Prevalence of Microorganisms of Public Health Significance in Ready-to-Eat Foods Sold in Developing Countries: Systematic Review and Meta-Analysis

Background The issue of microbial quality and safety of ready-to-eat foods has become a public health concern that needs to be addressed to protect the consumer's health. Contamination of ready-to-eat foods by enteric pathogens such as Escherichia coli, Salmonella, and Staphylococcus aureus bacteria is associated with potential health risks and can cause foodborne outbreaks. Thus, the systematic review and meta-analysis aimed at determining the overall evidence on the prevalence of microorganisms of public health significance in ready-to-eat foods based on previous studies. Methods The articles published from 2015 to 2020 were searched from multiple electronic databases such as PubMed, Google Scholar, MEDLINE, CINAHL, Science Direct, Web of Science, and the Directory of Open Access Journals. The JBI critical appraisal tool was applied to the included articles. To determine the heterogeneity among the included articles, I2 statistics were used while publication bias was evaluated using the visual funnel plot. A Forest plot using the random effect model for meta-analysis was used to estimate the pooled prevalence of E. coli, Salmonella, and S. aureus in ready to eat foods. Results The pooled prevalence of E. coli, Salmonella, and S. aureus in ready to eat foods was 33.8% (95% CI: 19.9, 51.2; Q value = 67.080, I2 = 89.56%), 26.0% (95% CI: 13.8, 43.6%; Q value = 83.67, I2 = 91.63%), and 46.3% (95% CI: 24.8, 69.4%, I2 = 94.9%), respectively. Conclusion The findings show that contamination of ready-to-eat foods with pathogenic microorganisms continues to be a public health risk. Thus, effective food hygiene and safety systems are necessary to protect the health of the consumers and the public as a whole.

. And foodborne illnesses are an important challenge to public health and cause a significant economic problem in many countries [11]. In developed countries, an estimated one-third of the population is affected by foodborne diseases each year. However, foodborne diseases are common and leading causes of illness in developing countries because of the prevailing poor hygienic and sanitation conditions or practices, weak food safety and regulatory systems, lack of resources, and lack of education [12,13]. In response to these problems, health and other concerned organizations are increasing their effort to improve the quality and safety of foods and to prevent foodborne disease [14]. Thus, it is essential to provide the overall evidence on the prevalence of microorganisms of public health concern in ready-to-eat foods.

Objective
The study aimed to systematically review and provide the overall current evidence on the prevalence of microorganisms of public health concerns in ready-to-eat foods in developing countries.

Method
This systematic review and meta-analysis was conducted under the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [15].
3.1. Eligibility Criteria. The articles that met the following inclusion criteria were included in the systematic review and meta-analysis: (i) Study Area. Research articles conducted in developing countries.
(ii) Study Design. Cross-sectional studies reported the prevalence of microorganisms of public health concern in ready-to-eat foods such as Salmonella, S. aureus, and E. coli.
(iii) Language. Full-text articles published in the English language.
(iv) Population. Articles conducted on any type of readyto-eat foods in developing countries.
(v) Publication Issue. Primary research articles published in peer-reviewed journals from 2015 to 2020.

Outcome
Measure. The outcome of this systematic review and meta-analysis is to determine the pooled prevalence of Escherichia coli, Salmonella, and Staphylococcus aureus in ready-to-eat foods in developing countries. The prevalence of selected microorganisms was calculated by dividing the number of positive ready-to-eat foods samples to the total number of ready-to-eat foods samples investigated or analyzed (total sample size) multiplied by 100. , and keywords. The searches were done using the keywords and Boolean logic operators as following: (Prevalence * OR Occurrence * OR frequency * OR Contamination) AND (Microorganisms * OR Bacterial * OR Microbial * OR foodborne pathogens) AND (Public * OR Health * OR Public health) AND (significance * OR concern * OR hazards * OR risk) AND (ready to eat * OR street * OR fast * OR cooked * OR processed * OR prepared * OR packed) AND (foods * OR meal * OR meat * OR fruit * OR fruit products * OR dairy product * OR vegetable * OR vegetable products, etc..) AND (Sub-Saharan * OR Developing * OR low income * OR middle income) AND (Countries * OR * OR region, Africa, etc.). The identified keywords and index terms were checked by authors (Mengistu DA and Tolera ST) across the included electronic databases. Additionally, manual searching for further studies was done by authors (Mengistu DA and Tolera ST) to cover other published articles. The last search was done on April 20, 2020.

Study Selection.
Duplicated studies were removed using the ENDNOTE software version X5 (Thomson Reuters, USA). The authors (Mengistu DA and Tolera ST) individually screened all identified articles based on their titles and abstracts by applying the eligibility criteria. Disagreements were solved by taking the mean score of the two reviewers (Mengistu DA and Tolera ST) after discussing the rationale on differences and repeating the review procedure.
3.5. Data Extraction. All required and relevant data were extracted from the included articles using a predetermined data abstraction form by authors (Mengistu DA and Tolera ST) independently. The extracted data include the name of the authors, sample size, the primary outcome (prevalence of selected microorganisms of public health significance), countries where the article was carried out, year of publication, and study design.
3.6. Quality Assessment. For articles met inclusion criteria, abstracts were checked to establish their relevance for the study. The quality of the included articles was assessed using independent appraisal tools (JBI Critical Appraisal tools) [16]. Then, the score was taken across all the studies and graded as high (75% and above score), moderate (50-75% score), and low (<50% score) quality. High quality in this work indicates a low risk of bias. Disagreements made among authors (Mengistu DA and Tolera ST) on what is to be extracted were solved by discussion. Finally, the authors found all articles with a low risk of bias.
3.7. Statistical Procedure. The required data were extracted using a Microsoft Excel format, 2016. After the extraction, the data were imported to the Comprehensive Meta-Analysis (CMA) V3 statistical package (software). Then, the characteristics of the original articles were described using texts, tables, and forest plots. Heterogeneity among 2 International Journal of Food Science the reported prevalence was checked by using a heterogeneity I 2 test.
The random-effect model of meta-analysis and forest plot was used to estimate the pooled prevalence of microorganisms of public health concern in ready-to-eat foods with 95% confidence intervals. The possibility of publication bias was assessed by visual funnel plots, and a p value of < 0.05 was considered as the evidence for publication bias. Moreover, subgroup analysis was done based on the countries where the included articles were conducted and publication year to minimize the random variations (heterogeneity) between the point estimates of the included articles.

Study Selection.
A total of 1221 articles published from 2012 to 2020 were identified using electronic databases and hand searching. After an initial screening of articles by their titles and abstracts, 149 duplicate articles were excluded, while 1044 studies were excluded based on the predetermined inclusion and exclusion criteria. Then, the full texts of the remaining 28 articles were further assessed to determine their eligibility for the systematic review and metaanalysis. Additionally, 16 articles were excluded as they failed to report the prevalence of selected microorganisms of public health concern. Twelve original articles that meet the predetermined inclusion criteria were included in the systematic review, of which 8 articles were included in quantitative analysis (meta-analysis) ( Figure 1).

Study Characteristics.
In this study, a total of 625 readyto-eat food samples were included in eight articles published from 2015 to 2020 that were conducted in seven different developing countries: one in Ethiopia, two in Nigeria, one in India, one in Pakistan, one in Sudan, one in Namibia, and one in South Africa. All the included articles were cross-sectional studies with a sample size ranging from 15 to 205 ready-to-eat foods. In addition, based on JBI Critical Appraisal tool [16], all the included articles had a low risk of bias (Table 1).

Prevalence of Microorganisms of Public Health
Concerns in Ready-to-Eat Foods. We conducted a metaanalysis using the Comprehensive Meta-Analysis (CMA) V3 statistical package to determine the pooled prevalence of microorganisms of public health concern in ready-to-eat foods.
Based on a subgroup analysis of included articles by country, the lowest prevalence [1.8% (95% CI: 0.7%, 4.9%) with a p value of < 0.001] of E. coli in ready-to-eat foods was observed in South Africa whereas the highest prevalence [88.8%, (95% CI: 62.5-97.40%) with a p value of 0.009] of E.coli in ready-to-eat foods was observed among the studies conducted in Nigeria.
Regarding the publication year, the pooled prevalence of E. coli was higher      (3) Prevalence of Staphylococcus Aureus. The pooled prevalence of Staphylococcus aureus in ready-to-eat foods of developing countries was found to be 46.3% (95% CI: 24.8, 69.4%, I 2 = 94:9%) (Figure 8).
A subgroup analysis was performed based on the countries where the articles were conducted and the year of pub-lication. Among the included countries, the lowest prevalence [7% (95% CI: 3.3%, 14.2%) with a p value of < 0.001] of Staphylococcus aureus in ready-to-eat foods was observed in Namibia whereas the highest prevalence [85%, (95% CI: 29.5-98.7%)] of Staphylococcus aureus in readyto-eat foods was observed in Nigeria. However, after subgroup analysis, the overall prevalence of Staphylococcus aureus was found to be [44.0% (95% CI: 39.3, to 48.8%) with a p value of < 0.014] prevalence ( Figure 9).
Subgroup analysis based on the publication year of included articles found the lower pooled prevalence of Staphylococcus aureus

Discussion
Various studies conducted across the world are strongly agreeing with the fact that most of the pathogenic        International Journal of Food Science microorganisms of public health concerns are introduced to foods during handling, processing, and preparation [25]. These microorganisms have potential health risks to consumers, particularly in developing countries, and need to be addressed to protect the health and wellbeing of the public. Among 8 articles included in this review, Ire and Imuh 2016 reported a higher prevalence of E. coli and S. aureus than other included articles, while Asiegbu et al., 2020, reported lower E. coli. However, the highest and lowest prevalence of Salmonella was reported by Ire and Imuh, 2016, and Asghar et al., 2018, respectively ( Figure 11).
The difference in microorganisms may be related to poor hygiene and safety practices or due to contamination of raw materials used or water supply used or the lifestyle of the community where the study was conducted.
The current review found the overall prevalence of positive ready-to-eat foods in terms of E. coli, Salmonella, and S. aureus that accounts for 172 (27.5%), 168 (26.9%), and 259 (41.4%), respectively. However, quantitative analysis of the included articles indicated the pooled prevalence of Escherichia coli in ready-to-eat foods accounts 33.8%, while the pooled prevalence of Salmonella and Staphylococcus accounts for 26.0% and 46.3%, respectively. Another quantitative analysis (meta-analysis) conducted in selected African countries was agreed with the current finding in terms of pooled prevalence of E. coli (31.6%). However, found lower pooled prevalence of Salmonella (21.7%) and Staphylococcus aureus (25.1%) among ready to eat foods than the current finding [26]. The variation may be due to poor hygiene, safety and sanitation conditions, low quality of raw materials used, lack or inadequate training on food hygiene, and safety for food handlers.

Conclusion
This systematic review and meta-analysis estimated the pooled prevalence of E. coli, Salmonella, and Staphylococcus aureus in ready-to-eat foods in developing countries. The  Figure 9: Subgroup analysis of the pooled prevalence of Staphylococcus aureus in ready-to-eat foods in developing countries by countries, 2020.

Group by year
Study name Statistics for each study Event rate and 95% CI Event rate

Lower limit
Upper limit Z-value p-value