Uptake of cervical cancer screening and the associated factors among women living with HIV in Northern Tanzania.

Background: HIV infection is a common risk for developing cervical cancer (CC). Routine screening for CC among women living with Human Immune Deficiency Virus (WLHIV) is recommended for early detection and control of pre-malignancies. Evidence on CC uptake and its associated factors is scanty among WLHIV in Tanzania similar to other sub-Saharan Africa (SSA) countries. This study therefore aimed to assess the uptake of CC screening and its associated factors among WLHIV in Tanzania. Methods: This cross-sectional study was conducted between June and September 2020 among WLHIV attending Care and Treatment Center (CTC) at the Kilimanjaro Christian Medical Center (KCMC). Data was collected through face to face interview using a pre-tested standardized questionnaire interviewed in Swahili. Analyses were conducted using descriptive statistics to establish the CC uptake and using regression analyses to characterize the CC screening uptake and factors associated with the CC uptake through SPSS version 23 software. Associations with P<0.05 were considered statistically significant. Results: A total of 341 WLHIV with mean age 45.6 years (SD 10.8) were recruited for interview. Of them, 184 (54%) WLHIV reported ever being screened for cervical cancer. After adjusting for confounders, knowledge of the screening methods was one of the factors associated with uptake of CC screening [AOR=15.61, (95% CI: 7.93-30.72), p<0.0001]. Other factors included living with HIV for at least 10 years since diagnosis [AOR=2.83; (95% CI: 1.11-7.26), P=0.030]; having knowledge of CC [AOR= 1.75, (95% CI: 1.02-3.01), p=0.041]; and having knowledge of the signs or symptoms of CC [AOR=1.95, (95% CI: 1.17-3.27), p=0.011]. Conclusion: More than four in ten WLHIV attending CTC at KCMC have never been screened for cervical cancer. Knowledge of the available screening methods, the disease condition, and duration since fist HIV diagnosis were


Introduction
Cervical cancer (CC) is the major global cause of death, responsible for nearly 570,000 new cases annually 1 . The burden of this disease is higher in developing countries compared to developed countries, especially in the Eastern and Western parts of sub-Saharan Africa 2 . The high prevalence of human papillomavirus (HPV) has been reported in the most African countries, and accounts for the majority of cases of CC, along with other risk factors such as human immunodeficiency virus (HIV) infection 3 .
Like in other countries in the Sub Saharan Africa, Tanzania has a high incidence of CC which remains the most common cause of deaths among women of reproductive age 4 . For example, while about 7,300 women in Tanzania are diagnosed annually, more than half of them die from the disease. Such unprecedented case fatality rates is caused by late diagnosis and therefore care and treatment 5,4 . The World Health Organization (WHO) reported limited access to effective screening tools for CC in most developing countries, resulting in late diagnosis with poor outcome and high mortality 6 .
The risks associated with CC have been well studied 4 . They include early first sexual intercourse, multiple male sexual partners, and HIV infection, among others 4 . These can lead to chronic HPV infection, transforming cervical epithelial lining, and therefore progression to CC. A higher incidence of HPV infection has been found among HIV positive women, which can develop into premalignant lesions of the cervix 7 . The WHO report highlights the needs for routine screening among WLHIV who are at a 10 years earlier risk of developing invasive CC compared to their counterparts 2 . In a nearby country, evidence suggested that WLHIV were twice as much more likely to have pre-cancerous pathology compared to their negative counterparts 8 .
Although it present with significant mortality risk, CC remains only gynaecological cancer that can be prevented during the pre-cancerous stage by providing early treatment 9 . The WHO recommends that all women aged 30-49 years need to be screened for CC regardless of their serostatus. Further, all sexually active girls and women should be screened once diagnosed as HIV positive 10 .
The age-standardized rate for CC mortality in Eastern Africa is higher than in other regions of the world 11 . Tanzania, has a striking 9772 new cases and 6695 deaths each year 12 . In Tanzania, most hospitals perform visual inspection with acetic acid to identify pre-cancerous lesion and use cryotherapy as the first modality in treatment. Despite having free cervical cancer screening services in Tanzania, the number of women accessing it is low 13 . To reduce CC mortality in Tanzania, women need to access CC screening services and have pre-cancerous lesion treated earlier 14 . Despite its sheer burden and mortality, little is known about CC screening services uptake and its associated factors in Tanzania general, and particularly among HIV infected women. Therefore, this study aimed to assess the uptake of CC screening and its associated factors among women living with HIV. Thus, findings emanating from this study are expected to help policy/decision-makers, health care providers, community service providers and non-government organizations (NGOs) design strategies and implement appropriate interventions accordingly.

Study Design and Population
This hospital-based cross-sectional study was conducted from June to September 2020 among WLHIV attending the Kilimanjaro Christian Medical Centre (KCMC) Care and Treatment Clinic (CTC). The study included women aged 18-70 years attending CTC and followed for at least 1 year from the date of HIV diagnosis.

Sample Size
The sample size was obtained using the precision approach with a single proportion 15 . The proportion of WLHIV who had ever been screened for CC was obtained from the previous study by Wanyenze et al. in 2017 in Uganda (30.3%) 7 . The minimum sample size was thus 325. To counteract the effect of incomplete variables, non-responses, and preserving the minimum sample size, an additional 25 participants were recruited giving a total of 350 WLHIV attending KCMC CTC.

Sampling Technique
A systematic sampling procedure was used to select eligible participants. All participants who presented at the CTC for a follow-up visit on every clinic day (Monday, Wednesday, Thursday and Friday) were checked for eligibility criteria at the entrance desk where sequential numbers were given according to the arrival order. Selection started with the first client of the day and used the sampling interval of two to select subsequent participants until the end of that clinic day. To avoid multiple enrollments of the same participants, researchers used stickers on each participant's file record.

Data collection tool and method
A pre-tested standardized questionnaire developed a prior with some questions adapted 16 , 17 , was employed in this study. The questionnaire included information on demographics (participant age, level of education, marital status, occupation and monthly income), awareness of CC (presence of the disease, risk factors, symptoms, and prevention) and participant's history of CC screening (date, screening method, and the number of screenings conducted). The questionnaire was prepared in English and translated to Swahili. We conducted two-day training for data collectors before the initiation of data collection. The training package included the study objectives, methods, and the study tool. We further pre-tested our tool with 15 women with the aim of correcting inappropriate responses. Four medical students in year 4, two diploma nurses and one master's degree graduate nurse were employed for data collection.
The participants were given an opportunity to ask questions and get whatever clarification they needed. Also, HIV data was collected (date of HIV infection, diagnosis, history of CD4 count measures, clinical stage and antiretroviral therapy use) from the CTC records of each participant.

Variables and measurements
The uptake of CC screening 'ever been screened for CC since confirmed HIV positive' was defined as a dichotomous dependent variable (yes/no). Independent variables were demographic characteristics, HIV clinical characteristics and variables related to awareness of CC (having heard about CC, being aware of the relationship between HPV and HIV, being aware of the CC prevention methods, and having been screened for CC). Participants' age was categorized as ≤45 years and >45 years. The age categorization was according to cut point of early reproductive age with its counterpart. Marital status was categorized as single/never married, married and divorced/widowed. The education level was categorized into three groups; primary or no formal education, secondary level for women who attend at least one class in secondary school, and university /college level. Participants' occupation was categorized into two-employed or unemployed. The number of children was dichotomized as having ≤ 2 and having ≥3.
Knowledge about the risk of CC was considered good if a respondent mentioned at least 3 from listed 10 known risk factors; otherwise, was considered to have poor knowledge on the risk of CC.
Knowledge on the signs/symptoms of CC was considered good if the participant mentioned at least 3 from listed 6 known signs/symptoms; otherwise, she was considered to have poor knowledge on the signs/symptoms of CC. Knowledge of CC prevention was considered good if the participant mentioned at least 3 out of 5 known prevention methods; otherwise, they were regarded as having poor knowledge about CC prevention. Knowledge about screening methods for CC was considered good if the participant reported having ever heard of at least 1 of the known methods, i.e. VIA or Pap smear, otherwise they were classified as having poor knowledge about screening methods of CC. The time since HIV diagnosed was categorized into three modalities; <5 years, 5-9 years and >= 10 years.
Current HIV staging was categorized according to WHO references as stage I, II, III, and 1V.

Data management and analysis
The completed questionnaires were reviewed and cross-checked before entry into SPSS 23.0 for statistical analysis. Descriptive analysis was reported using proportion and percentages. Bivariate logistic regression analysis was done for the crude odds ratio (COR) to determine the independent factors associated with uptake of CC screening. From the bivariate analysis, variables which showed statistically significant associations with the uptake at p<0.05 were run using multivariate logistic regression analysis for the adjusted odds ratio (AOR). Multivariate analysis was conducted to control for the possible attributed factors (confounders and modifiable factors). Variables with 95% CI that did not include 1 or P<0.05 were regarded as statistically significant factors for the uptake of CC screening.

Schematic presentation of the study participants
A total of 350 WLHIV attending CTC clinic at KCMC were enrolled in this study. Nine of them were not eligible and were excluded in the analysis. Thus, a total of 341 WLHIV were included in the final analysis. Focusing on screening assessment, 184 reported ever screened for CC in their lifetime since confirmed HIV positive; this indicates an uptake of 54%.

Characteristics of women living with HIV enrolled in the study
The participants had mean age of 45.6 years, with a standard deviation (SD) of 10.8 years, more than half 175 (51.3%) were aged above 45 years old. The marital status of the participants, 124 (36.4%) were married while 149 (43.7%) were widowed/divorced. The majority 219 (64.2%) had primary or no formal education. An average duration since confirmed HIV positive was 10 years with a standard deviation of 5 years. The majority (59.5%) had at least ten years since being confirmed HIV positive.
The length on ART was 9 years on average, with an interquartile range of 5-13 years since initiated ART. Majority 298 (87.4%) were on the first-line treatment.

Factors associated with cervical cancer screening uptake among WLHIV
Positive associations were observed among women aged above 45 years who had 1.24 times higher odds of screening uptake as compared to the younger group aged less than 45 years, but this was not screening uptake compared to women with lower income earners (<150,000 Tsh. per month), although this association did not reach a statistically significant level (p=0.840). The influence of partners seems to contribute to the uptake of CC screening; women with partners had 1.  (Table 1).   (Table 3). The current study also found that knowledge of the screening methods was the strongest factor associated with uptake of CC screening. This was linked with the health information's which are shared by health care providers every clinic day before provision of CTC service. This observation is consistent with the study in Uganda by Wanyenze which also reported that the knowledge of CC screening, where to go for screening, and low perception of risk were significant factors for CC screening 7 .

Limitations
This study's findings might be limited by self-reported information from our participants concerning the uptake of CC screening, however we verified by using cervical cancer screening cards. This study was conducted in an urban area with different characteristics compared to rural regions, including awareness and knowledge of CC. Therefore, this result may not be generalizable to other health facilities with different population characteristics from the current setting.

Conclusion and recommendations
The uptake of CC screening was low among WLHIV at the study site. Knowledge of the available screening methods was the strongest factor associated with uptake of CC screening. Other significant factors for the uptake of CC screening were the duration of HIV since diagnosis 10 years or above, and having good knowledge of CC and signs or symptoms of CC. There is a need to establish CC screening services at CTC that would help with easy counseling of WLHIV on CC screening including the increased risk of CC among WLHIV; this approach will strengthen the understanding of women on the importance of CC screening and reduce negative attitudes, and hence lower the morbidity and mortality associated with this disease. The ethical approval was sought from Kilimanjaro Christian Medical University College research ethics committee prior to the commencement of this study. This was followed by a letter to KCMC obstetrics and gynaecology department to get permission. Informed consent was written and signed by every participant prior to the interview and a unique identification number was used to enhance confidentiality. All methods were carried out in accordance with relevant guidelines and regulations of BMC.

Consent for publication
Not applicable

Availability of Data and Materials
The datasets generated and/or analysed during the current study are not publicly available as this study will progress into second phase however it can be available from the corresponding author on reasonable request.