Cervical cancer knowledge, attitudes, beliefs and practices of women aged at least 25 years in Harare, Zimbabwe

A sequential explanatory mixed methods design was used to generate evidence for this study. This design comprise of both quantitative and qualitative methods, with the quantitative study being the major study. Qualitative inquiry was conducted to aid deeper understanding of issues, explain surprising and unexpected results from the quantitative surveys [18]. A health facility based survey was conducted between January and April 2018 among women with cervical cancer. The second phase of the study consisted of a qualitative inquiry with key informants. These participants included health workers directly involved in cervical cancer control, WHO officials, Ministry of Health policy makers and NGO programme managers. Key informants were selected using snowball sampling technique [18, 19].

The study population consisted of women with cervical cancer, health workers and other stakeholders who are involved in cervical cancer interventions in Zimbabwe.

In the health facility based survey, 134 women with histologically confirmed cervical cancer were selected in a census at Island Hospice, Harare and Parirenyatwa Hospitals between January and April, 2018. This sample size was based on the follow of patients in these health facilities based on data from the previous year, 2017. A minimum of sample size of 80 had been determined before the study based on 2017 data; however 134 were ultimately enrolled to improve on precision of estimates. In the qualitative study the sample sizes were guided by the saturation principle [20]. The final sample size achieved for key informant interviews was 17 participants.

The health facility study sites were: Parirenyatwa Hospital, Harare Hospital and Island Hospice. The treating health facilities serve people from across the country; however, most of the patients recruited in this study were resident in Harare communities. All participants in the survey were aged at least 25 years (and no maximum age limit) and consented in writing before being enrolled in the study. Some of the key informants for qualitative study were selected from the same study sites as the survey. They were selected based on special characteristics such as knowledge of cervical cancer and unique experiences in the treatment of cervical cancer.

For the survey, a validated structured questionnaire [21] was administered by the researchers. The questionnaires had been programmed in SurveytoGo software (Dooblo, Israel) in android tablets. Electronic data collection allowed real time uploading of data, better quality control and reduced the time of data processing as there was no need for manual data entry. For the qualitative study, key informant interview guide (Additional file 1) was used for data collection. This study being a sequential mixed methods design, the qualitative tool was designed based on the outputs from the quantitative study analysis. The purpose of the qualitative study was to explain surprising results, outliers and put stories to data or associations in the survey.

Data analysis for the survey was conducted using univariate and multivariate methods. Univariate analyses were conducted to obtain proportions with respect to knowledge of causes, risk factors, prevention and treatment amongst participants. Since Likert scales were used on some variables, the responses were compressed to generate binary variables “Yes” and “No” for the univariate and bivariate analyses. The multivariate analysis outcome variables were also converted into binary variables as criteria for conducting logistic regression analysis. Logistic regression analysis was conducted to identify factors that were associated with correct knowledge of causes, prevention and treatment. The confounding variables that were adjusted for in the multivariate analysis were: sources of cervical cancer information and head of household’s occupation, education, radio listenership, watching TV, reading newspapers and accessing internet for general information. The identification of these confounders was based on literature from similar studies. All statistical analyses were conducted using STATA version 14 (StataCorp, Texas).

A total of 17 key informants were enrolled for the qualitative research. Participants were selected based on specific characteristics relevant to the research question and to ensure diversity of perspectives to strengthen the study outcomes.

A total of 143 participants responded to the survey these were women with cervical cancer. The mean age of participants enrolled in the survey was 52 years (SD = 12.2).

Table 1 show that women with cervical cancer had poor knowledge with respect to the causes of the disease. Only 23% of the respondents specifically reported HPV as the cause of cervical cancer. Seventy-three percent (73%) of participants had knowledge of at least one risk factor of cervical cancer. Cervical cancer patients who had knowledge about the availability of free screening services at local hospital or health facility and HIV testing being optional during screening were 73%. Eighty percent (80%) of women with cervical cancer knew about treatment modalities about the disease.

In Table 2, the majority (98%) of women reported that screening was important for early treatment of cervical cancer. Only 17 and 2% of women had negative attitudes that cervical cancer treatment was for people with money and the treatment was embarrassing respectively. With regards to beliefs, the majority of women had positive beliefs, however; 52 and 33% believed that cervical cancer treatment abroad was better and health workers abroad provided better treatment and care respectively. In addition, 29% of participants believed that patients treated abroad had better survival outcomes compared to those treated locally. Women who had access to a regular doctor were relatively low, 48%. However, participants who reported ever screening for cervical cancer in their lifetime were 95%. Proportion of participants who reported screening at least once in their lifetime was relatively low, 66%.

Table 3 shows that knowledge of causes of cervical cancer was negatively associated with being aged 45 years or more (p = 0.004), having no household income (p = 0.007), household income less than US$600 per month (p = 0.015), having middle class wealth (p = 0.032), watching TV daily (p = 0.007) and between once and 6 times per week (p = 0.045). Factors positively associated with correct knowledge of causes were listening to the radio daily (p = 0.001) and between once and 6 times per week (p = 0.010). Factors positively associated with cervical cancer prevention knowledge were listening to the radio daily (p = 0.022) and between once and 6 times per week (p = 0.018).

Knowledge about the causes of cervical cancer was low among cervical cancer patients. This finding suggests that there is still limited awareness or health education with regards to the basic information on cervical cancer. The assumption would be that cervical cancer patients because of their condition would tend to be exposed more to information or have positive attitude towards information about their condition. However, findings suggest that this may not be so. About 72% of women with cervical cancer were able to report at least one risk factor of cervical cancer (HPV, HIV/AIDS, STI, poor personal hygiene, multiple sex partners without protection, sex with uncircumcised partner, and use of herbs in the vagina). However, HPV as a cause of cervical cancer was only reported by a small proportion of participants, 23%, confirming that specific knowledge about the cause of this condition is still limited in Zimbabwe. Our findings suggest relatively high knowledge of risk factors of cervical cancer as compared to other studies. In an Ethiopian study, only 42% of participants reported at least one risk factor of cervical cancer [10]. Researchers in Nigeria reported knowledge of risk factors as low as 15.6% in a recent study [15]. In Tanzania, multiple sexual partners and history of HPV infection as risk factors for developing cervical cancer were reported by 47.4 and 43.1% of respondents respectively [6].

Knowledge of causes of cervical cancer reported was associated negatively associated with being aged 45 years or more (p = 0.004), having no income (p = 0.007), income less than US$600 per month (p = 0.015), being in the middle class wealth quintile (p = 0.032), watching TV daily (p = 0.007) and between once and 6 times per week (p = 0.045). However, listening to the radio frequently was positively associated with knowledge of causes of cervical cancer. These results suggest that radio messaging may have been effective in influencing cervical cancer awareness. The findings may also suggest that awareness or health education programmes may be biased towards radio programmes or that the packaging of messages is appropriate for radio audience. Programmes in television may not be properly packaged to influence knowledge of causes but may be reinforcing fear, stigma, myths and misconceptions in the population. However, this was not explicitly explored in our qualitative study and future studies should consider understanding the role of television in knowledge of causes and prevention of cervical cancer. Our findings support those of Mitiku & Tefera [10] who found that better income was associated with knowledge of cervical cancer. A Nigerian study reported that being younger than 30 years was associated with lack of knowledge of cervical cancer or its screening [22], and this is in contrast to our results which revealed that being 45 or older was associated with lack of knowledge. Our results did not find any association of knowledge of causes with levels of education and this directly contrast those of a similar study in Zimbabwe, which suggested that knowledge of cervical cancer was associated with having at least secondary education [23].

There was suboptimal knowledge of cervical cancer prevention reported among women with cervical cancer in our study, 71%. This suggests that lack of prevention knowledge for cervical cancer may be a possible risk factor for disease incidence [24, 25]. Mukama et al. [5], in Uganda reported a relatively lower knowledge of preventive measures (62%). The factor associated with preventive knowledge was listening to the radio frequently. These results also suggest that radio programmes may be a better platform for disseminating cervical cancer awareness and education. Packaging of messages for cervical cancer could be targeting radio audience. These findings contrasts those of Ayinde [22] who reported that married women were more likely to be aware of cervical cancer compared to single women. Our study contrast the findings from a Ugandan study which revealed that living in peri-urban and urban areas, having a higher monthly income and having had an HIV test were associated with knowledge about cervical cancer prevention [5].

Our study revealed that knowledge of treatment was relatively high among women with cervical cancer (80%). This high knowledge could be explained by frequent interface with health facilities and perhaps desperation for treatment for a disease commonly believed to be untreatable [26‐29]. It is imperative that awareness or health education be biased towards risk factors and causes to reduce the risk of developing cervical cancer which is costly to treat and is a significant threat to life [30]. As revealed in the qualitative phase of this study, there are still some gaps in the packaging of awareness or health education interventions for cervical cancer. In our setting, cervical cancer treatment is marred with myths and misconceptions though only a small proportion (8%) of the respondents did not believe that the disease was treatable. To our knowledge, this study is the first to report knowledge levels with regards to treatment of cervical cancer in our context. There is generally little information about the knowledge of treatment in developing countries with most studies focusing on primary and secondary interventions for cervical cancer.

Generally, women participants in our study had positive attitudes towards cervical cancer, its screening and treatment. The majority agreed that early screening was important for early treatment however; a few women reported that cervical cancer treatment was for people with money and treatment procedures were embarrassing. These negative attitudes have an influence on the uptake of treatment services [5]. Low screening uptake and poor health seeking behaviours may be explained by the negative attitudes towards treatment despite positive attitudes towards screening itself [5, 11‐14]. Our qualitative findings also reported negative attitudes as a barrier to both preventive and curative interventions for cervical cancer.

Cervical cancer treatment is marred with many beliefs, myths and misconceptions [31]. Most participants had positive beliefs with regards to cervical cancer, its screening and treatment. However, a relatively high proportion believed that treatment, care of health workers abroad and survival outcomes for those treated abroad were better. These findings suggest relatively low levels of confidence in local health services or local capacity for cervical cancer treatment. This belief could also be emanating from numerous newspaper and online reports of affluent members of society including senior government officials and politicians who seek and receive treatment of chronic diseases abroad. Medical tourism has been reported to have a negative effect on the perceptions and development of local health service delivery [32].

Our study showed that the majority of women with cervical cancer had used screening services in their lifetime. However, they had limited access to regular doctors for health services and this could be a predisposing factor to late presentation or limited access to treatment. All the women enrolled in the study had at least locally advanced disease (>stage 1B2). These findings suggest the use of screening services for “diagnosis” among women with overt symptoms of cervical cancer. Despite the relatively high levels of awareness of cervical cancer, its prevention and treatment, our study suggests suboptimal routine uptake of screening for cervical cancer among asymptomatic women in Harare. These results are in tandem with ZDHS (2015) findings which reported that only 24% of healthy women had ever screened for cervical cancer in their lifetime [17]. These results are unexpected given the relatively high levels of knowledge of cervical cancer, prevention and treatment in our context. Our qualitative study explained some of the reasons of low screening uptake among them: lack of knowledge, beliefs that cervical cancer was not treatable, misconception that cervical cancer was caused by witchcraft or avenging spirits, fear of the unknown, limited screening services and negative health worker attitudes. Lack of funds for transport and negative attitudes towards screening and poor health seeking behavior with a common attitude that a person should only use health facilities when they are sick or in pain could be contributing to low cervical cancer screening uptake in our settings. Social norms that necessitate women to seek husband or partner approval or financial support to get screened may also explain low uptake of screening services. Considering other low-income settings, in Nigeria uptake of screening services was reported to be as low as 23% and this was not surprising given the low levels of awareness of cervical cancer [15]. Again in Tanzania, a study amongst nurses revealed low screening uptake, only 15%, using the Pap smear technique [6]. Surprisingly in a similar study in Japan, a high-income country, only 14% of nurses had screened for cervical cancer in the previous 2 years [9]. In contrast in Thailand, Visanuyothin and colleagues reported a high Pap smear uptake of 77% amongst women in a recent study [16].

This study encountered some limitations which would deserve to be accounted for in interpreting the findings. Firstly, this study was conducted in Harare, the capital city of Zimbabwe where cervical cancer awareness, screening and treatment services are largely available compared to other parts of the country. Therefore, the results of this study may not be generalizable to other contexts and a follow-up national study would be imperative in future. Secondly, the study was conducted at a time health workers had series of strikes and therefore there may have been selection bias in the characteristics of patients who were scheduled for reviews or treatments particularly at Parirenyatwa and Harare Hospitals. This study comprised of a cross sectional survey whose outcomes cannot be used to infer causal relationships. Our qualitative study could not explore the underlying issues behind watching TV being negatively associated with knowledge of causes and prevention and further studies are recommended. However, this study had its own strengths; firstly this study used mixed methods with a diverse study population to understand the subject under investigation better. Most studies cited have used either qualitative or quantitative methods and for policy recommendations mixed methods provide better outcomes [18, 33]. Secondly, this study, to the best of our knowledge, is the first in Zimbabwe to comprehensively investigate knowledge, attitudes, beliefs and practices about cervical cancer, its screening and treatment. Many previous studies focused on primary and secondary prevention of cervical cancer but there is a general paucity of information on tertiary interventions in the context of low-income settings.