ABSTRACT
Objectives Courier jobs are characterised as high demand, low control with lack of support, meaning that job burnout is a serious concern. Burnout could have negative influences on safety, performance, turnover and work engagement. This paper aims to review existing evidence about the level of and risk factors for burnout among couriers.
Design Systematic review and narrative synthesis
Primary and secondary outcome measures Burnout, risk factors
The review followed PRISMA to search studies published in English and Chinese. English databases included MEDLINE, EMBASE, PsycINFO, Cochrane Library, Web of Science (Core Collection), Open Science Framework and Google Scholar. Chinese databases included CNKI, WANFANG data, Sino Med and CQVIP.
Results The search yielded nine papers, of which eight studied Chinese couriers and one Malaysian. Two of them were published in English language and seven were published in Chinese. Seven studies reported moderate to high levels of burnout, such as mean scores of 2.45 on a 1-5 scale, 4.02 on a 0-7 scale, and 49 where the total score is between 15 and 75. The included studies demonstrated that burnout is associated with low income, work stress, physical demands, emotional demands and organizational constraints. Some of them also measured factors that ameliorate the negative effect of work stress on burnout, such as job resources, decision latitude, social support and organizational support. Individual psychological processes, such as emotional regulation and psychological empowerment also play a role in ameliorating or exacerbating burnout. Level of burnout also varies significantly among different socio-demographic groups, such as age, income and marital status.
Conclusions Job burnout is moderate to high among couriers, but there is a dearth of literature in the English language. Interventions to prevent or reduce burnout are currently lacking but could include increasing job autonomy and offering more organizational support.
PROSPERO registration number: CRD42021247644
Strengths and limitations of this study
▸ Given the importance of China in global e-commerce, comprehensive search strategies were employed to review publications in both English and Chinese languages
▸ Outcome measures investigated have been clearly defined and referenced.
▸ This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses reporting guideline.
▸ However, due to the heterogeneity of included studies, it was not meaningful to perform meta-analysis.
BACKGROUND
The COVID-19 pandemic has highlighted the essential roles of healthcare workers and couriers in maintaining public health and keeping economies running during lockdowns. It is also known that occupations such as healthcare and last mile logistics are physically and psychologically demanding, a known factor for job burnout.[1, 2] However, while research about healthcare workers’ burnout has supplied good quality evidence for intervention development,[3-5] understanding about the level of burnout and relevant risk factors among couriers is lacking. The aim of the present research is to review systematically the literature on risk factors and burnout among couriers with a view to identifying targets for intervention.
The burnout issue among couriers has come to our attention not just because of the surged demand for home deliveries during the COVID-19 pandemic, but also the rapid expansion of global e-commerce in the last a few decades.[6] China has become the global leader of e-commerce and the fastest growing region. In 2019, China’s online retail sales reached $1,625.00 billion, a figure that grew by 16.5% from 2018 (MofCom.gov.cn). The phenomenal growth has triggered a large amount of research published in Chinese. The importance and resilience of the sector is supported by a large and growing work population, parcel and takeaway couriers who connect the last mile of the logistics. However, current working practices may present multiple mental health risks.[7] A recent road safety study from China revealed that 76.5% of interviewed couriers (n=480) had been involved in a traffic crash at least once, while the average length of staying in this occupation was only about 1.5 years.[8] Research about overwork among Chinese takeaway couriers also found that 41.74% respondents (n=1114) were at high risk of overworking (danger zone), 35.91% were at very high risk of overworking (acute danger zone) and only 7.9% were considered risk free (safe zone).[9] The Chinese language literature may therefore be a key area to search for relevant research on burnout in couriers and will go some way to extending the reach of previous systematic reviews of burnout, which to date have focused on English language databases, effectively ignoring vast sections of the global economy.
Even in European countries, such as the UK where health and safety standards are deemed high, 42% of surveyed couriers and taxi drivers (n=231) said they had been involved in a collision where vehicle had been damaged and 10% said that someone had been injured, usually themselves.[10] Parcel and food delivery is an essential part of the gig economy, where workers are categorised as self-employed and the platforms do not provide the traditional employer responsibilities, such as health, safety and other employee benefits. While this mode of working has the benefit of flexibility, analysis of large scale survey data in North America has shed additional light on the relationship between gig workers’ socio-economic status and work stress, that is the dependency on platforms (ie as the main job and only source of income) and financial strain exacerbate the mental health penalties associated with platform work.[11] These studies, although they do not focus directly on couriers’ burnout, depict a stressful work environment. Multiple drivers of burnout are at work, including excessive work load and time pressure, low job control, low safety standard, lack of organizational support and other socio-economic factors. Therefore, it is important that we have a clear understanding about the situation, so that relevant research can be developed to evaluate and mitigate high risks. This review aims answer the questions by reviewing evidence from the Chinese and English language literatures. Our review questions are:
Q1 What is the level of job burnout among couriers?
Q2 What are the main risk factors of job burnout among couriers?
There are several popular definitions of burnout[12] and this review adopts the one summarized by Schaufeli and Greenglass[13]: burnout is the state of physical, emotional and mental exhaustion that results from long-term involvement in work situations that are emotionally demanding. It has three dimensions: emotional exhaustion, depersonalization, and reduced personal accomplishment.[14] Please note the “depersonalization” dimension is also named “cynicism” or “disengagement” in some literature. They appeared to be referring to the same construct, hence this paper adopt these dimension names interchangeably. The similar applies to the “reduced personal accomplishment” dimension that is often named “low professional efficacy” or “low sense of achievement”.
METHOD
Search strategy
The initial searches were completed in April 2021, with additional searches carried out in September 2021 to catch the latest publications.
We searched the following electronic databases for English literature:
MEDLINE, EMBASE, PsycINFO and the Cochrane Library (from their respective inception dates to the current date) using OVID platform.
Web of Science (Core Collection)
Pre-print database: Open Science Framework
Grey Literature database: Google Scholar
We also searched the following Chinese electronic databases for Chinese literature:
WANFANG data
Sino Med
VIP (http://www.cqvip.com/)
For search terms on OVID, we started with Exp Burnout to catch as much as possible relevant terms, in addition, we used (burnout OR stress OR job strain OR fatigue OR exhaustion OR tired OR tiredness OR tiring OR weariness OR worn out) to include possibly related terms in social and health sciences. For the occupation, we used (courier* or deliverym*n or delivery worker* or delivery driver* or takeaway delivery) to catch all relevant population in focus. At the end, we used (work OR occupational OR job) to limit results to studies that are work-related. A full search strategy on all databases is available as supplemental material.
We also hand searched the reference lists of all the included studies, and of some excluded studies. We have reviewed peer-reviewed papers and postgraduate theses that are published in English or Chinese, with the full text available. No restriction on publication dates were applied.
Inclusion/Exclusion criteria
Reviews, RCT, CRT, observational and qualitative studies are all eligible.
Inclusion criteria: Studies assessing levels of occupational burnout in a population of couriers/delivery workers or reporting associated risk factors.
Exclusion criteria: Studies not focussing on work-related mental ill health or not in a sample of delivery workers or couriers; Conference abstracts; Commentaries; Studies not in English or Chinese language.
Couriers including parcel/goods couriers and takeaway food couriers who conduct the “last mile” deliveries will be the focus of the review. Hence, studies about postmen, public transport drivers and long-haul truck or lorry drivers were excluded.
Most studies used Maslach Burnout Inventory (MBI) to measure burnout. Due to the variation in relation to cut-off points of burnout in existing literature, we generally considered scores above the middle point as an indication of high burnout.
Patient and Public Involvement
Patient consent for publication is not required.
Review process
We followed a review process adapted from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses[15] (see Figure 1 PRISMA diagram below). A review protocol was pre-published on PROSPERO (ID: CRD42021247644). It was deemed meta-analysis or subgroup analysis was not meaningful due to the varied reporting standards and measurement tools used in the included studies, and qualitative synthesis was the most appropriate method. Five reviewers (HW, SGL, TO, MKY and SL) formed two teams to carry out systematic reviews in Chinese and English. SGL, MKY and SL were responsible for Chinese review and HW and TO English. Both teams went through the same steps, including searching the databases, screening the results, data extraction and quality assessment. The protocol was reviewed by and discussed among all the co-authors, with feedback actively taken into account to ensure quality of the process. HW and SGL cross checked both languages at each steps to maintain consistent standard. When both reviews were completed, HW translated data extracted from the Chinese review into English. The abstracts of all included Chinese studies were also translated into English by HW to enable discussions among all co-authors.
PRISMA diagram
For the step of title/abstract screening, we used a standardised title and abstract form. The form was developed in English by all five reviewers. Each team conducted a pilot exercise with the review team screening the same 10 abstracts to calibrate. Any conflicts were discussed and if disagreement remained, we would refer to the review team for further discussion until consensus was reached. Two reviewers of each team then screened the remaining abstracts independently with conflict resolution through discussion and referral to a third reviewer if necessary.
For full-text review, we used a standardised full text form. The form was developed in English by all five reviewers. Each team then conducted a pilot exercise with the same five articles reviewed by the review team to calibrate and test the form, following the same conflicts resolution process as depicted above. Two reviewers of each team then reviewed all full text articles independently with conflict resolution through discussion and referral to a third reviewer if necessary.
During data extraction, in each team, two reviewers independently extracted data using a piloted form. Again the form was developed in English by all five reviewers. Any discrepancies were checked, discussed and resolved. The same process was followed for risk of bias assessments. A version of the Newcastle Ottawa Scale (NOS) that was adapted for cross-sectional surveys were used for quality assessment.[16]
We will present a narrative synthesis of the data to answer the research questions. Tables were produced to highlight key findings based upon the research questions and outcomes of interest.
RESULTS
Search results
Databases searches returned 676 results after deduplication, 497 in English and 179 in Chinese. Following title and abstract screening 36 papers were included for full-text review, after which 17 papers were further excluded, leaving 19 papers for data extraction. During data extraction a further 10 papers were excluded as they did not report burnout as an outcome of interest. At the end 9 were included for reporting and quality assessment. All included studies are cross-sectional surveys. Figure 1 illustrates the number of studies excluded and reasons of exclusion at each stage.
Study characteristics
Table 1 describes the characteristics of studies included in this review from both Chinese and English language searches.
Study characteristics
Across databases, eight out of nine included studies were from China, with one coming from Malaysia. Two were published in English and seven were in Chinese. All included studies are small to mid-sized cross-sectional surveys (n=105-1,425, Mean=412). Samples were primarily male (−23% female on average), with the most frequent age bands reported being -20-40, one study did not report age or sex/gender statistics.
Quality assessment
Table 2 shows the results of various quality assessments carried out on the included studies, using the Newcastle-Ottawa adapted cross-sectional (NOS cross-sectional). Across studies, quality ranges from 3/10 to 7/10. The NOS adapted scale does not provide further interpretation of the scores. In this paper we arbitrarily attributed 1-4 to low quality, 5-7 to moderate quality and 8-10 to high quality.
Quality assessment of the included studies
Burnout among couriers
All included studies used a version of Maslach Burnout Inventory (MBI), or a subscale of MBI, or when scale was unspecified, the dimensions appeared to be consistent with MBI. However, language, reporting method and scale used varied significantly. In addition, most included studies did not determine cut-off points, nor reported percentages of participants fall within the different standards. Since the original MBI manual suggested cut-off points were 27 out of 54 as high in emotional exhaustion, 13 out 30 as high in depersonalisation and 21 out 48 as high in professional efficacy,[26] we then generally considered results above middle point of the scale as high burnout. The only study that specified a clinical cut-off point used a 15-item Chinese version MBI using Likert scale 1-5 and hence a total maximum score should be 75. They defined any score < 44 as low level of burnout, 45<score< 74 as high and 75 as severe.[24] They found nearly half of the respondents suffered a high level of burnout, 48.57% 45-74 (n=105).
Most of the included studies reported mean scores that were either near or above the middle point of the scale, indicating moderate to high burnout in this population. For example, Yoon et al [18] used a MBI-GS that was translated and validated in Bahasa Malaysia with 16 items and a four-point frequency scale from 1 (never) to 4 (always). They reported a mean score at 2.45, which indicates high level of burnout. Xiao [21] and Hu [23] both used Chinese version of MBI translated and validated by Li and Shi [27], but Xiao used a 0-7 scale whilst Hu used a 1-5 scale. In terms of reporting, Xiao reported a mean score of 4.02 ± 0.85, which indicates high level of burnout on a 0-7 scale. Hu reported mean scores of each subscales: Emotional exhaustion: 3.3210±0.72783, Disengagement: 2.9973±0.70436, and Low professional efficacy: 3.1319±0.58961. On a 1-5 scale, these scores indicate high level of burnout especially in the emotional exhaustion and low professional efficacy dimensions. Wu[25] conducted exploratory factor analysis when combined MBI with occupational stress and organizational support scales. They subsequently reduced it to 13 items and reported fairly high level of burnout among 325 respondents from confirmation factor analysis stage (Emotional exhaustion 3.7947, Depersonalisation 3.7272, Low professional efficacy 3.4997, calculated averages as the paper reported mean scores by gender). Two studies reported mean of total scores of each subscale[19] or the total mean score of the entire measurement.[22] Table 3 describes them in details. Two studies did not report the scores.[17, 20]
Burnout among couriers
Main factors of burnout
All included studies reported one or more risk or protective factors for job burnout among couriers (see details in Table 4), with five of them reported in both categories. The relationships between the factors and burnout were established using statistical methods such as Correlation Coefficient (CC), structural equation modelling (SEM), or standardized coefficient (SC).
Risk and protective factors for burnout
Among the exacerbating factors, four studies analysed the relationship between occupational stress and burnout --one found no significant association,[19] another found 70% of the participants selected it as main factors of burnout (measured by long hours, heavy workload, safety concerns), and two found significant correlations between occupational stress and job burnout.[20, 25] Other reported exacerbating factors included physical job demands,[18] customer behaviour,[17, 24] low income, career prospect concerns and organizational constraints.[22]. One dimension of Emotional Regulation (Expressive Suppression) was reported positively correlated with job burnout and two subscales (emotional exhaustion and disengagement).[23]
Among the ameliorating factors, perceived organizational support were reported by four studies.[17, 20, 23, 25] Another dimension of Emotional Regulation (Cognitive Reappraisal) was negatively correlated with burnout.[23] Other ameliorating factors included psychological empowerment,[21] job resources, social support and decision latitude.[18]
Xiao[21] compared burnout outcome among different socio-demographic groups and reported significant differences between age groups, the 26-33 years old group experienced higher level of burnout compared to the 18-25; the married group experienced higher level of burnout compared to the unmarried; and the >5000 (RMB Yuan) monthly income group experienced significantly higher burnout than the <3000 group.
DISCUSSION
This review has described the level of job burnout and associated risk factors among Chinese and Malaysian couriers. The included studies reported moderate to high level of burnout among couriers and certain risk or protective factors associated with burnout. The associations were established using statistical methods including correlation analysis, regression analysis and structural equation modelling. Factors related to individual psychological process such as emotional regulation could work both ways.
Since all studies used a translated or modified version of MBI to measure burnout, there were significant variations in language, scale and version used. Hence, it was not meaningful to carry out meta-analysis. In addition, as only one paper suggested cut-off points and reported percentages of participants in each category, we were not able to derive the prevalence of burnout. When interpreting the data from the other eight studies, we then arbitrarily applied a simplified standard by considering all scores that is above middle point as an indication of high burnout. This is because that the originally suggested MBI cut-off scores for high burnout were ≥27 in Emotional exhaustion (total 54), ≥13 in Depersonalisation (total 30) and ≤ 21 in Professional efficacy (total 48) on a 7-point frequency scale 0-6.[26] In addition, a burnout research conducted among nurses in Taiwan aimed to determine the cut-off points of a MBI-HSS Chinese version (0-6 scale) also suggested cut-off points for moderate and high burnout at ≤ 21 and ≤ 32 for Emotional (total 48), ≤ 23 and < 30 for Negative Personal Accomplishment (total 48), and ≤ 6 and ≤ 12 for Depersonalization (total 24).[28]
Psychological constructs that are closely related to burnout, such as occupational stress (OS) and turnover intention (TI) were also measured in a number of studies that were not included in our analysis as no burnout measure was reported. They have reported relatively high level of OS and TI among couriers. For example, Wen et al [29] used the Chinese version of Mobley TI scale and reported a mean score of 2.82 on a 1-5 scale. Xu et al [19] used a Chinese TI scale consists of three questions, with the question --“Looking for other job opportunities” received the highest mean score at 2.775. Xiao [21] defined scores between 12 and 18 as high level of TI and the mean score was 15.27. Multiple standard tools were used to measure occupational stress. Xu et al[19] and Wang et al[20] both used Chinese version Occupational Stress Indicator (OSI) but Xu used the short version with 7 items. Yang[30] and Yang and Mo[31] reported the same study and used the 39 items Enterprise Staff Work Stress Scale (ESWSS). The latter publication reported a mean score of 93.849, indicating a high level of work stress among couriers. One study reported an intervention program that was effective, evaluated by a longitudinal cohort study with one year follow-up.[32] The intervention was an Employee Assistance Program (EAP) that included a range of organizational support, such as stress management training, 7×24 counselling, forums and mini saloons, and a mental health booklet. It significantly lowered TI and OS response in the intervention group.
We found that most of the risk factors reported in the included studies are consistent with the predictors of burnout identified from broader occupational background.[2] However, road safety appears to be a common concern among this population but few studies have analysed it along with burnout. For example a UK study about worker who drive in the gig economy reported 42% of the participants said they had been involved in a collision and 10% said that someone had been injured, usually themselves.[10] Conditions could be worse in developing countries. Wang et al[8] conducted 600 observations on busy roads and interviewed 480 couriers in Tianjin, China. They reported that the rate of exceeding speed limit was 91.3%, 21.2% used cell phone when riding, and reduced use of helmet at night (64.7% compared to during the day 73.0%). 76.5% of interviewees had been involved in a traffic crash at least once, with an average employment time of 17.9 months. Overwork also appears to be serious among Chinese parcel and takeaway couriers.[9, 33] It may be worth considering these factors in future studies about couriers’ burnout in certain culture and economic sector where pressure for working hard at fast pace is high.
CONCLUSION
Based on the findings, we suggest further observational studies are needed to collect higher quality evidence. It is important to develop interventions that offer more organizational support and provide more job resources to couriers. Seven of the nine included studies were published in Chinese, suggesting the importance of including Chinese databases when conducting reviews in relation to couriers. We did not find any burnout studies about couriers conducted in Western country settings. The only relevant study we found was one that researched work stress during COVID-19 pandemic among delivery workers in France and that was assessed using one question.[34] This suggests a significant gap in English literature. We hope this review will provide a useful base for the development of research that aims to reduce job burnout and improve work-related health and safety among couriers in both developing and developed countries.
Contributors
MvT, HW, CA and TC conceived and designed the project. HW, SGL, TO, MKY and SL conducted reviews in English and Chinese languages and extracted the data. HW and SGL checked the data cross both languages at each steps to maintain consistent standard. HW and TO drafted the manuscript, with CA, TC, PW and MvT contributed to the writing and editing. All remaining authors extensively reviewed and approved the final manuscript.
Funding
The project is funded by the MRC Public Health Intervention Development Scheme (PHIND). Grant Ref: MR/T027215/1
Disclaimer
The funders had no role in the decision to publish or in the preparation of the manuscript. The content is the responsibility of the authors.
Competing interests
None declared.
Patient consent for publication
Not required.
Ethics approval
Not required
Provenance and peer review
Not commissioned; externally peer reviewed.
Data availability statement
Data are available on request. All data relevant to the study are included in the article or uploaded as supplementary information. Any further data are available on request.
Acknowledgments
The authors would like to thank professor Jiangmei Qin and Dr. Yanchun Zhang of China Health Economics Association for their advices that helped refine the aims and objectives of this article.
References
