- Email: [email protected]
Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs
Accepted Manuscript Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs Yongheng Yang, Lijuan Wei, Jianan Pei PII:
S0956-7135(18)30466-3
DOI:
10.1016/j.foodcont.2018.09.013
Reference:
JFCO 6313
To appear in:
Food Control
Received Date: 23 May 2018 Revised Date:
11 September 2018
Accepted Date: 12 September 2018
Please cite this article as: Yang Y., Wei L. & Pei J., Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs, Food Control (2018), doi: https:// doi.org/10.1016/j.foodcont.2018.09.013. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 2
Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs
3
Yongheng Yang a *, Lijuan Wei b, Jianan Pei a a
Department of Food Science and Environment, Dalian University of Technology, Liaoning, China b Instrumental Analysis & Research Center, Dalian University of Technology, Liaoning, China
RI PT
4 5 6 7
Abstract:
9
Despite effectiveness of HACCP-based FSM systems in food safety management
10
(FSM) and their common adoption in industrialized food businesses, doubts have
11
been raised on their feasibility in small and/or less developed food businesses (SLDBs)
12
because of barriers such as financial constraints, lack of technical expertise, and
13
insufficient personnel training faced by SLDBs. Uncertainty about their effectiveness,
14
as well as feasibility for SLDBs (including numerous Chinese SLDBs) made them
15
reluctant to adopt these FSM systems. In order to have clear understanding on the
16
effectiveness of HACCP-based FSM systems for SLDBs, this study presented
17
meta-analyses on associations between implementing HACCP-based FSM systems in
18
Chinese SLDBs and levels of their food safety status. Having HACCP-based FSM
19
systems implemented led to significantly higher odds of qualified numbers, with
20
pooled ORs (95% CI) being 10.58 (6.06, 18.49), 8.66 (4.88, 15.35), 6.39 (3.36, 12.18),
21
and 6.65 (4.29, 10.30), respectively, with regard to food products, food handlers’
22
hygiene, employees’ knowledge about food safety and hygiene, and food contacting
23
surfaces & operating environments. Solid evidence on effectiveness of HACCP-based
24
FSM systems in SLDBs should encourage food businesses to adopt these systems for
25
ensuring food safety.
26
Keywords: Food safety; HACCP; Effectiveness
27 28
* Corresponding author: [email protected]; 2 Dagong Rd, Panjin, Liaoning Province, China.
AC C
EP
TE D
M AN U
SC
8
1
ACCEPTED MANUSCRIPT 1. Introduction
30
Hazard Analysis & Critical Control Point (HACCP), a preventative system built on
31
the basis of prerequisite programs, has in consensus been regarded in the international
32
food safety community as a scientific and effective approach to food safety
33
management (FSM) (CAC, 2009; WHO, 1997, 2006). Due to its preventative
34
characteristics based on hazard analysis rather than end-product inspection, HACCP
35
(as well as other FSM systems adopting HACCP principles, e.g. ISO 22000), is
36
believed to be a cost-effective approach to FSM (WHO, 1997). Effectiveness of FSM
37
systems in ensuring food safety depend on various factors including, but not limited to,
38
top management support, access to expert knowledge for hazard analysis,
39
understanding of the systems, well-trained personnel, motivation and/or adhesion to
40
the system, as well as establishment and maintenance of prerequisite programs
41
(Garayoa, Vitas, Díez-Leturia, & García-Jalón, 2011; Jevšnik, Hlebec, & Raspor,
42
2006; Shih & Wang, 2011). It has been evidenced that more than half of food industry
43
managers do not sufficiently understand what HACCP is, and that knowledge, via the
44
mediating influence of attitude, actively influences FSM practices which will in turn
45
affect the effectiveness of HACCP-based FSM systems (Baş, Yüksel, & Çavuşoğlu,
46
2007; Ko & Wen-Hwa, 2013). In addition, degrees of compliance of HACCP
47
implementation
48
Commission (CAC) in a national scale varied between countries (Trafiałek, Lehrke,
49
Lücke, Kołożyn-Krajewska, & Janssen, 2015; Wallace, Holyoak, Powell, & Dykes,
50
2014; Wallace, Sperber, & Mortimore, 2010). Effectiveness of FSM systems would
51
be compromised if the necessary resources with regard to the infrastructures, work
52
environment, and the human aspect, which are the fundamentals of food enterprises to
53
plan, realize, and validate the production of safe products, are not provided by top
54
management (Psomas & Kafetzopoulos, 2015). Because of financial constraints, lack
55
of technical support, and insufficient personnel training that small and/or less
56
developed food businesses (SLDBs) are usually confronted with, as well as
57
complexity of documentation and records, implementation of HACCP-based FSM
TE D
M AN U
SC
RI PT
29
recommendations
stipulated
by
Codex
Alimentarius
AC C
EP
with
2
ACCEPTED MANUSCRIPT systems may not appear as effective as expected for them in ensuring food safety (W.
59
Dzwolak, 2014, 2016; Ramalho, Moura, & Cunha, 2015; Violaris, Bridges, & Bridges,
60
2008). And doubts have been raised on the feasibility of implementing FSM systems,
61
as well as their effectiveness for SLDBs in controlling foodborne hazards (Cerf &
62
Donnat, 2011; Jevšnik, et al., 2006; Ramalho, et al., 2015; Taylor, 2001).
63
Concepts of HACCP were introduced into China in late 1980s, and standards for
64
implementing HACCP-based FSM systems (e.g. GB/T 27341 and GB/T 22000) have
65
been developed and continuously updated since 2002 (Gao, Ma, & Yan, 2002).
66
However, implementation of these FSM systems are not obligatory, despite the fact
67
that many large and/or developed food companies voluntarily adopted these
68
HACCP-based FSM systems and have them certified in order to gain competitive
69
advantages, get access to new markets, and enhance consumer confidence.
70
Considering the huge amount of SLDBs in China, as well as disparities of them in
71
scale, financial resources, management, techniques, and human resources,
72
effectiveness of HACCP-based FSM systems in ensuring food safety may not be
73
sufficiently evident for them, which will lead to reluctance of these SLDBs to adopt
74
such FSM systems. To have Chinese SLDBs, especially those with constraints in food
75
safety knowledge and/or other resources, realize the importance of HACCP-based
76
FSM systems in controlling foodborne hazards along the whole chain of food
77
production, it is imperial to gather as much information as possible on the
78
effectiveness of HACCP-based FSM systems in ensuring safety of their products.
79
Meta-analysis is a set of quantitative techniques used to enhance the scientific rigor of
80
the review process, by integrating, contrasting and synthesizing data from a large
81
number of studies carried out by different research groups on the same question,
82
which is unlikely to be achieved using only a single study (Gonzales-Barron, Cadavez,
83
Sheridan, & Butler, 2013; Sutton, Abrams, & Jones, 2001). It has been widely
84
accepted as an effective tool to produce more reliable and precise estimation of effect
85
size of a particular treatment, cumulative knowledge of the research question, as well
86
as clear directions about the remaining research needs. Despite its origination from
AC C
EP
TE D
M AN U
SC
RI PT
58
3
ACCEPTED MANUSCRIPT evidence-based medicine, meta-analysis has been increasingly adopted to address a
88
broad range of research questions with regard to food safety, including but not limited
89
to, presence of microbial contamination in foods, risk assessment of foodborne
90
pathogens, occurrence of foodborne diseases, and barriers to implementation of FSM
91
systems (Gonzalesbarron & Butler, 2011; Hedges, 2015; Jevšnik, et al., 2006;
92
Noordhout, et al., 2014; Xavier, Gonzales-Barron, Paula, Estevinho, & Cadavez,
93
2014).
94
The objectives of this research were to compile all publicly accessible information on
95
the assessment of HACCP-based FSM systems in Chinese SLDBs and provide
96
evidence on the effectiveness of these FSM systems in SLDBs with a quantitative
97
approach.
98
2. Materials and methods
99
2.1 Methodology
M AN U
SC
RI PT
87
A meta-analysis technique was used to assess the overall effectiveness of
101
HACCP-based FSM systems in Chinese SLDBs. The problem statement, population,
102
intervention, and measured outcomes in this study were described in Table 1. Food
103
businesses were assumed as SLDBs unless they are in the section of dairy, functional
104
foods, food additives, or with a specified staff population over 300. Food products in
105
this study included raw materials, in-process and final food products. Employees’
106
knowledge about food safety & hygiene, food contacting surfaces & operating
107
environments, and food handlers’ hygiene were included in measured outcomes since
108
they were considered as important indexes to FSM levels of food enterprises, as well
109
as their food safety status. Odds ratio (OR), which referred to the degree to which the
110
events of interest (i.e., qualified numbers in this study) occurred in the population,
111
was used as effect size. Primary identification of suitable studies was conducted
112
through literature search for studies reporting effectiveness of HACCP-based FSM
113
systems in Chinese SLDBs, indexed since 1990. Literature search and selection of
114
included studies was performed according to procedures illustrated in Figure 1 (Box
115
A).
AC C
EP
TE D
100
4
ACCEPTED MANUSCRIPT 2.2 Quality assessment and data extraction
117
The quality of the included studies was evaluated and graded using a scoring
118
approach. A particular study was assigned with one point when meeting each of the
119
following criteria: I) year of study was specified; II) both intervention and control
120
included at least 100 monitoring objects in any of the four categories examined for
121
chemical or microbial indexes; III) monitoring indexes were clearly specified; IV)
122
methods for examination of particular monitoring objects as well as the criteria to
123
judge qualification status of the objects were clearly specified; V) unqualified
124
numbers based on single testing indexes and on all testing indexes were presented.
125
Whereas one point was subtracted when a particular study lacked information on
126
qualified numbers based on results of all testing indexes when more than one index
127
was involved for examination. Exclusion of studies was conducted prior to quality
128
assessment, and thus no papers were excluded in this stage according to the result of
129
quality assessment. According to the results, included studies were scored 0 to 5
130
points based on the number of criteria they met, and were assorted to three grades: 1)
131
studies considered of high quality with a total of 4 or 5 points, 2) studies considered
132
of moderate quality with a total of 2 or 3 points, and 3) studies with a total 0 or 1
133
points considered of low quality.
134
Information extracted from all included studies (where applicable) included year of
135
study, type of food businesses, study site (province), monitoring objects (e.g. hygiene
136
of handlers’ hands), as well as data presented in 2 × 2 contingency tables or indirectly
137
in the form of sample sizes associated with qualification rates. The worst case was
138
chosen to calculate ORs if qualified numbers of the objects, examined for more than
139
one testing indexes, were reported only based on single indexes, to avoid
140
overestimation of food safety levels both in intervention and control.
141
2.3 Statistical analysis
142
ORs from 2 × 2 contingency tables were calculated and forest plots generated using
143
statistical software Stata (Version 14, StataCorp LLC, USA), in which all zero
144
prevalence estimates were added with a standard correction of 0.5, with the
AC C
EP
TE D
M AN U
SC
RI PT
116
5
ACCEPTED MANUSCRIPT ‘Meta-Analysis’ package for each category of monitoring objects. Heterogeneity of
146
the studies was determined with Higgin’s test, and random effects models were
147
applied throughout because of the evidence for heterogeneity (I2>50%). Papers
148
included in this study were assumed to be free of publication bias if the P values were
149
above 0.05 when determined with Egger’s test. ORs higher than 1 indicated higher
150
odds of being qualified for monitoring objects when intervention (implementation of
151
HACCP systems) was applied in SLDBs to control potential foodborne hazards
152
and/or enhance food safety.
153
3. Results
154
A total of 2457 papers were retrieved from CNKI, CSPD, and VIP after eliminating
155
duplicates, 68 articles were selected after screening according to title or abstract, and
156
finally 24 studies were included in our research for meta-analysis after full-text
157
screening. Selection of included studies were conducted according to procedures and
158
criteria illustrated in Figure 1 (Box B). All of the 24 included publications were in
159
Chinese, and characteristics of the included studies were summarized in Table 2.
160
Briefly, 12 of the included studies were on effectiveness of HACCP-based FSM
161
systems in catering establishments (Guo, Hu, & Feng, 2013; X.-g. Jiang & Chen,
162
2016; Kou, 2013; Long, Wei, Chen, Lu, & Zhong, 2011; Ma, 2002; Pan, Chen, &
163
Xiao, 2009; Qi, Zhang, Gao, Li, & Li, 2013; Shao, Liu, Jin, & Gao, 2009; J.-s. Wei,
164
2012; L.-m. Wei & Li, 2010; Yang, Zhan, & Ma, 2014; Yu, Zheng, & Guo, 2015;
165
Yuan, Liu, Xin, Shi, & Zhang, 2015; Zhan & Ma, 2014), 4 studies in beverage and
166
bottled water (J.-z. Jiang, 2005; Z.-y. Jiang & Zhao, 2004; Li, Chen, & Chen, 2011;
167
Pan, et al., 2009), 3 studies in ready-to-eat meat (Cao, 2006; Guan, Wang, Bao, Li, &
168
Lv, 1994; Ma, 2002), 2 studies in bakery (S.-q. Liu, Huang, & Li, 2008; Wang, Xu,
169
Lu, & Yu, 1999), 2 studies in frozen and dry food (L.-d. Liu, Gao, Gao, & Liang,
170
1997; Sun, Li, Lin, & Chen, 2007), and 1 study in tableware sanitizing (Zhang &
171
Feng, 2013). Within the 24 included studies investigating effectiveness of
172
HACCP-based FSM systems in SLDBs, 11 studies examined only food products or
173
food contacting surfaces & operating environments, 1 study examined food contacting
AC C
EP
TE D
M AN U
SC
RI PT
145
6
ACCEPTED MANUSCRIPT surfaces & operating environments as well as employees’ knowledge about food
175
safety & hygiene, and the other 12 studies examined at least three of the four
176
mentioned aspects, with 9 of them containing food products.
177
Forest plots on the effects of HACCP-based FSM systems on food safety levels of
178
SLDBs in perspectives of food products, employees’ knowledge about food safety &
179
hygiene, food contacting surfaces & operating environments, as well as handlers’
180
hygiene were presented in Figure 2. Publication bias (P < 0.05) was found for three of
181
the above four meta-analyses, with exception of meta-analysis for association
182
between implementing FSM systems and levels of employees’ knowledge about food
183
safety & hygiene (P = 0.212). However, when studies involving more than 200
184
samples (in intervention or control) were excluded, or only studies focused on
185
catering section were involved, meta-analysis for association between implementing
186
FSM systems and food safety levels of food products was demonstrated as free of
187
publication bias, with P values of 0.103, and 0.078 respectively.
188
The estimated pooled effects [measured as ORs (95% CI)] of having HACCP-based
189
FSM systems implemented compared with controls were 10.58 (6.06, 18.49), 8.66
190
(4.88, 15.35), 6.39 (3.36, 12.18), and 6.65 (4.29, 10.30), respectively, in the
191
perspectives of food products, food handlers’ personal hygiene, employees’
192
knowledge about food safety & hygiene, and food contacting surfaces & operating
193
environments. As demonstrated by the overlapping 95% CIs of ORs (Table 3),
194
subgroup analyses on classification of food products into different categories
195
according to their characteristics did not reveal significant difference in the
196
effectiveness of FSM systems in SLDBs among these subgroups.
197
4. Discussion
198
Considering the wide application and acceptance of HACCP-based FSM systems for
199
ensuring food safety and doubts on their feasibility and effectiveness in SLDBs (Baş,
200
et al., 2007; Maldonado-Siman, Bai, Ramírez-Valverde, Gong, & Lara, 2014; Taylor,
201
2001), in this study a meta-analysis technique was used to assess the effectiveness of
202
HACCP-based FSM systems in Chinese SLDBs. Results of this study demonstrated
AC C
EP
TE D
M AN U
SC
RI PT
174
7
ACCEPTED MANUSCRIPT that HACCP-based FSM systems were effective in SLDBs, as demonstrated by
204
significant improvements (95% CI of ORs on right of 1) in any of the four measured
205
outcomes after implementing FSM systems, despite the heterogeneity of the included
206
studies in meta-analyses. These involved SLDBs varied in characteristics of their
207
products, scales of the businesses, education levels of their employees, degrees of
208
training their employees received, as well as managements’ and employees’ cognition
209
to food safety (Gao, et al., 2002; Ko & Wen-Hwa, 2013; Long, et al., 2011; J.-s. Wei,
210
2012; Zhan & Ma, 2014). Taking into consideration these diversities, it’s reasonable
211
to accept heterogeneity of the included studies in meta-analyses with random effects
212
models to estimate the compiled effectiveness of HACCP-based FSM systems in
213
SLDBs. Combining these diversities with the absence of studies determinately
214
reporting ineffectiveness of HACCP-based FSM systems, it was reasonable to
215
propose that it was not preferentially reporting studies demonstrating effectiveness of
216
FSM systems that led to publication bias demonstrated by results of Egger’s test. This
217
was evidenced by the findings that when studies involving more than 200 monitoring
218
objects, or focusing on non-catering food businesses were excluded, meta-analyses
219
were free of publication bias for association between implementing HACCP-based
220
FSM and odds of qualified food products. Thus, information generated in this study
221
was reliable to understand the effectiveness of HACCP-based FSM systems
222
implemented in SLDBs in ensuring food safety.
223
Improvements in the aspects of food products and food handlers’ hygiene, by
224
adopting HACCP-based FSM systems were statistically significant in all included
225
studies, demonstrated with 95% CIs of ORs locating completely on the right side of
226
the line representing a value of 1, as shown in Figure 2a and 2b. However, despite the
227
significant improvements at overall levels implied by meta-analyses (Figure 2c and
228
2d), 2 and 1 included studies demonstrated insignificant improvement after
229
implementing HACCP-based FSM systems, respectively, in the aspects of food
230
contacting surfaces & operating environments, and employees' knowledge about food
231
safety & hygiene. Large values of pooled ORs, as shown in Figure 2, demonstrated
AC C
EP
TE D
M AN U
SC
RI PT
203
8
ACCEPTED MANUSCRIPT that having HACCP-based FSM systems implemented in SLDBs resulted in a 6 –
233
10-fold increase in the ratios of qualified numbers to unqualified numbers of all
234
testing objects monitored in FSM. However, these findings also implied the huge gaps
235
in these food businesses between their current food safety levels and what would be
236
achieved when adopting HACCP-based FSM systems. Even though all the four
237
mentioned aspects of food safety were covered in FSM, the final goal of FSM was to
238
ensure safety of food products that would be eventually consumed, through
239
improvement in food handlers’ hygiene, employees’ knowledge in related fields, food
240
contacting surfaces of utensils and operating environments, as well as other inputs
241
(e.g. raw materials) that were not in the scope of this study (Djekic, et al., 2016; FAO,
242
1998; Ko & Wen-Hwa, 2013; WHO, 1997). The higher value of pooled OR (10.56) of
243
having HACCP-based FSM systems implemented compared with not having these
244
FSM systems in the perspective of food products than that in the perspective of food
245
handlers’ hygiene (8.66), food contacting surfaces & operating environments (6.65),
246
or employees’ knowledge on food safety & hygiene (6.39) indicated the cumulative
247
effects of FSM systems on safety status of final food products, which were gained
248
through improvements in all aspects of food production by adoption of FSM systems.
249
Despite the effectiveness of HACCP-based FSM systems in ensuring food safety, as
250
well as the associated benefits such as access to new markets (Al-Busaidi, Jukes, &
251
Bose, 2017; Maldonado-Siman, et al., 2014; Psomas & Kafetzopoulos, 2015), a large
252
portion of Chinese food businesses, especially SLDBs, were confronted with various
253
difficulties and barriers (such as financial constraints, lack of prerequisite programs,
254
and frequent staff turnover) for implementing HACCP-based FSM systems (Q. Jiang
255
& Batt, 2016; x.-d. Liu, Duan, & She, 2007). Uncertainty about effectiveness of
256
HACCP-based FSM systems in SLDBs also led to their reluctance to adopt these
257
FSM systems (Baş, et al., 2007; Cerf & Donnat, 2011). Significant improvement in
258
food safety levels by adopting HACCP principles in production of roast chicken
259
(Guan, et al., 1994; Ma, 2002) and school canteens (X.-g. Jiang & Chen, 2016; Long,
260
et al., 2011; Shao, et al., 2009; L.-m. Wei & Li, 2010; Yang, et al., 2014; Yu, et al.,
AC C
EP
TE D
M AN U
SC
RI PT
232
9
ACCEPTED MANUSCRIPT 2015) should reduce or eliminate doubts on the effectiveness of such FSM systems in
262
SLDBs. The estimated random effects ORs (Figure 2) provided evidence of the
263
effectiveness of HACCP-based FSM systems in SLDBs at an overall level, taken into
264
account the diversities of these food businesses in aspects such as business scales,
265
characteristics of the products, production environments, expertise and skills in food
266
safety management, education levels of employees, degrees of personnel training, and
267
knowledge about food safety & hygiene, as well as difficulties and barriers SLDBs
268
were commonly confronted with.
269
It had been suggested that efforts from various stakeholders were required to promote
270
implementing HACCP-based FSM systems in SLDBs, including but not limited to
271
technical
272
communication between governments and businesses and between businesses,
273
infrastructures and facilities assured by governments and businesses, sector-specific
274
industry guides developed industry associations, and provision of basic but adequate
275
knowledge about food safety & hygiene (WHO, 1999, 2006). Considering the
276
existence of numerous self-employed food businesses, as well as low level of food
277
handlers’ knowledge about food safety & hygiene (Gong, Wang, Yang, & Bai, 2016),
278
it was necessary and critical to develop simplified versions of HACCP-based FSM
279
systems for SLDBs where it is not feasible to completely implement FSM systems as
280
recommended for scaled and industrialized food businesses (W. Dzwolak, 2014, 2016;
281
Taylor, 2008). In such simplified HACCP, attention should be focused on potential
282
hazards and CCPs to control them (Waldemar Dzwolak, 2018). Such simple versions
283
would be more easily to he accepted by these numerous SLDBs, which would
284
certainly help ensure food safety of their product and thus protect consumer health.
285
In conclusion, implementing HACCP-based FSM systems in Chinese SLDBs
286
significantly increased the ratios of qualified to unqualified numbers of testing
287
objectives monitored in FSM; and evidence demonstrated that HACCP-based FSM
288
systems were effective in SLDBS. However, impedance of barriers, constraints, and
289
difficulties to adopting HACCP-based FSM systems shouldn’t be ignored when
governments
and
SC
from
industry
associations,
adequate
AC C
EP
TE D
M AN U
support
RI PT
261
10
ACCEPTED MANUSCRIPT 290
promoting application of FSM systems in SLDBs.
291
Acknowledgement
293
This work was supported by China Central Universities Fundamental Research Funds
294
(DUT16RC(3)025. and DUT16QY37.).
AC C
EP
TE D
M AN U
SC
RI PT
292
11
ACCEPTED MANUSCRIPT 295
Reference:
296
Al-Busaidi, M. A., Jukes, D. J., & Bose, S. (2017). Hazard analysis and critical control point
297
(HACCP) in seafood processing: An analysis of its application and use in regulation in
298
the Sultanate of Oman. Food Control, 73, 900-915. Baş, M., Yüksel, M., & Çavuşoğlu, T. (2007). Difficulties and barriers for the implementing of
300
HACCP and food safety systems in food businesses in Turkey. Food Control, 18(2),
301
124-130.
SC
RI PT
299
CAC. (2009). Codex Alimentarius: Food hygiene, basic texts. Fourth Edition. . World Health
303
Organization (WHO), Food and Agriculture Organization of the United Nations (FAO),
304
Rome, 2009, 23-33.
M AN U
302
Cao, Q.-j. (2006). Assessment of Effects of Applying HACCP on Controlling Microbial
306
Contamination of Unpackaged Cooked Meat Products. Chinese Journal of Public
307
Health Management, 22(5), 435-436.
TE D
305
Cerf, O., & Donnat, E. (2011). Application of hazard analysis – Critical control point (HACCP)
309
principles to primary production: What is feasible and desirable? Food Control, 22(12),
310
1839-1843.
EP
308
Djekic, I., Kuzmanović, J., Anđelković, A., Saračević, M., Stojanović, M. M., & Tomašević, I.
312
(2016). Effects of HACCP on process hygiene in different types of Serbian food
AC C
311
313 314
establishments. Food Control, 60, 131-137.
Dzwolak, W. (2014). HACCP in small food businesses – The Polish experience. Food Control,
315 316
36(1), 132-137. Dzwolak, W. (2016). Documenting HACCP in a small restaurant – a practical approach.
12
ACCEPTED MANUSCRIPT 317 318
Quality Assurance & Safety of Crops & Foods, 9(2), 1-8. Dzwolak, W. (2018). Developing a hazard analysis worksheet in a small food business with the
319
application of a T-shaped matrix diagram. Food Control, 87, 180-191. FAO. (1998). Food quality and safety systems: a training manual on food hygiene and the
321
Hazard Analysis and Critical Control Point (HACCP) system. Food Quality and
322
Standards Service Food and Nutrition Division, Food and Agriculture Organization of
323
the United Nations, Rome.
SC
Gao, P., Ma, F.-k., & Yan, Z.-c. (2002). Application and Reseach of HACCP Management in
325
M AN U
324
RI PT
320
China and other Countries. Chinese Journal of Public Health, 18(10), 1262-1264. Garayoa, R., Vitas, A. I., Díez-Leturia, M., & García-Jalón, I. (2011). Food safety and the
327
contract catering companies: Food handlers, facilities and HACCP evaluation. Food
328
Control, 22(12), 2006-2012.
329
TE D
326
Gong, S., Wang, X., Yang, Y., & Bai, L. (2016). Knowledge of food safety and handling in
330
households: A survey of food handlers in Mainland China. Food Control, 64, 45-53. Gonzales-Barron, U., Cadavez, V., Sheridan, J. J., & Butler, F. (2013). Modelling the effect of
332
chilling on the occurrence of Salmonella on pig carcasses at study, abattoir and batch
AC C
EP
331
333
levels by meta-analysis. International Journal of Food Microbiology, 163(2–3),
334 335
101-113.
Gonzalesbarron, U., & Butler, F. (2011). The use of meta-analytical tools in risk assessment
336
for food safety. Food Microbiology, 28(4), 823-827.
337
Guan, W.-j., Wang, M.-q., Bao, D.-y., Li, S.-q., & Lv, L.-t. (1994). Assessment of Effects of
338
Applying HACCP in Production and Vendition of Roast Chiken in Zhongshan District, 13
ACCEPTED MANUSCRIPT 339
Dalian
Chinese Journal of Food Hygiene(4), 1-3.
Guo, S.-y., Hu, R.-f., & Feng, W. (2013). Study on the Application of Hazard Analysis and
341
Critical Control Point Management System in Sanitary Supervision of University
342
Canteens. Practical Preventive Medicine(9), 1097-1099.
343
Hedges, L. V. (2015). The early history of meta-analysis. Research Synthesis Methods, 6(3),
344
284.
Jevšnik, M., Hlebec, V., & Raspor, P. (2006). Meta-analysis as tool for barriers identification
SC
345 346
during HACCP implementation to improve food safety. Acta Alimentaria(3), 319-353.
M AN U
347
RI PT
340
Jiang, J.-z. (2005). Investigation on Effects of HACCP in the Production of Purified Drinking
348
Water. Occupation and Health, 21(2), 236-237.
Jiang, Q., & Batt, P. J. (2016). Barriers and benefits to the adoption of a third party certified
350
food safety management system in the food processing sector in Shanghai, China.
351
Food Control, 62, 89-96.
352
TE D
349
Jiang, X.-g., & Chen, Y.-z. (2016). Study on Food Safety Intervention Strategy of Catering
353
EP
Jiang, Z.-y., & Zhao, X.-m. (2004). Investigation on Analysis and Preventive Measures for
AC C
354
distribution enterprise. Chinese Journal of Public Health Management(2), 247-249.
355
Hygiene Quality of Bottled Drinking Water in City of Linyi. Occupation and Health,
356
20(9), 64-65.
357
Ko, Wen-Hwa. (2013). The relationship among food safety knowledge, attitudes
358
and;self-reported HACCP practices in restaurant employees. Food Control, 29(1),
359
192-197.
360
Kou, L.-q. (2013). Applied Research in Effects of HACCP on Food safety of High School 14
ACCEPTED MANUSCRIPT 361
Canteens in a County Guide of China Medicine, 11(9), 36-38. Li, Z., Chen, J., & Chen, J.-q. (2011). A Study on Effects of HACCP on Controlling
363
microorganisms and Pesticide residues in Oolong Tea. Science & Technology for
364
Development(s1), 182-183.
RI PT
362
Liu, L.-d., Gao, J.-s., Gao, Y., & Liang, Y.-b. (1997). Assessment of Management Effects of
366
Implementing Hazard Analysis and Critical Control Points in Frozen Food Production
367
Preventive Medicine Tribune(1), 56-57.
Liu, S.-q., Huang, L.-z., & Li, Q. (2008). Evaluation of Application of HACCP System in
369
M AN U
368
SC
365
Process Production of Cake and Paste. China Tropical Medicine, 8(9), 1660-1661. Liu, x.-d., Duan, Q.-j., & She, R.-p. (2007). Problem Existing and Suggestions on
371
Implementation and Establishment of HACCP System in China. Food Science, 28(5),
372
369-373.
TE D
370
Long, P., Wei, X.-w., Chen, Y.-b., Lu, Z.-g., & Zhong, M. (2011). Analysis on the effectiveness
374
of establishing HACCP system in student canteens in High School District. Chinese
375
Journal of Food Hygiene, 23(2), 167-169. Ma, W.-m. (2002). Investigation of Effects of Applying HACCP Methodology on Controlling
AC C
376
EP
373
377
Hygiene of Cookded Bitern Food. Medical Journal of Communications, 16(3),
378
297-297.
379
Maldonado-Siman, E., Bai, L., Ramírez-Valverde, R., Gong, S., & Lara, R. D. (2014).
380
Comparison of implementing HACCP systems of exporter Mexican and Chinese meat
381
enterprises. Food Control, 38(4), 109-115.
382
Noordhout, C. M. D., Devleesschauwer, B., Angulo, F. J., Verbeke, G., Haagsma, J., Kirk, M., 15
ACCEPTED MANUSCRIPT 383
Havelaar, A., & Speybroeck, N. (2014). The global burden of listeriosis: a systematic
384
review and meta-analysis. Lancet Infectious Diseases, 14(11), 1073-1082. Pan, H.-j., Chen, S.-p., & Xiao, L.-j. (2009). Effect Assessment of Applying HACCP in
386
Inspection and Quarantine Supervision of Exported Health Protection Tea. Chinese
387
Frontier Health Quarantine(1), 45-47.
388
RI PT
385
Psomas, E. L., & Kafetzopoulos, D. P. (2015). HACCP effectiveness between ISO 22000
389
SC
certified and non-certified dairy companies. Food Control, 53, 134-139.
Qi, R.-j., Zhang, X.-j., Gao, G.-p., Li, X., & Li, x.-d. (2013). Combination Application of HACCP
391
and 6S Method in School Canteens Management. Chinese Journal of School Health,
392
34(12), 1475-1478.
393
M AN U
390
Ramalho, V., Moura, A. P. D., & Cunha, L. M. (2015). Why do small business butcher shops
394
TE D
fail to fully implement HACCP? Food Control, 49(49), 85-91. Shao, X.-n., Liu, Y.-q., Jin, X.-m., & Gao, R.-q. (2009). Influence and Effect on Food Hygiene
396
Management after Establishing HACCP System in Dining Enterprises. Chinese
397
Journal of Public Health Management, 25(3), 251-253. Shih, K.-m., & Wang, W.-k. (2011). Factors influencing HACCP implementation in Taiwanese
AC C
398
EP
395
399 400
public hospital kitchens. Food Control, 22(3–4), 496-500.
Sun, S.-l., Li, Z., Lin, M.-x., & Chen, J.-q. (2007). Effects of HACCP System on Controlling
401
Total Bacterial Colonies in Frozen-dried Green Union. Journal of
402 403
Henan University of
Science and Technology: Natural Science, 28(5), 51-53. Sutton, A. J., Abrams, K. R., & Jones, D. R. (2001). An illustrated guide to the methods of
404
meta-analysis. Journal of Evaluation in Clinical Practice, 7(2), 135-148. 16
ACCEPTED MANUSCRIPT 405
Taylor, E. (2001). HACCP in small companies: benefit or burden? Food Control, 12(4),
406 407
217-222. Taylor, E. (2008). A new method of HACCP for the catering and food service industry. Food
408
RI PT
Control, 19(2), 126-134. Trafiałek, J., Lehrke, M., L ü cke, F. K., Kołożyn-Krajewska, D., & Janssen, J. (2015).
410
HACCP-based procedures in Germany and Poland. Food Control, 55, 66-74.
411
Violaris, Y., Bridges, O., & Bridges, J. (2008). Small businesses – Big risks: Current status and
SC
409
412
M AN U
413
future direction of HACCP in Cyprus. Food Control, 19(5), 439-448.
Wallace, C. A., Holyoak, L., Powell, S. C., & Dykes, F. C. (2014). HACCP – The difficulty with
414
Hazard Analysis. Food Control, 35(1), 233-240.
Wallace, C. A., Sperber, W. H., & Mortimore, S. E. (2010). Food Safety for the 21st Century:
416
Managing HACCP and Food Safety throughout the Global Supply Chain:
417
Wiley-Blackwell, Oxford, Iowa.
TE D
415
Wang, C.-m., Xu, X.-m., Lu, L.-h., & Yu, M.-h. (1999). Assessment of Effects of Applying
419
HACCP in Production and Vendition of Decorative Cake Zhejiang Journal of
420
Preventive Medicine(3), 28-31.
AC C
421
EP
418
Wei, J.-s. (2012). Investigation of Effects of HACCP Establishment in
422 423
Nanchang Railway. World Health Digest Medical Periodieal(52), 10-11.
Wei, L.-m., & Li, H.-j. (2010). Application of HACCP in Hygiene Management of School
424 425
Apartment Canteen of
Canteens. Journal of Applied Preventive Medicine(2), 123-124. WHO. (1997). HACCP: Introducing the Hazard Analysis and Critical Control Point System.
426
WHO/FSF/FOS/97.2, 17
Geneva,
World
Health
Organization.
ACCEPTED MANUSCRIPT 427
http://www.who.int/iris/handle/10665/63610. WHO. (1999). Strategies for implementing HACCP in small and/or less developed businesses.
429
Report of a WHO consultation in collaboration with the Ministry of Health, Welfare and
430
Sports, The Netherlands, The Hague, 16-19 June.
431
RI PT
428
WHO. (2006). FAO/WHO guidance to governments on the application of HACCP in small
432
and/or less-developed food businesses. FAO
Food & Nutrition Paper, 86.
Xavier, C., Gonzales-Barron, U., Paula, V., Estevinho, L., & Cadavez, V. (2014). Meta-analysis
434
of the incidence of foodborne pathogens in Portuguese meats and their products.
435
Food Research International, 55(2), 311-323.
M AN U
SC
433
Yang, W.-c., Zhan, Z.-p., & Ma, H.-l. (2014). Analysis of Effects of HACCP System
437
Establishment in Canteens of Universities and Colleges. Medical Aesthetics and
438
Beauty(5), 554-554.
TE D
436
Yu, X.-m., Zheng, G.-w., & Guo, H.-m. (2015). Application research of HACCP combined with
440
six T practical management model in food safety of school canteen. China Modern
441
Medicine(30), 140-143.
Yuan, J., Liu, H., Xin, X., Shi, S.-g., & Zhang, G.-j. (2015). Application of HACCP in Preparing
AC C
442
EP
439
443
Hot Dishes in Catering Establishments Journal of Preventive Medicine of Chinese
444 445
People's Liberation Army(1), 39-40.
Zhan, Z.-p., & Ma, H.-l. (2014). Effectiveness of HACCP Implementation in Canteens of
446
Primary and Secondary Schools. Biotech World(10), 74-74.
447
Zhang, D.-s., & Feng, S.-x. (2013). The Implementation and Effects of HACCP Management
448
Principle in the Unit Concentratedly Disinfected Tableware. Modern Preventive 18
ACCEPTED MANUSCRIPT 449
Medicine, 40(15), 2925-2928.
AC C
EP
TE D
M AN U
SC
RI PT
450
19
ACCEPTED MANUSCRIPT Tables Table 1. Problem statement, population, intervention, and measured outcome specified for meta-analysis. Description
problem statement
Investigation of the overall effectiveness of HACCP-based FSM systems in of Chinese
population
SLDBs located in mainland of China
intervention
Implementation of HACCP-based systems
measured outcome
Qualified numbers of : 1) food products
SC
SLDBs
RI PT
Elements
2) employees’ knowledge about food safety & hygiene 3) food contacting surfaces & operating environments
AC C
EP
TE D
M AN U
4) food handlers’ hygiene
ACCEPTED MANUSCRIPT Table 2. Characteristics of included studies.
W.-J Guan, 1994 C.-M Wang, 1999 W.-M Ma, 2002 Z.-Y Jiang, 2004 J.-J Jiang, 2005
Study of
Categories of Food
Monitoring
Total
study
province
business
objects involved
points
1992-1993 1997
Liaoning Zhejiang
2000
Jiangsu
2002-2003 2003
RTE meat Bakery RTE meat
Shandong Jiangsu
Drinks/bottled water Drinks/bottled water
Q.-J Cao, 2006
2005
Jiangsu
RTE meat
S.-L Sun, 2007
NA
Fujian
Frozen/dry food
L.-D Liu, 1997 L.-Q Kou, 2013 S.-Q Liu, 2008 X.-N Shao, 2009 D.-S Zhang, 2013 J Yuan, 2015
Guangdong
NA
Fujian
1995
Drinks/bottled water
Shandong
NA
Sichuan
2005-2007 NA
NA
Guangdong
Guangxi Gansu
Frozen/dry food Catering
Shandong
2009-2011
Drinks/bottled water
Bakery Catering
2
F
a
3
F
a
5
F
a
3
F
a
4
F
a
3
Fa
3
F
a
1
F
a
1
a
b
c
3
a
b
c
2
a
b
c
3
SC
Z Lin, 2011
NA
M AN U
H.-J Pan, 2009
F
a
RI PT
Study ID
Year of
Tableware sanitizing Catering
F , H , S_E
F , H , S_E
F , H , S_E S_E
c
S_E
c
2 2
b
c
d
H , S_E , K a
b
c
2 d
4 2
P Long, 2011
NA
Xinjiang
Catering
F , H , S_E , K
S.-Y Guo, 2013
NA
Hunan
Catering
F a, H b, S_E c , K d
W.-C Yang, 2014 X.-G Jiang, 2016 X.-M Yu, 2015 R.-J Qi, 2013 L.-M Wei, 2010 Z.-P Zhan, 2014 a
NA
Jiangxi
2013
Liaoning
NA 2014 NA 2009 2014
Zhejiang
Catering
Catering
Catering
TE D
J.-S Wei, 2012
Jiangxi
Anhui
Guangxi
Liaoning
Catering
Catering Catering
Catering
AC C
EP
Final food products; Handlers’ personal hygiene; c Contacting surfaces of utensils and operating environments; d Employees’ knowledge about food safety and hygiene. b
a
b
c
d
3
a
b
c
d
4
F , H , S_E , K F , H , S_E , K b
c
d
2
b
c
d
2
H , S_E , K H , S_E , K a
c
F , S_E , K a
b
F , H , S_E c
S_E , K
d
d
c
1 4 2
ACCEPTED MANUSCRIPT Table 3. Results of the subgroup analyses examining the association of HACCP systems with food safety levels in the perspective of food products. Random effects pooled OR (95% CI) 7.58 (1.56, 36.88) 45.47 (0.88, 2351.65) 14.68 (2.34, 92.08) 12.67 (4.68, 34.35) 7.09 (2.52, 19.96) 9.67 (5.53, 16.92)
I2 89.10% 86.00% 73.40% 55.00% 88.70% 84.30%
AC C
EP
TE D
M AN U
SC
RTE meat Bakery Drinks/bottled water Frozen/dry food Catering Overall
Number of studies 3 2 4 2 6 17
Weight (%) 20.38 9.32 17.16 13.64 39.49 100
RI PT
Food category
ACCEPTED MANUSCRIPT Figure captions Figure 1. Literature search strategy and selection criteria for identifying suitable studies (Box A), and resulting papers included for meta-analysis (Box B). Figure 2. Meta-analysis of the association of HACCP-based FSM systems with food
RI PT
safety levels in the perspective of (a) food products, (b) food handlers’ hygiene, (c) food contacting surfaces & operating environments, and (d) employees' knowledge about food safety & hygiene.
Databases of CNKI, CSPD, and VIP
Primary
search under study Chinese terms for food safety/effectiveness/ hazard control
Box A
Requirements:
1) to be an original article;
M AN U
studies
SC
Figure 1.
2) microbial or chemical indexes involved in intervention and
Suitable studies
control; 3) ≥ 20 test samples involved;
TE D
Studies identified
2457 papers after eliminating
AC C
68 articles
EP
duplicates
Full-text screening
24 studies finally included for meta-analysis
Box B
2389 papers excluded according to title or abstract
17 studies (quantitative information on effectiveness of FSM systems was not found) 15 studies (sample size or clear differentiation of intervention and control was not specified ) 9 studies (data not presented in the required format) 2 studies (focused on occurrence of epidemic diseases in poultry farms) 1 study (reported effectiveness of HACCP in water supply system)
ACCEPTED MANUSCRIPT Figure 2.
% Weight
OR (95% CI)
W.-J Guan, 1994
23.27 (7.84, 69.14)
6.36
C.-M Wang, 1999
11.96 (4.29, 33.40)
6.54
W.-M Ma, 2002
2.36 (1.61, 3.46)
8.08
Z.-Y Jiang, 2004
46.74 (2.87, 761.80)
2.71
J.-J Jiang, 2005
25.72 (1.51, 439.38)
2.64
Q.-J Cao, 2006
9.60 (2.62, 35.21)
5.77
S.-L Sun, 2007
24.27 (7.14, 82.47)
5.98
H.-J Pan, 2009
30.34 (4.06, 226.55)
4.01
Z Lin, 2011
3.32 (1.84, 6.00)
7.67
8.62 (4.52, 16.41)
7.55
17.33 (8.73, 34.41)
7.45
237.47 (14.66, 3846.43)
2.71
18.55 (6.97, 49.36)
6.67
17.25 (5.19, 57.37)
6.04
4.67 (1.65, 13.24)
6.49
2.06 (1.39, 3.03)
8.06
13.95 (3.17, 61.34)
5.27
10.58 (6.06, 18.49)
100.00
SC
RI PT
Study
L.-D Liu, 1997 L.-Q Kou, 2013
P Long, 2011 S.-Y Guo, 2013 W.-C Yang, 2014 R.-J Qi, 2013 L.-M Wei, 2010 Overall
(I-squared = 84.5%, p = 0.000)
NOTE: Weights are from random effects analysis
1
3846
TE D
.00026
M AN U
S.-Q Liu, 2008
AC C
EP
a
ACCEPTED MANUSCRIPT
OR (95% CI)
% Weight
L.-D Liu, 1997
61.50 (17.26, 219.11)
8.90
L.-Q Kou, 2013
19.67 (6.27, 61.65)
9.71
S.-Q Liu, 2008
64.78 (8.80, 477.06)
5.41
J Yuan, 2015
7.67 (2.61, 22.54)
10.14
P Long, 2011
3.94 (1.43, 10.81)
10.61
S.-Y Guo, 2013
5.03 (1.83, 13.84)
10.58
SC
RI PT
Study
W.-C Yang, 2014 X.-G Jiang, 2016
L.-M Wei, 2010 Overall
(I-squared = 72.9%, p = 0.000)
NOTE: Weights are from random effects analysis
.0021
1
TE D EP
3.85 (1.82, 8.11)
12.47
3.62 (2.43, 5.40)
14.70
22.00 (2.05, 236.05)
4.26
8.66 (4.88, 15.35)
100.00
477
b
AC C
13.23
M AN U
X.-M Yu, 2015
7.43 (3.93, 14.06)
ACCEPTED MANUSCRIPT
% Weight
OR (95% CI)
D.-S Zhang, 2013
4.54 (2.98, 6.90)
8.59
Jian Yuan, 2015
14.75 (0.81, 269.34)
1.83
J.-S Wei, 2012
19.45 (8.26, 45.76)
6.88
L.-D Liu, 1997
24.11 (12.09, 48.08)
7.57
L.-M Wei, 2010
12.46 (4.87, 31.86)
6.53
L.-Q Kou, 2013
11.80 (5.56, 25.01)
7.32
P Long, 2011
7.37 (3.27, 16.62)
7.06
R.-J Qi, 2013
2.14 (1.57, 2.92)
8.91
S.-Q Liu, 2008
22.82 (5.33, 97.74)
4.60
S.-Y Guo, 2013
4.44 (1.95, 10.11)
7.01
SC
RI PT
Study
W.-C Yang, 2014 X.-G Jiang, 2016
X.-M Yu, 2015 Z.-P Zhan, 2014 Overall
(I-squared = 82.3%, p = 0.000)
NOTE: Weights are from random effects analysis
.00371
1
TE D EP
6.95
2.62 (1.16, 5.93)
7.05
4.09 (2.71, 6.19)
8.62
5.09 (0.89, 29.27)
3.75
6.65 (4.29, 10.30)
100.00
269
c
AC C
7.33
2.77 (1.20, 6.40)
M AN U
X.-N Shao, 2009
7.33 (3.47, 15.52)
ACCEPTED MANUSCRIPT
OR (95% CI)
% Weight
J Yuan, 2015
3.11 (1.65, 5.87)
12.18
P Long, 2011
7.72 (0.92, 64.69)
5.51
S.-Y Guo, 2013
12.25 (2.78, 53.99)
7.97
J.-S Wei, 2012
6.26 (2.48, 15.82)
10.74
W.-C Yang, 2014
17.94 (10.11, 31.84)
12.45
X.-G Jiang, 2016
3.07 (2.37, 3.97)
13.51
SC
RI PT
Study
X.-M Yu, 2015 R.-J Qi, 2013
Overall
(I-squared = 90.1%, p = 0.000)
NOTE: Weights are from random effects analysis
.0155
12.82
28.46 (14.84, 54.61)
12.10
64.7
TE D
d
EP
1.78 (1.10, 2.88)
6.39 (3.36, 12.18)
1
AC C
12.73
M AN U
Z.-P Zhan, 2014
5.55 (3.35, 9.20)
100.00
ACCEPTED MANUSCRIPT Highlights: HACCP-based systems in Chinese SLDBs significantly improved their food safety levels; Pooled ORs were in range of 6.39-10.58, regarding different aspects of food safety;
AC C
EP
TE D
M AN U
SC
RI PT
Evidence on effectiveness should encourage application of such FSM systems in SLDBs.
S0956-7135(18)30466-3
DOI:
10.1016/j.foodcont.2018.09.013
Reference:
JFCO 6313
To appear in:
Food Control
Received Date: 23 May 2018 Revised Date:
11 September 2018
Accepted Date: 12 September 2018
Please cite this article as: Yang Y., Wei L. & Pei J., Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs, Food Control (2018), doi: https:// doi.org/10.1016/j.foodcont.2018.09.013. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 2
Application of meta-analysis technique to assess effectiveness of HACCP-based FSM systems in Chinese SLDBs
3
Yongheng Yang a *, Lijuan Wei b, Jianan Pei a a
Department of Food Science and Environment, Dalian University of Technology, Liaoning, China b Instrumental Analysis & Research Center, Dalian University of Technology, Liaoning, China
RI PT
4 5 6 7
Abstract:
9
Despite effectiveness of HACCP-based FSM systems in food safety management
10
(FSM) and their common adoption in industrialized food businesses, doubts have
11
been raised on their feasibility in small and/or less developed food businesses (SLDBs)
12
because of barriers such as financial constraints, lack of technical expertise, and
13
insufficient personnel training faced by SLDBs. Uncertainty about their effectiveness,
14
as well as feasibility for SLDBs (including numerous Chinese SLDBs) made them
15
reluctant to adopt these FSM systems. In order to have clear understanding on the
16
effectiveness of HACCP-based FSM systems for SLDBs, this study presented
17
meta-analyses on associations between implementing HACCP-based FSM systems in
18
Chinese SLDBs and levels of their food safety status. Having HACCP-based FSM
19
systems implemented led to significantly higher odds of qualified numbers, with
20
pooled ORs (95% CI) being 10.58 (6.06, 18.49), 8.66 (4.88, 15.35), 6.39 (3.36, 12.18),
21
and 6.65 (4.29, 10.30), respectively, with regard to food products, food handlers’
22
hygiene, employees’ knowledge about food safety and hygiene, and food contacting
23
surfaces & operating environments. Solid evidence on effectiveness of HACCP-based
24
FSM systems in SLDBs should encourage food businesses to adopt these systems for
25
ensuring food safety.
26
Keywords: Food safety; HACCP; Effectiveness
27 28
* Corresponding author: [email protected]; 2 Dagong Rd, Panjin, Liaoning Province, China.
AC C
EP
TE D
M AN U
SC
8
1
ACCEPTED MANUSCRIPT 1. Introduction
30
Hazard Analysis & Critical Control Point (HACCP), a preventative system built on
31
the basis of prerequisite programs, has in consensus been regarded in the international
32
food safety community as a scientific and effective approach to food safety
33
management (FSM) (CAC, 2009; WHO, 1997, 2006). Due to its preventative
34
characteristics based on hazard analysis rather than end-product inspection, HACCP
35
(as well as other FSM systems adopting HACCP principles, e.g. ISO 22000), is
36
believed to be a cost-effective approach to FSM (WHO, 1997). Effectiveness of FSM
37
systems in ensuring food safety depend on various factors including, but not limited to,
38
top management support, access to expert knowledge for hazard analysis,
39
understanding of the systems, well-trained personnel, motivation and/or adhesion to
40
the system, as well as establishment and maintenance of prerequisite programs
41
(Garayoa, Vitas, Díez-Leturia, & García-Jalón, 2011; Jevšnik, Hlebec, & Raspor,
42
2006; Shih & Wang, 2011). It has been evidenced that more than half of food industry
43
managers do not sufficiently understand what HACCP is, and that knowledge, via the
44
mediating influence of attitude, actively influences FSM practices which will in turn
45
affect the effectiveness of HACCP-based FSM systems (Baş, Yüksel, & Çavuşoğlu,
46
2007; Ko & Wen-Hwa, 2013). In addition, degrees of compliance of HACCP
47
implementation
48
Commission (CAC) in a national scale varied between countries (Trafiałek, Lehrke,
49
Lücke, Kołożyn-Krajewska, & Janssen, 2015; Wallace, Holyoak, Powell, & Dykes,
50
2014; Wallace, Sperber, & Mortimore, 2010). Effectiveness of FSM systems would
51
be compromised if the necessary resources with regard to the infrastructures, work
52
environment, and the human aspect, which are the fundamentals of food enterprises to
53
plan, realize, and validate the production of safe products, are not provided by top
54
management (Psomas & Kafetzopoulos, 2015). Because of financial constraints, lack
55
of technical support, and insufficient personnel training that small and/or less
56
developed food businesses (SLDBs) are usually confronted with, as well as
57
complexity of documentation and records, implementation of HACCP-based FSM
TE D
M AN U
SC
RI PT
29
recommendations
stipulated
by
Codex
Alimentarius
AC C
EP
with
2
ACCEPTED MANUSCRIPT systems may not appear as effective as expected for them in ensuring food safety (W.
59
Dzwolak, 2014, 2016; Ramalho, Moura, & Cunha, 2015; Violaris, Bridges, & Bridges,
60
2008). And doubts have been raised on the feasibility of implementing FSM systems,
61
as well as their effectiveness for SLDBs in controlling foodborne hazards (Cerf &
62
Donnat, 2011; Jevšnik, et al., 2006; Ramalho, et al., 2015; Taylor, 2001).
63
Concepts of HACCP were introduced into China in late 1980s, and standards for
64
implementing HACCP-based FSM systems (e.g. GB/T 27341 and GB/T 22000) have
65
been developed and continuously updated since 2002 (Gao, Ma, & Yan, 2002).
66
However, implementation of these FSM systems are not obligatory, despite the fact
67
that many large and/or developed food companies voluntarily adopted these
68
HACCP-based FSM systems and have them certified in order to gain competitive
69
advantages, get access to new markets, and enhance consumer confidence.
70
Considering the huge amount of SLDBs in China, as well as disparities of them in
71
scale, financial resources, management, techniques, and human resources,
72
effectiveness of HACCP-based FSM systems in ensuring food safety may not be
73
sufficiently evident for them, which will lead to reluctance of these SLDBs to adopt
74
such FSM systems. To have Chinese SLDBs, especially those with constraints in food
75
safety knowledge and/or other resources, realize the importance of HACCP-based
76
FSM systems in controlling foodborne hazards along the whole chain of food
77
production, it is imperial to gather as much information as possible on the
78
effectiveness of HACCP-based FSM systems in ensuring safety of their products.
79
Meta-analysis is a set of quantitative techniques used to enhance the scientific rigor of
80
the review process, by integrating, contrasting and synthesizing data from a large
81
number of studies carried out by different research groups on the same question,
82
which is unlikely to be achieved using only a single study (Gonzales-Barron, Cadavez,
83
Sheridan, & Butler, 2013; Sutton, Abrams, & Jones, 2001). It has been widely
84
accepted as an effective tool to produce more reliable and precise estimation of effect
85
size of a particular treatment, cumulative knowledge of the research question, as well
86
as clear directions about the remaining research needs. Despite its origination from
AC C
EP
TE D
M AN U
SC
RI PT
58
3
ACCEPTED MANUSCRIPT evidence-based medicine, meta-analysis has been increasingly adopted to address a
88
broad range of research questions with regard to food safety, including but not limited
89
to, presence of microbial contamination in foods, risk assessment of foodborne
90
pathogens, occurrence of foodborne diseases, and barriers to implementation of FSM
91
systems (Gonzalesbarron & Butler, 2011; Hedges, 2015; Jevšnik, et al., 2006;
92
Noordhout, et al., 2014; Xavier, Gonzales-Barron, Paula, Estevinho, & Cadavez,
93
2014).
94
The objectives of this research were to compile all publicly accessible information on
95
the assessment of HACCP-based FSM systems in Chinese SLDBs and provide
96
evidence on the effectiveness of these FSM systems in SLDBs with a quantitative
97
approach.
98
2. Materials and methods
99
2.1 Methodology
M AN U
SC
RI PT
87
A meta-analysis technique was used to assess the overall effectiveness of
101
HACCP-based FSM systems in Chinese SLDBs. The problem statement, population,
102
intervention, and measured outcomes in this study were described in Table 1. Food
103
businesses were assumed as SLDBs unless they are in the section of dairy, functional
104
foods, food additives, or with a specified staff population over 300. Food products in
105
this study included raw materials, in-process and final food products. Employees’
106
knowledge about food safety & hygiene, food contacting surfaces & operating
107
environments, and food handlers’ hygiene were included in measured outcomes since
108
they were considered as important indexes to FSM levels of food enterprises, as well
109
as their food safety status. Odds ratio (OR), which referred to the degree to which the
110
events of interest (i.e., qualified numbers in this study) occurred in the population,
111
was used as effect size. Primary identification of suitable studies was conducted
112
through literature search for studies reporting effectiveness of HACCP-based FSM
113
systems in Chinese SLDBs, indexed since 1990. Literature search and selection of
114
included studies was performed according to procedures illustrated in Figure 1 (Box
115
A).
AC C
EP
TE D
100
4
ACCEPTED MANUSCRIPT 2.2 Quality assessment and data extraction
117
The quality of the included studies was evaluated and graded using a scoring
118
approach. A particular study was assigned with one point when meeting each of the
119
following criteria: I) year of study was specified; II) both intervention and control
120
included at least 100 monitoring objects in any of the four categories examined for
121
chemical or microbial indexes; III) monitoring indexes were clearly specified; IV)
122
methods for examination of particular monitoring objects as well as the criteria to
123
judge qualification status of the objects were clearly specified; V) unqualified
124
numbers based on single testing indexes and on all testing indexes were presented.
125
Whereas one point was subtracted when a particular study lacked information on
126
qualified numbers based on results of all testing indexes when more than one index
127
was involved for examination. Exclusion of studies was conducted prior to quality
128
assessment, and thus no papers were excluded in this stage according to the result of
129
quality assessment. According to the results, included studies were scored 0 to 5
130
points based on the number of criteria they met, and were assorted to three grades: 1)
131
studies considered of high quality with a total of 4 or 5 points, 2) studies considered
132
of moderate quality with a total of 2 or 3 points, and 3) studies with a total 0 or 1
133
points considered of low quality.
134
Information extracted from all included studies (where applicable) included year of
135
study, type of food businesses, study site (province), monitoring objects (e.g. hygiene
136
of handlers’ hands), as well as data presented in 2 × 2 contingency tables or indirectly
137
in the form of sample sizes associated with qualification rates. The worst case was
138
chosen to calculate ORs if qualified numbers of the objects, examined for more than
139
one testing indexes, were reported only based on single indexes, to avoid
140
overestimation of food safety levels both in intervention and control.
141
2.3 Statistical analysis
142
ORs from 2 × 2 contingency tables were calculated and forest plots generated using
143
statistical software Stata (Version 14, StataCorp LLC, USA), in which all zero
144
prevalence estimates were added with a standard correction of 0.5, with the
AC C
EP
TE D
M AN U
SC
RI PT
116
5
ACCEPTED MANUSCRIPT ‘Meta-Analysis’ package for each category of monitoring objects. Heterogeneity of
146
the studies was determined with Higgin’s test, and random effects models were
147
applied throughout because of the evidence for heterogeneity (I2>50%). Papers
148
included in this study were assumed to be free of publication bias if the P values were
149
above 0.05 when determined with Egger’s test. ORs higher than 1 indicated higher
150
odds of being qualified for monitoring objects when intervention (implementation of
151
HACCP systems) was applied in SLDBs to control potential foodborne hazards
152
and/or enhance food safety.
153
3. Results
154
A total of 2457 papers were retrieved from CNKI, CSPD, and VIP after eliminating
155
duplicates, 68 articles were selected after screening according to title or abstract, and
156
finally 24 studies were included in our research for meta-analysis after full-text
157
screening. Selection of included studies were conducted according to procedures and
158
criteria illustrated in Figure 1 (Box B). All of the 24 included publications were in
159
Chinese, and characteristics of the included studies were summarized in Table 2.
160
Briefly, 12 of the included studies were on effectiveness of HACCP-based FSM
161
systems in catering establishments (Guo, Hu, & Feng, 2013; X.-g. Jiang & Chen,
162
2016; Kou, 2013; Long, Wei, Chen, Lu, & Zhong, 2011; Ma, 2002; Pan, Chen, &
163
Xiao, 2009; Qi, Zhang, Gao, Li, & Li, 2013; Shao, Liu, Jin, & Gao, 2009; J.-s. Wei,
164
2012; L.-m. Wei & Li, 2010; Yang, Zhan, & Ma, 2014; Yu, Zheng, & Guo, 2015;
165
Yuan, Liu, Xin, Shi, & Zhang, 2015; Zhan & Ma, 2014), 4 studies in beverage and
166
bottled water (J.-z. Jiang, 2005; Z.-y. Jiang & Zhao, 2004; Li, Chen, & Chen, 2011;
167
Pan, et al., 2009), 3 studies in ready-to-eat meat (Cao, 2006; Guan, Wang, Bao, Li, &
168
Lv, 1994; Ma, 2002), 2 studies in bakery (S.-q. Liu, Huang, & Li, 2008; Wang, Xu,
169
Lu, & Yu, 1999), 2 studies in frozen and dry food (L.-d. Liu, Gao, Gao, & Liang,
170
1997; Sun, Li, Lin, & Chen, 2007), and 1 study in tableware sanitizing (Zhang &
171
Feng, 2013). Within the 24 included studies investigating effectiveness of
172
HACCP-based FSM systems in SLDBs, 11 studies examined only food products or
173
food contacting surfaces & operating environments, 1 study examined food contacting
AC C
EP
TE D
M AN U
SC
RI PT
145
6
ACCEPTED MANUSCRIPT surfaces & operating environments as well as employees’ knowledge about food
175
safety & hygiene, and the other 12 studies examined at least three of the four
176
mentioned aspects, with 9 of them containing food products.
177
Forest plots on the effects of HACCP-based FSM systems on food safety levels of
178
SLDBs in perspectives of food products, employees’ knowledge about food safety &
179
hygiene, food contacting surfaces & operating environments, as well as handlers’
180
hygiene were presented in Figure 2. Publication bias (P < 0.05) was found for three of
181
the above four meta-analyses, with exception of meta-analysis for association
182
between implementing FSM systems and levels of employees’ knowledge about food
183
safety & hygiene (P = 0.212). However, when studies involving more than 200
184
samples (in intervention or control) were excluded, or only studies focused on
185
catering section were involved, meta-analysis for association between implementing
186
FSM systems and food safety levels of food products was demonstrated as free of
187
publication bias, with P values of 0.103, and 0.078 respectively.
188
The estimated pooled effects [measured as ORs (95% CI)] of having HACCP-based
189
FSM systems implemented compared with controls were 10.58 (6.06, 18.49), 8.66
190
(4.88, 15.35), 6.39 (3.36, 12.18), and 6.65 (4.29, 10.30), respectively, in the
191
perspectives of food products, food handlers’ personal hygiene, employees’
192
knowledge about food safety & hygiene, and food contacting surfaces & operating
193
environments. As demonstrated by the overlapping 95% CIs of ORs (Table 3),
194
subgroup analyses on classification of food products into different categories
195
according to their characteristics did not reveal significant difference in the
196
effectiveness of FSM systems in SLDBs among these subgroups.
197
4. Discussion
198
Considering the wide application and acceptance of HACCP-based FSM systems for
199
ensuring food safety and doubts on their feasibility and effectiveness in SLDBs (Baş,
200
et al., 2007; Maldonado-Siman, Bai, Ramírez-Valverde, Gong, & Lara, 2014; Taylor,
201
2001), in this study a meta-analysis technique was used to assess the effectiveness of
202
HACCP-based FSM systems in Chinese SLDBs. Results of this study demonstrated
AC C
EP
TE D
M AN U
SC
RI PT
174
7
ACCEPTED MANUSCRIPT that HACCP-based FSM systems were effective in SLDBs, as demonstrated by
204
significant improvements (95% CI of ORs on right of 1) in any of the four measured
205
outcomes after implementing FSM systems, despite the heterogeneity of the included
206
studies in meta-analyses. These involved SLDBs varied in characteristics of their
207
products, scales of the businesses, education levels of their employees, degrees of
208
training their employees received, as well as managements’ and employees’ cognition
209
to food safety (Gao, et al., 2002; Ko & Wen-Hwa, 2013; Long, et al., 2011; J.-s. Wei,
210
2012; Zhan & Ma, 2014). Taking into consideration these diversities, it’s reasonable
211
to accept heterogeneity of the included studies in meta-analyses with random effects
212
models to estimate the compiled effectiveness of HACCP-based FSM systems in
213
SLDBs. Combining these diversities with the absence of studies determinately
214
reporting ineffectiveness of HACCP-based FSM systems, it was reasonable to
215
propose that it was not preferentially reporting studies demonstrating effectiveness of
216
FSM systems that led to publication bias demonstrated by results of Egger’s test. This
217
was evidenced by the findings that when studies involving more than 200 monitoring
218
objects, or focusing on non-catering food businesses were excluded, meta-analyses
219
were free of publication bias for association between implementing HACCP-based
220
FSM and odds of qualified food products. Thus, information generated in this study
221
was reliable to understand the effectiveness of HACCP-based FSM systems
222
implemented in SLDBs in ensuring food safety.
223
Improvements in the aspects of food products and food handlers’ hygiene, by
224
adopting HACCP-based FSM systems were statistically significant in all included
225
studies, demonstrated with 95% CIs of ORs locating completely on the right side of
226
the line representing a value of 1, as shown in Figure 2a and 2b. However, despite the
227
significant improvements at overall levels implied by meta-analyses (Figure 2c and
228
2d), 2 and 1 included studies demonstrated insignificant improvement after
229
implementing HACCP-based FSM systems, respectively, in the aspects of food
230
contacting surfaces & operating environments, and employees' knowledge about food
231
safety & hygiene. Large values of pooled ORs, as shown in Figure 2, demonstrated
AC C
EP
TE D
M AN U
SC
RI PT
203
8
ACCEPTED MANUSCRIPT that having HACCP-based FSM systems implemented in SLDBs resulted in a 6 –
233
10-fold increase in the ratios of qualified numbers to unqualified numbers of all
234
testing objects monitored in FSM. However, these findings also implied the huge gaps
235
in these food businesses between their current food safety levels and what would be
236
achieved when adopting HACCP-based FSM systems. Even though all the four
237
mentioned aspects of food safety were covered in FSM, the final goal of FSM was to
238
ensure safety of food products that would be eventually consumed, through
239
improvement in food handlers’ hygiene, employees’ knowledge in related fields, food
240
contacting surfaces of utensils and operating environments, as well as other inputs
241
(e.g. raw materials) that were not in the scope of this study (Djekic, et al., 2016; FAO,
242
1998; Ko & Wen-Hwa, 2013; WHO, 1997). The higher value of pooled OR (10.56) of
243
having HACCP-based FSM systems implemented compared with not having these
244
FSM systems in the perspective of food products than that in the perspective of food
245
handlers’ hygiene (8.66), food contacting surfaces & operating environments (6.65),
246
or employees’ knowledge on food safety & hygiene (6.39) indicated the cumulative
247
effects of FSM systems on safety status of final food products, which were gained
248
through improvements in all aspects of food production by adoption of FSM systems.
249
Despite the effectiveness of HACCP-based FSM systems in ensuring food safety, as
250
well as the associated benefits such as access to new markets (Al-Busaidi, Jukes, &
251
Bose, 2017; Maldonado-Siman, et al., 2014; Psomas & Kafetzopoulos, 2015), a large
252
portion of Chinese food businesses, especially SLDBs, were confronted with various
253
difficulties and barriers (such as financial constraints, lack of prerequisite programs,
254
and frequent staff turnover) for implementing HACCP-based FSM systems (Q. Jiang
255
& Batt, 2016; x.-d. Liu, Duan, & She, 2007). Uncertainty about effectiveness of
256
HACCP-based FSM systems in SLDBs also led to their reluctance to adopt these
257
FSM systems (Baş, et al., 2007; Cerf & Donnat, 2011). Significant improvement in
258
food safety levels by adopting HACCP principles in production of roast chicken
259
(Guan, et al., 1994; Ma, 2002) and school canteens (X.-g. Jiang & Chen, 2016; Long,
260
et al., 2011; Shao, et al., 2009; L.-m. Wei & Li, 2010; Yang, et al., 2014; Yu, et al.,
AC C
EP
TE D
M AN U
SC
RI PT
232
9
ACCEPTED MANUSCRIPT 2015) should reduce or eliminate doubts on the effectiveness of such FSM systems in
262
SLDBs. The estimated random effects ORs (Figure 2) provided evidence of the
263
effectiveness of HACCP-based FSM systems in SLDBs at an overall level, taken into
264
account the diversities of these food businesses in aspects such as business scales,
265
characteristics of the products, production environments, expertise and skills in food
266
safety management, education levels of employees, degrees of personnel training, and
267
knowledge about food safety & hygiene, as well as difficulties and barriers SLDBs
268
were commonly confronted with.
269
It had been suggested that efforts from various stakeholders were required to promote
270
implementing HACCP-based FSM systems in SLDBs, including but not limited to
271
technical
272
communication between governments and businesses and between businesses,
273
infrastructures and facilities assured by governments and businesses, sector-specific
274
industry guides developed industry associations, and provision of basic but adequate
275
knowledge about food safety & hygiene (WHO, 1999, 2006). Considering the
276
existence of numerous self-employed food businesses, as well as low level of food
277
handlers’ knowledge about food safety & hygiene (Gong, Wang, Yang, & Bai, 2016),
278
it was necessary and critical to develop simplified versions of HACCP-based FSM
279
systems for SLDBs where it is not feasible to completely implement FSM systems as
280
recommended for scaled and industrialized food businesses (W. Dzwolak, 2014, 2016;
281
Taylor, 2008). In such simplified HACCP, attention should be focused on potential
282
hazards and CCPs to control them (Waldemar Dzwolak, 2018). Such simple versions
283
would be more easily to he accepted by these numerous SLDBs, which would
284
certainly help ensure food safety of their product and thus protect consumer health.
285
In conclusion, implementing HACCP-based FSM systems in Chinese SLDBs
286
significantly increased the ratios of qualified to unqualified numbers of testing
287
objectives monitored in FSM; and evidence demonstrated that HACCP-based FSM
288
systems were effective in SLDBS. However, impedance of barriers, constraints, and
289
difficulties to adopting HACCP-based FSM systems shouldn’t be ignored when
governments
and
SC
from
industry
associations,
adequate
AC C
EP
TE D
M AN U
support
RI PT
261
10
ACCEPTED MANUSCRIPT 290
promoting application of FSM systems in SLDBs.
291
Acknowledgement
293
This work was supported by China Central Universities Fundamental Research Funds
294
(DUT16RC(3)025. and DUT16QY37.).
AC C
EP
TE D
M AN U
SC
RI PT
292
11
ACCEPTED MANUSCRIPT 295
Reference:
296
Al-Busaidi, M. A., Jukes, D. J., & Bose, S. (2017). Hazard analysis and critical control point
297
(HACCP) in seafood processing: An analysis of its application and use in regulation in
298
the Sultanate of Oman. Food Control, 73, 900-915. Baş, M., Yüksel, M., & Çavuşoğlu, T. (2007). Difficulties and barriers for the implementing of
300
HACCP and food safety systems in food businesses in Turkey. Food Control, 18(2),
301
124-130.
SC
RI PT
299
CAC. (2009). Codex Alimentarius: Food hygiene, basic texts. Fourth Edition. . World Health
303
Organization (WHO), Food and Agriculture Organization of the United Nations (FAO),
304
Rome, 2009, 23-33.
M AN U
302
Cao, Q.-j. (2006). Assessment of Effects of Applying HACCP on Controlling Microbial
306
Contamination of Unpackaged Cooked Meat Products. Chinese Journal of Public
307
Health Management, 22(5), 435-436.
TE D
305
Cerf, O., & Donnat, E. (2011). Application of hazard analysis – Critical control point (HACCP)
309
principles to primary production: What is feasible and desirable? Food Control, 22(12),
310
1839-1843.
EP
308
Djekic, I., Kuzmanović, J., Anđelković, A., Saračević, M., Stojanović, M. M., & Tomašević, I.
312
(2016). Effects of HACCP on process hygiene in different types of Serbian food
AC C
311
313 314
establishments. Food Control, 60, 131-137.
Dzwolak, W. (2014). HACCP in small food businesses – The Polish experience. Food Control,
315 316
36(1), 132-137. Dzwolak, W. (2016). Documenting HACCP in a small restaurant – a practical approach.
12
ACCEPTED MANUSCRIPT 317 318
Quality Assurance & Safety of Crops & Foods, 9(2), 1-8. Dzwolak, W. (2018). Developing a hazard analysis worksheet in a small food business with the
319
application of a T-shaped matrix diagram. Food Control, 87, 180-191. FAO. (1998). Food quality and safety systems: a training manual on food hygiene and the
321
Hazard Analysis and Critical Control Point (HACCP) system. Food Quality and
322
Standards Service Food and Nutrition Division, Food and Agriculture Organization of
323
the United Nations, Rome.
SC
Gao, P., Ma, F.-k., & Yan, Z.-c. (2002). Application and Reseach of HACCP Management in
325
M AN U
324
RI PT
320
China and other Countries. Chinese Journal of Public Health, 18(10), 1262-1264. Garayoa, R., Vitas, A. I., Díez-Leturia, M., & García-Jalón, I. (2011). Food safety and the
327
contract catering companies: Food handlers, facilities and HACCP evaluation. Food
328
Control, 22(12), 2006-2012.
329
TE D
326
Gong, S., Wang, X., Yang, Y., & Bai, L. (2016). Knowledge of food safety and handling in
330
households: A survey of food handlers in Mainland China. Food Control, 64, 45-53. Gonzales-Barron, U., Cadavez, V., Sheridan, J. J., & Butler, F. (2013). Modelling the effect of
332
chilling on the occurrence of Salmonella on pig carcasses at study, abattoir and batch
AC C
EP
331
333
levels by meta-analysis. International Journal of Food Microbiology, 163(2–3),
334 335
101-113.
Gonzalesbarron, U., & Butler, F. (2011). The use of meta-analytical tools in risk assessment
336
for food safety. Food Microbiology, 28(4), 823-827.
337
Guan, W.-j., Wang, M.-q., Bao, D.-y., Li, S.-q., & Lv, L.-t. (1994). Assessment of Effects of
338
Applying HACCP in Production and Vendition of Roast Chiken in Zhongshan District, 13
ACCEPTED MANUSCRIPT 339
Dalian
Chinese Journal of Food Hygiene(4), 1-3.
Guo, S.-y., Hu, R.-f., & Feng, W. (2013). Study on the Application of Hazard Analysis and
341
Critical Control Point Management System in Sanitary Supervision of University
342
Canteens. Practical Preventive Medicine(9), 1097-1099.
343
Hedges, L. V. (2015). The early history of meta-analysis. Research Synthesis Methods, 6(3),
344
284.
Jevšnik, M., Hlebec, V., & Raspor, P. (2006). Meta-analysis as tool for barriers identification
SC
345 346
during HACCP implementation to improve food safety. Acta Alimentaria(3), 319-353.
M AN U
347
RI PT
340
Jiang, J.-z. (2005). Investigation on Effects of HACCP in the Production of Purified Drinking
348
Water. Occupation and Health, 21(2), 236-237.
Jiang, Q., & Batt, P. J. (2016). Barriers and benefits to the adoption of a third party certified
350
food safety management system in the food processing sector in Shanghai, China.
351
Food Control, 62, 89-96.
352
TE D
349
Jiang, X.-g., & Chen, Y.-z. (2016). Study on Food Safety Intervention Strategy of Catering
353
EP
Jiang, Z.-y., & Zhao, X.-m. (2004). Investigation on Analysis and Preventive Measures for
AC C
354
distribution enterprise. Chinese Journal of Public Health Management(2), 247-249.
355
Hygiene Quality of Bottled Drinking Water in City of Linyi. Occupation and Health,
356
20(9), 64-65.
357
Ko, Wen-Hwa. (2013). The relationship among food safety knowledge, attitudes
358
and;self-reported HACCP practices in restaurant employees. Food Control, 29(1),
359
192-197.
360
Kou, L.-q. (2013). Applied Research in Effects of HACCP on Food safety of High School 14
ACCEPTED MANUSCRIPT 361
Canteens in a County Guide of China Medicine, 11(9), 36-38. Li, Z., Chen, J., & Chen, J.-q. (2011). A Study on Effects of HACCP on Controlling
363
microorganisms and Pesticide residues in Oolong Tea. Science & Technology for
364
Development(s1), 182-183.
RI PT
362
Liu, L.-d., Gao, J.-s., Gao, Y., & Liang, Y.-b. (1997). Assessment of Management Effects of
366
Implementing Hazard Analysis and Critical Control Points in Frozen Food Production
367
Preventive Medicine Tribune(1), 56-57.
Liu, S.-q., Huang, L.-z., & Li, Q. (2008). Evaluation of Application of HACCP System in
369
M AN U
368
SC
365
Process Production of Cake and Paste. China Tropical Medicine, 8(9), 1660-1661. Liu, x.-d., Duan, Q.-j., & She, R.-p. (2007). Problem Existing and Suggestions on
371
Implementation and Establishment of HACCP System in China. Food Science, 28(5),
372
369-373.
TE D
370
Long, P., Wei, X.-w., Chen, Y.-b., Lu, Z.-g., & Zhong, M. (2011). Analysis on the effectiveness
374
of establishing HACCP system in student canteens in High School District. Chinese
375
Journal of Food Hygiene, 23(2), 167-169. Ma, W.-m. (2002). Investigation of Effects of Applying HACCP Methodology on Controlling
AC C
376
EP
373
377
Hygiene of Cookded Bitern Food. Medical Journal of Communications, 16(3),
378
297-297.
379
Maldonado-Siman, E., Bai, L., Ramírez-Valverde, R., Gong, S., & Lara, R. D. (2014).
380
Comparison of implementing HACCP systems of exporter Mexican and Chinese meat
381
enterprises. Food Control, 38(4), 109-115.
382
Noordhout, C. M. D., Devleesschauwer, B., Angulo, F. J., Verbeke, G., Haagsma, J., Kirk, M., 15
ACCEPTED MANUSCRIPT 383
Havelaar, A., & Speybroeck, N. (2014). The global burden of listeriosis: a systematic
384
review and meta-analysis. Lancet Infectious Diseases, 14(11), 1073-1082. Pan, H.-j., Chen, S.-p., & Xiao, L.-j. (2009). Effect Assessment of Applying HACCP in
386
Inspection and Quarantine Supervision of Exported Health Protection Tea. Chinese
387
Frontier Health Quarantine(1), 45-47.
388
RI PT
385
Psomas, E. L., & Kafetzopoulos, D. P. (2015). HACCP effectiveness between ISO 22000
389
SC
certified and non-certified dairy companies. Food Control, 53, 134-139.
Qi, R.-j., Zhang, X.-j., Gao, G.-p., Li, X., & Li, x.-d. (2013). Combination Application of HACCP
391
and 6S Method in School Canteens Management. Chinese Journal of School Health,
392
34(12), 1475-1478.
393
M AN U
390
Ramalho, V., Moura, A. P. D., & Cunha, L. M. (2015). Why do small business butcher shops
394
TE D
fail to fully implement HACCP? Food Control, 49(49), 85-91. Shao, X.-n., Liu, Y.-q., Jin, X.-m., & Gao, R.-q. (2009). Influence and Effect on Food Hygiene
396
Management after Establishing HACCP System in Dining Enterprises. Chinese
397
Journal of Public Health Management, 25(3), 251-253. Shih, K.-m., & Wang, W.-k. (2011). Factors influencing HACCP implementation in Taiwanese
AC C
398
EP
395
399 400
public hospital kitchens. Food Control, 22(3–4), 496-500.
Sun, S.-l., Li, Z., Lin, M.-x., & Chen, J.-q. (2007). Effects of HACCP System on Controlling
401
Total Bacterial Colonies in Frozen-dried Green Union. Journal of
402 403
Henan University of
Science and Technology: Natural Science, 28(5), 51-53. Sutton, A. J., Abrams, K. R., & Jones, D. R. (2001). An illustrated guide to the methods of
404
meta-analysis. Journal of Evaluation in Clinical Practice, 7(2), 135-148. 16
ACCEPTED MANUSCRIPT 405
Taylor, E. (2001). HACCP in small companies: benefit or burden? Food Control, 12(4),
406 407
217-222. Taylor, E. (2008). A new method of HACCP for the catering and food service industry. Food
408
RI PT
Control, 19(2), 126-134. Trafiałek, J., Lehrke, M., L ü cke, F. K., Kołożyn-Krajewska, D., & Janssen, J. (2015).
410
HACCP-based procedures in Germany and Poland. Food Control, 55, 66-74.
411
Violaris, Y., Bridges, O., & Bridges, J. (2008). Small businesses – Big risks: Current status and
SC
409
412
M AN U
413
future direction of HACCP in Cyprus. Food Control, 19(5), 439-448.
Wallace, C. A., Holyoak, L., Powell, S. C., & Dykes, F. C. (2014). HACCP – The difficulty with
414
Hazard Analysis. Food Control, 35(1), 233-240.
Wallace, C. A., Sperber, W. H., & Mortimore, S. E. (2010). Food Safety for the 21st Century:
416
Managing HACCP and Food Safety throughout the Global Supply Chain:
417
Wiley-Blackwell, Oxford, Iowa.
TE D
415
Wang, C.-m., Xu, X.-m., Lu, L.-h., & Yu, M.-h. (1999). Assessment of Effects of Applying
419
HACCP in Production and Vendition of Decorative Cake Zhejiang Journal of
420
Preventive Medicine(3), 28-31.
AC C
421
EP
418
Wei, J.-s. (2012). Investigation of Effects of HACCP Establishment in
422 423
Nanchang Railway. World Health Digest Medical Periodieal(52), 10-11.
Wei, L.-m., & Li, H.-j. (2010). Application of HACCP in Hygiene Management of School
424 425
Apartment Canteen of
Canteens. Journal of Applied Preventive Medicine(2), 123-124. WHO. (1997). HACCP: Introducing the Hazard Analysis and Critical Control Point System.
426
WHO/FSF/FOS/97.2, 17
Geneva,
World
Health
Organization.
ACCEPTED MANUSCRIPT 427
http://www.who.int/iris/handle/10665/63610. WHO. (1999). Strategies for implementing HACCP in small and/or less developed businesses.
429
Report of a WHO consultation in collaboration with the Ministry of Health, Welfare and
430
Sports, The Netherlands, The Hague, 16-19 June.
431
RI PT
428
WHO. (2006). FAO/WHO guidance to governments on the application of HACCP in small
432
and/or less-developed food businesses. FAO
Food & Nutrition Paper, 86.
Xavier, C., Gonzales-Barron, U., Paula, V., Estevinho, L., & Cadavez, V. (2014). Meta-analysis
434
of the incidence of foodborne pathogens in Portuguese meats and their products.
435
Food Research International, 55(2), 311-323.
M AN U
SC
433
Yang, W.-c., Zhan, Z.-p., & Ma, H.-l. (2014). Analysis of Effects of HACCP System
437
Establishment in Canteens of Universities and Colleges. Medical Aesthetics and
438
Beauty(5), 554-554.
TE D
436
Yu, X.-m., Zheng, G.-w., & Guo, H.-m. (2015). Application research of HACCP combined with
440
six T practical management model in food safety of school canteen. China Modern
441
Medicine(30), 140-143.
Yuan, J., Liu, H., Xin, X., Shi, S.-g., & Zhang, G.-j. (2015). Application of HACCP in Preparing
AC C
442
EP
439
443
Hot Dishes in Catering Establishments Journal of Preventive Medicine of Chinese
444 445
People's Liberation Army(1), 39-40.
Zhan, Z.-p., & Ma, H.-l. (2014). Effectiveness of HACCP Implementation in Canteens of
446
Primary and Secondary Schools. Biotech World(10), 74-74.
447
Zhang, D.-s., & Feng, S.-x. (2013). The Implementation and Effects of HACCP Management
448
Principle in the Unit Concentratedly Disinfected Tableware. Modern Preventive 18
ACCEPTED MANUSCRIPT 449
Medicine, 40(15), 2925-2928.
AC C
EP
TE D
M AN U
SC
RI PT
450
19
ACCEPTED MANUSCRIPT Tables Table 1. Problem statement, population, intervention, and measured outcome specified for meta-analysis. Description
problem statement
Investigation of the overall effectiveness of HACCP-based FSM systems in of Chinese
population
SLDBs located in mainland of China
intervention
Implementation of HACCP-based systems
measured outcome
Qualified numbers of : 1) food products
SC
SLDBs
RI PT
Elements
2) employees’ knowledge about food safety & hygiene 3) food contacting surfaces & operating environments
AC C
EP
TE D
M AN U
4) food handlers’ hygiene
ACCEPTED MANUSCRIPT Table 2. Characteristics of included studies.
W.-J Guan, 1994 C.-M Wang, 1999 W.-M Ma, 2002 Z.-Y Jiang, 2004 J.-J Jiang, 2005
Study of
Categories of Food
Monitoring
Total
study
province
business
objects involved
points
1992-1993 1997
Liaoning Zhejiang
2000
Jiangsu
2002-2003 2003
RTE meat Bakery RTE meat
Shandong Jiangsu
Drinks/bottled water Drinks/bottled water
Q.-J Cao, 2006
2005
Jiangsu
RTE meat
S.-L Sun, 2007
NA
Fujian
Frozen/dry food
L.-D Liu, 1997 L.-Q Kou, 2013 S.-Q Liu, 2008 X.-N Shao, 2009 D.-S Zhang, 2013 J Yuan, 2015
Guangdong
NA
Fujian
1995
Drinks/bottled water
Shandong
NA
Sichuan
2005-2007 NA
NA
Guangdong
Guangxi Gansu
Frozen/dry food Catering
Shandong
2009-2011
Drinks/bottled water
Bakery Catering
2
F
a
3
F
a
5
F
a
3
F
a
4
F
a
3
Fa
3
F
a
1
F
a
1
a
b
c
3
a
b
c
2
a
b
c
3
SC
Z Lin, 2011
NA
M AN U
H.-J Pan, 2009
F
a
RI PT
Study ID
Year of
Tableware sanitizing Catering
F , H , S_E
F , H , S_E
F , H , S_E S_E
c
S_E
c
2 2
b
c
d
H , S_E , K a
b
c
2 d
4 2
P Long, 2011
NA
Xinjiang
Catering
F , H , S_E , K
S.-Y Guo, 2013
NA
Hunan
Catering
F a, H b, S_E c , K d
W.-C Yang, 2014 X.-G Jiang, 2016 X.-M Yu, 2015 R.-J Qi, 2013 L.-M Wei, 2010 Z.-P Zhan, 2014 a
NA
Jiangxi
2013
Liaoning
NA 2014 NA 2009 2014
Zhejiang
Catering
Catering
Catering
TE D
J.-S Wei, 2012
Jiangxi
Anhui
Guangxi
Liaoning
Catering
Catering Catering
Catering
AC C
EP
Final food products; Handlers’ personal hygiene; c Contacting surfaces of utensils and operating environments; d Employees’ knowledge about food safety and hygiene. b
a
b
c
d
3
a
b
c
d
4
F , H , S_E , K F , H , S_E , K b
c
d
2
b
c
d
2
H , S_E , K H , S_E , K a
c
F , S_E , K a
b
F , H , S_E c
S_E , K
d
d
c
1 4 2
ACCEPTED MANUSCRIPT Table 3. Results of the subgroup analyses examining the association of HACCP systems with food safety levels in the perspective of food products. Random effects pooled OR (95% CI) 7.58 (1.56, 36.88) 45.47 (0.88, 2351.65) 14.68 (2.34, 92.08) 12.67 (4.68, 34.35) 7.09 (2.52, 19.96) 9.67 (5.53, 16.92)
I2 89.10% 86.00% 73.40% 55.00% 88.70% 84.30%
AC C
EP
TE D
M AN U
SC
RTE meat Bakery Drinks/bottled water Frozen/dry food Catering Overall
Number of studies 3 2 4 2 6 17
Weight (%) 20.38 9.32 17.16 13.64 39.49 100
RI PT
Food category
ACCEPTED MANUSCRIPT Figure captions Figure 1. Literature search strategy and selection criteria for identifying suitable studies (Box A), and resulting papers included for meta-analysis (Box B). Figure 2. Meta-analysis of the association of HACCP-based FSM systems with food
RI PT
safety levels in the perspective of (a) food products, (b) food handlers’ hygiene, (c) food contacting surfaces & operating environments, and (d) employees' knowledge about food safety & hygiene.
Databases of CNKI, CSPD, and VIP
Primary
search under study Chinese terms for food safety/effectiveness/ hazard control
Box A
Requirements:
1) to be an original article;
M AN U
studies
SC
Figure 1.
2) microbial or chemical indexes involved in intervention and
Suitable studies
control; 3) ≥ 20 test samples involved;
TE D
Studies identified
2457 papers after eliminating
AC C
68 articles
EP
duplicates
Full-text screening
24 studies finally included for meta-analysis
Box B
2389 papers excluded according to title or abstract
17 studies (quantitative information on effectiveness of FSM systems was not found) 15 studies (sample size or clear differentiation of intervention and control was not specified ) 9 studies (data not presented in the required format) 2 studies (focused on occurrence of epidemic diseases in poultry farms) 1 study (reported effectiveness of HACCP in water supply system)
ACCEPTED MANUSCRIPT Figure 2.
% Weight
OR (95% CI)
W.-J Guan, 1994
23.27 (7.84, 69.14)
6.36
C.-M Wang, 1999
11.96 (4.29, 33.40)
6.54
W.-M Ma, 2002
2.36 (1.61, 3.46)
8.08
Z.-Y Jiang, 2004
46.74 (2.87, 761.80)
2.71
J.-J Jiang, 2005
25.72 (1.51, 439.38)
2.64
Q.-J Cao, 2006
9.60 (2.62, 35.21)
5.77
S.-L Sun, 2007
24.27 (7.14, 82.47)
5.98
H.-J Pan, 2009
30.34 (4.06, 226.55)
4.01
Z Lin, 2011
3.32 (1.84, 6.00)
7.67
8.62 (4.52, 16.41)
7.55
17.33 (8.73, 34.41)
7.45
237.47 (14.66, 3846.43)
2.71
18.55 (6.97, 49.36)
6.67
17.25 (5.19, 57.37)
6.04
4.67 (1.65, 13.24)
6.49
2.06 (1.39, 3.03)
8.06
13.95 (3.17, 61.34)
5.27
10.58 (6.06, 18.49)
100.00
SC
RI PT
Study
L.-D Liu, 1997 L.-Q Kou, 2013
P Long, 2011 S.-Y Guo, 2013 W.-C Yang, 2014 R.-J Qi, 2013 L.-M Wei, 2010 Overall
(I-squared = 84.5%, p = 0.000)
NOTE: Weights are from random effects analysis
1
3846
TE D
.00026
M AN U
S.-Q Liu, 2008
AC C
EP
a
ACCEPTED MANUSCRIPT
OR (95% CI)
% Weight
L.-D Liu, 1997
61.50 (17.26, 219.11)
8.90
L.-Q Kou, 2013
19.67 (6.27, 61.65)
9.71
S.-Q Liu, 2008
64.78 (8.80, 477.06)
5.41
J Yuan, 2015
7.67 (2.61, 22.54)
10.14
P Long, 2011
3.94 (1.43, 10.81)
10.61
S.-Y Guo, 2013
5.03 (1.83, 13.84)
10.58
SC
RI PT
Study
W.-C Yang, 2014 X.-G Jiang, 2016
L.-M Wei, 2010 Overall
(I-squared = 72.9%, p = 0.000)
NOTE: Weights are from random effects analysis
.0021
1
TE D EP
3.85 (1.82, 8.11)
12.47
3.62 (2.43, 5.40)
14.70
22.00 (2.05, 236.05)
4.26
8.66 (4.88, 15.35)
100.00
477
b
AC C
13.23
M AN U
X.-M Yu, 2015
7.43 (3.93, 14.06)
ACCEPTED MANUSCRIPT
% Weight
OR (95% CI)
D.-S Zhang, 2013
4.54 (2.98, 6.90)
8.59
Jian Yuan, 2015
14.75 (0.81, 269.34)
1.83
J.-S Wei, 2012
19.45 (8.26, 45.76)
6.88
L.-D Liu, 1997
24.11 (12.09, 48.08)
7.57
L.-M Wei, 2010
12.46 (4.87, 31.86)
6.53
L.-Q Kou, 2013
11.80 (5.56, 25.01)
7.32
P Long, 2011
7.37 (3.27, 16.62)
7.06
R.-J Qi, 2013
2.14 (1.57, 2.92)
8.91
S.-Q Liu, 2008
22.82 (5.33, 97.74)
4.60
S.-Y Guo, 2013
4.44 (1.95, 10.11)
7.01
SC
RI PT
Study
W.-C Yang, 2014 X.-G Jiang, 2016
X.-M Yu, 2015 Z.-P Zhan, 2014 Overall
(I-squared = 82.3%, p = 0.000)
NOTE: Weights are from random effects analysis
.00371
1
TE D EP
6.95
2.62 (1.16, 5.93)
7.05
4.09 (2.71, 6.19)
8.62
5.09 (0.89, 29.27)
3.75
6.65 (4.29, 10.30)
100.00
269
c
AC C
7.33
2.77 (1.20, 6.40)
M AN U
X.-N Shao, 2009
7.33 (3.47, 15.52)
ACCEPTED MANUSCRIPT
OR (95% CI)
% Weight
J Yuan, 2015
3.11 (1.65, 5.87)
12.18
P Long, 2011
7.72 (0.92, 64.69)
5.51
S.-Y Guo, 2013
12.25 (2.78, 53.99)
7.97
J.-S Wei, 2012
6.26 (2.48, 15.82)
10.74
W.-C Yang, 2014
17.94 (10.11, 31.84)
12.45
X.-G Jiang, 2016
3.07 (2.37, 3.97)
13.51
SC
RI PT
Study
X.-M Yu, 2015 R.-J Qi, 2013
Overall
(I-squared = 90.1%, p = 0.000)
NOTE: Weights are from random effects analysis
.0155
12.82
28.46 (14.84, 54.61)
12.10
64.7
TE D
d
EP
1.78 (1.10, 2.88)
6.39 (3.36, 12.18)
1
AC C
12.73
M AN U
Z.-P Zhan, 2014
5.55 (3.35, 9.20)
100.00
ACCEPTED MANUSCRIPT Highlights: HACCP-based systems in Chinese SLDBs significantly improved their food safety levels; Pooled ORs were in range of 6.39-10.58, regarding different aspects of food safety;
AC C
EP
TE D
M AN U
SC
RI PT
Evidence on effectiveness should encourage application of such FSM systems in SLDBs.