Knowledge and experience of a cohort of HIV-positive and HIV-negative Ghanaian women after undergoing human papillomavirus and cervical cancer screening

We conducted a mixed-methods study using convergent parallel design with quantitative and qualitative strands of the study implemented concurrently, kept independent during analysis and combined in interpretation [33]. We used a combination of interviewer-administered questionnaires and focus group discussions (FGDs).

Participants were sampled purposively from a parent-cohort study of women at the Cape Coast Teaching Hospital (CCTH) in Cape Coast, Ghana, (parent study N = 343, HIV-positive n = 173). The parent study was a comparative-cohort study of HIV-positive and HIV-negative women that investigated the epidemiology of HPV and cervical squamous intraepithelial lesions (SIL); evaluated a rapid HPV DNA screening test (careHPV, Qiagen, Gaithersburg, MD), and determined the performance and acceptability of self-sampling for HPV testing [30‐32]. At parent-study enrolment, women underwent gynaecological examination, HPV-DNA testing and cervical cytology, and socio-demographic and medical history data were collected. During this first visit, they were also given basic facts about HPV and cervical cancer. At 6-months participants were followed up in clinic with review of symptoms, gynaecological examination, and repeat cervical cytology. Women found to have abnormal cytology after the first screening were referred for gynaecological evaluation and treatment. At the 6-month parent-study follow up visit participants were informed about the sub-study and consenting individuals consecutively recruited.

A 32-item questionnaire was developed specifically for this study through literature review and collaboration with the parent-study investigators. Information on knowledge (HPV, cervical cancer and screening) and screening experiences of the women was gathered. The questionnaire was piloted with administering study-personnel (four nurses and one doctor) prior to initiation. The questionnaire was developed in English, but administered face-to-face in the preferred language of the respondent. Clinic staff were fluent in local languages and English and there was no formal assessment of their language skills. Specific translations of each question into local languages were chosen during piloting. Demographic data, HPV and HIV testing results were extracted from the parent-study.

Focus-group discussions were held after the questionnaire collection phase had ended, and were run in the School of Medical Sciences teaching building at CCTH. There was no financial incentive for participation, but light refreshments were served and costs of travel reimbursed. Participants were stratified into two FGDs by HIV-status (eight participants in each). The aim of this was to 1) reduce the risk of stigmatisation through inadvertent disclosure of HIV-status during discussions, and 2) examine differences/similarities in results. Participants were not informed that their group was of particular HIV-status, and were not aware of other members’ HPV or cytology results.

A semi-structured topic guide using pre-scripted open-ended questions was used, focusing on knowledge and understanding of HPV, cervical cancer and screening; and on participants’ experience of the screening process. Both FGDs were conducted in the local language (Fante) and each lasted approximately 1h. Two female facilitators fluent in local languages and with experience of conducting FGDs, were recruited from the Institute of Development Studies at the University of Cape Coast. FGDs were recorded onto digital audio recorders, and field notes recording interactions were taken. Transcripts were translated into English and transcribed verbatim, and checked for accuracy. Thematic analysis was done using the Framework Method [34]. After familiarization with the transcripts, initial codes were generated manually using an inductive approach. Transcripts were then imported into Nvivo10 (QSR International) and fully coded, with generation of additional codes as they emerged. A framework matrix was created, with themes and sub-themes generated from the codes making up the columns of the matrix, and cases (individual participants), the rows. Associations, explanations and relationships were explored through the framework.

Results of the knowledge domains are presented in Table 2. Questions with the highest proportion of correct responses for HIV negative versus HIV positive women include HPV being sexually transmitted (86.7100%), HPV being the cause of cervical cancer (91.9% vrs. 98.2%), condoms being partially protective (82.7% vrs. 94.6%), and cervical cancer being preventable (90.8% vrs. 94.6%). Questions with the highest proportion of incorrect responses also tended to be those with the highest proportions of “unsure” responses. Questions showing difference by HIV-status included whether HPV is sexually transmitted (p = 0.01) and whether cervical cancer is rare in Ghana (p = 0.005), with HIV-positive women giving more frequent correct answers. HIV-negative women were more often correct in identifying that cervical cancer has stages (p = < 0.001).

The symptoms that survey participants associated with cervical cancer are detailed in Fig. 1. Postcoital bleeding, offensive vaginal discharge and intermenstrual bleeding were commonly correctly identified as potential symptoms (90.8, 90.1, and 70.1%). Pruritus was also a commonly misidentified as a potential symptom with 56.6% of participants selecting it. Identifying that one could still have cervical cancer but no symptoms was less common at 18.3%. More HIV-negative women (49.6% vrs. 40.5%, p = 0.04) correctly identified offensive vaginal discharge as a symptom, but there were no other statistically significant differences between HIV-status groups.

The mean knowledge score for all survey participants was 11.6 (SD + 2.7; range: 3–18), with no evidence of association with HIV-status (p = 0.1). An “adequate” knowledge score (correct responses for ≥9 of 19 items) was reached by 87.8% of participants, with no difference by HIV-status (p = 0.11). On bivariate analysis having a current partner was associated with having an adequate knowledge score (OR 2.53, 95%CI 1.03–6.24, p = 0.04).

Four themes relating to women’s knowledge of HPV, cervical cancer and screening were constructed from the FGDs: “acquisition”, “nature of disease”, “protection” and “testing”.

That HPV is sexually transmitted was commonly expressed, and the majority of women were clear on this with few incidences of misconceptions. The manner in which participants described the acquisition of HPV was categorised into “stigmatised” and “non-stigmatised”, with “stigmatised” expressions putting a critical or “moral” judgment on sexual transmission. These were more common amongst HIV-positive women, and were mostly related to the expression that sexual promiscuity was necessary for acquisition of disease:

“The way through which we come to get it [HPV virus] is if we and men about three or four have sex.” (40-49yrs, HIV-positive, HPV-positive, cytology negative)

“Non-stigmatised” acquisition was expressed in the form of simple statements that HPV is sexually transmitted. Only one participant mentioned HIV-status affecting vulnerability to infection:

“I think when the doctor was teaching us she made me understand that it can attack everyone, but for those of us who specifically have HIV, it is easy that we can get it [HPV virus]” (30-39yrs, HIV-positive, HPV- and cytology-negative)

The most common misconception expressed about acquisition was the effect of abortion (only expressed by HIV-negative participants):

“I was told that women who cause abortions, those who have had cases of STDs, and those who have a lot of sexual partners are all at risk of getting infected.” (60-69yrs, HIV-negative, HPV-positive, cytology-negative)

The early asymptomatic phase of cervical cancer made it more difficult for women to define “what” cancer was; for them cancer was something you could see or feel, and identify by characteristic symptoms. This also caused difficulty in identifying “what” HPV was, and in reconciling its lack of symptoms with its harmful nature:

“At first, I thought cancer only affected the breast but we went for a workshop and heard that it can also affect the mouth of the womb, but that the cancer that affects the mouth of the womb has no symptoms because for the breast cancer you can see that your breast has lump in it and look some style (does not look normal) but they said for the cervical cancer it will be there without showing so we should start doing some test” (40-49yrs, HIV-positive, HPV- and cytology-negative)

Only two (HIV-negative) women correctly expressed that HPV is an infection that can lead to cancer if it persists. When asked to describe the symptoms of cervical cancer and HPV they were again described interchangeably. Several women believed that HPV infection would cause vaginal discharge was expressed. One woman described the symptoms of advanced cervical cancer as:

“vaginal discharge with very bad smell, abdominal pains, a lot of complications, loss of appetite, sores in the mouth, etc. The cancer affects every part of the body.” (50-59yrs, HIV-negative, HPV-positive, cytology-negative).

A strong sub-theme of “nature of the disease” was the effect of HPV on men. Women queried whether it would also cause disease in men, or whether there would be any visible signs on men that they were infected with HPV:

“Does it mean that this HPV it can be on [infect] males and it can be on females? But for the men when they have it doesn’t it show on their bodies or it does it not give them any problems?” (40-49yrs, HIV-positive, HPV- and cytology-negative)

Some women expressed the belief that HPV would cause visible symptoms in an infected man:

“Me too I think that if you are there with someone and he has it, you will see that his skin is changing” (40-49yrs, HIV-positive, HPV positive and cytology negative).

There were also feelings of frustration from participants that men could engage in sexually “careless” behaviour without fear of consequences as:

“the disease will affect you the woman” (60-69yrs, HIV-negative, HPV-positive, cytology-negative)

Condoms were mentioned frequently and emphatically by participants as being important for protecting themselves and this was phrased in absolute terms:

“I learnt that when you have sex and you protect yourself you will not be infected, usually through condom use.” (20-29yrs, HIV-negative, HPV- and cytology-positive).

Modifying sexual behaviour by limiting one’s number of sexual partners, or exhibiting sexual ‘restraint’ was also expressed as an important factor in prevention, and was commonly referred to as “being careful” or “taking care”:

“Eeeem … some people do not like condom so they will have to, reduce the men they do sex with and all that” (40-49yrs, HIV-positive, HPV- and cytology-negative)

“All we have to do is to take care of ourselves from this disease. Like pulling ourselves away from some things that when we do, it will not go and bring us this problem.” (30-39yrs, HIV-positive, HPV- and cytology-negative)

Participants expressed feelings of powerlessness in protecting themselves from disease when discussing their partners. “Fidelity” and “sexual negotiation” were predominant sub-themes. Male partners’ infidelity was expected:

“Because it is got from men, its prevention will be difficult. A woman cannot advise her husband to not have sex with other women.” (50-59yrs, HIV-negative, HPV- and cytology-positive).

Women also believed that because in general men did know about HPV or cervical cancer they would not be taking any steps to reduce the risk of transmitting HPV to their female partners, and that education of men would be an important element in protecting women:

“If education could be given on radio and TV for me, I believe it would help us.” (50-59yrs, HIV-negative, HPV-positive, cytology-negative)

Responses to the Likert-type questionnaire items are presented in Table 3. Responses did not vary by HIV-status. Among HIV negative versus HIV positive women, a third of women (30.7% vrs. 35.0%) agreed that screening was embarrassing but most women did not find screening painful (85.3% vrs. 85.0%).

The themes of “protection” and “testing” also emerged from the FGD aspects focussed on experience of screening, but the theme “fear and anxiety” was additionally constructed. Women were unanimous in their expression that learning about HPV and cervical cancer was frightening; and for many this related to simultaneously learning of the disease, and that they were at risk, for the first time:

“Please, as for me it got me really scared, because me, myself, I had not heard some. Before I knew it is only breast that cancer affects, so me it got me very scared” (40-49yrs, HIV-positive HPV-positive, cytology-negative)

Other elements that generated anxiety were: the “not knowing” that one could be infected with HPV due to its asymptomatic nature; what this meant for one’s relationship to one’s partner (and the question of infidelity); and media messages about cervical cancer:

“What really pains me a little about it, is that when you get it you won’t get any symptoms that this is what is happening to you” (40-49yrs, HIV-positive, HPV- and cytology-negative)

“ The way the doctor said it on TV made it sound scary.” (40-49yrs, HIV-negative, HPV-positive, cytology-negative)

When asked how they felt whilst waiting for their results, women commonly expressed that they found it difficult but receiving information and counselling from health workers alleviated fears and anxiety as well as the fact that testing would help discover any problems at an early stage:

“I was scared, waiting for the result was very scary. I could not even sleep.” (60-69yrs, HIV-negative, HPV-positive, cytology-negative)

“It was scary at first but later I became relaxed. I also was glad about the fact that I took the test because then if there had been something it would be found and stopped.” (40-49yrs, HIV-negative, HPV-positive, cytology-negative).

When asked if they had sought information from other sources between initial screening and follow up, under half said “yes” (41.4,40.0%). The source of information with the greatest proportion of responses was “other healthcare professional” at 45% of responses (Table 4.)

In FGDs women also talked about the cost of testing. Some women were aware of cervical screening prior to joining the parent-study but said they had not availed of it due to cost. It was also mentioned that the government should increase the availability of screening by reducing the cost:

“The government should also try and reduce the cost for us” (50-59yrs, HIV-negative, HPV-positive, cytology-negative).

In FGDs, the view that screening was protective against developing disease through both informing and educating women and detecting disease early was frequently expressed:

“If a test is done it would help prevent any further damage the disease would have caused to the womb if the result is positive. If it is negative, then you will be educated on how to stay safe or protect yourself.” (50-59yrs, HIV-negative, HPV-positive, cytology-negative)

The sub-theme of “testing imperative” related to multiple expressions that testing was something that must be done if the opportunity presents itself; and that other women should seek testing. However, this was only voiced by women who had had a positive HPV result:

“Whether morning or afternoon, wherever they call you for the test you will have to do it.” (40-49yrs, HIV-positive, HPV-positive, cytology-negative)

“This should be of greater concern to all women so that from time to time we can run the tests.” (40-49yrs, HIV-negative, HPV-positive, cytology-negative)

This study is the first in West Africa that examines the knowledge and experiences of women who have undergone cervical screening with two screening methods (HPV-DNA testing and cervical cytology). It provides a unique perspective on the impact of health education from screening, and psychological experiences of screening in women of mixed HIV-status. Both the questionnaire and scoring system used were developed specifically for this study and were unvalidated. There was no pre-existing validated tool suitable for use with this particular population.

Furthermore, we did not assess knowledge levels before participation in the study and cannot assume that knowledge “gains” were made, although it seems likely based on participant statements during the FGDs and comparison with existing research in similar settings. In addition, since these women had undergone screening and received some education from healthcare workers, it is acknowledged that they do not necessarily reflect the general population hence their knowledge may not represent women across Ghana and differences between HIV-positive and negative women needs to be explored further.

Ideally, questionnaire responses would have been used to develop the FGD topic-guide, and more FGDs should have been conducted to verify if saturation was achieved. However, this was not possible due to time constraints. As a result, there were aspects of each study component that were not explored in the other, causing some loss of cohesiveness.