly erythromycin), fluoroquinolones (mainly ciprofloxacin), and tetracycline [10,15,16]. Nevertheless, there is an escalating number of Campylobacter isolates resistant to these drugs [17,18] due to the immeasurable and misuse of antimicrobials [19], not only in animals but also in humans [20]. Several factors have been associated with occurrence of Campylobacter infections. ey include consumption of different food items like undercooked poultry meat and pork, red meat at barbecue, grapes, and drinking unpasteurized milk, having a chronic illness [21][22][23], drinking contaminated water, type of water source, animal contact, young age, eating prepared salad, latrine usage, bottle feeding, and nutritional status [24][25][26]. ere is a wide range of natural reservoirs for Campylobacter including chicken and other poultry, wild birds, pigs, dogs, cats, sheep, and cows [27,28]. Consequently, colonization of different reservoirs by Campylobacter poses an important risk for humans through shedding of the pathogen in livestock waste and water sources contamination, environment, and food [29,30].
In LMICs, studies on thermophilic Campylobacter species are few due to limited capacity in laboratory diagnosis [31] and lack of surveillance of enteric diseases [32]. e objective of this review was to gather information on the prevalence, risk factors, and antimicrobial resistance profiles of thermophilic Campylobacter species in humans and animals in SSA. e findings of this review are expected to provide evidence for policy formulation, prevention, and control of Campylobacter infections and increase awareness of the AMR issue.

Methods
e data were collected by searching articles published in English from electronic databases, namely, PubMed, Google Scholar, Research4life-HINARI Health, and Researchgate.net. e search terms were "thermophilic Campylobacter," "Campylobacter jejuni," "Campylobacter coli," "diarrhea/ diarrhoea," "antimicrobial resistance," "antibiotic resistance," "humans," "animals," "Sub-Saharan Africa," and "a specific country name." Initially, a total of 614 articles were identified, and the lists of references were screened in which 22 more articles were identified. After screening, 33 articles on humans and 34 on animals and animal products were included in this review (Figure 1). e reviewed articles were those published from 1997 to 2018. During the review process, the data extracted included title, country, sex and age distribution, sample size, isolation and identification methods, isolation rates, and antimicrobial resistance profiles. Articles for which the sample size was not shown or which used archived Campylobacter cultures were excluded from this review.

Campylobacter Infections in Humans.
Of the 47 SSA countries [33], data on human campylobacteriosis were available from 15 (31.9%) countries. e prevalence of thermophilic Campylobacter in humans was reported in 33 articles (Table 1). Nigeria reported the highest overall prevalence of thermophilic Campylobacter (62.7%); followed by Malawi (21%) and South Africa (20.3%). Kenya reported the highest prevalence (16.4%) of Campylobacter infections in under-five children; followed by Rwanda (15.5%) and Ethiopia (14.5%). e mean prevalence in all ages and under-five children was 18.6% and 9.4%, respectively. Burkina Faso and Mozambique had the lowest prevalence of campylobacteriosis for all ages (2.3%) and under-five (1.7%), respectively. Of the 33 articles reviewed, 16 (48.5%) presented data on distribution of Campylobacter infections by sex but the difference was not statistically significant. Of these 16 articles, campylobacteriosis was more prevalent among males (22.7%; n � 3966) than females (17.7%; n � 3705). Culture methods on selective media, biochemical tests, molecular, and biotyping techniques were used for identification of Campylobacter (Table 1). Of the 33 articles, 27 studies were carried out at clinical settings (hospitals and health centres) while 6 were community-based studies. Probability sampling methods were adopted in 5 articles while the remaining used convenience sampling. Although C. jejuni and C. coli were isolated in the mentioned articles, 15 articles reported other enteric pathogens as probable aetiologies of diarrhoea. Furthermore, more than 85% of the articles considered diarrhoeic cases while the remaining included even asymptomatic participants.
Of the 33 articles, only four reported on risk factors of campylobacteriosis in humans. In Tanzania, Campylobacter infections were associated with sex, young age, poultry meat consumption, and eating of salads [26,38]. In Ethiopia, human campylobacteriosis was significantly associated with nonuse of latrines, water source, drinking unboiled water, bottle feeding, nutritional status, and exposure to domestic animals including cats, dogs, poultry, and pigeons [25]. In Burkina Faso, Campylobacter infections were most common among under-fives and those aged 21-40 years with more pet contacts [57].

Campylobacter spp. in Animals and Contamination of
Animal Products. Of the 34 articles from which data on animals were extracted, 17 collected faeces from live animals, while 16 collected samples from meat or caeca at abattoirs. In 2 articles, samples were collected from both markets and 2 International Journal of Microbiology Total articles retrieved (n = 614) Additional articles from searching of reference lists (n = 22) Relevant articles whose abstracts were further scrutinized (n = 121) Articles considered in this review (n = 67) Articles excluded with reasons (n = 54) Outside the study period (n = 25) Using archived specimens (n = 5) Not showing prevalence (n = 13) Not showing data for thermophilic Campylobacter species (n = 11) Excluded (n = 515) due to Articles on humans (n = 33) Articles on animals and animal products (n = 34) Google Scholar (512) (i)  International Journal of Microbiology abattoirs. Probability sampling methods were used in 6 articles while the remaining used convenience sampling. Data on Campylobacter in cattle were obtained from ten articles published from studies conducted in six countries. e overall mean prevalence was 17.6% and C. jejuni had higher prevalence (70%) than C. coli (23.5%). e highest [64] and the lowest overall prevalence [52] were reported from Tanzania. Furthermore, Tanzania and Ghana showed higher prevalence for C. jejuni and C. coli, respectively ( Table 2).
Data on Campylobacter in goats were reported in three articles from three different countries. e overall mean prevalence was 31.2%, and C. jejuni presented with a higher prevalence (56.2%) than C. coli (38.5%). e highest and lowest prevalence were reported from the Democratic Republic of Congo (DRC) [83] and Ghana [83], respectively. Ethiopia [70] and DRC [83] had the highest frequencies for C. jejuni and C. coli, respectively ( Table 2). For sheep, data were reported in four articles from three countries. e overall mean prevalence was 31.8%, with C. jejuni being reported at a higher frequency (56.7%) than C. coli (35.4%). e highest and lowest prevalence were reported from Ethiopia [70] and Ghana [70], respectively. Ethiopia [70] and Tanzania [75] had the highest prevalence for C. jejuni and C. coli, respectively (Table 2).
Data on presence of thermophilic Campylobacter in pigs were available from six articles from five countries. e overall mean prevalence was 45.5% and contrary to other animals, C. coli occurred at a higher prevalence (70.1%) than C. jejuni (27.2%). e highest and lowest prevalence were reported from Nigeria [60] and South Africa [78], respectively. Ethiopia [69] had both higher and lower values for C. jejuni and C. coli (Table 2).
Data on thermophilic Campylobacter in chickens were obtained from 11 articles from five different countries. In this review, the number of articles on chickens was the highest compared to other reservoirs. e overall mean prevalence was 62.6% which was the highest in all animal reservoirs documented in this review. Campylobacter jejuni was reported in higher prevalence (81.0%) than C. coli (18.1%). e highest and lowest prevalence rates were reported in Ethiopia [70] and South Africa [18], respectively ( Table 2).
As regards to animal products, data on cattle meat were reported in three articles from three countries. e overall mean prevalence was 5.5%, and C. jejuni had higher prevalence (95.2%) than C. coli (4.8%). e highest and lowest prevalence rates were reported in Ethiopia [72] and Kenya [73], respectively. For cattle carcasses, data were reported by two articles from two countries with a mean prevalence of 15.9%. Ghana [68] reported a higher prevalence of C. jejuni while Tanzania [74] observed a higher prevalence of C. coli (Table 2).
Data on sheep meat were reported by a single article from Ethiopia [72] with the prevalence of 10.5%. In sheep carcasses, the mean prevalence was 23.3% computed using two articles from two countries. Ghana [68] showed a higher prevalence of C. jejuni while Ethiopia [76] reported a higher prevalence of C. coli. In pork, the prevalence was 8.5% in one article from Ethiopia [72] with C. coli being more prevalent than C. jejuni. In pig carcasses, the prevalence was 36.3% from one article reporting a study carried out in Ghana [68]. In chicken meat, the mean prevalence was 49.4% reported by two articles from two countries. Dadi and Asrat in a study conducted in Ethiopia [72] indicated a higher prevalence for C. jejuni while a study in Kenya [73] found a higher prevalence for C. coli. For chicken carcasses, the prevalence was 50% from one article in Burkina Faso [79] and all isolates were C. jejuni. In goat meat, the mean prevalence was 22.5% reported by only one article from Ethiopia [72]. In goat carcasses, the mean prevalence was 16.7% reported by two articles from two countries. A study conducted in Ghana [76] reported a higher prevalence for C. jejuni while that in Ethiopia [76] found a higher prevalence for C. coli (Table 2). e overall prevalence of thermophilic Campylobacter in cats [84] and dogs [84,85] were 18.3% and 20%, respectively. Of the reviewed articles, some presented data on companion, wild, and other animals ( Table 3).

Antimicrobial Resistance Profiles of C. jejuni and C. coli in
Humans and Animals. In humans, the AMR profiles, determined using disk diffusion, were available in 4 articles from four different countries (Figure 2), while the remaining did not specify the species. e antimicrobials considered in this review for the ease of comparison were ampicillin (AMP), erythromycin (ERY), tetracycline (TET), cefalotin (CF), nalidixic acid (NAL), azithromycin (AZM), gentamicin (GEN), ciprofloxacin (CIP), chloramphenicol (CHL), and trimethoprim-sulfamethoxazole (TM-SFX). e percentage of antimicrobial resistant isolates ranged from 2-100% for C. jejuni and 0-100% for C. coli. e AMR data for CIP and ERY, which are drugs of choice for treating Campylobacter infections, showed that Ghana [61] and Tanzania [26] reported higher values for both C. jejuni and C. coli. Resistance of Campylobacter jejuni to GEN was similar for both Tanzania and Ghana while for C. coli, it was higher in Tanzania compared eith that of Ghana [26,61]. Higher frequencies of resistance were also reported for TET and AMP which have been in use for many years. In general, higher levels of AMR were reported in C. jejuni than C. coli.
In animals, the percentage of resistant isolates varied from 0-100%. Resistance to CIP was in the range of 0-80.5% and 0-68.8% for C. jejuni and C. coli, respectively. Resistance to ERY varied from 0-99.5% and 0-100% for C. jejuni and C. coli, respectively. Resistance to GEN was <55.6% for both C. jejuni and C. coli. e highest resistance to most of the drugs was seen in Ghana [68] while the lowest resistance was observed in Tanzania [74,89]. Resistance to nalidixic acid was high for both C. jejuni and C. coli in a study conducted in Tanzania [75]. Data on multidrug resistance were available from three studies in which values ranged from 23.3% to International Journal of Microbiology 63.3% for C. jejuni [18,59,74] and from 0-25% for C. coli [59,70]. ere were variations in resistance levels to commonly used antimicrobials in animal species depending on the species tested.

Discussion
e overall mean prevalence of thermophilic Campylobacter in humans ranged from 9.6-18.5% and is within the ranges reported elsewhere in LMICs [31] and in Poland [90]. However, the prevalence was higher than that reported from Korea [91], and was lower than that reported from the USA [92]. is variation may be attributed to the fact that campylobacteriosis is hyperendemic in LMICs probably due to poor sanitation and close proximity of humans and domestic animals [31]. e risk factors for human infections highlighted in this review partly explain this. ey include consumption of poultry meat, drinking surface water, and animal contact, which is in agreement with other studies with consumption of poultry being the major risk factor [24,93]. e prevalence of thermophilic Campylobacter in animals varied between 1.2% and 80%. e mean prevalence    International Journal of Microbiology recorded in chickens (60.3%) concurs with findings from other LMICs such as ailand [94], Sri Lanka [95], and Vietnam [96]. e mean prevalence of thermophilic Campylobacter in pigs was comparable to what was reported in Spain and Vietnam [30,96] but lower than those reported in Norway and the Netherlands [97,98]. e prevalence of Campylobacter in goats and sheep was slightly higher than the prevalence reported in Germany and Trinidad [99,100] but lower than the prevalence reported in Spain [30]. e prevalence in cattle (17.6%) was lower than those reported in the USA and Iran [101,102] but higher than the prevalence reported in another paper in the USA [103].
Although thermophilic Campylobacter species are frequently isolated from animal faeces, this review showed that they are also present in considerable amounts in a number of animal products. e reported prevalence of Campylobacter in cattle and goat carcasses in sub-Saharan Africa was higher compared to the prevalence in Poland for cattle [104] and in Canada for goat [105]. e contamination of carcasses may result from contact with gut contents during manual skin removal, cleaning, and processing in the slaughter house [106]. e prevalence rates in beef, pork, and mutton were slightly higher compared to those observed in other countries [107][108][109]. e variation could be influenced by the differences in husbandry practices which determine exposure of the animals to the bacteria. Partly, this could also be attributable to slaughter and animal product handling practices which enhance the contamination of the products.
Campylobacter jejuni and C. coli were the most frequently encountered species from both human and animals. Similar observations have been reported by other authors [30,110]. e predominance of C. jejuni in various animals, other than pigs, in sub-Saharan Africa has been previously reported [31,111]. e possible explanation is that most of the studies rely on culture and biochemical tests which may not correctly identify some species. Another reason is the use of selective media containing antibiotics to which some other Campylobacter species are sensitive to. Furthermore, higher incubation temperatures may limit the growth of some thermophilic Campylobacter species like C. lari and C. upsaliensis [112,113].
In pigs, C. coli showed higher prevalence (67.4%) than C. jejuni (27.2%) which is in agreement with reports in Canada and the USA that C. coli is a normal flora of pigs' intestines [114,115]. Furthermore, some studies show that C. jejuni and C. coli may cohabit in pigs but usually C. jejuni is always present in lower frequencies than C. coli [116,117]. e results on AMR in both humans and animals highlight that resistance to mostly used antimicrobials is frequent. e resistance ranged from 0 to 100%, and higher resistance rates were reported in C. jejuni than in C. coli. e antimicrobials to which resistance was high included AMP, TET, ERY, and TET. e findings concur with the reports from other studies in both LMICs and high-income countries showing an increment in the number of Campylobacter strains resistant to most of the antimicrobials used in treating human campylobacteriosis [118][119][120]. e increase in resistance to most antimicrobial agents and emergence of MDR isolates could be associated with extensive use of antimicrobials not only as therapeutic agents for human infections [20] but also for prophylaxis and growth promotion in animal husbandry [68]. However, there are challenges in surveillance, differences in design and predominance of the disk diffusion method and not using globally accepted methods. ese may cause differences within and between countries and certainly limit comparability with data reported in other parts of the world. e resistance to TET was comparable with the findings reported from Poland [121] and the USA [122] and the pooled estimate prevalence worldwide (94.3%) [120]. is resistance   may be due to wide use of tetracycline in both human and veterinary medicine [20]. e proportion of isolates resistant to macrolides (ERY) ranged from 0 to 100% in both humans and animals for C. jejuni while the range was from 0 to 92.3% for C. coli. e frequency of isolates resistant to fluoroquinolone was relatively lower in humans which is comparable to rates described in Western Europe [118,121]. e resistance to both erythromycin and ciprofloxacin is of public health concern as there are currently limited options in the choice of treatment of Campylobacter infections. e proportion of multidrug resistance (MDR) isolates varied between 23.3 and 63.3% (Figure 3) which falls within the range of 37-90% from studies in China, Korea, and France [123][124][125].
ere are no internationally agreed criteria of susceptibility testing and breakpoint assessment for Campylobacter spp. [126]. erefore, it is difficult to interpret the available data and draw conclusion. Several laboratory standards have been applied for the susceptibility testing of Campylobacter species. Although disk diffusion was used in some studies, it should be used only as a screening method for resistance to erythromycin and ciprofloxacin [127].

Conclusion
is review indicates that C. jejuni and C. coli are frequently isolated from humans, food animals, and animal products in sub-Saharan Africa. Isolates from the different sources display varying degrees of resistance to commonly used antimicrobial agents. e findings of this review suggest that the disease burden due to thermophilic Campylobacter species in SSA is of public and economic importance. erefore, routine diagnosis of C. jejuni and C. coli, appropriate use of antimicrobials, educating communities on hygienic practices, establishment of both national and regional multisectoral antimicrobial resistance standard surveillance protocols are necessary to curb both the campylobacteriosis burden, and increase of antimicrobial resistance in the region.