Knowledge, perceptions, and breast cancer treatment practices and associated factors among traditional medicine practitioners in Ghana.

Introduction
Breast cancer continues to be among the leading causes of cancer morbidity and mortality in women around the world. The increasing global incidence, morbidity, and mortality rates suggest increasing challenges for public health systems. The World Health Organization (WHO) reported that in 2022 alone, 2.3 million women were diagnosed with breast cancer, and 666,103 deaths were reported globally, indicating that the global burden of BC remains huge [1].
Data from different countries over the last decade and a half show that Africa, particularly sub-Saharan Africa, is emerging as one of the hardest hit regions with breast cancer in terms of incidence, prevalence, and mortality rates. For example, the data from Global Cancer Observatory for 2022 show that 198,553 new cases and 91,252 BC deaths [1]. The annual incidence of breast cancer in Africa is projected to increase by almost 54% from the 2020 diagnosis of 186,598 cases by 2040 [2]. Similarly, breast cancer-related mortality is projected to reach 1,037,639 by 2040 in the African region [2]. This may be due to increasing diagnoses coupled with limited resources, resulting in increased morbidity and mortality.
The increasing trend of breast cancer burden in terms of incidence and mortality rates is being experienced in Ghana, and one of the major public health challenges in the country, breast cancer, is one of the leading causes of death among women. For example, 5,026 new cases of BC were diagnosed among Ghanaian women, constituting 31.4% of all 15,987 new cancer diagnoses among women in 2022 [2]. The same data reported that there were 2 369 (24%) BC deaths in Ghana in 2022 alone among women [2]. The regions in Ghana with the highest BC incidence rates were the Eastern region (52.3 per 100,000), followed by the Greater Accra (48.9), Ashanti (38.3), Central (34.7), and Volta (31.7) regions, on the basis of 2012 data [3]. Importantly, these confirmed and reported cases are not considered undiagnosed or unreported cases.
Sufficient evidence shows that the chances of cure and/or successful treatment and improved survival rates are high when BC is detected at the earliest stages of development [4]. For example, when noninvasive cancer detected early at the localized stage, the 5-year relative survival rate is 99%, whereas the 5-year relative survival rate is 31% for those diagnosed at later stages [5]. However, in developing countries such as Ghana, challenges abound that make early detection, diagnosis, and initiation of appropriate treatment an illusion for many patients. Evidence from literature suggests that the pervasive use of traditional healers (THs), herbal medicines (HMs), and other alternative medicines prior to orthodox treatments, leads to late diagnosis at stages II and III [6]. Researchers have reported that TMPs and HMs constitute Ghana’s most employed alternative source of healthcare, with the TMP-patient ratio far better than the medical practitioner-patient ratio, especially in rural communities throughout Ghana [7]. For example, the Upper East, Upper West and Western regions in Ghana had one doctor attending to more than 10,000 people in 2021 according to the Ministry of Health (MoH) annual health sector report [8]. On the other hand, the TMP-patient ratio in many rural communities in Ghana is 1:200, indicating the abundant availability of and easy access to TMPs in Ghana [7].
There are various traditional medicine practitioner associations in Ghana, including thousands of practitioners across the length and breadth of the country. However, limited data or research has examined the knowledge and literacy level of TMPs about BC and their involvement its treatment and care in Ghana. The objectives of this study, therefore, were to examine BC knowledge and perceptions among TMPs in Ghana; to examine the involvement of TMPs in BC treatment, and referral practices; and to determine the potential role of TMPs in BC early reporting and treatment in Ghana.

Methods and materials

Study design and setting
A descriptive cross-sectional study design was used to gather quantitative data from 170 TMPs, popularly known as herbalists in Ghana, from six(6) randomly selected regions of Ghana, including the Greater Accra, Bono, Central, Eastern, Northeast, and Volta regions. This design was appropriate to gather large number at a single point in time.

Sample size determination
A sample size formula for a finite population from OpenEpi, Version 3, open-source calculator—SSPropor was used to determine the sample size for the study based on a proportion of a population [9].
Where:
Deff is the design effect = 1, for random samples.
N is the population of the Traditional Medicine Practitioners (TMP) in 6 regions of Ghana, that is, 2000.
p is the hypothesized proportion of TMPs.
d is the absolute precision (Confidence limits as % of 100(absolute +/- %) = 7.5%.

For this study, the sample size was 158, with 5% non-response rate, making a total sample size of 166. The final sample size taken was 170.

Sampling and sample
A combination of simple random and purposive sampling was used to recruit participants into the study. This approach was necessary to help ensure representativeness and generalizability of the findings of the study to TMPs across Ghana. The six regions above were randomly selected from the 16 administrative regions. To be included in the study, a participant had to be an adult of at 20 years and had been practicing as a TMP for at least one year. Like the sampling design, the six regions were randomly selected to ensure that all types of TPMs were represented in the study and for the results to be representative of Ghanaian TMPs involved in BC treatment. The website of the Ghana Federation of Traditional Medicine Practitioners (GHAFTRAM) was used to identify potential participants in addition to snowballing. Others were recruited from local communities by the researchers, especially those who are not members of GHAFTRAM.

Data collection and analysis
A semi-structured questionnaire was used to collect data on sociodemographic variables, breast cancer literacy levels, perceptions, attitudes, treatment practices, and channels for collaboration with biomedical oncologists to promote early breast cancer diagnosis and treatment in Ghana. Breast CAM (Caner Research UK, 2009) [10], a standardized breast cancer literacy instrument in public domain, from Cancer Research UK was adopted and modified to satisfy the cultural context and to add questions on breast cancer diagnosis and treatment practices among TMPs in Ghana. The modified instrument, Survey of Breast Cancer Literacy and Treatment Practices of Traditional Medicine Practitioners (Adam, Appiah, and Bonsu, 2024) [11], was pilot tested with ten (10) TMPs in Oti Region who did not participate in the actual study. Both closed-ended and open-ended questions allow appropriate and adequate responses depending on the type or nature of the question. The finalized questionnaire was used to develop a KoboCollect form that was deployed to collect the data. KoboCollect form was used to ensure ease of access to TMPs whose locations were far so that they could use mobile devices to complete the surveys. It was also to make it convenient for those who could complete the forms without having to meet the researchers face-to-face. The KoboCollect form was also pre-tested and necessary changes effected prior to deployment for data collection. The questionnaires were typed in English and were administered through the researcher-assisted administration or self-administered. The researcher-assisted administration of the survey was done with TMPs who could neither read nor write in English to ensure the accuracy and completeness of the data. Sessions were scheduled with TMPs at their convenient times to provide guidance for TMPs to complete the questionnaire.

Primary outcomes and their measurement
The primary outcomes of this study included knowledge, perception, BC treatment history and referral practices among TMPs who participated in the study. We examined TMPs level of knowledge and perceptions of BC to determine if they have comprehensive knowledge of BC that could help them realize the urgency and need to refer their clients to hospitals for diagnosis and treatment. BC knowledge was measured with 29 questions which covered BC risk factors, screening, symptoms, prevention, and treatment options. A total knowledge for each participant was calculated by creating a composite variable by assigning correct answers a score of one(1) and wrong answers a score of zero (0) and totaling it. Descriptive statistics was then run where the median score for all participants was established. Those that whose total score was above the median score were deemed to have good knowledge and those whose total were below the median score were categorized as having poor knowledge. For perception of BC among the TMPs, eight questions were included including perceived origin or cause of BC, misconceptions, orthodox versus herbal medicine, and how BC patients should be treated. A composite variable, perception index, was created whereby positive answers were scored one and negative answers scored zero. More positive answers meant positive perception and more negative answers meant negative perception. The BC treatment history determined the involvement of TMPs in treating BC cases or suspected BC cases in their various practices and how many and how often they get BC clients. TMPs were if they treated patients with BC, how many they treated in the last two years, and perceived outcomes of the treatments for the patients. The referral practices sought to find out if TMPs refer BC suspected cases to hospitals or other TMPs for diagnosis and the processes used to refer clients. A secondary outcome examined the potential of TMPs to help promote early BC reporting and diagnosis through collaborative strategies with trained healthcare professionals and the primary health delivery system. Participants were asked if they ever referred a client with BC to a hospital, at what stage they referred them, what necessitated the referral, and how the referral was done. Frequencies were calculated and presented in a table.

Data analysis
Data were analyzed using STATA version 17. Descriptive statistics was used to analyze and present categorical data in percentages and frequency distribution tables. A chi-square (X2) test of significance was used to determine the sociodemographic variables that were significantly associated with the dependent variables. Binary and multiple logistic regression models were used to determine variables that significantly predicted the dependent variables.

Ethics review, approval, and informed consent
The study proposal was reviewed and approved by University of Health and Allied Sciences Ethics Review Committee (ERC)with approval number UHAS-REC B.10[135]24–25. This was not a clinical trial and no registration and clinical trial number are applicable. Standard informed consent procedures were followed throughout in participant recruitment and participation in the study. Detailed information sheet that clearly introduced the researchers, explained the objectives of the study, and laid out the procedures for participation was provided to all participants. All participants voluntarily consented either verbally or written.

Results

Sociodemographic characteristics of the respondents
A total of 170 TMPs were included in the data analysis, comprising 110 (64.7%) women and 60 (35.3%) men. The participants’ response rate in this study was therefore over 102% as the sample size calculated was 166 TMPs. The majority of the respondents were between the ages of 51–60 (35.3%) and female (64.7%), with 31.2% having no formal education and 21.7% holding a bachelor’s degree. Most are married/living together (76%), with Akan (35.3%) being the largest ethnic group represented in this study, followed by Ewe (21.2%). In terms of membership in herbal associations, 34.1% are members of the Plant Medical and Traditional Healers Association. With respect to regional representation, the Central Region has the highest representation (34.1%). The results are presented in Table 1 in Appendix 1.

Knowledge of breast cancer
The results showed that over 90% identified key symptoms such as lumps (92.9%), nipple discharge (98.2%), and redness (94.7%). For risk factors, most respondents correctly reported that a past history of breast cancer (94.7%), having a close relative with breast cancer (59.4%) and smoking (78.2%) increased the risk of breast cancer. Additionally, 97.6% recognized screening as a method for the early detection of breast cancer. However, there we some misconceptions about BC. For example, 84.7% believed that without pain in the breast, change in size or shape of the breast, there is no need for breast screening. Again, 83.5% believed that a woman who feels healthy should not perform CBE, BSE, or mammogram. Also, 97.1% of TMPs did not know that BC can be prevented or cured if it is detected early. Only 44.1% and 48.8% correctly reported that clinical breast examination (CBE) should start between the ages of 30–40 years and that mammography should start at 40 years. the results are presented in Table 2 in (Appendix 2). A total knowledge score was derived by creating a composite variable in which all 29 questions were graded and scored. A median score of 19 was established, which is the 50th percentile when the scores are arranged in ascending order. It shows that half the participants scored below and half scored above. The median score is used because the data did not follow a normal distribution and also effect of outliers does affect median scores compared to mean scores [12]. A score above the median showed good BC knowledge while a score below the median showed poor BC knowledge.
The analysis revealed that 92 (54%) and 78 (46%) of the participants had good and poor BC knowledge respectively. This is presented in Fig. 1 below.

Multivariate logistic regression model to predict breast cancer knowledge
Logistic analysis shows that education, ethnicity, and locality (region) significantly influenced BC knowledge, as was found in the cross tabulations in the crude logistics model. Although educational level is correlated with high knowledge at the crude stage, the effect weakens after adjustments, so we fail to reject our null hypothesis (H0) that there is no statistically significant association between the level of education and the level of BC knowledge among TMPs in Ghana. Logistics analysis also showed that belonging to the Ewe ethnicity is significantly associated with good of knowledge of BC (AOR = 5.3, CI: 1.1–26.5, p = 0.041), while TMPs in Northeast Region was strongly associated with poor BC knowledge (AOR = 0.05, CI: 0.0-0.91, p = 0.043). The other demographic variables and background characteristics had no significant association with BC knowledge. The results are presented in Table 3 attached in Appendix 3.

Perception and attitudes toward breast cancer
Analysis of the results revealed that there is generally positive perception and/or attitude among the TMPs towards BC. For example, 130 (76.5%) and 62.9% of the participants disagreed that breast cancer is a curse from God for sinful behavior and a result of witchcraft, respectively. Additionally, 145 (85.3%) disagreed that herbalists know more about BC than trained doctors, while 85.3% of TMPs reported advising their patients to report to hospitals rather than first reporting to herbalists. However, there were negative perceptions about BC. For example, 139 (81.8%) of the participants believed that BC patients should be isolated and treated, whereas 137 (80.6%) believed that orthodox medicine worsened BC for patients. Again, 40 (23.5%) and 63 (37.1%) believed that breast cancer is a curse and the result of witchcraft, respectively.
On the basis of the results, a perception index was derived by creating a composite variable in which all the 8 questions were graded and scored. Positive responses were scored 1 and negative responses scored zero. A median score of 7 and a mean score was 6.16 (SD = 2.28)was observed from the analysis average. Using the median score as basis for positive perceptions towards BC, 59.3% score 7 and above while 40.7% demonstrated negative perception towards BC. All the above are presented in Table 4 as Appendix 4.

Associated factors for breast cancer perception
A multivariate logistics model was used to determine the variables that significantly predicted positive perceptions of breast cancer among the TMPs, and the results are summarized in Table 5 as Appendix 5. The results of the analysis indicate that certain demographic characteristics are significantly associated with the outcome variable perception of breast cancer. Education, ethnicity, and regional affiliation significantly influence BC perception. Higher education, particularly a bachelor’s degree (AOR = 9.5, CI: 1.8–49.4, p = 0.008), is strongly associated with positive perceptions and those with bachelors’ degree nine times more likely to have positive perceptions towards BC compared to those without a bachelor’s degree. Ethnicity significantly predicted positive breast cancer perceptions, with Ga-Adangbe TMPs being significantly less likely to have positive perceptions towards (AOR = 0.07, CI: 0.01–0.5, p = 0.006). Membership in the Ghana National Association of Traditional Healers (GNATH) was also strongly associated with positive perceptions towards BC (AOR = 12.4, 95% CI: 1.8–83.6, p = 0.010) and GNATH members were 12 times more likely to have positive perceptions than those who belonged to other associations or had no membership in any association. Finally, the region where a TMP is based also had strong association with having positive perception of BC with being a TMPs in Northeast Region showing the strongest association positive perception BC (AOR = 123.5, CI: 4.2-3615.5, p = 0.005). The high AOR(123.5) and the wide confidence interval suggest model instability or sparse data resulting in substantial imprecision showing direction of association but not causality.

Breast cancer treatment and referral history
Determining the level of involvement of TMPs in the treatment and management of breast cancer or breast-related conditions in Ghana was also a key objective of this study; for this objective, several questions were included in the questionnaire. The results are presented in Table 6, which is attached in Appendix 6. First, the results show that TPMs in Ghana are consulted by women with breast-related complaints or problems, as most of the TMPs 142 (84%) in this study reported receiving and treating clients with breast-related problems. With actual treatment of breast-related problems in addition to being consulted, 126 (74.1%) reported that they had treated women with breast cancer. Only 44 (26%) indicated that they had never treated a client with breast-related problems. Among those with experience treating breast cancer, 60 (35.88%), 26 (15.29%), and 40 (23.53%) reported treating 1–5, 6–10, and more than 10 cases of breast cancer, respectively. These numbers are of paramount importance, as they confirm anecdotal reports that TMPs are involved in breast cancer treatment and management in Ghana. This could account for the high percentage of late-stage reports of breast cancer cases and the associated poor outcomes in Ghana in terms of survival and mortality rates.
The participants were also asked if they had encountered cases of breast problems that they realized they could not treat; 144 (85%) responded affirmatively, whereas 26 (15%) had not encountered cases beyond their practice. The reasons given among those who indicated that they could not handle such cases included TMP not having medicine for breast cancer 79 (47%), the client came late 45 (27%), the client was too sick 91 (54%), and the TMP was too busy 21 (12%).
In terms of referrals of breast cancer-related cases, 99 (58%) and 107 (63%) of the participants in this study indicated that they had referred their clients to other TMPs and hospitals, respectively. With respect to referrals, the TMPs were asked about the procedures they used in referring to their latest referred clients. The results revealed that 21 (12.4%), 16 (9.4%), 16 (9.4%), and 117 (69%) of the TMPs called the doctor/hospital, the TMPs gave the client names of the doctor/hospital, the TMPs gave written referrals, and the TMPs gave verbal advice to the hospital.

Self-reported breast cancer treatment outcomes
Finally, the participants who reported having treated women with breast cancer were asked to indicate the treatment outcomes of at least one woman they treated. In response to the treatment outcome question, 38 (22.4%), 42 (25%), 21 (22%), and 53 (31.2%) reported that their last treated client was cured, treated and discharged, died, and referred to the hospital, respectively. From the above, it is clear that at least the TMPs in this study are emersed in informal healthcare delivery systems, specifically in the treatment and management of breast-related problems. When asked if there was a hospital or doctor confirmation of a cure among those who reported having cured their patients, none of the 38 TMPs could answer affirmatively. Therefore, the claim of a cure cannot be guaranteed.

Determinants of referral of BC patients to hospitals by TMPs
A regression model was developed to predict variables that predict the referral of breast-related clients to hospitals or medical doctors among the TMPs who reported that they had referred. In addition to the demographics, the total knowledge score for breast cancer patients was added to the model as a predictor variable. At the crude level, many variables appeared to be significantly associated with the referral of clients to qualified doctors or hospitals. However, at the adjusted odds ratio level, only educational attainment, ethnicity and region of practice were significantly associated with referral of patients to qualified doctors/hospitals. Earning a bachelor’s degree was strongly associated with referral of BC patients (AOR = 9.5, CI: 1.8–49.5, p = 0.00. TMPs of the Ewe ethnicity were more likely to refer patients (AOR = 5.7, CI: 0.9–32.5, p = 0.050) than those of other ethnicities. On the other hand, TMPs based in Northeast China had the lowest odds of referring women with breast-related problems to doctors/hospitals (AOR = 0.05, CI:0.00-1.02, p = 0.052). Therefore, we fail to reject our null hypothesis (H0) that there is no statistically significant relationship between the educational attainment of TMP and the willingness to refer BC patients to hospitals or trained medical doctors among TMPs in Ghana. All other sociodemographic variables and knowledge of breast cancer did not predict the referral of clients with breast-related problems in this study. A summary of these analyses is presented in Table 7 as Appendix 7.

Promoting early diagnosis and treatment of breast cancer
One of the major concerns of breast cancer in Ghana is late-stage reporting and diagnosis, which contributes to negative treatment outcomes for breast cancer patients in Ghana. The participants in this study were asked what roles they could play in promoting early reporting, diagnosis, and treatment of breast cancer, with the goal of improving treatment outcomes for patients and their families. The participants provided five (5) important areas or strategies they felt could contribute to early diagnosis. These included investing and promoting nation-wide targeted breast cancer screening campaigns sponsored by government and nonprofit organizations, especially in rural areas and those far from reaching communities. This was suggested by 161 (94.71%) of the participants who felt that more effort from the public, nonprofit and private sectors was needed to initiate targeted campaigns instead of episodic screenings. A second important proposal by 106 (62.35%) TMPs in Ghana is to be trained on comprehensive knowledge of breast cancer in terms of risk factors, signs and symptoms, screening, prevention, and treatment options. There was a nearly unanimous (98%) willingness and readiness of TMPs in this study to be trained on breast cancer to help them recognize early signs, educate women about prevention and screening methods, and refer suspect cases quickly. TMPs, in turn, educate women in their communities. 91% of the participants also indicated the importance of including TMPs in breast cancer awareness creation nationwide, as TMPs are in every community in Ghana and can reach a large number of people with awareness messages. Another critical area of potential contribution of TMPs to breast cancer diagnosis proposed by 135 (79.4%) is early referral of suspected cases of breast cancer by TMPPs to qualified doctors and hospitals for correct diagnosis and initiation of treatment.
Finally, TMPs emphasized the initiation of formal collaboration between TMPs with experience treating breast-related conditions and biomedically trained oncologists in Ghana to work together to improve breast cancer treatment in Ghana. In specific areas of collaboration, 150 (88.24%) participants suggested joint breast cancer awareness and education programs throughout the year, not just in October. Mutual recognition of the expertise and roles of TMPs and biomedical professionals was suggested by 134 (78.82%) of the participants, which can lead to mutual respect and trust in the common good. Joint workshops and training and joint research on effective and safe treatment were suggested by 126 (74.12%) and 78 (45.88%) participants as important steps toward early diagnosis. The above recommendations and/or proposals indicate that achieving early diagnosis and treatment of breast cancer in Ghana is feasible and that faithful consideration of the suggestions can result in positive outcomes, increase survival rates, and improve the quality of life of breast cancer survivors in Ghana.

Discussion
This study was conducted with three broad objectives in mind: to assess breast cancer awareness and knowledge and perceptions among TMPs in Ghana; to examine breast cancer treatment history and practices; and to determine the potential contribution of TMPs in the early reporting, diagnosis, and treatment of breast cancer in Ghana. A number of important findings were made after the analysis of the results in line with the objectives.

TMPs’ awareness and knowledge of breast cancer
Importantly, those who claim to treat and cure any disease must first demonstrate their knowledge, skills, and proficiency in the disease or condition to be entrusted with such responsibility. The results revealed universal awareness of breast cancer but generally basic rather than comprehensive knowledge among the TMPs in this study. Basic knowledge of breast cancer in this study was defined as having the foundational understanding of breast cancer, including being able to identify and enumerate signs such as lumps in the breast, redness, swelling, and nipple discharge, and to list a few risk factors but still have many misconceptions about the origin, types, and treatment, for which reason, more knowledge acquisition is needed. On the other hand, comprehensive knowledge beyond basic knowledge is needed to gain a detailed understanding of the risk factors, symptoms, and treatment options at advanced levels. Among the 29 questions used to measure knowledge of breast cancer, the majority (92 [54%]) reported basic knowledge described above, whereas 78 (46%) reported poor knowledge. In terms of identifying specific signs of breast cancer, most participants correctly did so, and symptoms such as lumps (92.9%), nipple discharge (98.2%), breast redness (94.7%), and breast pain (91.2%) were notably high. Additionally, 97.6% believe that early screening is essential, and 94.7% acknowledge that having a family history of breast cancer heightens risk. The level of knowledge of breast cancer reported in this study is greater than that reported among TMPs and other nonmedical health professionals in other places. For example, among 114 traditional healers (THs) in Burkina Faso, the THs identified 52 plant species that they believed could effectively treat breast cancer but demonstrated poor knowledge of breast cancer [13]. Additionally, in a study assessing the origin of cancer, causes, and treatments of breast cancer in Uganda, researchers reported poor knowledge among THs, with several misconceptions about breast cancer, including the belief that cancer is a sexually transmitted disease (STD) resulting from a spiritual attack [14]. In Uganda, researchers reported that 62% and 24% of 119 traditional healers identified swelling and breast pain as symptoms of breast cancer [15], which are much lower than our findings in this study. In a sociocultural study of the roles of THs in the African healthcare industry, researchers reported poor knowledge and misconceptions about cancer among the TMPs, including believing that cancer is a satanic illness that can be cured only through herbs and prayers to God [16]. In this study, 78.2% and 59.4% of the participants correctly identified smoking tobacco products and a family history of breast cancer as risk factors for breast cancer, respectively. This finding is inconsistent with findings in Ethiopia and Uganda, where researchers reported that THs were unaware of the main risk factors, including lifestyle factors such as smoking, alcohol consumption, and diet [17, 14].
However, there were some concerns about BC among the participants as majority believed that breast screening was not necessary unless physical pain, swelling, or changes in breast occurred. With respect to prevention and the possibility of a cure, most (97.1%) did not believe that BC can be prevented or cured. These findings clearly demonstrate that while the majority of the participants were aware of and literate about BC, comprehensive knowledge of BC was lacking in this study group. Increasing breast cancer knowledge among TMPs in Ghana is therefore crucial for the fight against late diagnosis of BC in Ghana, as TMPs are key nonmedical healthcare professionals for whom thousands of people consult for critical health matters such as breast cancer.

Perception of breast cancer
An important finding of this study that we believe is critical for promoting early reporting and diagnosis and improving breast cancer treatment outcomes is that there were generally positive perceptions of breast cancer among the TMPs. The majority (59.3%) of the participants demonstrated positive perceptions and attitudes by responding positively to questions meant to measure perceptions and attitudes. For example, 130 (76.5%) and 62.9% of the participants disagreed that breast cancer is a curse from God for sinful behavior and a result of witchcraft, respectively. Additionally, 145 (85.3%) disagreed that herbalists know more about breast cancer than medically trained doctors do, while 85.3% advise patients to report to hospitals rather than first reporting to herbalists. These findings are consistent with incorrect perceptions by TMPs, who perceive breast cancer to be a sexually transmitted disease and the result of dirt accumulation [14]. This finding is also inconsistent with an earlier finding in a scoping review in Ghana that breast cancer is a test from God (religious belief), a result of ancestral curses and punishment (cultural belief), and from supernatural forces (spiritual belief) [18]. This finding is again consistent with findings in Ethiopia, where THs believed that breast cancer is a punishment for sinful behaviors and wrongdoers, could not describe or identify the cause of cancer, and believed that cancer could not be cured except by God or very experienced herbalists [19].
The sociodemographic variables that showed strong and significant association with BC perception included educational attainment, ethnicity, membership in the TMP association, region of practice, whereby having a bachelor’s degree, membership with Ghana National Association of Traditional Healers (GNATH), and being based in Northeast Region showed strong association with positive perception towards BC. However, these strong associations are interpretated with caution because though the adjusted odds ratios are high with significant p-values, the wide confidence intervals associated with them suggests model instability or overfitting, sparse data, and therefore, indicate imprecision. Particularly, the association of being in Northeast region to with positive BC perception. Therefore, the strong and significant associations indicate direction of relationships but do not necessarily mean causality. The public health policy and practice implication is a wholistic approach to understanding BC perceptions among TMPs should be adopted. Nonetheless, these findings may provide guidance regarding factors that help build positive perceptions of BC among TMPs. Taking advantage of positive perceptions and working to change negative perceptions are crucial for achieving positive outcomes in the fight against breast cancer in Ghana and beyond.

Breast cancer treatment and referral history
A major finding of this study is that TMPs are employed in the treatment and management of breast cancer and breast cancer-related conditions in Ghana. A large majority (74.1%) of TMPs reported that they had received clients with breast-related problems who presented with breast cancer symptoms described earlier and had provided various treatments, predominantly herbal products. The results confirm the findings of earlier studies that reported the pervasive use of traditional/herbal medicine for breast cancer treatment among Ghanaians [20–22]. This finding is consistent with literature outside of Ghana that reports the use of herbal or traditional medicine for the treatment of cancer around the world [19, 23]. These findings align with previous research that underscores the widespread reliance on traditional medicine for primary healthcare, particularly in low-resource settings where access to conventional medical facilities may be limited [24].
A critical and concerning finding in this study was the treatment outcomes for the clients. We found that 38 (22.4%), 42 (25%), 21 (22%), and 53 (31.2%) of the TMPs reported that their last treated clients were cured, treated and discharged, died, and referred to the hospital, respectively. The veracity of the report of a cure for breast cancer patients could not be verified, and the finding that only 31.2% of the 74.1% who reported that they had been treated for breast cancer were referred to hospitals is concerning. These findings have significant implications for public health policy and practice in the areas of BC related morbidity and mortality and its management in Ghana. BC patients may be convinced of false prognosis and cure and which will eventually lead to late-stage diagnosis and poor treatment outcomes. Strategic policy formulation to actively engage TMPs in research and training is needed at the community levels to promote early reporting and management of BC cases.
More concerning also was that more than 99(58%) of the participants reported that they had referred clients to other TMPs, 54% of whom referred patients only after attempting treatment.
The ultimate goal of referring clients is to avoid delays in reaching appropriate and well-resourced health facilities or professionals, but the pattern of referrals reported in this study could prolong delays in diagnosing and initiating appropriate treatment with the risk of negative or poor treatment outcomes. This finding is consistent with a study by Mwaka and colleagues above [16], which revealed that late-stage presentation of breast cancer in sub-Saharan Africa is partly due to patients initially seeking care from traditional healers before accessing biomedical treatment. This finding is also consistent with findings in Nigeria that traditional healers often attempt treatment for prolonged periods before being referred to patients, contributing to poor prognostic outcomes [25]. However, it is inconsistent with findings in South Africa that the majority of the traditional healers interviewed engaged in patient referrals to conventional healthcare systems [26].
The variables we found with strong associations to referral of BC patients to hospital included ethnicity and region of practice where TMPs who were Ewes were more likely to refer clients while TMPs based in Northeast Region were less likely to refer clients to hospitals. Although knowledge of breast cancer did not significantly predict referral directly, it may have played a confounding or mediating role in that the Eve tribe demonstrated the highest level of knowledge of breast cancer and was also most likely to refer clients to qualified doctors or hospitals. On the other hand, TMPs in Northeast Region had the least knowledge and were also the least likely to refer clients to qualified doctors or hospitals. Engaging the public and especially TMPs on the critical need for early reporting and early initiation of treatment for BC to improve treatment outcomes should be an important public health goal.

Critical areas for promoting early diagnosis and treatment of breast cancer via TMPs
Valuable insights were provided by the participants with respect to potential avenues for TMPs in Ghana to promote early reporting, diagnosis, and treatment of breast cancer. The critical areas of contribution were their role in public awareness creation and education, training, immediate referral of suspected breast cancer cases, and mutual respect and recognition of expertise and experience on the part of TMPs and trained oncologists. For example, 94.7%, 98%, 79.4%, and 88% expressed support for involving TMPs in continuous public awareness creation, breast cancer knowledge training for TMPs, immediate referral of suspected breast cancer cases to hospitals, and willingness to collaborate with orthodox medical practitioners to support breast cancer initiatives, respectively. The above findings show that TMPs understand the gravity of breast cancer and their expressed willingness and readiness to support it in the fight against this disease. A similar interest was reported in a study in Ghana that explored opportunities for integrating traditional medicine into the mainstream health system [27]. These insights also align with global discussions on integrating traditional and modern healthcare systems to optimize health outcomes [24, 28]. In South Africa, researchers have demonstrated that structured training programs enhance collaboration between traditional and biomedical practitioners, leading to improved patient outcomes [29, 30]. This finding is, however, inconsistent with the finding that some traditional healers are reluctant to refer patients due to fear of losing clientele or skepticism toward biomedical treatment [29].

Conclusion
The findings have shown that TMPs in Ghana are deeply involved in the treatment and management of breast-related conditions including BC, generally have positive perceptions towards BC but lacks comprehensive knowledge of BC. There is high willingness among TMPs to collaborate with biomedical professionals and have the potential to contribute significantly to promotion of early reporting and diagnosis of BC in Ghana. Unverifiable claims of cure of BC patients among TMPs should raise public health concern and interest in conducting broader and nationwide research on traditional medicine use for BC treatment and management. Recognizing the role of TMPs and prioritizing their integration into BC programming, including awareness creation, education, training, formal referral guidelines, research and public health policy formulation, have high probability of improving BC treatment outcomes in Ghana and beyond.

Supplementary Information