Introduction

Meat is a nutrient-rich food which provides vital amount of proteins, vitamins and minerals with greater bioavailability than other food sources (McAfee et al., 2010). However, it has been recognized as the main vehicle for the transmission of foodborne pathogens to humans (EFSA, 2013). The water activity of fresh meat and its optimum pH play the major role for the growth of microorganisms. As a result, meat is considered as highly perishable foodstuff. Cross contamination of carcasses and meat products occur during subsequent handling, processing, preparation and distribution (Dave & Ghaly, 2011).

The safety of meat may be affected by many biological, chemical and physical hazards; although the biological hazards pose the highest foodborne risk for meat consumers (Norrung & Buncic, 2008). The pathogenic microorganisms possess greater socioeconomic impact due to their potential to contaminate meat and meat based products (Buzby et al., 2001). From the biological hazards, bacterial pathogens are the most serious concern regarding the issues of meat safety to consumers (Sofos, 2008).

Contamination of meat with foodborne pathogens is a major public health issue. Hence, the quantitative synthesis of studies is important to estimate the level of contamination of meat. In this meta-analysis, the population is defined as meat and meat-based products surveyed at abattoirs and retail establishments/markets in Ethiopia. The primary outcome of interest is the prevalence of pathogens, while the antibiotic resistance status of the pathogen is considered as a secondary outcome.

In Ethiopia, studies have been conducted on the prevalence of bacterial pathogens on meats in different parts of the meat chain and settings. These individual studies alone would not, however, show the nationwide burden of bacterial pathogens in meat unless evidence is generated from pooled estimation of the results of primary studies to provide a common national figure. Therefore, this systematic review and meta-analysis was aimed to estimate the overall prevalence of bacterial pathogens and their antimicrobial resistance profile in meat and meat products in Ethiopian abattoirs and retail establishments.

Methods

Study protocol

The identification of records, screening of titles and abstracts as well as evaluation of eligibility of full texts for final inclusion was conducted in accordance with the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) flow diagram (Moher et al., 2009a). PRISMA checklist (Moher et al., 2009b) was also strictly followed while conducting this systematic review and meta-analysis. The study protocol is registered on PROSPERO with reference number ID: CRD42018106361 and Available from: http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42018106361

Data sources and search strategies

The literature search was carried out through visiting electronic databases and indexing services. The PubMed/MEDLINE, Google Scholar, and WorldCat were used as main sources of data. Besides, other supplementary sources including Research Gate, Science Direct and University repositories were searched to retrieve relevant data. Excluding the non-explanatory terms, the search strategies included important key words and indexing terms: Meat (MeSH), “meat products”, meat*, bacteria (MeSH), bacterial* “antimicrobial resistance”, “antibacterial resistance”, “antimicrobial susceptibility”, and “Ethiopia”. The Boolean logical connectors (AND and OR), and truncation were applied for appropriate search and identification of records for the research question.

Inclusion and exclusion criteria

The papers with original article written in English language, possessed approved microbiological methods for pathogen detection and contain sufficient and extractable data were included in the meta-analysis. Having assessed all the information from the recovered publications, online records available from 2008 to June, 2018 were considered as appropriate for eligibility assessment. Furthermore, only studies focusing on meat and meat-based products were included. All review articles and original articles conducted outside Ethiopia, articles with irretrievable full texts (after requesting full texts from the corresponding authors via email and/or Research Gate) and records with unrelated outcomes of interest were excluded during screening and eligibility assessment.

Screening and eligibility of studies

Records identified from various electronic databases, indexing services and directories were exported to ENDNOTE reference software version 8.2 (Thomson Reuters, Stamford, CT, USA) with compatible formats. Duplicate records were identified, documented and removed with ENDNOTE. Some duplicates were addressed manually due to variation in reference styles across sources. Thereafter, two authors (AZ and MS) independently screened the title and abstracts with predefined inclusion criteria. Two authors (AZ and MS) also independently collected full-texts and evaluated the eligibility of them for final inclusion. In each case, the rest authors played a critical role in solving discrepancies arose between two authors to come up to consensus.

Data extraction

With the help of standardized data abstraction format prepared in Microsoft Excel, authors independently extracted important data related to study characteristics (study area, first author, year of publication, study design, slaughtered animals, sample source, sample type, sample size) and outcome of interests (number of positive samples (prevalence) per bacterium and number of resistant isolates (if any) per bacterium in each positive sample).

Quality assessment of studies

The quality of studies was evaluated according to Newcastle-Ottawa scale adapted for cross-sectional studies (Newcastle- Ottawa, 2016) and graded out of 10 points (stars). For ease of assessment, the tool included important indicators categorized into three major sections: (1) the first section assesses the methodological quality of each study and weighs a maximum of five stars (2) the second section considers comparability of the study and takes 2 stars (3) the remaining section assess outcomes with related statistical analysis. This critical appraisal was conducted to assess the internal (systematic error) and external validity of studies and reduce the risk of biases. The mean score of two authors were taken for final decision and studies with score greater than or equal to five were included.

Outcome measurements

The primary outcome measure is the prevalence of clinically relevant bacterial isolates in meat and meat products sampled in abattoir and retail establishments in Ethiopia. The pooled prevalence was calculated per bacterium. The calculation was conducted for both gram positive and gram negative bacterial isolates including Staphylococcus spp., L. monocytogenes, E. coli O157:H7, and Salmonella spp. The secondary outcome measure is the antimicrobial resistance status of the above-mentioned bacteria against selected antimicrobials from different categories (ceftriaxone, gentamicin, ciprofloxacin and ampicillin). Subgroup analyses were also conducted based on the spatial source of meat and slaughtered animal type.

Data processing and analysis

The relevant data were extracted from selected studies using format prepared in Microsoft Excel and exported to STATA 15.0 software for analyses of pooled estimate of outcome measures and subgroup analyses. Subgroup analysis for the primary outcome (prevalence of selected pathogens) was done by sample source (abattoirs, butcher and market), and slaughtered animal types. Considering variation in true effect sizes across population, Der-Simonian-Laird’s random effects model was applied for the analysis at 95% confidence level.

The significance of heterogeneity of the studies was assessed using I2 statistics (based on Cochran's Q test, I2 returns the percent variation across studies. The formula is: I2 = 100% * (Q – df)/Q, Where: Q = Cochran’s Q and df = degrees of freedom. Comprehensive Meta-analysis version-3 software (Biostat, Englewood, New Jersey, USA) was used for publication bias assessment. The presence of publication bias was evaluated by using the Begg’s and Egger’s tests and presented with funnel plots of standard error of Logit event rate (proportion) (Begg & Mazumdar, 1994; Egger et al., 1997). A statistical test with a p-value less than 0.05 (one tailed) was considered significant.

Results

Search results

A total of 189 potentially relevant studies were identified from several sources including PubMed/MEDLINE, Google Scholar and WorldCat. From these, 18 duplicated articles were removed with the help of ENDNOTE and manual tracing. The remaining 171 records were screened using their titles and abstracts and 113 of them were excluded. Full texts of 58 records were then evaluated for eligibility. From these, 31 articles were excluded due to the outcome of interest was found missing, insufficient and/or ambiguous. Finally, a total of 27 articles fulfilled the eligibility criteria and quality assessment and thus included for systematic review and meta-analysis (Fig. 1).

Fig. 1
figure 1

PRISMA flow chart describing the selection process

Full size image

Study characteristics

Table 1 summarizes the characteristics of 27 eligible studies with 7828 samples which were considered for determining the prevalence of bacterial pathogens and their antimicrobial resistance status. The studies were published in the year between 2008 and 2018. All the selected studies were cross-sectional study design in nature. The majority of meat samples were investigated from beef only (Abdissa et al., 2017; Alemu & Zewde, 2012; Atnafie et al., 2017; Bedasa et al., 2018; Beyi et al., 2017; Dagnachew, 2017; Garedew et al., 2015a; Garedew et al., 2015b; Gebretsadik et al., 2011; Abunna et al., 2016; Kore et al., 2017; Mengistu et al., 2017; Muluneh & Kibret, 2015; Wabeto et al., 2017; Adugna et al., 2018). The rest animal species were goat (Dulo, 2014; Dulo et al., 2015; Ferede, 2014), sheep (Mulu & Pal, 2016), and others (Ejo et al., 2016; Kebede et al., 2014; Senait & Moorty, 2016; Azage & Kibret, 2017). Samples of meat from two or more animals were also taken in four studies (Bekele et al., 2014; Hiko et al., 2008; Kebede et al., 2016; Zewdu & Cornelius, 2009). The foodborne pathogens such as Staphylococcus spp. and L. monocytogenes were the outcomes/pathogens with the fewest observations retrieved: Staphylococcus spp. (with only four published studies) and L. monocytogenes (with only three published studies) as their presence in meats have not been widely surveyed. The average quality score of included studies ranges from 6.5 to 9 as per the Newcastle-Ottawa scale adapted for cross sectional studies (Table 1).

Table 1 Characteristics of the studies describing the prevalence of selected bacterial pathogens in meat and meat products in Ethiopia
Full size table

The antimicrobial resistance profile of common bacterial isolates against four major antimicrobial agents (ampicillin, gentamicin, ciprofloxacin and ceftriaxone) is summarized in Table 2. Out of 722 positive samples, 475 of them were tested for susceptibility. Regardless of the nature of bacterial pathogens, 73, 25, 17 and 15 bacteria were found resistant to ampicillin, ceftriaxone, gentamicin and ciprofloxacin, respectively.

Table 2 The antimicrobial resistance profile of bacterial isolates obtained from meat and its products in Ethiopia
Full size table

Study outcomes

Primary outcomes: Prevalence of bacterial isolates

The study showed that different bacterial pathogens have been detected in meat and meat products in Ethiopia at different level of occurrence (Table 1). The forest plot indicated that the pooled prevalence of Salmonella in meat and meat products was found to be 9% (95% CI: 6.0, 12.0) (Fig. 2).

Fig. 2
figure 2

Forest plot depicting the prevalence of Salmonella spp. obtained from meat and meat products in Ethiopia

Full size image

The highest prevalence was observed in goat meat 18% (95% CI: 13.0, 22.0) followed by chicken meat, 14% (95% CI: 10.0, 19.0), whereas the least prevalence was observed in fish meat 2% (95% CI: 0.00, 5.00) (Table 3). The prevalence of Salmonella in butcher, market and abattoirs was 36% (95% CI: 26.0, 44.0), 11% (95% CI: 6.0, 16.0) and 6% (95% CI: 3.0, 9.0), respectively (Table 4).

Table 3 Subgroup analysis of bacterial prevalence in meat samples based on the slaughtered animals
Full size table
Table 4 Subgroup analysis of the prevalence of bacterial isolates in meat by sample source
Full size table

The pooled estimate of E. coli O157:H7 was found to be 5% (95% CI: 4.0, 7.0) (Fig. 3) and subgroup analysis indicated that the highest prevalence was recorded in beef and sheep meat with value of 6% in each (Table 3). The prevalence of E. coli O157:H7 in meats collected from market, abattoir and butcher was 8% (95% CI: 4.0, 12.0), 5% (95% CI: 3.0, 7.0) and 6% (95% CI: 2.0, 9.0), respectively (Table 4).

Fig. 3
figure 3

Forest plot of E. coli O157:H7 prevalence in meat and meat products

Full size image

The pooled estimate of Staphylococcus spp. isolated from meat samples was 21% (95% CI: 12, 30) (Fig. 4). Comparable pooled estimates were observed across spatial sources of meat (21%, 20% and 22% from abattoir, butcher and market, respectively) (Table 4). The overall prevalence of L. monocytogenes in meat samples was 4% (95% CI: 2.0, 6.0) (Fig. 5). Beef and sheep meat were the only sources of this bacterium with 4.1% prevalence in each (Table 3). The highest prevalence (6%; 95% CI: 3.0, 7.0) of L. monocytogenes was reported from meat samples collected from butcher (Table 4).

Fig. 4
figure 4

Forest plot depicting the prevalence of Staphylococcus spp. in meat and meat products

Full size image
Fig. 5
figure 5

Forest plot showing the prevalence of L. monocytogenes in meat and meat products

Full size image

Secondary outcomes: Antimicrobial resistance profiles of bacterial isolates

The bacterial isolates showed different antimicrobial resistance profile against selected agents. About 25% (95% CI: 10.0, 40.0) of the Salmonella spp. were found resistant to ampicillin. Besides, 9% (95% CI: 2.0, 15.0) of Salmonella spp. and 2% (95% CI: 0.0, 5.0) of E. coli O157:H7 isolates were found to be resistant to ceftriaxone. The pooled estimate indicated that 10% of E. coli O157:H7 isolates were resistant to ciprofloxacin. Salmonella spp. (6%), L. monocytogenes (5%) and E. coli O157:H7 (2%) were resistant to gentamicin (Table 5).

Table 5 Subgroup analysis of bacterial pathogens resistance profile against selected antibiotics
Full size table

Publication bias

Funnel plots of standard error with Logit event rate (prevalence of bacterial isolates) supplemented by statistical tests confirmed that there is some evidence of publication bias on studies reporting the prevalence of bacterial isolates from meat and meat products in Ethiopia (Begg’s test, p = 0.003; Egger’s test, p = 0.000) (Fig. 6).

Fig. 6
figure 6

Funnel plot depicting publication bias of studies reporting the prevalence of bacterial isolates obtaied from meat and meat products in Ethiopia

Full size image

Discussion

Out of 27 original studies with 7828 meat samples included in this study, the pooled prevalence of Salmonella in meat and meat products was 9%. This result is in concordance with the meta-analysis conducted in Portugal where the prevalence of Salmonella spp. in meats was 6% (95% CI: 4, 9%) (Xavier et al., 2014). The finding is much higher than the report made by United States Department of Agriculture, Food Safety and Inspection Service (USDA-FSIS: United States Department of Agriculture, Food Safety and Inspection Service, 2014) which showed that the Salmonella prevalence in ground beef was 1.9% in United States. This difference might be due to the fact that the presence of poor food handling practice, lack of slaughtering facility and poor animal health management at primary production and substandard transport of animal meat contributing to high prevalence bacterial pathogen in Ethiopia. Furthermore, reduced prevalence of Salmonella spp. might be attributed to effective management strategies of pathogens at different stages of production in developed countries.

In a meta-analysis conducted in Portugal, the prevalence of Salmonella in raw and minced beef were 1.9% (95% CI: 0.5, 7.2%) and 1.5% (95% CI: 0.3, 7.8%), respectively (Xavier et al., 2014). Compared to studies conducted in developed countries, the subgroup analysis indicated that the pooled estimate of Salmonella in beef meat is much higher in Ethiopia, 10.0% (95% CI: 6.0, 12.0). The highest Salmonella prevalence was observed in goat meat 18% (95% CI: 13.0, 22.0). The prevalence of Salmonella on chicken meat (14%) is also higher than the European surveys which indicated that the overall pooled estimate of Salmonella spp. in poultry meat was 7.10% (95% CI: 4.60, 10.8%) (Gonçalves-Tenório et al., 2018). Generally, this finding supports the conclusion made by Islam et al. (Islam et al., 2014) who identified slaughtered animal species as one of the sources of variation when estimating the prevalence of bacterial pathogens.

The least prevalence of Salmonella was observed in fish meat, 2.0% (95% CI: 0.0, 5.0). In line with this result, in United States, the prevalence of Salmonella in domestic fish and its products as well as imported fish and its products was 1.3% and 7.2%, respectively (Olgunoğlu, 2012). Animal waste can be introduced directly through bird droppings in ponds or indirectly through runoff. Fish and fish products may carry Salmonella spp., particularly if they are caught in areas contaminated with fecal pollution. Moreover, unsafe handling and packaging may contribute to its contamination.

Our study indicated that the pooled prevalence of E. coli O157:H7 isolated from meat and meat products was 5% which is much higher than a study conducted by Hill et al. (Hill et al., 2011) who reported that E. coli O157:H7 was detected on 0.25% of ground beef and 0.82% of trimmed beef meats in USA. Similarly, very low (1.7%) E. coli O157:H7 prevalence was detected on manufactured beef collected at the processing facility in Australia (Kiermeier et al., 2011).

The highest prevalence was recorded in beef and sheep meats with estimates of 6% in each, whereas the lowest prevalence (3%) was recorded in goat meat. Similarly, Jacob et al. (Jacob et al., 2013) reported that the prevalence of E. coli O157:H7 on goat carcasses was 2.7% (95% CI: 0.8, 4.5%) in United States. In this regard, ruminants, particularly cattle, are considered as the primary reservoirs for E. coli O157:H7, where the organism typically colonizes the lower gastrointestinal tract (Low et al., 2005). In Ethiopia beef is most commonly consumed foods, however, the risk of acquiring E. coli O157:H7 from beef meat appears higher than the risk from meats of other animal species. Many outbreaks of E. coli O157:H7 are usually associated with foods from cattle or their fecal contamination (CDC, 1991).

The pooled estimate of Staphylococcus spp. was found to be 21% in meat and meat products which is in trajectory with the prevalence of S. aureus in Portuguese meat product samples, 22.6% (95% CI: 15.4, 31.8%) (Xavier et al., 2014). The high occurrence of Staphylococcus spp. in meat and meat products is an indicator of hygiene deficiency during processing of meat (Rajkovic, 2012).

In this study, the overall prevalence of L. monocytogenes isolated from beef and mutton meat was 4%. Comparable estimate was reported in Ireland where the prevalence of L. monocytogenes in meat products was 4.2% (Leong et al., 2014). However, much higher prevalence of L. monocytogenes (18.7%) was reported in raw meat and raw meat products in Estonia (Kramarenko et al., 2013). The live animals may contribute little to the total contamination of the abattoir. Nevertheless, the L. monocytogenes may be introduced from potential environment and dirty transport crates into the meat production chain at different level. The contamination of carcass by L. monocytogenes is likely to occur due to poor handling by retailers and abattoir workers.

Most of bacterial pathogens were more prevalent in meats samples collected from retails as compared to meat samples collected from abattoirs. Correspondingly, the bacterial pathogen prevalence was globally lower in carcasses at the slaughter house level and higher in meat cuts and minced beef at retail (ECDC, 2013; Stevens et al., 2006). The temperature fluctuation during distribution, meat contamination by handlers, lack of hygiene and unsafe loading and unloading practices might have contributed for slight increment of meat contamination in retail outlets (Rajkovic, 2012). The high cost of cold storage equipment can also be key factors impeding the transportation of meat under refrigeration conditions in developing countries. Likewise, Gill et al. (Gill & McGinnis, 2000) reported that raw beef sold at retail outlets is subjected to a long chain of slaughtering and transportation where each step poses a potential risk of microbial contamination. Whereas, in abattoirs, a variety of decontamination measures might be employed during carcass processing in order to reduce the microbial load and contamination of carcass with pathogens.

Meta-analysis was conducted for antimicrobial susceptibility profile of bacterial isolates from subset of studies to which the secondary outcome measures were considered. The antimicrobial resistance profile of bacterial isolates from meat and meat products was found less than 10% in majority of estimates. However, slightly higher resistance profile (25% of Salmonella isolates) was recorded against ampicillin. To this end, higher prevalence (38%) of antimicrobial resistance Salmonella isolates against ampicillin was reported from chicken meat and their processing environment in Brazil (Medeiros et al., 2011). In the present study, 10% of E. coli O157:H7 isolates were resistant against ciprofloxacin. Despite a temporal variation, a study conducted in China in 2010 noted 4.1% antimicrobial resistance E. coli isolates against ciprofloxacin (Jiang et al., 2012). Antimicrobial resistance profile of meat borne pathogens might vary spatially and temporally due to sample type, environmental contamination and exposure, farm management system and antimicrobial use.

Implication and limitation of the study

According to the evidence generated from the meta-analysis, the contamination of meat and meat products requires stringent management on the area of food safety in meat sector in Ethiopia. The national food, medicine and health care administration and control authority and policy makers could make use of the estimates as inputs to enforce food safety measures. In this study, sufficient data was not found to assess the seasonal effect on the prevalence of bacterial pathogens in meat. Likewise, the risk factors of meat and meat products contamination along the production chain were not addressed. In most of the studies, there was lack of enumeration or bacterial load determination which indicates the actual safety status of food/ meat. Besides, very few reports were available for some pathogens from meat and meat products.

Most of the retrieved studies were carried out in slaughterhouses and markets in urban area of the country where most abattoirs are located therefore, the pooled prevalence estimates of contaminated meat items should not be generalized for rural and smaller settings of the country. All these limitations are clear gaps for further research in the area of meat safety in Ethiopia.

Conclusion

Relatively high prevalence of bacterial pathogens observed in meat and its products in Ethiopia, as highlighted in this review, may possibly be considered as potential sources of human foodborne illnesses. The results justify the need for strict measures to reduce contamination of carcasses in meat throughout the entire supply chain. The antibiotic resistance profiles of bacterial isolates in meat and its product was found lower. Relatively, Salmonella spp. showed high resistance against ampicillin.