Introduction

Cervical cancer is the fourth most common cancer in women, and globally, approximately 7.5% of cancer-related deaths among women are attributed to it [1–3]. Cervical cancer is caused by persistent human papillomavirus (HPV). Human papillomavirus (HPV) is one of the most common sexually transmitted diseases worldwide. HPV has more than 130 low- and high-risk serotypes, and high-risk serotypes (HPV 16 and 18) can cause cancers such as cervical cancer, although the low risk causes benign warts [4]. There is a risk for all women that HPV infection may become chronic and that precancerous lesions may progress to invasive cervical cancer [1].

In 2020, an estimated 604,000 women were diagnosed with cervical cancer, and approximately 342,000 deaths from the disease occurred worldwide, with low- and middle-income countries accounting for more than 90% of the deaths [5, 6]. In sub-Saharan Africa (SSA), cervical cancer is the second leading cause of cancer in women, accounting for 22.5% of all cancer cases. Moreover, approximately 117,316 females are diagnosed annually with the disease [7]. In SSA, the overall prevalence of cervical cancer screening is low [8]. Tanzania is one of the five countries with the highest cervical cancer rates in Africa [9]. The prevalence of cervical cancer in Tanzania is a significant health issue, with high global rates and the third highest mortality rate among countries eligible for support from the GAVI Alliance [10]. Tanzania has a prevalence of 89,693 (0.14%) cancer cases by 2022, with an annual incidence of 44,931(0.07%) and 29,743 (0.04%) cancer-related deaths [11]. However, the prevalence of cervical cancer in Tanzania is significantly higher, with 10,868 cases (24.2%), making it the most common cancer [11]. Moreover, the prevalence of cervical cancer in Malawi was lower than that in Tanzania (4701 (23.7%]). Cervical cancer predominantly affects young, tobacco smokers, those with a history of multiple sexual partners, those who have sexually transmitted diseases, immunocompromised, and uneducated females living in less developed countries where access to information about the disease, screening, and treatment is inadequate [6].

HPV vaccination and regular screening are recommended to reduce the burden of cervical cancer in low-income and middle-income countries [12]. However, ample gaps exist due to patient-level factors, health systems, and providers [13]. Owing to the higher risk of HPV positivity at the age of 30, the World Health Organization (WHO) recommends initiating cervical cancer screening every 5–10 years. Moreover, it encourages a minimum of two lifetime screens with a high-performance HPV test by the age of 35, and again by the age of 45 [1]. Lugol’s iodine (VILI), acetic acid (VIA), and Papanicolaou tests (Pap smear) are the simplest visual tests for cervical cancer screening [14]. They have all been proven to be effective for cervical cancer screening, although there are differences in the screening methods [14, 15].

In Africa, cervical cancer screening is available only in selected public health facilities, which presents the challenge of accessing women in rural areas [9]. Although cervical cancer screening in Tanzania is free of charge in government and church-based hospitals using VIA [16], the uptake of the screening program remains low, particularly in rural populations and those with low socioeconomic status. The low uptake of cervical cancer screening ranges from 7.9% to 8.2% in Tanzania [17, 18].

The WHO has established–90-70-90 targets for 78 low-income and lower-middle-income countries. These goals include (1) ensuring that 90% of girls receive proper HPV vaccination, (2) screening 70% of women for cervical cancer, and (3) providing treatment to 90% of women diagnosed with cervical cancer by the end of 2030. All countries have committed to eliminating cervical cancer as a public health problem by 2030, with the aim of screening 70% of women with a high-quality test by the ages of 35 and 45 [1]. Since 2002, the Tanzanian Ministry of Health and Social Welfare [19] has been working in collaboration with various partners, including the WHO, the International Agency for Research on Cancer (IARC), the international non-profit Jhpiego, and numerous local and international non-governmental organizations, to increase efforts aimed at preventing cervical cancer among Tanzanian women [3]. Cervical cancer screening is a critical factor in addressing the burden of the disease. Therefore, evidence on the determinants of cervical cancer screening is a burning issue. However, there is a paucity of data on individual- and community-level factors in cervical cancer screening in Tanzania. Prior studies have revealed that poor knowledge of cervical cancer screening, unfavorable religious attitudes, and the unavailability of screening services at women’s usual healthcare points significantly impact the uptake of cervical cancer screening [20, 21]. However, existing studies have limitations such as small sample sizes, data derived from facility-based surveys lacking national scope/relevance, and failure to consider community-level factors. This study aimed to examine the relationship between individual, household, and community factors and cervical cancer screening uptake among women of reproductive age in Tanzania, using multi-level regression analysis. These findings could aid in designing evidence-based interventions and measures aimed at improving the availability and accessibility of cervical cancer screening in low-resource settings (Tanzania). Policymakers, healthcare providers, and researchers in Tanzania can use these findings to make evidence-based decisions and implement interventions that help improve the screening uptake level.

Materials and methods

Results

Study participant’s descriptive characteristics

A total of 15,140 women of reproductive age were included in this study (Table 1). Of these, 5742 (37.93%) were aged 15–24 years and 8075 (53.34%) were women who attained primary education. Approximately 9736 (64.31%) patients were included in the study. Regarding household wealth status, around 3940(26.03%) were the richest, and 14,257(94.17%) had no health insurance. More than half (55.07%) had media exposure, 9004 (59.47%) owned mobile phones, and the majority (85.66%) had never used the Internet. Nearly half of the women (53.01%) visited the health facility in the 12 months and received permission to seek medical care. The ease of money to seek medical care and distance to health facilities were not significant problems for 14061(92.87), 9655(63.77), and 10780 (71.20%) women in Tanzania, respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Individual and community level characteristics of the study participants in Tanzania.

https://doi.org/10.1371/journal.pone.0315438.t001

Regarding the community-level characteristics of the study participants, 9739 (64.32%) were from rural areas, 4353 (28.75%) were from the Eastern region, 4385 (.28.97%) were from Lake, and 518 (3.41%) were from Zanzibar. Autonomy in decision-making was high in 10952 (72.34%) women in the community (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Proportion of cervical cancer screening uptake among women’s of reproductive age (15–49) in Tanzania, 2024.

https://doi.org/10.1371/journal.pone.0315438.t002

Random effect analysis

Null model.

This model is an intercept-only model without predictors. We examined whether the multilevel binary logistic regression model was more significant than the single-level binary logistic regression model by using the Likelihood Ratio (LR). The LR test results were statistically significant (p<0.05), indicating that the multilevel binary logistic regression model best fitted the single-level binary logistic regression analysis. Therefore, the LR-test test suggests using a multilevel binary logistic regression model. Four random-effects models were fitted, and the final model was chosen because it had the lowest deviance value (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Multilevel analysis of factors associated with cervical cancer screening among reproductive age women in Tanzania.

https://doi.org/10.1371/journal.pone.0315438.t003

Table 2 shows that the Intraclass Correlation (ICC) in the null model was 20.2%, indicating that 20.2% of the overall variability in cervical cancer screening was related to the variation between clusters. Moreover, the Median Odds Ratio (MOR) for cervical cancer screening was 2.39, indicating variability between clusters. If we randomly selected individuals from two different clusters, those in the cluster with higher cervical cancer screening had 2.39 times of odds of cervical cancer screening than those in the cluster with lower uptake of cervical cancer screening. The best-fit model was chosen based on the lowest deviance value (6354.6) (Table 3).

Discussion

Although cervical cancer is highly prevalent in Tanzania, it has not received the necessary attention. The national health policy lacks a dedicated screening strategy for cervical cancer and instead focuses primarily on infectious diseases. Even though the Tanzanian Ministry of Health and Social Welfare introduced visual inspection with acetic acid (VIA) for cervical cancer screening along with cryotherapy at more than 300 locations across the country [22], the uptake of cervical cancer screening among women of reproductive age was 7.28% [95% CI: 6.87%, 7.70%], which is far from the WHO 2030 target. Moreover, rural resident women in Tanzania have a lower uptake of cervical cancer screening than urban resident women. A possible reason could be that rural regions face various obstacles that hinder access to cervical cancer screening services, as well as broader healthcare options. Factors such as low health literacy, inadequate awareness of the available screening services, and poor infrastructure contribute to these challenges. Additionally, cultural beliefs significantly influence health-seeking behaviors, affecting the likelihood of individuals seeking cervical cancer screenings [23]. This finding is in line with studies by Ethiopia [24] and Uganda [25]. Furthermore, studies have shown that a lack of awareness and understanding of cervical cancer and its screening options significantly contribute to the low rates of screening uptake [26].

The uptake of cervical cancer screening in Tanzania (7.28%) is lower than the uptake of cervical cancer screening reported from five SSA countries [27], systematic review and meta-analysis in SSA [28], Zimbabwe (13.4%) [29] and Ethiopia (13.4%) [30]. A possible explanation might be the lack of political commitment (leaders are not prioritizing healthcare programs, and funds and resources for screening programs are transferred to other programs) [26], and sociocultural influences, such as myths about cervical cancer screening, may affect their enthusiasm to undergo screening [31]. Our finding is higher than the prevalence of cancer screening in Ghana (2.4%) [32]. This may be because there is no national policy program regarding cervical cancer screening in Ghana, which could result in a low prevalence of screening in Ghana [33].

Further, in the final model, we found that women’s age, educational status, employment, covered by health insurance, household wealth status, visiting health facilities in the last 12 months, owned mobile phones, and Internet use were significantly associated with higher odds of cervical cancer screening, whereas region was associated with lower odds of cervical cancer screening.

Women aged ≥ 25 years had higher odds of cervical cancer screening than women aged less than women aged less than 25 years. This is consistent with findings reported in Ethiopia [30], Kenya [34], Zimbabwe [29]. A possible reason might be that at this age, early signs or symptoms of cervical lesions may have led them to seek cervical cancer screening before the invasive stages [35, 36]. Moreover, at this age, most women give birth and receive health education about reproductive and sexual health issues, including cervical cancer screening.

Women’s educational status had a positive effect on the uptake of cervical cancer screening. Women who had attained primary, secondary, and higher education had higher odds of being screened for cervical cancer than women with no formal education. This is consistent with findings from Ethiopia, Nigeria, and Ghana [30, 37, 38]. A possible explanation might be that education is associated with better access and utilization of healthcare services, including cervical cancer screening [39, 40]. Moreover, having a high number of educated women may ease the dissemination of information to women with poor educational status by using different community spaces and social networks [41].

Employment was significantly associated with increased odds of cervical screening. This is in line with the findings of Ghana [42, 43]. The employment of women not only promotes financial independence, but also leads to enhanced reproductive healthcare services.

Women with a high household wealth index had increased odds of cervical screening compared with women from low households. This is consistent with the findings of Zimbabwe [29] and Kenya [34]. The uptake of cervical cancer screening services and the burden of cervical cancer are inexplicably dispersed among poor women globally. A possible explanation could be that the poorest women could struggle financially, which is a hurdle to accessing cervical cancer screening [34].

Women with a history of health facility visits in the last year had higher odds of receiving cervical cancer screening services than women with no history of health facility visits. This result is consistent with the findings of Kenya [34] and Peru [44]. A possible explanation could be that to effectively access screening services, it is essential for women to have a regular point of contact for healthcare [45, 46]. This is especially important in resource-poor settings, where visiting health facilities and interacting with healthcare professionals can increase women’s awareness of related health issues and motivate them to take preventive measures [47]. Studies have also consistently shown that a healthcare provider’s recommendation plays a vital role in determining whether a woman should undergo cervical cancer screening [48].

Women who own mobile phones have increased odds of being screened for cervical cancer compared to their counterparts. This could be because women who own mobile devices demonstrate better health-seeking behaviors, which may include preventive measures such as cervical cancer screening [49].

Women who resided in urban areas had higher odds of cervical screening than those from rural areas. This is in line with the studies conducted in Nepal [50], Ethiopia [30], and Uganda [25]. The possible explanation Urban residents may have important information about cervical cancer screening and access to health care services. Moreover, women from urban dwellers may have better educational backgrounds than their counterparts.

The availability of cervical screening services may be hindered in rural areas compared to urban areas because of the absence of screening centers within walking distance, limited transportation options, and cost of transportation [51, 52]. Additionally, urban areas may offer better opportunities for residents to access health information because of the accessibility of information sources, which could result in differences in screening utilization rates between urban and rural populations.

Women from the Southern Highlands administrative Zones of Tanzania were significantly associated with higher odds of being screened for cervical cancer compared to Eastern Zones; however, women from the Southern and Zanzibar Zones were significantly associated with lower odds of cervical cancer screening. A possible reason for this difference might be that women from the southern highlands of Tanzania are socioeconomically valuable [53], which will help them access education and healthcare services compared with the southern and Zanzibar administrative zones.

This study had both strengths and limitations. This study utilized a weighted pooled nationally representative TDHS survey in Tanzania, and a multilevel binary logistic regression analysis was conducted to obtain reliable estimates and standard errors. The sample size was sufficient to detect the true effects of the independent variables. However, readers should keep in mind the limitations of this study, as it cannot establish a causal relationship between cervical cancer screening and predictors owing to the use of cross-sectional data. Additionally, the outcome variable, cervical cancer screening, was self-reported. This may have introduced the potential for recall bias, which could underestimate the effect size in those who were not able to accurately recall having undergone cervical screening. There is also the potential for reporting bias, resulting in women reporting cervical cancer screening, thereby potentially overestimating the effect sizes.

Conclusion

Cervical cancer screening among women of reproductive age in Tanzania is low. Our findings identified both individual- and community-level factors associated with cervical cancer screening in women of reproductive age in Tanzania. Women’s age, educational status, employment, household wealth status, visiting health facilities in the last 12 months, owning mobile phones, place of residence, and region were significant predictors of cervical cancer screening uptake. Enhancing awareness campaigns and health education are essential for improving access to cervical cancer screening, particularly in rural areas and among low-income populations. Additionally, integrating cervical cancer screening into existing maternal health services such as contraceptive programs could significantly enhance participation and overall effectiveness. Moreover, to boost cervical cancer screening rates, policymakers and stakeholders must consider the risk factors associated with screening among women, especially in areas where the uptake of cervical cancer screening is low.

Acknowledgments

We would like to thank the TDHS program for providing the datasets.