Strengths and limitations of this study

This study may be the first to examine hepatitis B-related risk behaviours among rural-to-urban migrant workers in western China, with a relatively large sample of 1299 participants.

The survey questionnaire was constructed according to the theory of planned behaviour, and was further developed by adding two innovative modules: experience of behaviour and regret feeling.

Causal inference based on the associations observed in the study might be limited by the cross-sectional study design.

Selection bias may have arisen, leading to an imbalance in occupation distribution because of the non-random sampling.

Reporting bias, mainly assumed as an underestimate of risk behaviours, may be inevitable because of social desirability and participants’ personal privacy concerns.

Introduction

Hepatitis B, a leading cause of liver cirrhosis and liver cancer, remains a major threat to global public health, particularly in the Asian-Pacific and sub-Saharan African regions.1 There are nearly 2.57 billion hepatitis B surface antigen (HBsAg)-positive people, and 887 000 people infected with hepatitis B dying of liver damage and complications worldwide, according to the WHO in 2015. There is an urgent need to increase investment in hepatitis B elimination, especially in low-income and middle-income countries.2 China is cited as a country with high hepatitis B endemicity by the WHO, and has more than 90 million hepatitis B patients and 100 000 new hepatitis B virus (HBV) infections annually.2 The National Health Commission of China has included the hepatitis B vaccine in its national immunisation programme since 2002. From that time, the vaccine has been administered to newborns, infants and unvaccinated children under 15 years of age in rural and urban China.3 As a result, the HBsAg-positive rate has declined to 2.08% in the Chinese population aged 1–14 years, according to the National Hepatitis Seroepidemiology Survey in 2006. However, the HBsAg-positive rate was 8.57% in the population aged 15–59 years in 2006,4 and remained above 8% in Chinese adults till 2016.5

HBV is usually transmitted through blood and other body fluids of infected individuals.5 Mother-to-child transmission, sexual transmission and blood transmission are the main modes of HBV transmission in China.6 For adults, sexual intercourse (including men who have sex with men (MSM)) and using contaminated needles are the major routes for HBV infection.7 8 Therefore, risk behaviours related to hepatitis B refer to unprotected sexual activity, soliciting services from sex workers, promiscuity and needle-sharing among drug users.8 9

Rural-to-urban migrant workers account for most of the internal migrant population in China. They are defined as people who have left their rural area (usually their hometown) to seek better employment opportunities and higher incomes in towns and cities.10 The latest Migrant Workers Monitoring Investigation Report of China showed that the number of migrant workers had increased to 288.36 million people by 2018.11 Most migrant workers have low education levels, are engaged in low-income and low-skilled jobs, and live stressful lives.12 Most are sexually active but are single or living apart from their spouse.12 Further, the majority lack sex-related knowledge, self-protection awareness, social support and basic health care.10 Therefore, migrant workers are more likely to have unprotected sex or solicit services from a sex worker, increasing the risk of infection of sexually transmitted diseases (STDs) and HBV. Previous surveys have shown that 40.0% of construction workers had unprotected sex, 14.9% solicited services from a sex worker, 7.9% were promiscuous, 8.4% were involved in blood selling and 0.7% used drug.13 14 A survey of 2462 migrant workers in central China revealed an HBsAg-positive rate of 11.66%, higher than the average rate among Chinese adults.15 Previous studies also found that migrant workers had relatively greater susceptibility to HBV infection compared with non-migrants and local dwellers.16 In addition, the frequent flow of migrant workers may increase the possibility of spreading HBV to the general population and facilitate regional transmission across China.17 Therefore, it is important to understand hepatitis B-related risk behaviours and have insight into their determinants among migrant workers.

The theory of planned behaviour (TPB, Ajzen, 1991)18 ,19 is a model widely used to understand behaviours. TPB describes three conceptual modules that determine behaviour: (1) attitude towards the behaviour (AB) refers to favourable (or unfavourable) appraisal of the behaviour; (2) subjective norms (SN) refers to perception of social pressure to perform (or not perform) the behaviour; and (3) perceived behavioural control (PBC) refers to the perceived ease (or difficulty) of performing the behaviour. TPB postulates that AB, SN and PBC lead to the formation of a behavioural intention (BI). Favourable AB and SN and great PBC predict a strong intention of performing a behaviour, and consequently a high likelihood of carrying out the action.18 19 Previous studies also argued that some independent variables, like experience and/or regret about performing a behaviour, would directly or indirectly influence the BI and should be taken into account to improve the TPB framework.20 21 TPB variables have been widely adopted to address the issue of health behaviours, for example, interpreting HIV/AIDS-related behaviours, particularly for highly susceptible groups such as sex workers and MSM.22 23 TPB has also been used to explore determinants of smoking, drinking and health-seeking behaviours,24 25 but to our knowledge, studies of the hepatitis B-related risk behaviours of rural-to-urban migrants have seldom been reported.

Thus, this study was the first attempt to (1) understand the status of risk behaviours related to hepatitis B undertaken by migrant workers, and (2) detect and describe factors that motivate and influence workers’ BIs and practical behaviours on the basis of TPB.

MethodsStudy sites and sampling

Chongqing, in southwestern China, is the largest municipality directly under the Chinese central government. It is regarded as ‘miniature China’ because its geographic characteristics, urban–rural distribution and socioeconomic profile are close to the national average.26 The city area of Chongqing, a popular destination for migrant workers, consists of nine administrative districts with an area of 5472.82 km2 and a population of 8.65 million, of which migrants constitute about 23.5%.26 27 The HBsAg-positive rate among migrant workers in the city area of Chongqing is estimated as 8.6%28 and there were nearly 26 000 new viral hepatitis infections in 2016, according to the Health Statistic Yearbook of Chongqing.27

Two-stage stratified cluster sampling was used to recruit participants between June 2018 and January 2019. First, nine districts of Chongqing’s city area were stratified into three layers by economic development, geographic background and population density: more developed, moderately developed and less developed. Three districts were randomly selected to represent each stratification. Second, two enterprises were purposively sampled in each district, including the manufacturing, construction, wholesale and retail industry, transportation industry, hotel and catering industry, and community services. The Local Center for Disease Control and Prevention, Health Supervision Institute and Urban–Rural Development Committee helped to coordinate arrangements with the sampled units. The inclusion criteria for participants were (1) at least 18 years of age, (2) having been in the city area of Chongqing for at least 6 months, (3) not registered as a Chongqing urban resident and (4) engaging mainly in secondary or tertiary industries. Given a considerable proportion of migrant workers were illiterate or only had primary school education, trained investigators assisted in explaining the questions item-by-item in both Mandarin and Chongqing dialect to those who found it difficult to understand the questionnaires. Participants were reassured that all responses would be anonymously recorded, and written informed consent was obtained from each participant. The surveys were administered in relatively undisturbed environments, outside of peak working hours to maximise the quality of the data collected. Each questionnaire was double-checked by investigators for completeness.

Study instrument

The questionnaire was constructed based on the TPB and health-related behaviours and perceptions reported in published studies.22 23 Experts in epidemiology and hepatology assisted in modifying the logic and wording of the questionnaire (online supplemental file 1). A pilot survey was conducted with 90 migrant workers in nearby restaurants and construction sites. The final version consisted of nine modules with Cronbach’s alpha coefficients ranging from 0.759 to 0.968, and confirmatory factor analysis showing a good fit (χ2/df=1.859, root mean square error of approximation=0.039, goodness-of-fit index (GFI)=0.900, adjusted GFI=0.883, comparative fit index =0.969, incremental fit index =0.969). TPB variables were assessed using a 5-point semantic differential scale, and the average item score for each module was computed for use as the scale score. Higher scores indicated more risk. The definitions of each module and variable scale are shown in table 1. Items of different dimensions and positive and negative items were sorted at intervals to reduce social bias through desirable responding or picking an initial scale for each item.

Table 1

Definition and item scales for each module in the hepatitis B-related behaviour questionnaire

Participants and public involvement

This study was designed by our research team with assistance from experts in epidemiology and hepatology. The item pool of the questionnaire and outcome measures was generated through interviews with nine migrant workers in a nearby factory. Recruitment of the participants was conducted by staff of local health institutions and a coordinator at each survey site. Results and conclusions to help decision makers improve health policies for migrant workers were summarised in a study report given to the local health institutions.

Data analysis

Survey data were double-checked and entered into a database using EpiData V.3.1 (The EpiData Association, Odense, Denmark). All data were analysed using IBM SPSS (V.22.0, SPSS Institute). Categorical data were assessed by the number and proportion of respondents. Continuous socio-demographic variables, such as age, years of being a migrant worker and working hours per day, were converted into categorical variables and then described by number and proportion of respondents. Respondents’ knowledge levels were divided into poor (scores less than 7), medium (8–10) and good (11–13).29 TPB variables were used as continuous variables with average scores. For the association analyses, independent variables were identified as the variables in the modules of sociodemographics, hepatitis B knowledge and TPB framework, and the dependent variables were identified as BI and hepatitis B-related risk behaviours. For the dependent variables, BI was dichotomised into ‘never had an intent’ and ‘had an intent for at least one behaviour’ and hepatitis B-related risk behaviours were dichotomised into ‘never had risk behaviour’ and ‘had at least one risk behaviour’. Univariate analyses were performed with independent variables of sociodemographic and hepatitis B knowledge with the two dependent variables by χ2 tests. Independent variables with p values less than 0.10 in the univariate analyses were subsequently inputted into the logistic regression models (α=0.05, β=0.10) along with variables in TPB modules to detect factors possibly influencing the two dependent variables. Binary logistic regressions were fitted with the dependent variables by entering three blocks of variables: block I: sociodemographics and knowledge level; block II: TPB variables; and block III: demographics, knowledge level and TPB variables. BI was not included in blocks II and III when it was regarded as dependent variable. Dummy variables were coded for variables with more than two values, and variables were entered stepwise into the models. Adjusted ORs and 95% CIs were computed, and p values less than 0.05 were deemed statistically significant.

ResultsBasic characteristics

A total of 1528 migrant workers were screened and 229 were excluded because they failed to complete the questionnaires. Thus, 1299 (85.02%) respondents completed the questionnaire, of which 758 (58.35%) were women. The median age of respondents was 30.58±21.18 years, ranging from 18 to 68 years. A total of 901 (69.36%) respondents were married or in a relationship, but 670 (51.58%) were not living with their spouse/partner. Meanwhile, 43 (26.33%) were single and 55 (4.23%) were divorced/widowed. Approximately one-third of respondents (443, 34.1%) had a monthly income above 4000 RMB and 451 (43.53%) regularly sent money back home to their families. There were 626 (48.19%) respondents educated to junior school level or below, 246 (18.93%) respondents who drink alcohol and 921 (70.9%) respondents with low hepatitis B knowledge levels (table 2). The top three ways of accessing hepatitis B knowledge and information were from friends/family members (53.88%), by television or radio (38.12%) and the internet or mobile phone (27.95%) (table 3).

Table 2

Characteristic differences in hepatitis B-related behaviours and behavioural intention

Table 3

Access to hepatitis B knowledge (N=1299)

Hepatitis B-related BI

In the 6 months prior to the survey, 1111 (85.52%%) respondents had intentions to conduct risk behaviours (table 2). More than half of the sample (55.65%) reported intending being promiscuous/unfaithful, while 36.26% planned to solicit services from sex workers. Almost a quarter (26.1%) planned to engage with MSM/anal sex, 26.1% to take drugs with shared needles, 31.18% to sell/transfuse blood illegally and 53.5% to share towels/toothbrushes. More than half of the respondents (62.36%) were very willing to wear a condom when engaged in high-risk sexual behaviour (table 4).

Table 4

Migrant workers’ behavioural intention and risk behaviours for hepatitis B

Hepatitis B risk behaviour status

Table 2 shows that 384 (29.56%) migrant workers had performed hepatitis B-related risk behaviours in the 6 months prior to the survey. Of all the respondents, 133 (10.24%) conducted casual sexual behaviour, 61 (4.7%) solicited services from sex workers and 40 (3%) had MSM/anal sex. Among the 842 respondents reporting risky sexual behaviours, more than half (58.19%) never used a condom. To explore the reason for this situation, 210 (42.6%) respondents indicated that they used other methods of contraception, 167 (33.87%) reported that it was uncomfortable to wear a condom and 43 (8.74%) respondents felt embarrassed to purchase condoms. About one-fifth (265 participants, 20.4%) of the respondents shared toothbrushes or towels with friends or family members, 13 (1%) respondents shared needles for intravenous drug use and 16 (1.23%) sold or transfused blood illegally (table 4).

Factors influencing BI

The univariate analyses, shown in table 2, indicate that the BI of respondents differed significantly in terms of gender, age, education background, whether living with spouse/partners, type of accommodation, type of work, working hours per day, monthly personal income, smoking, alcohol drinking and level of hepatitis B knowledge (p<0.05). Binary logistic regression detected that migrant workers with an education level of junior middle school (OR=2.16, 95% CI 1.25 to 3.73), aged from 18 to 30 (OR=3.49, 95% CI 1.91 to 6.39) and from 31 to 40 (OR=2.06, 95% CI 1.13 to 3.77), were more intent on undertaking risky behaviours. In contrast, women (OR=0.61, 95% CI 0.39 to 0.95) were less likely to have the BI in block Ⅰ. The scores of AB (OR=9.36, 95% CI 5.32 to 16.46), SN (OR=2.20, 95% CI 1.54 to 3.17), EB (OR=1.92, 95% CI 1.43 to 2.58) and RF (OR=1.20, 95% CI 1.05 to 1.38) modules had positive associations with behaviour intention for hepatitis B-related risk behaviours in block II. In block Ⅲ, migrant workers were younger (OR=2.77, 95% CI 1.41 to 5.43) and with poor levels of knowledge (OR=2.10, 95% CI 1.03 to 4.28). They were more intent on undertaking hepatitis B-related behaviours, and the scores of AB (OR=9.49, 95% CI 5.32 to 16.91), SN (OR=2.06, 95% CI 1.44 to 2.95), EB (OR=2.17, 95% CI 1.60 to 2.94) and RF (OR=1.23, 95% CI 1.06 to 1.42) were positively associated with BI (table 5).

Table 5

Factors associated with intention to undertake hepatitis B-related behaviours and actual behaviours of migrant workers

Factors influencing hepatitis B-related risk behaviours

As table 2 shows, univariate analyses indicated that risk behaviour of respondents differed significantly by gender, age, working hours per day, monthly personal income, smoking and alcohol consumption (p<0.05). Binary logistic regression detected that smoking (OR=1.43, 95% CI 1.01 to 2.03) and drinking (OR=1.63, 95% CI 1.18 to 2.26) were positively associated with risk behaviours in block Ⅰ. In block Ⅱ, modules of AB (OR=1.29, 95% CI 1.02 to 1.63), BI (OR=1.38, 95% CI 1.07 to 1.76) and EB (OR=1.29, 95% CI 1.07 to 1.56) were positively associated with risk behaviours. In block Ⅲ, adjusted for sociodemographics, TPB modules of BI (OR=1.42, 95% CI 1.10 to 1.82), EB (OR=1.23, 95% CI 1.01 to 1.50) and RF (OR=1.13, 95% CI 1.02 to 1.25) were positively associated with risk behaviours (table 5).

Discussion

The Action Plan for Prevention and Treatment of Viral Hepatitis in China (2017–2020) underlined the significance of preventing and controlling viral hepatitis, particularly for those who are susceptible to the disease and may then increase the transmission of HBV.30 In a meta-analysis of 411 studies by Zou et al, the prevalence of viral hepatitis among rural-to-urban migrants was 0.45%, a 38.5 higher OR of infection than in the general population in China. The increased rate is probably related to risky sexual behaviours like soliciting services from sex workers and illicit drug use.16 Our finding that approximately one-third of participants had demonstrated at least one hepatitis B-related risk behaviour in the 6 months before the survey is also consistent with previous studies. However, the proportion of migrant workers reporting promiscuity (10.24%) and soliciting services from sex workers (4.7%) was lower than those reported in studies conducted in Shanghai (15.22%) and Zhejiang (5.7%).31 In addition, 2.8% of migrant workers reported MSM behaviours, of relevance to this research because MSM is risky for STD transmission.22

Although the Chinese Ministry of Health has advocated condom use to prevent STDs since 2006,32 our study found that a considerable proportion of migrant workers (58.19%) never use condoms. This finding is in line with the low rate of condom use by migrant workers in Hefei, China (52.68%).14 Potential barriers to condom use included use of other contraceptive methods, discomfort and not being required by the partner. These main reasons indicate that condom use was only viewed as contraception rather than as protection from STDs by migrant workers. Sexual enjoyment was prioritised over disease transmission. The limited knowledge and awareness of STDs among migrant workers described in previous studies is also consistent with the generally low levels of hepatitis B knowledge among respondents in our study. Protected sex with a condom has been proved to be significantly practical and cost-effective in preventing STD transmission.22 Therefore, extensive publicity for condom use should be targeted at migrant workers. Vending machines for condoms or even free distribution stations could be established near migrant workers’ living areas to ease the embarrassment mentioned by the respondents.33

As expected, there were a small number of respondents with a history of needle sharing for drug use (1%) and illegal blood selling/transfusion (1.23%), in line with studies of migrant workers in eastern China.14 31 Blood transmission, along with sexual transmission, is a significant route of HBV infection that should not be neglected. In addition, one-fifth of participants in our study reported sharing personal hygiene products like toothbrushes and/or towels, amplifying the possibility of HBV infection through damaged skin.34 35 Therefore, health education targeting these issues is necessary for migrant workers.

Approximately one-third of participants admitted to undertaking hepatitis B-related risk behaviours. Moreover, nearly 90% reported that they intend to perform risk behaviours on occasion. BI indicates the potential for a person to perform the actual behaviour. Therefore, educational interventions to alter BI and self-protection cognition are crucial, in addition to direct regulation of risk behaviours.

Logistic regressions suggested that migrant workers who are men, younger, with lower educational backgrounds and knowledge levels, would be more intent on carrying out hepatitis B-related risk behaviours.20 Compared with women, most men perceived lower disease risk and overestimated their own health status.36 Younger people may experience more sexual demands than older counterparts.31 Compared with highly educated people, those with lower education tend to neglect disease prevention.37 People with limited hepatitis B knowledge may lack understanding of the disease and be less aware of self-protection compared with those who are more informed.36 Consistent with a study of HIV-related behaviours in northwest Ethiopia, there was a positive association between drinking and risk behaviours. Drinking may create more opportunities for promiscuity/infidelity and unprotected sex for migrant workers.38

After adjustments for sociodemographics, migrant workers who scored higher on AB and SN were more intent on undertaking risk behaviours, and those with higher scores for BI were more likely to have performed hepatitis B-related risk behaviours. These findings can be interpreted using the standard TPB framework—AB and SN, derived from behavioural and normative beliefs, will act on BI, and then work together with BI to trigger the behaviour.19 That is, if migrant workers have a more favourable attitude towards hepatitis B-related risk behaviours and there is less perceived social pressure against the behaviours, they will be more intent on proceeding and actually realising the behaviour.18

To strengthen the interpretability of the actual behaviours of migrant workers, two socio-psychological modules—EB and RF—were introduced into the typical TPB framework. As expected, the two variables were positively associated with both BI and practical risk behaviours. Previous studies argued that daily decision making will be affected by the actual emotional experience, and successful implementation of risk behaviours in the past appears to render migrant workers more likely to repeat them in the future.39 In contrast, strong regret felt about a behaviour may lead to less intent in future and lower likelihood of performing risk behaviours. Regret represents a negative consciousness and an emotional reaction to a person’s intention or behaviours.20 Given migrant workers’ poor perception and self-protection against HBV infection, health educational campaigns are necessary to improve their cognition and behaviours. Peer education may be effective considering friends/family members, television/radio and internet/cell phone applications are the main sources of migrant workers’ health knowledge. Health education should combine new and traditional media. Only half the migrant workers in our study had been inoculated using the hepatitis B vaccine. This may be because free hepatitis B immunisation is currently not offered to people over 15 years of age in China.4 Extra financial support should be provided to hepatitis B-susceptible adults, including migrant workers, to expand the coverage of the vaccine.

Some limitations of our study should be considered. First, causal inference based on the associations observed might be limited because of the cross-sectional design. Second, selection bias may have given rise to an imbalance in occupation distribution between the sampled participants and Chongqing’s population of migrant workers, because of the non-random sampling. In addition, reporting bias—reflected as an underestimate of actual risk behaviours—may be inevitable as respondents wanted to maintain personal privacy and social desirability, although their anonymity was assured.

Conclusions

Our study found that one-third of migrant workers undertook hepatitis B-related risk behaviours and 90% indicated they intended to act in this risky way. Migrant workers who were men, less educated, with limited hepatitis B knowledge and younger, had stronger intentions of hepatitis B-related risk behaviours, and drinking alcohol was positively associated with realisation of the risk behaviours. TPB framework was enhanced by the innovative variables of EB and RF, and served well in interpreting the influencing factors, showing that migrant workers have a more positive attitude and fewer SN towards risk behaviour intent and the actual behaviours. Accordingly, more attention should be paid to improving the disease perception and self-protection awareness, and helping migrant workers to regulate their behaviours. Theory-grounded interventions should be combined with new and traditional media and peer education to address the key influencing factors proposed by the analyses.

Data availability statement

No data are available. Not applicable.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by Institutional Review Broad of Chongqing Medical University (No. 2018016). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We would like to thank local health institutions and the Urban-Rural Development Committee for their kind assistance and coordination throughout the field study. All the migrant workers who participated in the study are much appreciated. We would like to thank Lisa Zhao for thorough polishing of the wording of the manuscript.