Study setting and design

This study was conducted in Shandong Province, which is located in eastern China, with a population of more than 100 million.17 In 2016, the gross regional product of Shandong Province amounted to CNY6803 billion (US$1024 billion), ranking it as the third largest economy within China.17 Jining Prefecture, located in the south-west of Shandong Province with a population of 8.35 million, had the gross regional product of CNY430 billion (US$63 billion) in 2016.18

Our study design drew on descriptive phenomenology,19 concepts of acceptability and appropriateness in health implementation research20 and was framed by options for change in policy or practice, as identified by local health managers and experts. These helped identify qualitative tools and categories of inquiry, deemed useful to understand knowledge and attitudes affecting women’s and family’s decision-making in relation to screening. The qualitative data collection methods included semi-structured in-depth interviews with 21 women eligible for free screening aged between 35 and 64 years and with five healthcare providers. These were supplemented by focus group discussions with healthcare providers of varying disciplines (described below) aiming to use discussion to elicit contrasting disciplinary perspectives. To gain a partner perspective, key informant interviews were conducted with husbands of women eligible for screening.

Study participants and sample

Two townships (Tangma and Xingcun) were randomly selected from within the mainly rural areas of Sishui and Yutai counties in Jining Prefecture in eastern China. Purposive sampling was used to focus research on those who were representative of women less likely to take up screening, identified by local health managers. The selection criteria included: women aged between 35 and 64 years (those eligible for free screening and at higher risk), resident in the study townships, not in formal employment and those who had either never been screened or who had failed to attend a second or third follow-up screening. Twenty-one women (10 from Xingcun and 11 from Tangma) were recruited in collaboration with township community partners. Fourteen healthcare providers (seven from Xingcun and seven from Tangma) including hospital managers, public health directors and medical practitioners whose responsibilities included contact, recording and managing data for screening and screening service provision, were purposefully selected by discipline mix. Five providers took part in semi-structured in-depth interviews and nine in two focus group discussions (four in the first and five in the second group). Four key informant interviews (two from Xingcun and two from Tangma) were conducted with screening-eligible women’s husbands who were purposefully selected and were not related to the women participants. The numbers of interviews and focus group discussions were designed to enable a theoretical saturation of qualitative themes and data collection ceased at a point when no new information was being obtained (as assessed by rapid review during field operations).

Data collection

All interviews and discussions were conducted over a 1 week period in November 2017 by four researchers from Shandong University who had extensive knowledge of cervical cancer and were trained in conducting qualitative research and interviews on sensitive topics. Women participants were interviewed in healthcare institutions identified by community partners. Individual interviews in private rooms gave participants the opportunity to speak freely and comfortably. Before interviews, a short questionnaire4 including 17 items was used to assess knowledge on cervical cancer, risk factors, symptoms and the objectives and processes of cervical cancer screening. Consistent with earlier studies, responses were scored as correct (one point) and incorrect or ‘do not know’ (zero points), with numbers of correct scores used to categorise respondents into two groups: ‘high level of knowledge’ (score 9 to 17) or ‘low level of knowledge’ (score 0 to 8).21 22 The semi-structured in-depth interviews with women used open questions in categories informed by our design, with detailed probes to deepen responses (box 1). Topics covered included sociodemographic characteristics, screening experience, purpose of cervical cancer screening, perception of barriers to screening uptake and suggestions for overcoming barriers.

In-depth interviews and focus group discussions with providers were conducted close to the workplace and lasted approximately 30 and 45 min, respectively. The in-depth interviews and focus group discussions guide (box 2) explored providers’ experiences in providing screening services, perception of barriers to cervical cancer screening uptake for women and suggestions for overcoming barriers. Key informant interviews with husbands were conducted in their home. Interview guides included questions on their knowledge about cervical cancer and attitudes towards screening for cervical cancer and potential barriers to uptake for women.

Verbal informed consent using a standard script was obtained from all participants prior to interviews. All interviews and focus group discussions were digitally recorded. Confidentiality was ensured through using code numbers rather than names. Researchers took care to identify and reflect on any bias relating to differential status between themselves and interviewees, using introductory explanations on themselves their neutrality and the study purpose, ensuring simple comprehensible language, and maintaining careful non-judgemental listening, in order to minimise bias during the interview process. Each participant was given a vacuum beverage cup to compensate for their time commitment to the study.

Data analysis

All interviews and focus group discussions were transcribed verbatim and subjected to thematic analysis.23 We analysed the data across all stakeholder groups collectively. HY collated interview and questionnaire data across all sites, ensuring consistency supervised by SL for data integrity. HY and SL jointly read all transcripts and developed themes (both pre-determined by our design and emerging), a topic index and code structure. HY, SL and QC undertook coding and thematic consolidation, with any differences discussed and resolved through consensus. These analyses were performed in Chinese and then translated into English by HY for further review by SL, QC and CM. Care was taken to ensure data validity in the translation of dialect and colloquialisms.

Ethical approval

This study was approved by the Medical Ethics Committee of Medical School, Shandong University (LL-201401048), and conforms to the ethics guidelines of the Declaration of Helsinki.

Patient and public involvement

Participants were not involved in the study design and conduct, but their experiences and preferences informed the development of the research questions to generate major themes for analysis. Public involvement including community partners was involved in the recruitment process. A policy brief containing findings and implications will be provided to local policy makers and managers, and to community leaders for dissemination to study participants.

Sociodemographic characteristics of participants

There were no refusals of consent or dropouts during participation, with a total of 21 rural women participating. The mean age of participants was 48.7±6.4 years, with a range from 37 to 60 years, and with more than half (57.1%) aged between 46 and 55 years. Most participants (95.2%) were married and most (90.5%) had an education level of primary school or below. About half of the participants (52.4%) had previously been screened for cervical cancer. Table 1 shows details of the characteristics of participants. The mean knowledge scores were 4.4±2.3 (range from 0 to 10).

Among the 14 healthcare providers, the mean age was 42.6±9.0 years and most (64.2%) were aged between 41 and 50 years. They had an education level of junior college and above. Most providers were medical practitioners (50.0%). The mean age of four husbands who undertook key informant interviews was 50.7±6.3 years and all of them were small-hold farmers.

Key themes

Thematic analysis generated five major themes: (1) gaps in knowledge of cervical cancer and health awareness, (2) fear of cancer and screening outcomes, (3) cultural barriers, (4) influence of close contacts and (5) inconvenience. Each of these is elaborated below, including quotations from participants with their identification number and age.

Gaps in knowledge of cervical cancer and health awareness

Knowledge gaps and lack of access to reliable information on cervical cancer is reported as a major barrier to women’s participation in cervical cancer screening. Findings from these interviews showed that the majority of women participants had a low level of cervical cancer knowledge. Most women indicated that they had heard about cervical cancer from their friends or television, however, almost none knew that HPV infection was the cause. Some cited inaccurate information and most reported that their knowledge, especially about causes and symptoms, was speculative. For example, one reported misconception was that ‘cervical cancer will never happen after menopause’. Limited knowledge seemed linked to lack of awareness of being at any personal risk of cervical cancer happening. When asked why she had not participated in cervical cancer screening, one woman stated:

I think my health is in a good condition. I’ve never had anything serious or any symptoms. No symptoms, no examination. We only go to the hospital when we are uncomfortable or feel sick. (Woman 11, 53 years)

One of the health provider focus groups also mentioned:

Cervix is a dumb organ. There are no symptoms nor is it typical in the early stage. It doesn’t affect their work or daily life; there is even no bleeding. No symptoms, nothing at all. They don’t think they are infected, so there is no need to go through the screening. (FGD 2, provider 05, 41 years)

There were gaps in disease prevention understanding, as well as limited knowledge of the process, cycle, purpose and importance of screening. The importance of early screening prior to symptoms was poorly recognised. These factors were reported to limit women’s acceptance of or commitment to screening, hindering or postponing their participation in cervical cancer screening.

A woman who had been notified to attend cervical cancer screening said:

I was told to do the screening yesterday. Well, I didn’t want to. I thought I always have a good appetite and never feel uncomfortable. There’s no point doing it. (Woman 04, 48 years)

Another woman who participated only once in cervical cancer screening stated:

I did the examination once and the doctor said I was good. So I never came back. Why bother when I’m good? (Woman 15, 55 years)

Healthcare provider viewpoints echoed this:

These women have low healthcare awareness. They wouldn’t go to the hospital even if they have other common diseases. All they care about is if they could still do the housework. They don’t have much understanding of medicine, so they couldn’t know the significance of early screening if they got the disease. (FGD 1, provider 03, 52 years)

Fear of cancer and screening outcomes

The majority of women in the interviews expressed a deep fear of cancer. Some did not want to know if they had cervical cancer; a common reason offered for why they were unwilling to attend the screening was that they did not want to learn the results. This fear extended to any form of cancer terminology. Women wanted to avoid the anxiety and distress that would be associated with being told of a cancer-related lesion:

It’s a terrible disease. I’m afraid something bad might show. If I don’t do the screening I wouldn’t know if I had it, I will at least have a good appetite. If the results were bad, I wouldn’t feel comfortable knowing that. Appetites and good sleep would be gone for sure. I would have more years to live freely if I didn’t know it. (Woman 01, 37 years)

People become upset whenever they talk about cancer. In the first 2 years of the screening campaign, we couldn’t put ‘cancer screening’ on the banner and used ‘health examination’ instead. People couldn’t take it if they saw the word ‘cancer’. (FGD 2, provider 01, 47 years)

A husband also mentioned, during key informant interview:

They were all afraid they might have it. Older women would say ‘why do the screening? I would be upset if I knew I had it’. (Husband 02, 63 years)

Fear of a positive screening test was also related to the financial implications. Some women with fewer economic resources reported avoiding screening because they were worried they would not have the money for treatment if they were diagnosed with the cancer, noting that at present, while screening is free of charge, treatment is not. This suggests that family finances are likely to influence their participation, with families weighing up the impact of debt against that of the disease:

We wouldn’t afford the treatment even if we found out we had it, so it’s better we don’t do it at all. Adults need money and so do children. The money would be wasted if we went through the treatment and not cured. No need to spend money on a disease that can’t be cured. We don’t want to add burden to the family. (Woman 10, 54 years)

Two focus groups also noticed this:

People only see nearly 1 00 000 yuan for the surgery and it’s a very heavy burden. The disease would drag the already poor family into debt. They can’t take it. (FGD 1, provider 02, 44 years)

There are indeed financial problems, especially for rural families. They think they don’t feel anything right now, if they find out they are infected after the screening they have to go through the treatment and spend money. So it’s better not to do the screening at all. (FGD 2, provider 03, 41 years)

Cultural barriers

A common report, especially among older women, was a reluctance to remove clothing or allow genital examination, especially being exposed in front of non-family members. It was mentioned by nearly all women that they felt embarrassed about this step. Informants related this to cultural norms in their rural setting, and among Chinese women more generally.

Felt like a very private part and I don’t want other people to see it. That’s the thought. Yes it’s embarrassing. I’m too old. Don’t want to be a joke. So ashamed. (Woman 09, 55 years)

When I went to the gynaecology for the screening, I went inside and the doctor told me to take off my pants, and I suddenly didn’t want to do the screening. (Woman 02, 48 years)

Participants were specifically asked about willingness to accept screening by male doctors. For some participants, the gender of the service provider seemed a barrier; nearly half indicated that they would not undertake screening if performed by a male doctor.

I wouldn’t go if a man was doing the examination. It’s less acceptable than a woman. It’s so private and men are so improper for that. (Woman 14, 41 years)

A female service provider mentioned:

They are very conservative. Some women with vaginitis go to the clinic. We tell them to take off the pants and they are embarrassed. Some even put up with the disease until they can’t. So they would be more uncomfortable if male doctors are here. (Provider 01, 63 years)

Influence of close contacts

It was found that close contacts (such as mothers, sisters or neighbours) were an important factor influencing women’s participation in cervical cancer screening. Knowing other people who had been screened or who had been diagnosed with cervical cancer and undertaken treatment was reported to increase participation in screening. The converse was also true, with reports that if close contacts were not willing to participate, this reduced women’s acceptance of screening. This was also reported to generate clusters of women who would refuse screening.

I wouldn’t know it if they (the neighbours) asked me to come along. If the others (the neighbours) don’t come, I wouldn’t either. It’s good to have someone with you. I wouldn’t come by myself. (Woman 05, 55 years)

I would for sure not come by myself. (Woman 06, 48 years)

A service provider mentioned a similar situation during the screening campaign:

We met this situation once when we were doing the screening in a town. Several women came together and they learnt they need to take off the pants. One said ‘This is what it is. Forget it. Let’s go’. And all of them felt reluctant and left together. It’s a group mentality. (Provider 03, 43 years)

Inconvenience

Difficulty in scheduling or otherwise allocating time for screening was considered to be a significant barrier to women’s participation. Many women eligible for screening work in disparate locations throughout the year may only return home at the time of Spring Festival. Such holidays are busy with limited time available for activities such as routine health checks. One working woman stated that leave is only feasible for actual illness.

I work elsewhere and only ask for leave when they have an emergency or special occasions at home. It’s hard to ask for leave and I lose some salary for that. Usually I come home once a year for only 5 or 6 days. It’s not enough. (Woman 21, 40 years)

Many rural people go to other places to work and don’t come home often. We call them to come back but they can’t. They don’t get paid if they ask for leave and it’s too expensive on the road. So they don’t want to come back. (Provider 02, 44 years)

Women who do not travel for work also report many demands on their time including family business, child-care, housework and farming duties that are often seasonal. Lack of time for routine healthcare is reported as a barrier to cervical cancer screening either. When asked why not attend the screening, a woman who had never gone through the examination said:

I’m always busy with the kids going to school every day and housework never seems to end. Family and housework are the most important for me. So I somehow forget about it (the screening). (Woman 08, 46 years)

I tell you we farmers are pretty busy. When we are not, we want to do some small business to make money. So no time for examinations. (Woman 19, 56 years)