Utilisation of cervical cancer screening among women living with HIV at Kenya’s national referral hospital

Background In 2009, Kenyatta National Hospital (KNH) integrated cervical cancer screening within HIV care using visual inspection with acetic acid (VIA) and Pap smear cytology. Objectives We evaluated utilisation of cervical cancer screening and human papillomavirus (HPV) vaccination among women living with HIV (WLHIV) receiving HIV care at KNH. Method From November 2019 to February 2020, WLHIV aged ≥ 14 years were invited to participate in a survey following receipt of routine HIV services. We assessed awareness of cervical cancer, uptake of cervical cancer screening, uptake of the HPV vaccine, and barriers to utilisation of these services. In a subset of survey participants, focus group discussions (FGDs) were also conducted to identify screening barriers. Results Overall, 305 WLHIV participated in the survey. Median age was 36 years (interquartile range [IQR]: 28–43), 41% were married, and 38% completed secondary education. Most (90%) had HIV RNA < 1000 copies/mL. Awareness of cervical cancer was high (84%), although only 45% of WLHIV had screened for cervical cancer at the referral hospital and only 13% knew how to prevent high-risk HPV. No participants had received an HPV vaccination. Older age, higher education, and knowledge of the HPV vaccine were associated with higher likelihood of cervical cancer screening (P < 0.05). In FGDs, barriers to utilising the services included user fees, fear of the procedure impacting fertility, age and gender of the provider, and long waiting times. Conclusion Despite integration with HIV services, the utilisation of cervical cancer screening was low among WLHIV and implementation barriers contributed to low utilisation.


Utilisation of cervical cancer screening among women living with HIV at Kenya's national referral hospital
Read online: Scan 13 In Kenya, the cervical cancer screening uptake is higher among WLHIV (14%) compared to 3% among HIV-negative women. 14 The Kenyan HPV vaccination programme launched in 2019 targeting HIV-negative girls aged 10-15 years and adolescent girls and young adult WLHIV aged 10-24 years. The type of HPV vaccine available to protect against HPV infection is Gardasil which is a quadrivalent vaccine and protective against HPV types 6, 11, 16 and 18. 13 More data are needed to understand utilisation within the context of integrated service delivery in HIV care programmes to guide implementation improvement efforts.
According to Kenyan national guidelines, annual cervical cancer screening for WLHIV 25-49 years is recommended; however, age is not limiting and any sexually active WLHIV is eligible for screening. 13 In 2009, Kenyatta National Hospital (KNH) integrated cervical cancer preventive services within its Comprehensive Care Center (CCC) for HIV care and treatment.
To date, no evaluation of cervical cancer service utilisation has been conducted at the referral hospital. We aimed to evaluate the utilisation of cervical cancer prevention services and identify barriers to utilisation among WLHIV enrolled for care through a cross-sectional survey complemented by focus group discussions (FGDs) among a subset of participants. The overall aim of our evaluation is to guide implementation efforts to improve delivery and utilisation of cervical cancer prevention services by WLHIV in SSA.

Research methods and design Study design
This was a mixed methods study that included both exploratory qualitative analysis and a cross-sectional quantitative survey among consecutive sample of WLHIV.

Study site, population and recruitment
Based on programmatic data, the clinic annually serves approximately 10 000 patients on HIV treatment, with roughly half being WLHIV of reproductive health age. All WLHIV seeking services at the CCC who were ≥ 14 and < 25 years, aware of their HIV status, and willing and able to consent were eligible to enrol in our survey.

Data collection and analysis Quantitative survey
Structured, pre-tested paper-based questionnaires captured socio-demographic characteristics and factors influencing the utilisation of cervical cancer services. Knowledge, attitudes, and current practices regarding cervical cancer screening were also assessed (Appendix 1). The questionnaire was piloted among WLHIV seeking care at the study site before study recruitment started. During the pilot period, the appropriateness of the tool was assessed by how well respondents understood the questions and if the responses were congruent with the information we planned to obtain. The questionnaire was administered in English and clarifications were made by the RA when requested by respondents. Utilisation was defined as self-reported prior receipt of HPV vaccination (for participants < 25 years) or cervical cancer screening at CCC using VIA or Pap smear cytology. The satisfaction of participants with healthcare workers' (HCW) counselling services was defined using a numerical rating scale: 1 = poor, 2 = average, 3 = good, 4 = excellent. The satisfaction of the HCW's counselling was based on the current HIV clinic visit at KNH CCC. Clinical data were abstracted from medical records upon obtaining consent. The relationships between participant characteristics and utilisation of cervical cancer services were evaluated using multivariate logistic regression models. We determined a priori to adjust all multivariate models for age (years) and education level (primary and below vs above primary education). Data were analysed using Statistical Package for Social Sciences (SPSS) version 23.0.

Qualitative focus group discussions
Four (4) FGDs, each with six participants and lasting 30-40 min, were conducted, recorded, and transcribed. English and Kiswahili languages were used interchangeably and later translated to English. Qualitative data were deductively analysed and categories were constructed for content themes. Patterns of the content appearing repeatedly in the data formed the basis for themes. Themes were grouped to provide an integrated explanation of why participants utilised or did not utilise cervical preventive services at the CCC.

Ethical considerations
An application for full ethical approval was made to the University of Nairobi, Kenyatta National Hospital Ethics and Review Committee (UON/KNH ERC) and ethics approval was received in October 2019. The ethics approval number is P109/02/2019. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committees and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Participants < 18 years signed an assent form after obtaining parental or guardian written consent. Participants aged 18 years and above provided written informed consent. Consent forms were available in participants' preferred language, either English or Kiswahili.

Results
In total, 525 WLHIV were approached and screened for eligibility. Of the 371 eligible WLHIV, 305 consented and were enrolled into the study ( Figure 1).
Of the 371 WLHIV who were eligible to participate, 66 WLHIV did not consent to participate in our study as they had no time or declined without a reason.
The median age was 36 years (interquartile range [IQR]: 28-43), 10.5% of participants were < 18 years, 41% of participants were married, and 38% had a secondary level of education (as shown in Table 1   Several participants were not comfortable with male providers, indicating that their culture did not allow them to be seen by a male provider for such invasive procedures. One participant reported: 'It matters whether it's a male or a female [provider] because if a female tells you she will do the screening, you will be at ease and consent to screening as opposed to if the screening was done by a male as you feel victimised. Old mature ladies take their time to address most of your concerns before they start attending to you, unlike the young service providers'. (Participant 9, female, 41 years old) In some instances, (e.g. stock-outs, etc.), screening is not conducted in CCC and WLHIV are referred to the another unit in the referral hospital. This additional navigation through the hospital increases loss to follow-up as expressed by one participant: 'I was sent by the doctor to be screened at a different clinic from CCC. I went and waited for two and a half hours and still, no one came to attend to me. Later, I was rebooked to come for screening 10 days later. My work schedule isn't flexible'. (Participant 3, female, 38 years old) Some women did not perceive themselves to be at risk due to the absence of cervical pain or abnormalities. Adolescents believed that cervical cancer was a disease of older women. Others claimed they had only one sexual partner and thus were not at risk for cervical cancer. Many women believed that cervical cancer is a curse from God with no cure or a punishment from the ancestors due to the wrongdoings as expressed by one participant: 'Cervical cancer is a curse from God, even if you know you have it, nothing will change and no treatment will fight it. You will end up being completely depressed'. (Participant 2, female, 49 years old)

Discussion
Our study of utilisation of cervical cancer prevention and screening services integrated within HIV care at Kenya's largest national referral hospital found relatively low uptake among WLHIV and identified key barriers to utilisation. In our study population, less than half of WLHIV had screened for cervical cancer despite implementation of integrated cervical cancer screening since 2009. This is higher than previous studies in Nairobi (19.0%) 18 and a recent review of screening coverage rates in SSA which found ranges from 2.0% to 20.0% in urban areas and 0.4% -14.0% in rural areas, 19 but lower than South Africa where cervical cancer screening rates were 54.0%. 19 The higher screening rate in South Africa could be the result of the implementation of approaches where WLHIV self-collected their samples for examination. 20 Our results add to growing evidence that more efforts are needed to improve integrated delivery of cervical cancer screening within HIV care and to increase utilisation of services among WLHIV in SSA, a population disproportionately affected by cervical cancer. Further, periodic audits of screening activities and monitoring of facilities for quality assurance could help improve screening practices. In Kenya, the cervical cancer screening is higher (14%) among WLHIV compared to the HIV-negative women (3%), based on prior studies. 14,21 This may be attributed to higher engagement in care, with the majority of WLHIV on ART which requires routine visits to health facilities for medication refills.
In our study, the obstacles to screening utilisation included fear of painful pelvic procedures associated with VIA and Pap smears, avoidance of invasive procedures, encountering male healthcare providers when prompted to screen, and interference with future fertility, similar to other studies. 22,23 Utilising novel approaches to cervical cancer screening with self-sampling may increase acceptability of cervical cancer screening in this setting. The WHO recommends molecular testing for HPV as primary screening for cervical cancer in low-and middle-income countries using platforms such as Xpert HPV ® which has better sensitivity than cytology and VIA, even with self-sampling. 24,25 To date, Xpert HPV ® has not been evaluated as a strategy to improve cervical cancer screening utilisation in Kenya and could be one approach to addressing implementation barriers identified in our study. However, the governments's support is required in the implementation and sustainability of this novel approach. This will ensure access to free or subsidised HPV screening among WLHIV. The fear of screening, fatalism and fertility †, Cervical cancer screening was done using either using VIA or Pap smear cytology. ‡, All multivariate models were adjusted for age (years) and level of education (primary and below or above primary).
concerns can be addressed through routine health education, peer to peer education, counselling to address psychological barriers and media campaigns similar to what has been used to encourage populations to screen for HIV. 26 General anxiety tools such as a questionnaire need to be considered as they can play vital roles in assessment of fear, anxiety and concerns, thus ensuring these barriers are addressed promptly. 22 Having an STI in the last 6 months was not significantly associated with cervical cancer screening contrary to other studies that have been conducted in the region. However, diagnostic testing for STIs was not conducted in our studies and STIs rates are likely higher.
Similar to prior studies, 27 we found that older age and higher education were associated with increased likelihood of screening. Older women with HIV have accessed the health system longer than younger women and may be more sensitised to cervical cancer, thus increasing their likelihood of screening compared to younger women. 28 Our findings support that sensitisation of cervical cancer prevention and screening implementation should be prioritised for AGYW living with HIV with interventions to raise awareness tailored to this population. There was no difference in screening frequency among married and unmarried WLHIV which differs from prior studies among African women. 29 In Kenya, women from some regions marry at earlier ages and therefore married women may reflect a younger age group than prior studies outside of Kenya. 30 Cost was also a barrier to utilising cervical cancer services, especially HPV vaccination. There is a need to address modifiable factors contributing to low uptake of prevention services among young women and those with financial constraints. Cost of the HPV vaccination was reported as the key barrier to accessing the vaccine and no eligible study participants had been vaccinated. These findings are comparable with those of women recruited in a community-based cervical cancer screening programme in rural western Kenya where screening and HPV vaccination uptake was poor. 31 One approach to reducing barriers to HPV vaccination uptake among young Kenyan WLHIV is offering free or incentivised cytology (when indicated) and vaccination options combined with targeted health education. 32 The KEN-SHE study evaluated single-dose HPV vaccine efficacy among AGYW in Africa and has the potential to guide public health policy and increase HPV vaccine coverage by reducing the number of required doses. 33 Provision of free services does not guarantee increased uptake, particularly in SSA. Other additional costs such as transport to the health facility, lost wages and childcare cost could explain why free services are still resulting in poor uptake. 28 Women with lower social economic status have also been shown to have a negative attitude towards cervical screening even if freely offered, thus warranting the need for more health information to the specific target groups based on factors like particular age or social economic status before implementation of the interventional strategies. 34 Introducing HPV vaccination programmes for all children aged 9-15 years, ideally before sexual debut, would be especially useful for preventing HPV.
Our study has limitations. Our funding allowed us to hire only one RA to recruit participants which limited our ability to enrol all eligible clients and extended our recruitment period to attain the minimum required sample size. However, consecutive sampling was used and this limitation likely did not bias our results.
We did not distinguish between screening via VIA or Pap smear when ascertaining screening utilisation, so we were unable to determine the factors associated with each approach. Although representative of WLHIV receiving HIV care at the CCC, our sample included only a limited number of adolescents and therefore our results specific to that age group (e.g. HPV vaccine uptake) should be interpreted with caution. We did not ascertain timing or frequency of prior cervical cancer screening. These data would help elucidate the proportion of WLHIV who meet the national guidelines for cervical cancer screening. We only asked if participants were screened for cervical cancer at the CCC and not within a certain timeframe. It is possible that screening could have taken place at another facility outside of the study site. However, our eligibility criteria for participants ≥ 25 years included being a patient at the hospital for ≥ 1 year. The satisfaction with the HCW's counselling was based on the current HIV clinic visit; however, we did systematically ascertain reasons why participants were not satisfied with the counselling they received. Anecdotally, participants reported that incomplete information was provided on the availability of prevention services. This study was conducted at a single referral hospital in an urban setting and hence may not be generalisable to Kenyan WLHIV in rural settings. Additionally, self-reported data may be influenced by social desirability and recall bias. Our study did not interview the healthcare providers or other stakeholders in the programme who play a key role in the service provision and sustainability.

Conclusion
In this mixed methods evaluation of cervical cancer screening utilisation among WLHIV receiving HIV care at Kenya's national referral hospital, utilisation was relatively low. Barriers to utilisation such as fear of painful invasive procedures could potentially be addressed by integrating self-sampling approaches to cervical cancer screening such as Xpert HPV ® . Negative attitudes towards screening could be addressed by awareness campaigns via mass media. There is a need to improve screening recommendations offered by HCW to WLHIV at diagnosis, ideally at each healthcare visit to avoid missed opportunities.

Data availability
The data sets used and analysed during the current study are available from the corresponding author, K.J.M., on reasonable request.

Disclaimer
The statements made and views expressed are solely the responsibility of the authors and not an official position of the institution or funder.