Abstract

Background:

Breast and cervical cancers are among the most frequent malignancies in women worldwide and represent a growing public health concern in sub-Saharan Africa (SSA). In Gabon, national screening programs have been established, but their effectiveness remains poorly documented, particularly outside the capital city, Libreville. This study, therefore, aimed to describe the screening outcomes and associated factors for breast abnormalities and suspected cervical lesions at the Centre Hospitalier Universitaire Amissa Bongo (CHUAB) in Franceville, Gabon.

Methods:

We conducted a retrospective cross-sectional study that included all women who attended routine screening at CHUAB between 2015 and, 2016 and, 2018 and 2021. Data were retrieved from hospital registries and medical records. All women underwent both clinical breast examination (CBE) and visual inspection with acetic acid (VIA). Sociodemographic, reproductive, and clinical variables were analyzed to determine their associations with screening outcomes using univariable and exploratory multivariable logistic regression models.

Results:

A total of 1,328 women were included in the study, with a mean age of 35.4 ± 10.8 years. In univariable analysis, women aged 45–73 years had lower odds of cervical screening positivity compared with those aged 14–25 years (OR = 0.52, 95% CI: 0.29–0.90; p = 0.021). In exploratory multivariable analysis, non-Gabonese nationality and employment in the public sector were associated with cervical screening positivity. Regarding breast abnormalities, an age at first pregnancy between 30 and 39 years was associated with higher odds of screening positivity in univariable analysis (OR = 3.90, 95% CI: 1.07–11.4; p = 0.021); however, no factors remained statistically significant in the adjusted model.

Conclusion:

This study offers one of the first accounts of routine breast and cervical cancer screening outcomes in a semi-urban setting in south-eastern Gabon. VIA and CBE detected a measurable proportion of women with suspected cervical lesions and breast abnormalities, underscoring the value of these simple and accessible screening approaches in resource-limited settings. Strengthening follow-up systems, improving data management, and expanding screening coverage could further enhance cancer control among women in Gabon.

1 Introduction

Globally, breast and cervical cancer are among the most commonly diagnosed cancers in women and represent a major public health problem. The World Health Organization estimates that in 2022, approximately 666,000 women died from breast cancer and 350,000 from cervical cancer worldwide. A disproportionate number of these deaths occurred in Low and Middle Income Countries (LMICs) (around 71% for breast cancer, 91% for cervical cancer). A significant portion of these deaths could be prevented with increased public awareness and early-stage diagnosis (1).

The natural histories of breast and cervical cancers are interconnected, as they are both partly influenced by a woman’s sexual and reproductive choices, as well as exposures during early life (2). Consequently, a unified global approach to reducing these cancers is both feasible and necessary. In High-income countries (HICs), mortality from these cancers has been successfully reduced through organized screening strategies, including mammographic and cytology-based screening, often implemented through proactive public health policies.

By contrast, in LMICs, these trends are often reversed. Rapid changes in lifestyle and urban environments, driven by economic growth and the increasing participation of women in the workforce, have influenced breast cancer risk factors, including delayed childbearing, fewer children, excess weight, and physical inactivity. These evolving risk profiles are now aligning more closely with those of HICs (3, 4). As a result, in sub-Saharan African countries, both breast cancer incidence and mortality remain high, in part because of late-stage diagnosis and limited access to comprehensive treatment center services, leading to a poor prognosis for affected women (2).

Regarding cervical cancer, urbanisation and social changes are contributing to an earlier sexual debut and an increasing number of lifetime partners among women. In this context of lax screening policies, these factors contribute to increasing or at least maintaining the incidence of cervical cancer.

Inexpensive and feasible screening tests are available for these cancers. As recommended by the WHO, clinical breast examination (CBE) and visual inspection of the cervix after application of acetic acid (VIA) are suitable methods for early detection approaches in low-resource settings, particularly where mammography or cytology-based programmes remain difficult to implement at scale (5).

In Gabon, recent population-based cancer data from the Grand Libreville cancer registry showed that breast cancer is the leading cancer among women, followed by cervical cancer, together accounting for nearly half of all female cancers recorded in this urban setting (6). A national screening program for breast and cervical cancers, based on CBE and VIA, respectively, was established in Libreville, Gabon, in 2013. In 2015, authorities extended these screening services to Franceville, the capital of Haut-Ogooué province, located at the eastern border of the country, neighboring Congo. These units perform cervical and breast cancer screenings using visual inspection of the cervix and clinical breast examination. According to the Ministry of Health guidelines, women aged 20 years and older, as well as all sexually active women, are eligible for screening. However, to date, only partial data from Libreville, Gabon’s capital, have been published. Data from semi-urban and provincial settings remain scarce.

Thus, this study aims to provide data on women who attended the Franceville screening unit between 2015 and 2021 to assess the positivity rates of breast and cervical cancer screening and identify associated sociodemographic factors in the Haut-Ogooué province.

2 Materials and methods2.1 Study area

The study was conducted at the Centre Hospitalier Universitaire Amissa Bongo (CHUAB) in Franceville, a semi-urban city and the provincial capital of Haut-Ogooué in southeastern Gabon. As the third most populous province, Haut-Ogooué is a region of significant mining activity, primarily for uranium and manganese. CHUAB is the main referral hospital in Haut-Ogooué and also serves as a regional referral centre for southeastern Gabon, receiving patients from Franceville and the surrounding areas, including neighboring localities.

The hospital hosts a dedicated screening unit where women may be referred or attend voluntarily for opportunistic screening. This unit serves women from Franceville and the surrounding areas within the province. In the absence of a fully organised population-based screening programme, screening activities at CHUAB represent an important entry point for the early detection of women’s cancers in this semi-urban, resource-constrained setting.

This setting is particularly relevant for the present study, as it provides access to routine screening data from a region of Gabon where published information on breast and cervical cancer screening remains limited.

2.2 Study design

This retrospective cross-sectional study was conducted at Centre Hospitalier Universitaire Amissa Bongo (CHUAB) in Franceville, which serves as the regional referral hospital and the main cancer screening center for the province of Haut-Ogooué. The study included the records of all women who attended the CHUAB screening unit for cervical and breast cancer screening during the study period (2015–2016 and 2018–2021). As part of the routine screening procedure, all women underwent both cervical cancer screening by visual inspection with acetic acid (VIA) and breast cancer screening by clinical breast examination (CBE).

Records were eligible if they contained screening information and could be matched to an individual patient file. Duplicate records and records with incomplete key information were excluded from the final analysis. The sample size corresponded to all eligible screening records available during the study period.

2.3 Study procedures and data collection

Data collection was conducted in two phases. First, information was extracted from the screening unit registry. Second, the corresponding individual medical records were retrieved and reviewed. Full names and registration numbers were used to match entries between the registry and patient files. To improve data reliability, extracted information was cross-checked between the registry and the individual medical records.

A standardized data extraction sheet was used to ensure consistency during data collection and to reduce transcription errors. Data were subsequently entered into a structured Microsoft Excel database, cleaned, and checked for consistency before statistical analysis. The following variables were collected: socio-demographics (screening site, year of survey, year of birth, age, nationality, origin, department, occupation, and marital status); family history variables (breast cancer, cervical cancer, diabetes, high blood pressure); obstetrical and reproductive history (pregnancy, parity, live child, abortion, menarche, age of first procedure) and screening test result.

Screening examinations were routinely performed by trained healthcare providers working in the breast and cervical cancer screening unit. For breast cancer, screening was based on clinical breast examination (CBE), including visual inspection and palpation of the breasts. A screening result was considered abnormal when a suspicious palpable mass, skin retraction, nipple discharge, or another clinically suspicious breast lesion was identified. For cervical cancer, screening was based on Visual Inspection after applying acetic acid (VIA). A cervical screening result was considered positive when acetowhite lesions suggestive of suspected cervical lesions were observed after acetic acid application.

Two screening outcomes were analysed separately in this study:

-cervical screening positivity, defined as a positive VIA result suggestive of suspected cervical lesions; and

-breast screening positivity, defined as an abnormal CBE suggestive of breast abnormalities.

After screening, the results were recorded on a standardized form and in the service registry. A copy of the form was provided to the patient.

The selection of patients’ records was conducted in two phases. First, the data were extracted from the unit registry. Second, we processed the individual medical records of each screened woman. The full names and registration numbers were used to identify and match records between the registry and patient files. To improve data reliability, the extracted information was cross-checked between the registry and the individual medical records. Duplicate entries and records with incomplete key information were excluded from the final analysis.

2.4 Data analysis

Data were analysed using RStudio. Categorical variables were summarized using counts and percentages, while continuous variables were described using means and standard deviations or medians and interquartile ranges, as appropriate. The proportion of missing data was assessed for all variables and reported in the descriptive analyses.

Associations between categorical variables and screening outcomes were assessed using Pearson’s chi-square test when expected cell counts were sufficient, and Fisher’s exact test when sparse data were present. Continuous variables were compared using either the Student’s t-test or the Mann–Whitney U test, depending on the distribution of the data. Correlations between continuous variables were assessed using Pearson’s or Spearman’s correlation coefficients, as appropriate.

For data processing, records with missing key screening outcome information were excluded from the analysis. For other variables, analyses were conducted using available data. For selected family history variables, missing values were recoded as “No” under the assumption that non-reporting reflected absence of known history; however, this approach may have introduced non-differential misclassification. Implausible values (below 10 years) for age at menarche and age at first pregnancy, representing less than 3% of observations, were considered invalid and treated as missing values. Variables with a high proportion of missing data (>50%) were handled with caution and were not prioritised in multivariable modelling.

Associations between explanatory variables and each screening outcome were first explored using univariable logistic regression models to estimate crude odds ratios (ORs) and their 95% confidence intervals (CIs).

Exploratory multivariable logistic regression models were then fitted separately for each screening outcome to assess independent associations. However, because of substantial missing data for several variables and sparse data in some categories, these models showed instability, including convergence issues and imprecise or non-interpretable estimates. Multivariable findings were therefore interpreted cautiously and considered exploratory rather than confirmatory. Geographic variables reflected screening locations rather than intrinsic individual characteristics and were therefore interpreted with caution as explanatory variables. Multiple imputation approaches were explored to address missing data, but did not substantially improve model stability. The primary analyses were therefore based on complete-case data. A p-value <0.05 was considered statistically significant.

2.5 Ethics statement

The study was carried out in accordance with the ethical principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the institutional ethics committee of Centre Hospitalier Universitaire Amissa Bongo (CHUAB). As this was a retrospective study based on routinely collected screening records, no direct patient contact was involved. All data were anonymised before analysis, and confidentiality was strictly maintained throughout the study.

3 Results3.1 Demographics of study participants

Initially, 1,479 medical files were recorded in the register. After removing duplicates and files without screening results, a total of 1,328 medical files were included in the study between 2015 and 2021. The main socio-demographic and reproductive features of the enrolled sample are displayed in Tables 1 and 2.

CharacteristicCervical cancerp-valueNegative
(N = 368)Positive
(N = 38)Age[14-25]12% (44)7.9% (3)0.696[25-35]35% (130)42% (16)[35-45]25% (93)29% (11)[45-73]27% (101)21% (8)NationalityGabonese95% (349)89% (34)0.255Non-gabonese5.2% (19)11% (4)Marital statusUnmarried61% (224)58% (22)0.855Married39% (144)42% (16)City of screeningFranceville68% (252)74% (28)0.790Leconi1.4% (5)0% (0)Moanda7.6% (28)2.6% (1)Others23% (83)24% (9)DepartmentLemboumbi-Leyou8.7% (32)5.3% (2)0.629Mpassa82% (301)89% (34)Others9.5% (35)5.3% (2)Province of screeningHaut-Ogooué97% (357)95% (36)0.348Others3.0% (11)5.3% (2)Age at menarche[7-12]3.3% (12)0% (0)0.682[12-16]74% (272)74% (28)[20- +]23% (84)26% (10)PregnancyYes100% (368)100% (38)p=<0.001Age at first pregnancy10-1963% (231)55% (21)0.45620-2935% (127)45% (17)30-392.7% (10)0% (0)

Sociodemographic and reproductive factors associated with cervical cancer screening positivity.

P-values were calculated using Fisher’s exact test when expected frequencies were <5; otherwise, the Chi-square test was used.

CharacteristicBreast cancerp-valueNegative
(N = 439)Positive
(N = 21)Age[14-25]12% (51)9.5% (2)0.477[25-35]33% (145)43% (9)[35-45]28% (125)14% (3)[45-73]27% (118)33% (7)NationalityGabonese93% (409)95% (20)p=1.000Non-gabonese6.8% (30)4.8% (1)Marital statusUnmarried60% (262)57% (12)p=0.997Married40% (177)43% (9)City of screeningFranceville66% (289)86% (18)0.280Leconi1.6% (7)0% (0)Moanda11% (49)0% (0)Others21% (94)14% (3)DepartmentLemboumbi-Leyou13% (55)0% (0)0.122Mpassa79% (345)86% (18)Others8.9% (39)14% (3)Province of screeningHaut-Ogooué97% (425)95% (20)0.509Others3.2% (14)4.8% (1)Age at menarche[7-12]3.4% (15)0% (0)0.572[12-16]73% (321)86% (18)[20- +]23% (103)14% (3)PregnancyYes100% (439)100% (21)p=<0.001Age at first pregnancy10-1962% (272)62% (13)0.15020-2936% (157)29% (6)30-392.3% (10)9.5% (2)

Sociodemographic and reproductive factors associated with breast cancer screening positivity.

Overall, most women were aged 25–35 years, Gabonese, and screened in Franceville, mainly in the Mpassa department. Most participants reported a first pregnancy between 10 and 19 years and menarche between 12 and 16 years. The distribution of sociodemographic and reproductive characteristics was broadly similar between women with negative and positive screening results for both cancers. Suspected cervical lesions and breast abnormalities were detected among 9.36% and 4.56% women, respectively.

3.2 Sociodemographic and reproductive factors associated with suspected cervical lesions3.2.1 Univariable analysis

To further explore factors associated with cervical screening positivity, univariable logistic regression analyses were performed for selected sociodemographic and reproductive characteristics (Table 3). In univariable analysis, age was the only variable significantly associated with suspected cervical lesions.

CharacteristicNOR95% CIp-valueAge[14-25]1,017Reference——[25-35]0.880.56, 1.390.6[35-45]0.900.56, 1.470.7[45-73]0.520.29, 0.900.021NationalityGabonese869Reference——Non-gabonese1.170.47, 2.500.7Marital statusUnmarried1,002Reference——Married/engaged/widowed0.850.60, 1.190.4City of screeningOthers1,048——Franceville0.830.57, 1.230.3Leconi0.340.02, 1.820.3Moanda0.810.39, 1.600.6Occupationnactive/unemployed909——Skilled professions0.980.56, 1.65>0.9Informal/self-employed sector0.640.19, 1.670.4Public sector1.200.77, 1.840.4Age at menarche[7-12]825——[12-16]1.440.60, 4.290.5[20- +]1.740.68, 5.370.3PregnancyNo827——Yes2.281.10, 5.560.043Age at first pregnancy (years)10–19638——20–291.160.57, 2.340.730–390.00>0.9Family history of cancerNo1,048Reference—Yes0.680.16, 2.010.5Family history of cervical cancerNo1,048——Yes1.110.25, 3.500.9

Univariable logistic regression analysis of factors associated with cervical cancer screening positivity.

CI, confidence interval; OR, odds ratio. Bold values indicate statistically significant associations (p < 0.05).

Compared with women aged 14–25 years, those aged 45–73 years had significantly lower odds of a positive screening result (OR = 0.52, 95% CI: 0.29–0.90; p = 0.021). No statistically significant associations were found for age at first pregnancy, nationality, place of screening, professional status, marital status, age at menarche, or family history of cancer (all p > 0.05).

Although having had at least one pregnancy was associated with higher odds of a positive screening result (OR = 2.28, 95% CI: 1.10–5.56; p = 0.043), this variable was retained in the results but not emphasized in the interpretation, as all women with a positive screening result had experienced at least one pregnancy.

3.2.2 Multivariable analysis

Variables showing evidence of association in univariable analysis, as well as those considered epidemiologically relevant, were subsequently entered into the multivariable logistic regression model (Table 4). In the adjusted model, non-Gabonese nationality and professional status remained significantly associated with cervical cancer screening positivity.

CharacteristicNOR95% CIp-valueAge[14-25]1,177——[25-35]1.140.56, 2.460.7[35-45]0.870.39, 1.980.7[45-73]0.860.37, 2.020.7NationalityGabonese967——Non-gabonese1.250.37, 3.250.7Marital statusUnmarried1,162——Married/engaged/widowed0.720.41, 1.230.2City of screeningOthers1,209——Franceville1.370.71, 2.910.4Leconi2.620.38, 11.00.2Moanda0.610.14, 2.050.5Occupationnactive/unemployed1,056——Skilled professions0.730.25, 1.750.5Informal/self-employed sector0.420.02, 2.000.4Public sector1.510.80, 2.740.2Age at menarche[7-12]967——[12-16]0.550.22, 1.640.2[20- +]0.390.13, 1.340.11