Risk factors associated with leprosy in Bongo District, Ghana: a mixed methods study

Leprosy remains endemic in parts of Ghana despite global elimination targets. Bongo District in the Upper East Region has recorded rising case numbers (6 in 2021, 10 in 2022, 19 in 2023), suggesting ongoing challenges in diagnosis and case detection. We investigated biomedical, socio-economic, environmental and cultural factors associated with diagnosis. We conducted an explanatory sequential mixed-methods study comprising a matched case-control component followed by qualitative focus group discussions. Seventeen adults diagnosed with leprosy between 2019 and 2023 were compared with 30 community controls matched as closely as feasible on age, sex, community of residence, marital status and occupation. Five focus group discussions (n=25) involved persons affected by leprosy, family members, health workers, community health volunteers and community leaders. Quantitative data were analyzed using STATA v17 using descriptive, bivariate and exploratory multivariable logistic regression analyses. Qualitative data were analyzed thematically in NVivo following Braun and Clarke’s six-phase approach. A visible BCG scar was observed in 23.5% of cases and 80.0% of controls and was associated with lower odds of leprosy in the exploratory adjusted analysis (adjusted odds ratio 0.10; 95% confidence interval 0.01-0.98; p=0.048). Open defecation was more frequent among cases (94.1%) than controls (60.0%, p=0.017), although the adjusted estimate was imprecise (adjusted odds ratio 10.39; 95% confidence interval 0.36-303.64). Only 17.0% identified bacteria as the cause of leprosy; misconceptions included heredity (36.2%), witchcraft (19.2%) and divine punishment (10.6%). Among controls, 56.7% had high stigma scores. Qualitative findings explained how misconceptions, financial and geographical barriers and anticipated social exclusion contributed to delayed biomedical care. Absence of a visible BCG scar, poor sanitation, limited leprosy knowledge and stigma were associated with leprosy or delayed care in Bongo District. The findings support strengthened routine immunization, integrated community education, early case detection and anti-stigma interventions. Because of the small number of cases, adjusted estimates should be interpreted as exploratory and hypothesis-generating.

Leprosy (Hansen disease) is a chronic neglected tropical disease caused mainly by Mycobacterium leprae. It primarily affects the skin and peripheral nerves and may result in disability when diagnosis and treatment are delayed. Although multidrug therapy has substantially reduced registered prevalence, new cases continue to be detected in more than 120 countries, indicating persistent transmission. The World Health Organization’s current strategy therefore emphasizes interruption of transmission, prevention of disability and elimination of stigma (1).

Ghana continues to report new leprosy cases, with substantial subnational variation. In Bongo District in the Upper East Region, routine program records indicated increasing case notifications during 2021–2023. These notifications, together with the high proportion of multibacillary disease and disability observed among diagnosed patients, raise concern about delayed detection and ongoing community transmission (2, 3). Previous Ghanaian research has provided important evidence on individual clinical and social dimensions of leprosy. A case-control study in Sene District examined selected demographic and exposure-related factors, while qualitative studies in southern Ghana documented stigma, discrimination and courtesy stigma among affected persons and caregivers (2, 4). International and regional evidence has also linked leprosy to close contact, poverty, poor housing and sanitation, limited knowledge and delayed healthcare seeking (5, 6). BCG vaccination provides partial protection against leprosy, but evidence on effective coverage among people in endemic and hard-to-reach communities remains limited (4, 6). However, few studies in Ghana have integrated biomedical, environmental, socioeconomic and sociocultural evidence in a single mixed-methods investigation. Examining these factors together is important because statistical associations alone cannot explain how community beliefs, stigma and access barriers shape exposure, diagnosis and care seeking. We therefore investigated factors associated with leprosy in Bongo District using a matched case-control study followed by enquiry to explain and contextualize findings.

We used an explanatory sequential mixed-methods design comprising a matched case-control study followed by focus group discussions. Participants in the quantitative component were selected according to outcome status: adults diagnosed with leprosy served as cases and adults without history of leprosy served as controls. The subsequent qualitative component was used to explain and contextualize patterns identified in the quantitative data.

The study was conducted in Bongo District, Upper East Region, Ghana. Bongo is predominantly rural with 143 communities and an estimated population of 127,990. At the time of the study, the district had one hospital, seven health centers and 58 Community-based Health Planning and Services compounds. Leprosy diagnosis and treatment were provided mainly through the district hospital and health centers.

Quantitative component: participant, control selection and matching.

All registered adults aged 18 years or older who had been diagnosed with leprosy between 2019 and 2023, resided in Bongo District and consented to participate were eligible as cases. Because only 17 eligible cases were available, all were invited and enrolled. Controls were community residents without a previous diagnosis of leprosy. For each case, research staff sought up to three controls from the same community who were comparable on sex, age (within five years), marital status and occupation. A complete 1:3 ratio could not be achieved because eligible controls meeting all criteria were unavailable in some communities; consequently, 30 controls were recruited for 17 cases. The achieved matching was therefore variable rather than a fixed ratio. Matching participants by community was selected primarily to reduce confounding arising from geographical differences, although this strategy may have reduced variability in environmental exposures and potentially attenuated estimates of environmental risk factors.

Data collection and measures

Trained research assistants administered a structured questionnaire through face-to-face interviews. The instrument covered socio-demographic characteristics, housing, water and sanitation, health literacy, attitudes, stigma and social exclusion. BCG scar status was assessed by inspection for a characteristic scar on the upper arm and should therefore be interpreted as BCG-scar status rather than documented vaccination history. Disability among cases was graded using World Health Organization criteria (grades 0-2). Health literacy assessment: Health literacy was assessed using the Health Literacy Questionnaire (HLQ), consisting of 44 items scored on a 4-point Likert scale. Total scores ranged from 44 to 176, with participants scoring ≥ 132 (75th percentile) classified as having high health literacy. The HLQ has been validated in sub-Saharan African settings.

Stigma assessment: Stigma was measured using the Explanatory Model Interview Catalogue (EMIC) stigma scale, a 15-item instrument assessing perceived and enacted stigma. Each item was scored on a 3-point scale (0=no, 1=maybe, 2=yes), with total scores ranging from 0 to 30. Participants scoring ≥ 15 were classified as having high stigma scores, consistent with previous studies in Ghana (2, 4). The EMIC has been validated in leprosy-endemic settings (7).

Sample size

An initial calculation, based on an expected difference in knowledge between cases and controls, suggested a larger sample than was available in the district. We therefore conducted a census of all eligible registered adult cases and recruited the maximum feasible number of matched community controls. The resulting sample was sufficient for descriptive and selected bivariate comparisons but provided limited precision for multivariable modeling.

Qualitative component:

Five focus group discussions were conducted with five participants in each group (total n=25): persons affected by leprosy (n=5), family members or close contacts (n=5), health workers (n=5), community health volunteers (n=5) and community leaders (n=5). Participants were purposively selected to provide complementary perspectives on lived experience, household responses, service delivery, community health promotion, beliefs and social norms. Each focus group was homogeneous by stakeholder category to encourage open discussion and minimize power imbalances.

Semi-structured guide explored recognition of early symptoms, pathways to care, barriers to diagnosis and treatment, community beliefs, stigma and possible prevention strategies. Discussions were conducted in Gurune, audio-recorded with permission, transcribed verbatim and translated into English. Forward translation from Gurune into English was performed by bilingual research assistants, and English transcripts were independently reviewed against the original recordings to ensure conceptual equivalence. Any discrepancies were resolved through discussion among the research team before analysis.

Reflexivity statement: The research team comprised public health researchers with experience in neglected tropical diseases and qualitative research. Reflexive discussions were held throughout data collection and analysis to minimize the influence of researchers’ prior assumptions on data interpretation.

Data analysis

Quantitative analysis: Quantitative data were collected in KoboCollect, exported to Microsoft Excel and analyzed in STATA version 17. Frequencies, proportions, medians and interquartile ranges (Q1–Q3) were calculated. Cases and controls were compared using appropriate bivariate tests. Crude odds ratios and 95% confidence intervals were estimated for selected exposures. Because the achieved matching ratio varied and the number of cases was small, unconditional logistic regression was used, with matching factors and selected a priori confounders considered in model specification. The full model included age, sex, education, income, BCG scar status, open defecation, and housing type. To reduce over-parameterization, interpretation focused on a limited set of epidemiologically plausible variables, and adjusted estimates were treated as exploratory. Wide confidence intervals were interpreted as evidence of substantial statistical uncertainty rather than absence of an association. A two-sided p-value below 0.05 was used as a descriptive threshold, but interpretation considered effect size and precision.

Qualitative analysis: Qualitative transcripts were coded inductively in NVivo. Analysis followed Braun and Clarke’s six-phase thematic approach (8). Two investigators independently coded all transcripts. Initial coding produced 84 codes, which were subsequently organized into 12 categories and five overarching themes. Coding disagreements were resolved through discussion until consensus was reached. Data collection continued until thematic saturation was achieved, whereby no substantially new themes emerged from successive focus group discussions.

Integration occurred during interpretation and reporting: quantitative results were compared with qualitative themes to identify convergence, complementarity and divergence, following established mixed-methods principles. The study is reported in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines.

Ethical considerations

Ethical approval was obtained from the University of Health and Allied Sciences Research Ethics Committee (UHAS FNBSPH-REC M.10 [011]24-25). Permission was also secured from the Upper East Regional Health Directorate and Bongo District Health Directorate. Written informed consent was obtained from all participants. Transportation was provided, and participants received GHS 30.00 for time and travel. Confidentiality was ensured through anonymization.

Participant characteristics

Forty-seven participants were enrolled: 17 cases and 30 controls. Selected characteristics of the cases and controls are presented in Table 1. The median age was 52 years (IQR 44–62), 53.2% were male and 65.9% had no formal education. Monthly income was below GHS 500 for 59.6% of participants; this was more frequent among cases than controls (82.4% versus 46.7%), although this difference did not reach conventional levels of statistical significance (p=0.057). Among cases, 88.2% had multibacillary leprosy, and 47.0% had grade 2 disability.

BCG vaccination status and environmental conditions

A visible BCG scar was present in 4 of 17 cases (23.5%) and 24 of 30 controls (80.0%; p<0.001). In the exploratory adjusted analysis, BCG scar presence was associated with lower odds of leprosy (adjusted odds ratio 0.10; 95% confidence interval 0.01–0.98; p=0.048). Awareness that BCG may provide protection against leprosy was low (6.4% overall).

Open defecation was reported by 94.1% of cases and 60.0% of controls (p=0.017). Its adjusted estimate was large but very imprecise (adjusted odds ratio 10.39; 95% confidence interval 0.36–303.64), reflecting sparse data and instability of the multivariable model. This estimate should not be interpreted as a precise measure of effect.

Community health literacy, stigma and social exclusion

Only 25.5% of participants met the study’s threshold for high health literacy (HLQ score ≥132). Seventeen percent correctly identified bacteria as the cause of leprosy. Reported misconceptions included heredity (36.2%), witchcraft (19.2%) and divine punishment (10.6%). Only 6.4% identified loss of sensation as an early sign. Although 74.5% knew leprosy could be treated, 57.5% believed it could be cured. Among controls, 56.7% were classified as having high EMIC score ≥15). Ninety percent believed leprosy could hinder marriage, 76.7% believed it could hinder employment and 90.0% expressed reluctance to buy food from a person affected by leprosy. In addition, 73.3% would not allow an affected person to care for their children and 76.7% would not accept their child marrying a person affected by leprosy.

Qualitative findings

Five themes were developed from the focus group discussions, supported by verbatim quotations.

Theme 1: Limited recognition of early symptoms

Participants described limited recognition of early symptoms. Numbness and painless skin lesions were frequently normalized or initially attributed to non-biomedical causes, contributing to prolonged pathways to diagnosis.

“I noticed a patch on my skin that was not painful, but I thought it was just a normal skin problem. More than a year passed before I went to the hospital and was told it was leprosy.” (Male participant affected by leprosy, FGD 1)

“When the numbness started in my hand, I did not think it was serious. I continued my work until I could not feel things properly.” (Female participant affected by leprosy, FGD 2)

Theme 2: Barriers to biomedical care

Participants identified barriers to biomedical care, including transport costs, distance to facilities, competing household needs and the perceived burden of prolonged treatment. These barriers were particularly important for people living in remote communities.

“The hospital is far from our community. To go there, I have to pay for transport, and sometimes I do not have the money. I also have to leave my farm work.” (Male participant affected by leprosy, FGD 1)

“I heard the treatment takes many months. I was worried about how I would manage to go to the hospital for that long.” (Family member, FGD 2)

Theme 3: Stigma and social exclusion

Stigma influenced disclosure and care seeking. Affected persons described avoidance, restricted social participation and family separation. Community members’ concerns about marriage, food handling and employment were consistent with these experiences.

“People refused to shake my hand because they thought they would catch the disease. Some of my friends stopped visiting me.” (Female participant affected by leprosy, FGD 1)

“When my husband was diagnosed, some people said our children would also get it. My relatives advised me to leave him.” (Family member, FGD 2)

“I would not allow someone with leprosy to prepare food for my children because I am afraid.” (Community leader, FGD 5)

Theme 4: Spiritual and traditional beliefs

Explanations involving witchcraft, curses and wrongdoing encouraged some individuals to seek traditional or spiritual care before attending a health facility. Health workers and volunteers considered these beliefs important contributors to delayed diagnosis

“At first I believed it was a spiritual illness, so I went to the traditional healer before visiting the hospital. I spent money there but did not get better.” (Male participant affected by leprosy, FGD 1)

“Some people think leprosy is caused by a curse from ancestors or a punishment from God. So they go to the priest or traditional healer first.” (Community health volunteer, FGD 4)

Theme 5: Limited community education about leprosy and BCG

Participants reported limited community education about leprosy and BCG. None of the affected persons or family participants recalled receiving information about BCG’s possible protective effect. Health workers and community volunteers recommended sustained education through primary healthcare services, schools, community leaders and World Leprosy Day activities.

“I did not know that BCG could protect against leprosy. Nobody ever told me that.” (Male participant affected by leprosy, FGD 1)

“We need more education in the communities. Many people do not know the early signs of leprosy or that it can be treated for free.” (Health worker, FGD 3)

“The community leaders can help spread the message. If the chief and elders talk about leprosy, people will listen.” (Community leader, FGD 5)

Mixed-methods integration of findings.

The two data strands were mainly convergent and complementary. Quantitatively, low recognition of bacterial causation and high levels of misconceptions were accompanied qualitatively by accounts of spiritual attribution and initial use of traditional care. High stigma scores among controls converged with affected persons’ reports of avoidance and restricted social participation. The quantitative association between BCG scar absence and leprosy was complemented by widespread qualitative unawareness of BCG’s protective role. Open defecation and low income identified structural vulnerability quantitatively, while participants described transport costs and geographical barriers that delayed diagnosis. No major contradiction between the strands was identified; rather, qualitative findings explained mechanisms through which limited knowledge, stigma and poverty may contribute to delayed detection and continued transmission.

Principal findings and integrated interpretation

This study combined a matched case-control analysis with qualitative enquiry to examine factors associated with leprosy in Bongo District. A visible BCG scar was substantially less common among cases than controls, while open defecation, limited leprosy knowledge and stigma were also prominent. The qualitative findings provided explanatory context by showing how misconceptions, anticipated exclusion and access barriers shaped pathways to care.

The inverse association between BCG scar presence and leprosy is consistent with evidence that BCG provides partial protection against leprosy [4,7,10,11]. A meta-analysis of 26 studies estimated BCG efficacy against leprosy at 61% (95% CI 51–70) (5). However, scar inspection is an imperfect proxy for vaccination because scars may fade or fail to form. Studies have shown that up to 20% of vaccinated individuals may not develop a visible scar (9). The qualitative finding that affected persons and families were largely unaware of BCG’s possible protective role complements the quantitative association. Together, the findings support strengthening routine childhood immunization and zero-dose outreach, while avoiding the unsupported assumption that absence of a scar always means absence of vaccination.

Open defecation was more common among cases, but the adjusted estimate was highly unstable. Previous research has linked leprosy with poverty, inadequate housing and sanitation, although these variables may act as markers of broader deprivation rather than direct transmission pathways (10, 11). In this study, the quantitative sanitation and income patterns were complemented by qualitative accounts of transport costs and distance. This convergence suggests that structural disadvantage may increase exposure, impede early diagnosis and cluster multiple risks within households and communities.

Low knowledge and misconceptions were evident in both study components. Quantitative responses showed poor recognition of bacterial causation and early sensory loss, while qualitative participants described attribution to witchcraft and curses and initial use of traditional or spiritual care. These findings align with studies from Ethiopia and Nigeria where traditional beliefs were associated with diagnostic delay (10, 11). This integrated evidence indicates that limited health literacy affects not only knowledge scores but also diagnostic delay. Community education should therefore focus on early signs, curability, transmission, free multidrug therapy and referral pathways, and should involve trusted traditional, religious and community leaders.

Stigma was also strongly convergent across the data strands. More than half of controls had high stigma scores, and attitudes concerning marriage, employment and food handling were echoed by affected persons’ accounts of social avoidance. Ghanaian studies have similarly documented enacted, anticipated and courtesy stigma (12, 13). Studies from Ethiopia and Nepal have reported that stigma frequently leads to delayed diagnosis, social isolation and mental health problems (12). Evidence from stigma-reduction interventions suggests that education alone may be insufficient (14); contact-based approaches involving people affected by leprosy, counseling and community dialogue may be more effective (14, 15).

Comparison with other studies. Our finding of low health literacy (25.5%) is consistent with studies in other leprosy-endemic settings. A study in rural Ethiopia found that only 22% of participants could correctly identify bacteria as the cause of leprosy (16). Similarly, research in Nigeria reported that misconceptions about leprosy causation were common, with 31% attributing it to witchcraft (9). The high stigma scores observed in our study (56.7% of controls) are comparable to findings from Ghana’s Southern Region, where 62% of community members expressed stigmatizing attitudes toward persons affected by leprosy (2).

The association between BCG scar absence and leprosy observed in our study (aOR 0.10, 95% CI 0.01–0.98) is consistent with the protective effect reported in meta-analyses (5, 7). However, the wide confidence interval reflects the limited sample size and underscores the need for larger confirmatory studies.

Strengths and limitations

A key strength was the integration of epidemiological comparisons with perspectives from affected persons, families, service providers, volunteers and leaders. Recruiting all eligible registered adult cases maximized use of available district data, and community controls improved contextual comparability. The qualitative component, reported according to COREQ guidelines, provided rich explanatory data that enhanced interpretation of quantitative findings.

The study also has important limitations. Only 17 cases were available, limiting statistical power and the number of variables that could be included reliably in regression models. The adjusted estimates were therefore exploratory, and several confidence intervals were very wide. Matching was incomplete and variable, and unconditional regression may not fully recover the efficiency of a completely matched analysis. Additionally, matching participants by community may have reduced variability in environmental exposures and attenuated the estimated associations. Residual confounding and over fitting remain possible. BCG scar inspection may misclassify vaccination status because scars can fade or may not develop. Registered cases may underrepresent people with mild, undiagnosed or unreported disease, and retrospective measurement of knowledge may be affected by learning after diagnosis. Finally, focus group findings are context-specific and may have been influenced by social desirability. Future studies should use larger samples, clearly defined matched sets or unmatched designs with adequate sample sizes, and documented vaccination histories. Penalized likelihood methods such as Firth logistic regression or exact logistic regression would be appropriate for sparse data.

Implications for policy and research

The findings support four priorities: strengthening routine BCG delivery and outreach to zero-dose communities; integrating leprosy education and early case detection into primary healthcare and community platforms; implementing contact-based anti-stigma interventions with affected persons and community leaders; and addressing transport, sanitation and other structural barriers through cross-sectoral action. Larger multicenter studies should use clearly defined matched sets or unmatched designs with adequate sample sizes, parsimonious prespecified models and documented vaccination histories. Future mixed-methods studies could use joint displays to make integration more explicit.

In Bongo District, absence of a visible BCG scar, poor sanitation, limited leprosy knowledge and stigma were associated with leprosy or delayed care. Quantitative and qualitative findings converged in showing that biological, structural and sociocultural factors operate together. Integrated immunization, community education, early detection and anti-stigma action are warranted, while the exploratory adjusted estimates require confirmation in larger studies.