Characteristics of Nontyphoid Salmonella Isolated from Human, Environmental, Animal, and Food Samples in Burkina Faso: A Systematic Review and Meta-Analysis

Salmonella is one of the world’s leading causes of zoonotic and foodborne illnesses. Recently, antimicrobial resistance (AMR) has become one of the most critical challenges to public health and food safety. Herein, we employed a meta-analysis to determine the pooled prevalence and spatiotemporal distribution of serovars and antimicrobial resistance in NTS in Burkina Faso. To find eligible articles, a comprehensive literature search of PubMed, African Journals Online, ScienceDirect, Google Scholar, and the gray literature (university libraries) in Burkina was conducted for the period from 2008 to 2020. Studies meeting the inclusion criteria were selected and assessed for risk of bias. To assess the temporal and spatial relationships between serotypes and resistant strains from humans, animals, food, and the environment, a random-effects statistical model meta-analysis was carried out using the Comprehensive Meta-Analysis Version 3.0 program. The NTS prevalence rates were 4.6% (95% CI: 3–7) and 20.1% (95% CI: 6.6–47.4) in humans and animals, respectively, and 16.8% (95% CI: 10.5–25.8) and 15.6% (95% CI: 8.2–27.5) in food and the environment, respectively. Most NTS serovars were S. Derby, reported both in food and animals, and S. Typhimurium, reported in humans, while S. Croft II, S. Jodpur II, and S. Kentucky were the most prevalent in the environment. NTS isolates were highly resistant to erythromycin, amoxicillin, cefixime, and cephalothin, with a pooled prevalence of multidrug resistance of 29% (95% CI: 14.5–49.5). The results of this review show a high diversity of Salmonella serotypes, as well as high antibiotic resistance in Salmonella isolates from animal, human, food, and environmental samples in Burkina, calling for a consolidated “One Health” approach to better understand the drivers of pathogen emergence, spread, and antimicrobial resistance, as well as the formulation of intervention measures needed to limit the risk associated with the disease.


Introduction
Urbanization and the impact of socioeconomic development have become the main determinants of the health of the world's population and the hygiene of the environment in which they live [1].A lack of hygiene thus provides bioecological conditions favorable for the development of pathogenic germs responsible for numerous diseases, particularly gastroenteritis, the main manifestation of which is diarrhea [2].There are multiple causes of gastroenteritis, including bacteria, viruses, and parasites [3].Among bacterial etiological agents, Salmonella spp. is among the most problematic food-borne and zoonotic pathogens threatening general health and well-being [4].
Salmonella is the main cause of acute gastroenteritis in many countries, and salmonellosis remains a major public health problem worldwide, particularly in developing countries [5].It manifests primarily as mild diarrhea, also known as food poisoning [6].
The global burden of nontyphoidal Salmonella gastroenteritis is estimated to be 93.8 million cases of gastroenteritis each year, with 155,000 deaths [7].Moreover, it remains the second most reported agent of gastrointestinal infections in humans in Europe and North America [8].
In Africa, it has consistently been reported as a leading cause of bacteremia in immunocompromised individuals, infants, and newborns [9].Nontyphoidal Salmonella serotypes (NTS) are among the most frequent causes of bloodstream infections in sub-Saharan Africa, with approximately 680,000 deaths per year, mainly in children under five years of age [10].Although more than 2700 serotypes of Salmonella enterica have been identified [11], the serotypes most commonly implicated in invasive disease are S. Typhimurium, S. Enteritidis, and S. Dublin [12].Nevertheless, in recent years, data have revealed the emergence of infection due to the consumption of raw fruits and vegetables associated with rare serotypes from the environment [13].
Prevention of salmonellosis (caused by nontyphoidal Salmonella) is difficult due to its complex epidemiology and multiple modes of transmission [14].Human disease can result from exposure to numerous sources, such as infected animals, contaminated foodstuffs, contaminated water, and direct contact with an infected environment or direct contact between humans [15].
Poultry and pork meats, eggs, dairy products, and green vegetables contaminated with manure or irrigation water are the agents or risk factors most cited in the transmission of this bacterium [9].
The public health problem associated with nontyphoidal Salmonella is antimicrobial resistance.The emergence of multiple antibiotic-resistant Salmonella strains adds another important dimension to the challenge of controlling Salmonella, as resistant variants can compromise the ability to treat human infections, a particularly important issue in the case of systemic infections [14].Several clones of multidrug-resistant Salmonella emerged in the late 1990s and early 2000s, and since then, their prevalence in humans, domestic animals, and other wildlife has spread globally [16][17][18].Recently, the increasing prevalence of Salmonella multi-resistant to clinically important antimicrobials, such as fluoroquinolones and third-generation cephalosporins, has become an emerging problem worldwide [19][20][21][22].
Despite the knowledge of the prevalence of Salmonella and its antimicrobial resistance profile, which is mostly reported by individual and local surveillance study(s), comprehensive and robust studies of the prevalence and antimicrobial resistance pattern in Burkina are poorly characterized [23][24][25][26].Therefore, concise information from a systematic review is more helpful for scientific users to identify gaps for additional studies and for policymakers to develop prevention and control strategies based on the scientific information provided.Thus, this meta-analysis includes a comprehensive evaluation of the scientific literature published between 2008 to 2022 on the state of knowledge on the prevalence, serovars, and antimicrobial resistance phenotypes of NTS strains from studies carried out in Burkina Faso.

Overview of the Selected Studies
A total of 85 articles were found in our initial research, of which 70 publications were from Google Scholar, PubMed, Science Direct, and Research Gate, and 15 were from other sources.Of the total, 28 publications (33%) considered suitable for inclusion in this review were identified, including 23 eligible studies from a database search (85.71%) and 5 from the literature searches at university libraries (17.86%).A flow diagram of the literature search and selection of eligible studies is presented in Figure 1.

Overview of the Selected Studies
A total of 85 articles were found in our initial research, of which 70 publications were from Google Scholar, PubMed, Science Direct, and Research Gate, and 15 were from other sources.Of the total, 28 publications (33%) considered suitable for inclusion in this review were identified, including 23 eligible studies from a database search (85.71%) and 5 from the literature searches at university libraries (17.86%).A flow diagram of the literature search and selection of eligible studies is presented in Figure 1.

Study Characteristics of Eligible Studies
The following are the characteristics of the studies that are eligible: articles published primarily on the quantitative prevalence of Salmonella serovars/isolates in humans, animals (chickens, ducks, cows, pigs, sheep, etc.), foods (lettuce, sandwiches, beef, mutton, poultry, etc.), and the environment in Burkina Faso; the type of samples and method of diagnostics used; the exact number of samples, as well as the number of positives tested; articles reported in English only; and antibiotic resistance.All of the journal articles were published between 2008 and 2022 (    All of the studies included in this review were carried out in towns located in four regions on the country, Centre, Nord, Hauts-Bassins and Boucle du Mouhoun.Several studies were from Ouagadougou (n = 24), followed by Area 1 (Gourcy and Boromo (n = 3)); Area 2 (Bobo-Dioulasso and Ouagadougou (n = 2)); Area 3 (Ouagadougou, Gourcy and Boromo (n = 1)); Nanoro (n = 1); and Nouna in the province of Kossi (n = 1) (Figure 2).The most common method for determining the AR profiles of Salmonella serotypes isolated from all of the studies included in this systematic and meta-analysis study was disk diffusion.
Of the twenty-eight included studies, twelve (12) were from food sources, eleven (11) were from clinical sources (mainly children under 5 years of age), two (02) were from animal sources, one (01) was from environmental sources, and one (01) included both animal and environmental sources (Figure 3).One study examined food, human, and environmental hosts, and one study included hosts of both food and animal origin (Figure 1).Of the twenty-eight included studies, twelve (12) were from food sources, eleven (11) were from clinical sources (mainly children under 5 years of age), two (02) were from animal sources, one (01) was from environmental sources, and one (01) included both animal and environmental sources (Figure 3).One study examined food, human, and environmental hosts, and one study included hosts of both food and animal origin (Figure 1).

Publication Bias
The Begg and Mazumdar rank correlation test demonstrated no significant publishing bias for all parameters (p value > 0.05 (Table 2 and Figure 5).

Publication Bias
The Begg and Mazumdar rank correlation test demonstrated no significant publishing bias for all parameters (p value > 0.05 (Table 2 and Figure 5).

Discussion
Salmonella is globally one of the leading causes of human death among infectious diarrheal diseases, and antimicrobial-resistant strains of Salmonella are a worldwide health concern [49][50][51].Understanding the epidemiological status of Salmonella is thus crucial for controlling this foodborne pathogen [52].In Burkina Faso, primary research on the prevalence of Salmonella in humans, animals, food, and the environment, as well as its antimicrobial resistance status, have been reported.This meta-analysis focused on comprehensive and robust assessments.
This study examined 28 published articles from Burkina Faso, and the majority of the studies included were published between 2008 and 2022.This study revealed that the highest pooled prevalence of Salmonella was 23% (95% CI: 21.8-24.4) in Ouagadougou, followed by area 2 [18.4% (95% CI: 14.6-22.8)]and area 1 [6.4% (95% CI: 4.9-8.2)].The number of studies included might be a possible reason for the region wise pooled prevalence variation.For instance, only a single study each was found in area 3, Nanoro and Nouna, which may not reflect the true pooled prevalence.
The pooled prevalence in the present systemic review and meta-analysis was 4.6% in humans, 20.1% in animals, 16.8% in food, and 15.6% in the environment.One possible reason for the higher prevalence of Salmonella in animals and food may be associated with unhygienic husbandry practices [53,54].Moreover, meat handling practices in slaughterhouses and butcheries are generally unhygienic.The survival of Salmonella in both soil and water may be a potential cause of its high prevalence in the environment [55].

Discussion
Salmonella is globally one of the leading causes of human death among infectious diarrheal diseases, and antimicrobial-resistant strains of Salmonella are a worldwide health concern [49][50][51].Understanding the epidemiological status of Salmonella is thus crucial for controlling this foodborne pathogen [52].In Burkina Faso, primary research on the prevalence of Salmonella in humans, animals, food, and the environment, as well as its antimicrobial resistance status, have been reported.This meta-analysis focused on comprehensive and robust assessments.
This study examined 28 published articles from Burkina Faso, and the majority of the studies included were published between 2008 and 2022.This study revealed that the highest pooled prevalence of Salmonella was 23% (95% CI: 21.8-24.4) in Ouagadougou, followed by area 2 [18.4% (95% CI: 14.6-22.8)]and area 1 [6.4% (95% CI: 4.9-8.2)].The number of studies included might be a possible reason for the region wise pooled prevalence variation.For instance, only a single study each was found in area 3, Nanoro and Nouna, which may not reflect the true pooled prevalence.
The pooled prevalence in the present systemic review and meta-analysis was 4.6% in humans, 20.1% in animals, 16.8% in food, and 15.6% in the environment.One possible reason for the higher prevalence of Salmonella in animals and food may be associated with unhygienic husbandry practices [53,54].Moreover, meat handling practices in slaughterhouses and butcheries are generally unhygienic.The survival of Salmonella in both soil and water may be a potential cause of its high prevalence in the environment [55].
According to the subgroup analysis, the pooled prevalence of Salmonella in animals was 20.1%, which is lower than that reported in Chikwawa and Malawi in pigs (24.6%) [61], but higher than that observed in Nigeria (8.3%) [62], Australia (4%), and Japan (0.5%) [63,64].The levels of Salmonella contamination in animals can vary depending on the country, the nature of the production system and the specific control measures in place [39].In Burkina Faso, the majority of these animals are still raised under nonmodern conditions.Animals are grazed in an environment with feed of poor microbiological quality.This is more common in poultry farming.These practices undoubtedly lead to Salmonella contamination in animals [65].
The pooled prevalence estimate of Salmonella in food was similar to, but slightly greater than, that reported for Mali (12.80%) in poultry carcasses [66].Other studies have reported prevalence rates of 17.7% in Pakistan and 14.4% in Iraq [67].These results point to food contamination due to noncompliance with good hygiene practices [68].Salmonella, which is usually present in the intestinal tract of animals, can contaminate food during processing procedures and can survive after a cooking process where the cooking time is too short or the temperature is not high enough [69,70].A study conducted in Burkina Faso showed that 31 of the raw ingredients used to prepare sandwiches were contaminated with Salmonella from the water used to wash the lettuce [38].This contamination can be of environmental, animal, or human origin during the cultivation, harvesting, or handling of plants prior to consumption [38,71].
The environmental samples identified in our meta-analysis consisted mainly of water (well water, canals, taps, fountains, and reservoirs).The reported clustered prevalence of Salmonella corroborates that which was reported in Senegal in irrigation water, which was 17% [72].Increasing evidence indicates that irrigation water is a source (or a vehicle) for the transmission of Salmonella [73][74][75][76].This result is worrying, given that in Ouagadougou, all kinds of wastewater (urban or industrial wastewater) are used in an unplanned and uncontrolled way in urban and peri-urban agricultural activities, including the irrigation of vegetables consumed raw (lettuce in particular) [32].Moreover, the lack of mineral matter in the soil and the impossibility for some producers to obtain chemical fertilizers have led them to use animal manure as fertilizer, which is more economical and beneficial for the environment.However, several studies have shown that animal excrement is a major vector for bacterial contamination of garden products [77].
The measurement of trends in serovars over time can provide information about emerging serotypes and about the efficacy of prevention [78].
In this study, the majority of strains were isolated from animals (49.8%, 458/919) and food (38.41%, 353/919).This finding is in agreement with a previous report in South Africa [79].In the present study, NTS from both humans, the environment, food, and animals dominated by 7.69% in S. Kentucky, followed by 7.57% in S. Cubana, 6.67% in S. Poona, and 4.34% in S. Senftenberg.Salmonella has a broad host range, and hence is considered a universal pathogen.Each serovar has a different ability to adapt to the host environment and cause virulency.Some Salmonella serovars are restricted to one host, whereas some have a broad host spectrum [80].For instance, S. derby was isolated from animals and food in 23.92% of the patients.S. Croft II and S. Jodhpur II were observed only in the environmental samples.In this review, we found that S. Typhi was highly prevalent in humans and was also reported in Ethiopia [81], China [82], the Middle East, and North Africa [58].According to reports from sub-Saharan African nations, S. Typhimurium is one of the invasive forms of NTS, especially among susceptible people, such as those with HIV, malnourished children, and malaria [10,83].Furthermore, MDR strains mainly existed in S. Typhimurium.This serovar was also detected in both animal and food samples, even though it was not highly prevalent.Notably, no environmental cases of S. Typhi were found.
The main serotypes found in both animal and food samples in this study included S. Derby, S. Hato, and S. Chester.Poona was the most common serotype detected in both human and environmental samples.Salmonella, regardless of serovar, can persist in dry environments as well as in water for many weeks to months.Moreover, environmental factors may influence the survival of serovars and could contribute to within-and betweenhost species differences [84].Therefore, the ability of a pathogen to spread disease in populations is influenced by host adaptation in many ways [52,85].
Our meta-analysis revealed high NTS resistance against erythromycin (98.3%), amoxicillin (68.69%), cefixime (56.24%), cephalothin (54.65%), and amoxicillin and clavulanic acid (42.06%) for isolates from all source-specific samples.These levels of resistance may be due to the indiscriminate application of antibiotics in human and animal health and food production, followed by the leaching of antibiotics into the environment [86,87].High resistance of Salmonella to erythromycin was observed by Ramatla et al. in a systematic review in South Africa [88].This is probably due to its veterinary use as a feed supplement and/or treatment, particularly in the poultry sector [89,90].Amoxicillin is one of the most commonly used antibiotics worldwide for treating salmonellosis [79].
High resistance to this antibiotic was found in a study carried out in the Middle East, where the rate of resistance to amoxicillin was 100% [91].
However, with the emergence of resistant isolates, traditional antibiotics have been replaced by cephalosporins [92].In the present study, resistance to cefixime and cephalothin was 56.24% and 54.65%, respectively.A meta-analysis carried out in Saudi Arabia reported a high prevalence of 90.9% for cephalothin [93] and 9% for cefixime in Iran [94].This difference could be explained by the level of antibiotic use, which varies from one country to another.
Ten studies reported multidrug resistance to antimicrobials in the present study.Our results (29%) are in line with those of a previous study [88] in which multidrug resistance (MDR) (28.5%) was detected.It has been documented that infections caused by multidrugresistant strains are more severe than those caused by susceptible strains [95].Thus, MDR strains present in food (Typhimurium, Brancaster, Enteritidis, Kentucky, Chester, and Derby) and animals (Typhimurium, Hato, Urbana) could directly threaten human health, as they can cause diseases that are difficult to cure.In addition, integrons, which are mobile genetic elements often associated with multiresistant Salmonella phenotypes, are thought to play a key role in the spread of antimicrobial resistance genes among Gramnegative bacteria [96].The present study shows the urgent need to control the use of antibiotics in veterinary and human medicine to limit the spread of multidrug-resistant Salmonella strains.
This study has the following limitations.The number of studies from some localities was very high compared with that from others, which may have influenced the overall estimate.In addition, few studies on this topic in the environment are available.Only one study was included, which limited comparisons of the prevalence and resistance profiles.Some studies reported the total number of isolates without showing the number of individual serotypes.Resistance genotypes were not included in this meta-analysis due to the small number of studies carried out.

Study Design
The current systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for selection criteria, literature search, statistical analysis, and data extraction [97].

Search Strategy
The literature search was performed on four databases, namely, PubMed, Google Scholar, Science Direct, and African Journals Online, and manually by obtaining hard copies of locally published articles directly from authors and local libraries.
The following keywords were used to search for the articles: antibiotic resistance AND antibiotic AND drug resistance AND bacterial resistance AND multidrug resistance AND Salmonella species AND human OR animal AND environment OR food AND nontyphoid Salmonella AND Burkina Faso.We conducted our last search on November 2022.

Inclusion and Exclusion Criteria
All of the eligible studies included met the following inclusion criteria in the review: articles published primarily on the prevalence of nontyphoidal Salmonella spp. in the environment, foods, animals, and humans in Burkina Faso; the type of samples and method of diagnostics used (approved for the detection and antibiogram of Salmonella spp.); the exact number of samples as well as the number of positives tested; and articles reported in English only on serotypes and antimicrobial resistance published between 2008 and August 2022.Studies were excluded if they were not performed in Burkina Faso.In addition, reviews with smaller sample sizes (≤50) and articles not reported in English were excluded.

Study Selection
Records identified from various sources with the search strategies were exported to Endnote reference manager software version 8. Duplicate records were identified, recorded, and removed.The titles and abstracts of journal articles were examined and downloaded.Two authors independently screened the title and abstracts with the predefined inclusion criteria.The full versions of potentially relevant articles were obtained to evaluate their eligibility for final inclusion.During the screening process, an additional author was on standby to resolve any discrepancies that may arise with study selection.

Data Extraction and Data Collection
Data including names of authors, publication year, location, total sample size, and standard methods to detect the antibiotic resistant, number of positive samples were collected from each publication independently.Then, they were entered into a spreadsheet (Microsoft Excel ® ), tables, and a word document template.Only Salmonella species/isolates/serotypes specific journal articles were included in the meta-analysis.The studies with insufficient data were excluded.In addition, review articles and studies with an abstract, but without a full text, were excluded.The papers that included the drug susceptibility test, and number of multidrug resistance strains were considered and included in this study.

Data Analysis and Assessment of Risk of Bias
For each study, the prevalence, effect size, and 95% CIs were calculated as point estimates.A comprehensive meta-analysis was used for all of the statistical analyses, version 3.0 (CMA) program (https://www.meta-analysis.com/,accessed on 19 November 2022).The software was used to generate the pooled estimates, Cochran's Q, p values, and forest plots.I 2 (level of inconsistency) was used to assess the heterogeneity of the studies (Cochran's Q).I 2 values of 25-50%, 51-75%, and >75% were considered to represent low, medium, and high heterogeneity, respectively.Low I 2 values suggest that variability between estimates is consistent with random variation [13].Interstudy heterogeneity was considered significant if the p value of Cochran's Q test was less than 0.05.The Begg and Mazumdar test was used to investigate the possibility of propagation bias.Funnel plots were used to assess publication bias [98].

Conclusions
This systematic review and meta-analysis report the trends and distributions of Salmonella serotypes and antibiotic resistance in humans, animals, and food in Burkina Faso.It also highlights the lack of published scientific data on antibiotic resistance of Salmonella strains in the environment.
The pooled prevalence of NTS was greatest for animals and food.The predominant NTS serotypes were S. Cubana, S. Typhimurium, S. Derby, S. Hato, S. Chester, and S. Poona.This predominance of emerging isolates is a real public health problem.NTS isolates also were highly resistant to erythromycin, amoxicillin, cefixime, and cephalothin, with a fairly high rate of multidrug resistance.In view of this situation, a consolidated "One Health" approach to the human, animal, food, and environmental sectors is needed to establish a safety net against salmonellosis in Burkina Faso.

Figure 1 .Figure 1 .
Figure 1.PRISMA flowchart illustrating the process of identifying, screening, and selecting the eligible articles used in this study.

Figure 2 .
Figure 2. Map of Burkina Faso showing the study area.Figure 2. Map of Burkina Faso showing the study area.

Figure 2 .
Figure 2. Map of Burkina Faso showing the study area.Figure 2. Map of Burkina Faso showing the study area.

Figure 3 .
Figure 3. Distribution of sample types.

Antibiotics 2024 ,
13, x FOR PEER REVIEW

Figure 4 .
Figure 4. Forest plot showing the pooled estimates of multidrug resistance in studies conduc Burkina Faso.

Figure 4 .
Figure 4. Forest plot showing the pooled estimates of multidrug resistance in studies conducted in Burkina Faso.

Figure 5 .
Figure 5. Funnel plot with 95% confidence limits of the pooled prevalence of the studies conducted in Burkina.

Figure 5 .
Figure 5. Funnel plot with 95% confidence limits of the pooled prevalence of the studies conducted in Burkina.

Author
Contributions: K.A.T., A.R.A.-P., P.R. and N.B. conceived this study; K.A.T. developed the study protocol with the help of A.R.A.-P.and S.C.B.; K.A.T. implemented the review under the supervision of the P.R. and N.B.; K.A.T., A.R.A.-P.and S.C.B. performed the search, screening, and data extraction under the guidance of K.A.T.; S.C.B., J.B.O., A.K., P.R. and N.B. provided content expertise for this review.All authors have provided comments on the final manuscript before publication.K.A.T. is the guarantor of this review.All authors have read and agreed to the published version of the manuscript.

Table 1 .
Descriptive characteristics of the included studies.

Table 2 .
Pooled prevalence of Salmonella spp.from the environment, animals, food, and humans and sampling sites.

Table 3 .
Pooled prevalence and distribution of Salmonella serotypes.

Table 4 .
Multidrug-resistant Salmonella serotypes from different sources *.: The black box indicates multidrug-resistant Salmonella serotypes.* Data reported from the 28 references. Legend

Table 5 .
Percentage of pooled resistance rates of antimicrobials to Salmonella isolates from humans, animals, food, and the environment.