Food handling practice and associated factors among food handlers in public food establishments of Ethiopia: a systematic review and meta-analysis

Abstract

Objectives

Food handlers are individuals who are involved in food preparation, storage or service of food in a food facility. Unless they keep their personal hygiene, they are highly responsible for contaminating food or transmitting microbes to consumers. Thus, the main aim of this review was to pool the level of good food handling practice among food handlers working in public food establishments in Ethiopia.

Design

A systematic review and meta-analysis using the Joanna Briggs Institute (JBI) Critical Appraisal tool.

Data sources

PubMed, Google Scholar and Advanced Google were searched up to the end of February 2021.

Eligibility criteria for selecting studies

Publication status, publication language, type of study participants and the type of article were used to screen the article.

Data extraction and study quality assessment

All reviewers collected data independently and merged them together. A tool called the JBI Meta-Analysis of Statistics Assessment and Review Instrument was used to assess the quality of each research article. The assessment was performed by two reviewers.

Result

A total of 16 research articles were included in the review. The pooled level of good food handling practice was 50.72% with 95% CI (43.84% to 57.6%). Training (adjusted OR (AOR)=3.4, 95% CI (2.33 to 4.95)), attitude (AOR=3.62, 95% CI (2.23 to 5.85)) and medical check-up (AOR=5.12, 95% CI (2.93 to 8.95)) were identified as factors affecting good food handling practice with 95% CI at p<0.05.

Conclusion

The level of good food handling practice among food handlers working in public food establishments of Ethiopia was very low compared with literature, and variables such as food handler training, attitude towards good food handling practice and the existence of regular medical check-ups were factors affecting good food handling practice. Provision of training that could change the attitude and regular medical check-ups for food handlers has to be in place.

Strengths and limitations of this study.

• The investigators of this review were doing their best to include all available shreds of evidence regarding the issue under review but still, there might be works of literature that were not published and hung on by authors.

• Though the investigators also did their best to include articles of all study designs, still the available literature was obtained only with cross-sectional design and this might still have its influence on the quality of the review.

• A little bit different operational definitions were used to define good food handling practice by authors of individual articles, therefore, this might have its implication on the generalisability of the findings.

Introduction

Ethiopian Public Health Proclamation No. 200/2000 defined food as any substance, whether processed, semiprocessed or raw, which is intended for human consumption and includes drinks, chewing gum, and any substance which has been used in the manufacture, preparation, or treatment of food, but it does not include tobacco, cosmetics or substances used only as drugs.¹

Experts in different national and international organisations believed that food has to be promoted as part of essential components of primary healthcare.² The food that we eat should be safe for consumption and should not endanger the health of the consumer via contamination or intoxication. Moreover, it should be available in sufficient quantity with adequate nutritional content.² The food is wholesome and safe for consumption implies that the food we are going to eat has to be free of microbiological or chemical contamination that could bring ill-health.^{3 4}

Many stakeholders can be involved in bringing food to the dining table including producers, harvesters, shippers, processors, distributors, handlers and others.⁵ Usually, food handlers are individuals who are involved in food preparation, storage or service of food in a food facility. Unless they keep their personal hygiene, they are highly responsible for contaminating food there by exposing it to foodborne diseases. Foodborne disease (also referred to as foodborne illness or food poisoning) is any illness that results from the consumption of contaminated food, contaminated with pathogenic bacteria, viruses or parasites; therefore, maintaining good food handling practice by food handlers is very essential.⁶ A food handler is defined as anyone who handles packaged or unpackaged food directly as well as the equipment and utensils used to prepare or serve food and/or surfaces that come into contact with food.⁷ According to Ethiopian public health law proclamation, every food handler working in a food establishment should have a medical check-up at least once in every 3 months.¹

Research findings abroad indicated the level of food handling practice among food handlers was from very low (59.3%) to high (90%) (Indonesia, 90%⁸; Saudi Arabia, 80.29%⁹; Malaysia, 59.3%¹⁰; Jordan, 89.43%¹¹; Nigeria, 78.2%¹²). Similarly, good food handling practice among food handlers in Ethiopia also ranged from a very low 27.4% in Addis Ababa to 72% in Dessie.^13–28

Researchers, policymakers and other stakeholders need research evidence for making decisions. However, the shreds of evidence available are inconsistent, ranging from 27.4%¹³ in Addis Ababa to 72% in Dessie,²⁸ which would challenge the users of the evidence to make the best choice from the available evidence. Moreover, in the research articles included in the review,^13–28 the highest magnitude of good food handling practice was obtained from the small sample size,²⁸ while the smallest effect size was obtained from a relatively large sample size.¹³ Hence, a systematic review and meta-analysis is needed to overcome the limitations of small sample sizes and evaluate effects in different subsets of participants. Therefore, the main aim of this review is to pool the level of good food handling practice among food handlers working in public food establishments in Ethiopia.

Methods

Protocol and registration

Usually, researchers are advised to maintain prior registration of their systematic review and meta-analysis in organisations like PROSPERO (http://www.crd.york.ac.uk/PROSPERO/), which could make the process transparent and hence reduces duplication of efforts.²⁹ Therefore, this review has been registered at PROSPERO with registration identification number CRD42020223348.

Literature search

In this review, the step-by-step guide for conducting a systematic review and meta-analysis³⁰ was used. A preliminary search was done to see the existence of a similar article thereby reducing duplication of efforts. Individual articles were searched up to the end of February 2021 from databases such as PubMed, Google Scholar and Advanced Google Search using keywords and medical subject headings (‘‘Food’’, ‘‘Handling’’, ‘‘Practice’’, ‘‘Handlers’’ and ‘‘Ethiopia’’). After having individual research articles, the title and abstract were screened. The protocol has been written and approved by each member and registered to PROSPERO.

Study selection

Selection of studies was done through removing duplicate studies, selecting researches that meet the inclusion/exclusion criteria based on the abstracts, and then making the final selection of studies based on their full text. The eligible articles for the review were selected independently by each reviewer and brought together for ensuring the consistency of our search. Differences between two investigators (AD and DA) regarding a single research article have happened and the third investigator (MTM) brought the issue for discussion and finally, all of the reviewers agreed to include the article after a big debate moderated by the third investigator.

Eligibility criteria

Publication status, publication language, type of study participants and the type of article were used to screen the article. Moreover, title and full-text downloading were also used to screen the eligible article as well. Therefore, this review included both published and unpublished original articles written in English, which were conducted from 2010 to the end of February 2021 among food handlers working in public food establishments in Ethiopia. However, this review excluded articles (n=329 out of 345) losing originality (an article is considered original research if it is the report of a study written by the researchers who actually did the study), local or governmental reports, conference abstracts and articles that were very difficult to access the full text.

Definition of terms

All of the articles included in this review thought that good food handling practice is the practice of maintaining a high degree of personal cleanliness, wearing clean outer garments and effective hair restraints, and refraining from smoking, eating, and drinking in the food preparation and service areas. In addition, a favourable attitude is the positive feeling that food handlers want to practice during food preparation, storage and transportation.^13–28

Patient and public involvement

No patient involved.

Data extraction and study quality assessment

After collecting eligible individual research articles using criteria described under the heading ‘eligibility criteria’, variables (study region, study setting, study year, publication status, publication year, sample size, study design and sampling technique) more frequently used by individual research articles were collected as data. The collection of these data was performed independently by all reviewers and finally merged after reconciling disagreements created during data extraction.

A tool called Joanna Briggs Institute (JBI) Meta-Analysis of Statistics Assessment and Review Instrument adapted for both cross-sectional/case–control study designs³¹ was used to assess the quality of each research article (table 1). The assessment was performed by two reviewers, namely AD and DA. The two assessors (AD and DA) agreed and ranked accordingly. The JBI Critical Appraisal Checklist for analytical cross-sectional studies was used as criteria to include articles. Answers include the following: 1=not applicable, 2=not clear, 3=no and 4=yes.

Table 1.

Result of the JBI Critical Appraisal Checklist for analytical cross-sectional studies using eligibility criteria

Lead author and study year	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Decision
Abdi et al, 201713	4	4	1	4	4	3	4	4	Included
Tessema et al, 201323	4	4	1	4	4	3	4	4	Included
Chekol et al, 201816	4	4	1	4	4	3	4	4	Included
Yenealem et al, 201926	4	4	1	4	4	3	4	4	Included
Legesse et al, 201515	4	4	1	4	4	3	4	4	Included
Henok et al, 201820	4	4	1	4	4	3	4	4	Included
Lalit et al, 201524	4	4	1	4	4	3	4	4	Included
Reta et al, 201717	4	4	1	4	4	3	4	4	Included
Admasu and Kelbessa, 201828	4	4	1	4	4	3	4	4	Included
Gizaw et al, 201414	4	4	1	4	4	3	4	4	Included
Meleko et al, 201420	4	4	1	4	4	3	4	4	Included
Adane et al, 201429	4	4	1	4	4	3	4	4	Included
Nigusse and Kumie, 201125	4	4	1	4	4	3	4	4	Included
Derso et al, 201327	4	4	1	4	4	3	4	4	Included
Lema et al, 201918	4	4	1	4	4	3	4	4	Included
Azanaw et al, 201821	4	4	1	4	4	3	4	4	Included

JBI, Joanna Briggs Institute.

1. Were the criteria for inclusion in the sample clearly defined?

2. Were the study subjects and the setting described in detail?

3. Was the exposure measured in a valid and reliable way?

4. Were objective, standard criteria used for measurement of the condition?

5. Were confounding factors identified?

6. Were strategies to deal with confounding factors stated?

7. Were the outcomes measured in a valid and reliable way?

8. Was appropriate statistical analysis used?

Data synthesis and statistical analysis

Data were collected and entered into Excel and finally exported to Stata V.14 (StataCorp, College Station, Texas, USA) for analysis. Eyeball testing using forest plots, χ² test, and I² was used to identify and measure heterogeneity. Moreover, subgroup analysis and meta-regression were also employed to explore the existence of heterogeneity between research articles. I² reflects the percentage of total variation across studies that were attributable to heterogeneity rather than chance. Heterogeneity was quantified as low, moderate, and high, with upper limits of 25%, 50%, and 75% for I², respectively.³² Moreover, subgroup analysis is usually defined as the process of comparing a treatment effect for two or more variants of an intervention.³³

In this review, publication bias was determined by statistical methods such as drawing funnel plots and statistical testing (Egger’s regression test). Egger’s test with a p value of less than 0.1 indicates the presence of publication bias.³⁴

Individual research articles were collected from all regions in the country which were conducted from 2010 to the end of February 2021 in the country. Investigators of this review believed that the true estimate of food handling practice could vary from region to region due to differences in access to education and medical check-up (Ethiopian public health law proclaims every food handler working in a food establishment should have a medical check-up at least once in every 3 months). Moreover, the true estimate could also vary due to the size of the study participants enrolled in each study. Therefore, the random-effects model which could address this issue³⁵ was used during analysis, and ORs with their 95% CI were used to present the pooled effect sizes. Meta-regression was also done to examine the effect of characteristics of studies against the effect size that is good food handling practice with a 95% CI at p<0.05.

Result

Search results

The investigators of this review tried to search for databases such as PubMed, Advanced Google Search and Google Scholar. The research articles were filtered using key terms described above and articles fulfilling the eligibility criteria were included in the review (figure 1).

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart diagram describing the selection of studies.

Characteristics of studies

A total of 16 articles were included in the review. Out of which 12 (75%) of these articles collected data from food handlers at work within their institutions, while 4 (25%) of the articles collected data from food handlers off their work. All of the research articles reviewed were cross-sectional in study design. All of the research articles were conducted from 2011 to 2019 and published (15 of 16) from 2012 to 2020. Regarding the distribution of the included articles, 10 (62.5%) of them were conducted in the Amhara region followed by Addis Ababa and Tigray which contributed 2 (12.5%) each, while the remaining 2 (12.5%) were contributed by South nations nationalities and peoples region (SNNPR) and Benishangul-Gumuz regions. The mean value (343) of the study participants included in the review was calculated and more than half (62.5%) of the articles included >343 study participants. Study participants were sampled by simple random sampling by 66.67% (10 of 16) of the articles, systematic random sampling by 20% (3 of 16) of the articles, and census by 13.33% (2 of 16) of the articles (table 2).

Table 2.

Characteristics of included studies in the systematic review and meta-analysis

Serial no	First author and study year	Region	Study design	Sampling technique	Sample size	Response rate	Good food handling practice (%)
								1	Abdi et al, 201713	Addis Ababa	CS	SRS	394	95.2	27.4
2	Tessema et al, 201323	Amhara	CS	Census	406	94.4	52.5
3	Chekol et al, 201816	Amhara	CS	SRS	416	98.6	40.1
4	Yenealem et al, 201926	Amhara	CS	SyRS	214	95.5	66.4
5	Legesse et al, 2015 15	SNNPR	CS	SRS	383	99	32.6
6	Henok et al, 201820	Amhara	CS	SRS	423	100	49.6
7	Lalit et al, 201524	Tigray	CS	SRS	369	96.9	53.1
8	Melese et al, 201717	Amhara	CS	SRS	288	100	46.5
9	Admasu and Kelbessa, 201828	Benishangul-Gumuz	CS	SRS	355	100	67.8
10	Gizaw et al, 201414	Amhara	CS	SyRS	403	100	30.3
11	Meleko et al, 201420	Addis Ababa	CS	Census	302	100	47.7
12	Adane et al, 201429	Amhara	CS	SyRS	116	85.9	72
13	Nigusse and Kumie, 201125	Tigray	CS	SRS	277	97.5	63.9
14	Derso et al, 201327	Amhara	CS	SRS	417	98.8	67.6
15	Lema et al, 201918	Amhara	CS	SRS	394	97.8	46.7
16	Azanaw et al, 201821	Amhara	CS	SRS	338	88	49

CS, cross-sectional; SRS, simple random sampling; SyRS, systematic random sampling.

Pooled level of good food handling practice

The fixed-effects model was used to pool the level of good food handling practice but showed a very high level of heterogeneity. Therefore, this review employed a random-effects model for analysis and the pooled level of good food handling practice among food handlers of public food establishments was found to be 50.72% with 95% CI (43.84% to 57.6%) (figure 2). By observing the forest plot and considering I² (I²=96.6%), a high level of heterogeneity between studies was observed (figure 2).

Figure 2.

Forest plot showing the pooled level of good food handling practice (FHP) among food handlers working in public food establishments of Ethiopia, 2021.

Subgroup analysis

The test for subgroup analysis suggests that there is a statistically significant subgroup effect (p<0.0001). The pooled level of good food handling practice is higher among articles conducted before 2016 (52.34% with 95% CI (41.62% to 63.05%)) compared with articles conducted after 2016 (49.12% with 95% CI (39.75% to 58.5%)). However, there is substantial unexplained heterogeneity between the articles within each of these subgroups (conducted before 2016: I²=97.1%; conducted after 2016: I²=96.4%) (figure 3).

Figure 3.

Subgroup analysis of good food handling practice (FHP) by year of study.

Similarly, there is a statistically significant subgroup effect (p<0.0001). The pooled level of good food handling practice is higher among articles having a sample size of ≤343 (57.36% with 95% CI (48.98% to 65.73%)) compared with articles having a sample size of <343 (46.76% with 95% CI (37.65% to 55.86%)). However, there is substantial unexplained heterogeneity between the articles within each of these subgroups (sample size of ≤343: I²=91.4%; sample size of ≤343: I²=97.3%). Therefore, the validity of the good food handling practice estimate for each subgroup is uncertain, as individual article results are inconsistent (figure 4).

Figure 4.

Subgroup analysis of good food handling practice (FHP) by size of study participants.

Publication bias

A funnel plot was drawn and evaluated subjectively by investigators and evidence of publication bias was not observed. Publication bias can be suspected if the plot shows an asymmetric shape, with no points on one side of the graph (figure 5). Moreover, the Egger’s test for small-study effects was also performed but unable to show evidence of the existence of publication bias at p=0.949.

Figure 5.

Funnel plot, in which the vertical line indicates the effect size whereas the diagonal line indicates the precision of individual studies with 95% confidence limit.

Meta-regression

Potential effect modifiers such as study year and size of participants involved in each article were regressed against good food handling practice and none of the effect modifiers showed the existence of association against good food handling practice (table 3).

Table 3.

Meta-regression output for assessing causes of heterogeneity among studies included

Variable	Category	Coefficient	P value>|t|	95% CI
Year of study	After 2016	Reference	Reference	Reference
	Before 2016	3.184339	0.663	−12.14331 to 18.51198
Size of study participants	≤343	Reference	Reference	Reference
	>343	−10.69023	0.143	−25.47183 to 4.091367

Determinants of good food handling practice

At the individual level of analysis, 43.8% (7 of 16) of the articles identified training as a factor that could determine good food handling practice among food handlers working in public food establishments. Consequently, the pooled estimate indicated that the odds of having good food handling practice were 3.4 times higher among trained food handlers compared with non-trained food handlers (adjusted OR (AOR)=3.4, 95% CI (2.33 to 4.95)) (figure 6).

Figure 6.

Forest plot of the adjusted ORs (AORs) with corresponding 95% CIs of studies on the association of training against good food handling practice.

Twenty-five per cent (4 of 16) of the research articles included in the review identified that there was a significant association between food handlers’ attitude and good food handling practice. Similarly, the pooled estimate indicated that those food handlers having favourable attitudes were 3.62 times more likely to have good food handling practice compared with those who had unfavourable attitudes (AOR=3.62, 95% CI (2.23 to 5.85)) (figure 7).

Figure 7.

Forest plot of the adjusted ORs (AORs) with corresponding 95% CIs of studies on the association of attitude against good food handling practice.

Regarding medical check-up, 18.8% (3 of 16) of the included individual research articles indicated that the regular medical check-up of food handlers was associated with good food handling practice. The pooled estimate also indicated that food handlers who had regular medical check-ups were 5.12 times more likely to have good food handling practice compared with those food handlers who did not have regular medical check-ups (AOR=5.12, 95% CI (2.93 to 8.95)) (figure 8).

Figure 8.

Forest plot of the adjusted ORs (AORs) with corresponding 95% CIs of studies on the association of medical check-up against good food handling practice.

Discussion

In individual studies, the level of good food handling practice among food handlers working in public food establishments ranges from 27.4% in Addis Ababa to 72% in Dessie.^13–28 However, the pooled level of good food handling practice among food handlers working in public food establishments was 50.72%. This finding is lower than the findings from Indonesia, 90%⁸; Saudi Arabia, 80.29%⁹; Malaysia, 59.3%¹⁰; Jordan, 89.43%¹¹ and Nigeria, 78.2%.¹² The possible explanation for the finding could be related to the availability of training and medical check-up regularly. In Ethiopia, most establishments are opened without having trained food handlers and even without establishing a mechanism for having a regular medical check-up.

The odds of having good food handling practice were 3.4 times higher among trained food handlers working in public food establishments compared with non-trained food handlers. This finding was supported by findings from different parts of the world suggesting that the provision of training positively influences the food handling practices of food handlers.^36–38 Likewise, the pooled estimate also indicated that food handlers having favourable attitudes were 3.62 times more likely to have good food handling practice compared with those who had unfavourable attitudes. Evidence also suggested that attitude was playing a significant role in promoting good food handling practice.^{11 39 40} Usually, training is given to food handlers seeking for two major changes. One is to add knowledge and skill, and the second is to help the employees develop a positive attitude towards their job. Hence, a trained employee can know more about the job and develop a positive attitude which could help him/her to have good food handling practice.

The pooled estimate also indicated that food handlers who had regular medical check-ups were 5.12 times more likely to have good food handling practice compared with those food handlers who did not have a regular medical check-up. The probable explanation for the finding could be during regular medical check-up, healthcare professionals might counsel regarding safe handling of food, consequences of contaminated food and other related issues which could trigger positive attitude towards good food handling practice.

Conclusion

The level of good food handling practice among food handlers working in public food establishments of Ethiopia was very low and variables such as food handler training, attitude towards good food handling practice and the existence of regular medical check-ups were factors affecting good food handling practice among food handlers working in public food establishments of Ethiopia. Provision of training that could change the attitude towards good food handling practice and regular medical check-ups for food handlers has to be in place to improve good food handling practice among food handlers working in public food establishments of Ethiopia.

Data availability statement

Data are available upon reasonable request. The Excel data are available from the corresponding author upon request.

Data availability statement

The Excel data is available from the corresponding author on request. The Excel data are available from the corresponding author upon request.

Ethics statements

Patient consent for publication

Not required.

Ethics approval

This research was conducted as part of the routine educational programme and was not presented to the review board.