BC: Breast Cancer; BSE: Breast Self-Examination; CBE: Clinical Breast Examination; COR: Crud Odds Ratio; EB: Ethiopian Birr; FGD: Focus Group Discussion; LMIC: Low and Middle Income Countries; SPSS: Statistical Package for Social Studies; SSA: Sub-Saharan African; WHO: World Health Organization

The incidence of Breast Cancer (BC) is quiet high in industrializedcountries, but the international burden of the disease is increasinglyshifting to unindustrialized worlds [1]. Over 70% of BC patients inadvanced nations are diagnosed at early stage of cancer, whereas in lowand middle income countries, only 20%-60% of patients are diagnosedin early stages of the illness [2]. The major problems in developingcountries were non-establishment of disease searching, monitoringand controlling mechanisms of health systems. Limited establisheddata of this world showed that the morbidities and mortalities relatedto BC were higher in developing countries than that of developedcountries. This was demonstrated by the fact that, some previouslyconducted studies showed that majority (69%) of all BC deaths occurin developing world [3].

Another tragic issue in developing world was mortality to incidenceratio related to BC was higher than that of developed world; mostly dueto the fact that early detection of the disease was minimal and patientswere coming with disease at more progressive stages [4]. Similarinformation was also reported by different scholars; huge difference wasobserved between the two settings by the diagnosed cases and deathsof breast cancer. Diagnosed cases of BC in developing countries weretwofold as many as that of developed countries for reproductive age group women (15-49 years). It is a prime cause of cancer mortality indeveloping countries of the world including Africa which is evidencedby 7 out of 10 women newly identified with BC die in low resourcecountries whereas 2 out of 10 die in high resource countries [5,6].

Breast cancer knowledge in developing countries is not welldocumented, and what is known is far from encouraging 7 ascomparatively few women in these areas have adequate knowledgeof BC, its sign and symptoms, risk factors and preventive measuresor screening techniques for early detection. The lack of knowledgeand incorrectly held beliefs about BC prevention among females areresponsible for the negative perception of the curability of cancer and ofthe efficacy of the screening tests [7,8]. According to American cancersociety, the five-year survival rate for early screened BC approachesto 100%, however, if the cancer has not been early screened and thenspread to wider part of the breast, the survival rate is only 60%. Thisis depended on the knowledge of risk factors and especially sign andsymptoms of BC [9].

In Ethiopia, BC is typically a fatal disease with high mortality [10],unlike the experience of the Western world where BC is treatableand with lower mortality [11]. Ethiopia has set comprehensive BCprevention, diagnosis, and treatment interventions and availablefor women [12]. But stigma toward cancer, poor knowledge of BCrelated signs and its treatability, and system overload continue toaccount for delays in reaching care [13]. An important componentof the knowledge-action chain is to understand Ethiopian women’srecognition of sign and symptoms, risk factors of BC and theirmotivations for taking action. Ethiopian women typically presentfor care at a late stage in the disease [10], where treatment is mostineffective, and while system-related barriers to care account for aportion of that delay, women’s attitudes and lack of awareness of BCalso account for a stalled initiation of action [14].

In Ethiopia, breast cancer is typically a fatal disease with highand mortality [10], unlike the experience of the Western worldwhere BC is treatable and with lower mortality [11]. Ethiopia hasestablished comprehensive BC prevention, diagnosis, and treatmentinterventions and available for women [12]. But stigma toward cancer,poor knowledge of BC related signs and its treatability, and systemoverload continue to account for delays in reaching care [13]. Animportant component of the knowledge-action chain is to understandEthiopian women’s recognition of sign and symptoms, risk factors ofBC and their motivations for taking action. Ethiopian women typicallypresent for care at a late stage in the disease [10], where treatment ismost ineffective, and while system-related barriers to care account fora portion of that delay, women’s attitudes and lack of awareness of BCalso account for a stalled initiation of action [14]. So, this study wasdesigned to evaluate the knowledge of breast cancer and associatedfactors among reproductive age women in Bale zone, Ethiopia.

Study design, area and period

A community based cross-sectional study was conducted fromMarch to May/2017 in Bale Zone. The Bale zone has 20 districts. Robetown is the capital city of Bale zone which far 430 km from AddisAbaba, the capital city of Ethiopia.

Sample size determination

Single population proportion formula was used to determine thesample size for the quantitative data. the women’s knowledge score ofgreater than or equal to the mean knowledge score was 34.7% [15],the desired precession 5% with 95% confidence level, design effect of2, and 10% non-response rate was considered and the final sample sizecalculated to be 761.

For qualitative data six Focus Group Discussions (FGD), two focusgroup discussions for each districts, composing of 9-10 members ineach group, was conducted. Totally, 59 participants were participatedin the discussions.

Sampling procedure

A multi-stage sampling technique was used to select the studyparticipants. In the first stage three districts were selected randomly.The kebeles in the selected districts were identified and stratified intourban and rural kebeles. The selected districts have 7 urban and 70rural kebeles. Then three urban kebeles from the seven (one from eachdistrict) and nine rural kebeles from 70 were selected by using lotterymethod. The selection of kebeles was depended on their distance fromthe capital town of the district, taking the hospital as a center. For the three districts, furthest kebeles in average were at the distance of 45 km.From the total kebeles closest to the town, those were at about 15 km,three kebeles; from the middle distant kebeles those were at the second15 km, three kebeles; and from the furthest (third 15 km) alsothree kebeles were selected randomly. Lists of all households witheligible mothers were identified; finally, we used a sampling frame toselect the study participants using simple random sampling method.

For the qualitative data convenience sampling technique was usedto select participants. From each selected districts, two groups of childbearingage women, being that study population and not included inthe quantitative study were participated in the focus group discussion.

Data collection and data quality control

For quantitative data collection, an interviewer administersstructured and pre tested questionnaire was adopted from differentliteratures and modified for the purpose of the study. Interview guidewas used to conduct FGD for qualitative part of data collection, andtape recorder was also used [16,17]. The questionnaire and interviewguide were originally prepared in English language then translated tothe local language (Afan Oromo), and translated back to English tocheck the consistency.

Quantitative data was collected using a face-to-face interviewmethod with twelve Diploma Nurses for data collectors and threesupervisors were recruited and given two days training. The training ofdata collectors and supervisors mainly focused on issues such as datacollection tools, field methods, inclusion–exclusion criteria and recordkeeping. The investigators coordinated the interview process, andreviewed the completed questionnaire on a daily basis to ensure thecompleteness and consistency of the data collected. The questionnairewas pre-tested on 5% of the sample outside the selected district for thisstudy.

Each focus group discussion was conducted by two trained femalediploma nurses; one did moderate the discussion and the other tooknotes and recorded tape. One gate keeper (non- health professional) foreach FGD was assigned. Totally, two individuals with one gate keeperwere assigned for each group (3x6=18) people to handle the discussion.

Data analysis

The quantitative part of the data was inserted into EPI info version3.5.3, for data clearance and observation of data consistency and itwas exported to SPSS Version 20 for data analysis. First, descriptivestatistics like frequency, percentages, mean and standard deviationwere carried out to describe the data. Then, simple logistic regressionanalysis was done by taking each independent variable with dependentvariable to examine the association between the two variables. Onsimple logistic regression analysis significant variables at p-value<0.25were retained for subsequent multiple logistic regression. Hosmer-Lemeshow goodness-of-fit statistic was used to assess multicollinearity.Strength of association was tested using Adjusted Odds Ratio (AOR)and 95% Confidence Interval (CI). The significance level considered formultiple logistic regressions was p-value<0.05.

For qualitative data, tape recorded and transcribed qualitative datawas organized in narrative forms in congruent with the respondents’own words and analysed under selected themes based on the questionguide and summarized manually.

Operational definition

Knowledge: was assessed by asking 14 questions related toknowledge about breast cancer. Each correct answer was scored ‘Yes’ and each incorrect answer were scored ‘No’. Thus, the total scores wereranked from 0-14. Those reproductive age group women answered ‘Yes’equal to mean and above for the questions was labeled as knowledgeableand those reproductive age group women answered, ‘No’ for less thanmean for the questions was labelled as not knowledgeable. We usedmean for categorizing the knowledge by reviewing different previouslydone researches [15,17,18].

Ethical consideration

Administrative approval was obtained before conducting the studyand ethical considerations was respected. Ethical clearance letter wasobtained from Madda Walabu University ethical clearance committee.Official letter of collaboration was written to Bale zone administrationGinnir, Sinana and Madda Walabu districts administration to obtainformal permission. Informed consent was obtained from eachinterviewee and they were also given the choice to refuse to participatein the study.

Socio-demographic characteristics

A total of 750 respondents were included in the study with the meanage of 31.09 ± 7.34 SD years. Nearly half 46.8% of the study participantswere illiterate. The majority, 83.7% of the study participants weremarried. Regarding occupation of the mothers around ninety percentof them were house wives. Above half 54.7% of their husbands werecompleted primary school. Concerning monthly income the majority56.0% of the respondents were below poverty line (<1311 EB) thatwas $ 1.90. Around fifty seven percent of the study participants hadtelevision or radio (Table 1).

Table 1: Socio-demographic characteristics of the respondents in Bale zone, Southeast Ethiopia, 2017.

Factors associated with knowledge of breast cancer

The overall knowledge score of the respondents were ranged from0 to 14, with the mean knowledge score of 7.77 (SD=3.18) out of the14 questions designed to assess knowledge of breast cancer. Aroundfifty six percent of the respondents were knowledgeable as they had aknowledge score of greater than or equal to the mean.

About 90% of women had at least one visit to the health facilities.However, the majority, 81.6% of the study participants had not anyinformation about BC. Half of the participants responded that theydid hear about breast cancer from different sources other than healthworkers.

This finding supported by qualitative study as two civil servantand four merchant mothers’ reported, “…in our setting there has notbeen any awareness creating activities or education by doctors on breastmatter (presence of the disease, its consequences, its symptoms and itsoption of treatment). We hear some information from television. Someof us are hearing even the presence of its management now from thisdiscussion.”

In FGD a 33 years merchant reported that, “…I heard my neighbourcomplaining breast disease; we hear also the disease is cancer, manypeople have been suffered of breast disease, yet, I have not seen breastdisease on myself… When I was a child I had heard a woman of ourneighbour died because of breast disease. People was talking about thatdisease by saying it was cancer. Her breast was wounded and she wasreferred and taken far out of this area to get better treatment in hospitals.She had visited not only one hospital, but many hospitals. Finally shedied as the result of that problem.”

In focus group discussion (FGD) of mothers about BC, manymothers stated that breast cancer is unknown among them. Forexample, 36 years merchant women reported, “…We heard about breastcancer recently, before this recent time even if the disease was present wedid not know. As she (by pointing to a 33 years old civil servant woman)rightly said many women have breast disease, …,I know a woman herbreast problem was arisen from small infection and it was transferred tovery severe problem, went to health facility and improved at this time”.

Another FGD participant (a 37 years civil servant woman)reported, “…this breast disease is certainly present in its massive form,I have not experienced this disease on myself, it has hurt many women,some people say it became “hola” on a woman, some say cancer and someother say another thing. Even at the moment, there is a women with breast problem, she has birth recently, her breast has not have milk, shehas severe breast ache”.

The majority 60.4% of the respondents had at least one symptomof breast cancer. Lumps in the breast were the highest symptom amongthe respondents 20.5%.

This finding supported by qualitative study as a 35 years healthextension worker reported, “…breast disease is known, especially whenit reaches severe form, women can know on themselves. For example,if it is said how, women can palpate her breast. Even if we cannot sayit is cancer as your saying, we can differentiate on palpation using ourhand or fingers; it seems dhullaa or a small seed in the breast, it becomehard, we fear breast problems which have such symptoms, because thegreat problem of our breast is this one which we believe it does not havetreatment. Something that seems what has just been mentioning forexample, ‘michi’ as the result of child mouth is not like what I’m sayingnow. It does not have a small seed like things inside the breast and it isnot become hard, simply it is swelling which has ache and itching.”

A 38 years house wife in FGD also reported, “…it seems glandinside the breast, hard on palpation and painful, if something alike togland is present, it can be differentiated during breast feeding, because itis painful to touch.”

Regarding knowledge about risk factors, more than half 52.9% ofthe respondents did not know any risk factors of breast cancer.

During FGD about knowledge of BC, some women stated its riskfactors from different aspects. For example, a 31 years house wife said,“…if a woman gives birth and goes out from her house within ten days,fifteen days, one month or forty days starting from birth day, cold canbe the cause of breast disease. On some other women the disease ariseswithout any known reason, it swells and becomes painful, people say theproblem is engorgement, some say again it is because of another thing, inreality we do not know the cause.”

A 40 years merchant woman also reported, “…God (Rabbi) sendsdisease to human being; I think no one can know what God brings tohuman being. What God brings to human being arises from the bodyof the person itself. Usually, a woman can get breast disease when shemarries and gives birth for baby. Milk fill the breast, when it remainsinside the breast for long time, it become curdle, then changed to pus. Ifshe doesn’t get treatment either from traditional healer or health facility,the problem becomes worst and changed to severe breast disease.”

In addition, a 29 years civil servant also stated, “…what is said inour tradition, if child eat delicious food and suck the breast, it can bediseased as the result of ‘Michi’ and the belching of child on it.”

More than three-forth 75.9% of the respondents knew that breastcancer is common in Ethiopia. Nearly one third 28.6% of the studyparticipants recognized that all women have the chance of acquiringbreast cancer. The majority 87.6% of the women stated that breast canceris communicable disease. Around 87.8% participants mentioned breastcancer is a killer disease if it is not detected and treated early. Majority,78.6% of the participants identified that clinical breast examination isone of the screening method of breast cancer. Regarding treatability ofbreast cancer about 77.1% of respondents believed with treatability ofBC if the cancer is at the early stage. Medical treatments were preferredby the majority 76.2% of the respondents and 23.8% respondents prefertraditional treatments like herbal and holy water (Table 2).

Table 2: Factors associated with knowledge of breast cancer in Bale zone, Southeast Ethiopia, 2017.

In agreement to the above concept, qualitative part of the studyalso showed more interest to medical treatment than traditional. Forexample, a 23 years house maker stated, “…we, as this community,know and believe the presence of good treatment at hospitals. Inaddition, private health facilities also place where we visit most of thetime even more than hospitals for their expeditious services. Many peopleleft to go to traditional treatment and it has been reduced. Nowadays, wehave awareness from the health education given by our health extensionworkers to go to health facilities for any sickness that we may face.”

A 36 years civil servant woman stated, “…we go to health facility,I myself had breast disease in 2002, and it was very severe problemthat had severe pain. I went to health facility and got some injectionsand tablets. It was improved on that treatment, until now I have notexperienced such problem.”

In contrary to the above idea traditional treatment also mentionedby a few participants as urgent relief and even sometimes as ultimatecure. For instance a 39 years house maker said, “I had pain of breastrecently, I went to health facility, I was told the problem was engorgementand told to boil water and making it in high land water container andto apply on it. I did that, as it contained pus nothing was changed, Icouldn’t get improvement. The severity of the pain was increased andlastly I went to a traditional healer, he tied some medicine on that area,with this it was burst. After long period of time I got relief of that and atthe moment even the child is also sucking”.

And 30 years civil servant also stated, “…even I myself feel breastpain sometimes. I boil water and make it in high land water containerand apply on it or I apply kerosene (gas) on it, and then I get relief…Ihave never visited health facility.”

A 45 years merchant woman reported, “…we do not go for healthfacility (seeking treatment), but we go to traditional healers”, sherepeated naming the same remedies by adding when we apply habukurtoand hapeta on it we get some improvement, but these treatments cannever cure the problem.”

A 25 years house wife said; “…as our tradition what is being said,if child eat delicious food and suck breast can be infected as the resultof “Michi”. This time we feel breast itching and tubing pain. For this weapply some leave that we know traditionally as medicine for “michi”.

Logistic regression analysis of knowledge of breast cancer

Rural mothers had less knowledge compared to those living inurban, AOR=0.29 (95% CI: 0.13, 0.65). Those mothers who havemonthly family income greater than 1311 were more likely to haveknowledge of breast cancer compared to their counterparts, AOR=3.52(95% CI: 1.73, 7.17). Respondents who consider breast cancer is a killerdisease were more likely to have knowledge of breast cancer comparedto their complements, AOR=3.29 (95% CI: 1.49, 7.28). Similarly,participants who know treatability of breast cancer were more likelyto have knowledge compared to those do not know the treatability ofbreast cancer, AOR=3.81 (95% CI: 1.85, 7.79) and mothers who gotinformation or health education at health facilities were more likelyto have knowledge compared to those mothers who did not getinformation or health education at health facilities, AOR=2.65 (95%CI: 1.28, 5.49) (Table 3).

Table 3: Logistic regression analysis on knowledge of breast cancer in Bale zone, Southeast Ethiopia, 2017.

In this study the mean breast cancer knowledge was 7.77 ± 3.18SD for participants had ever heard about breast cancer. Around fifty six percent of the respondents were knowledgeable as they hada knowledge score of greater than or equal to the mean. Much lowerbreast cancer knowledge level was observed in a study conducted onhousehold women in Northern Ethiopia which indicated 34.7% ofrespondents had a knowledge score of greater than or equal to themean knowledge score [15]. This difference could be due to the fact thattype and number of questions prepared to assess the knowledge scorefor the two researches were not exactly the same. The four year time gapbetween the two researches could also be the other possible explanationfor the variation of knowledge between the two settings.

This study showed that rural participants were less knowledgeablecompared to urban AOR=0.29 (0.13-0.16) and it was in line withthe result reported from Zambia [19]. This was probably due to theexistence of different types of health facilities and various healthpromotion programs that use mass media in urban than ruralresidences. In addition to this, urban respondents seem more educatedand have more awareness than rural. Moreover, rural participantsare more influenced by harmful traditional practices and reluctant toaccept such sensitive health care services than urban.

In this study thinking about the treatability of BC was significantlyassociated with the knowledge of BC. Those mothers who did think BCis treatable were more knowledgeable than their counterpart AOR=3.81(1.85-7.79). It is in agreement with the study conducted innorthern Ethiopia which showed household head women who thoughtthe treatability of BC had more knowledge of BC than household headwomen who did not think the treatability of BC [15]. This could be dueto the probability of mothers who have some information on BC, herethe treatability of BC; implicit to have additional information aboutbreast cancer. This consideration of treatability of BC could be alsofrom the previous exposure to breast disease related problem/s that hadbeen treated either for themselves or other women.

Another predictor for BC in the present study was previousinformation or attending health education at health facility. Motherswho had previous information or health education on health matters athealth facility were more knowledgeable about BC than those motherswho had not have such previous exposure to information AOR=2.65(1.28-5.49). This may be due to that information or education obtainedfrom health professionals at health facility include many aspects ofhealth issue that could address breast health and mothers benefittedfrom that. Additionally this information or education might be directlygiven for mothers at the service delivery point during the managementof some problems that the mothers came for [20].

The other predictor that was significantly associated with theknowledge of BC was monthly family income. Mothers who hadmonthly family income greater than 1311 birr were more knowledgeablethan the mothers whose monthly family income was less than 1311 birrAOR=3.52 (1.73-7.17). The finding was consistent with the study donein South Africa [21], but it was not in line with the study done in Addis Ababa-Ethiopia which showed there was no difference in mothers’knowledge depending on their different monthly income [22].

Lastly, but not the least, thinking BC as a killer disease was alsosignificantly associated with BC knowledge. Those women who didthink BC as a killer disease were more knowledgeable than the womenwho did not think BC as a killer disease AOR=3.29 (1.49-7.28). It is inagreement with the study conducted in western Amazon which showedwomen who did think BC as a killer disease were more knowledgeablethan their counterpart [23]. This could be due to the probability ofmothers who have some information on BC, here BC as a killer disease;implicit to have additional information about breast cancer. Thisunderstanding of BC as a killer disease could be also from the previousexposure to problems of BC either on themselves or other women.

The study revealed that almost half of the reproductive age womenhad low breast cancer knowledge compared to other study. Smallproportion of the respondents had ever heard about breast cancer. Thisis evidenced by the fact that adequate information of health educationwas not being given by health professionals at different levels. The majorfactors for knowledge of breast cancer were residence, monthly income,knowing the seriousness of breast cancer, knowing the treatability ofbreast cancer and previous information or health education on breastcancer. So, it is very essential to develop health service programs thatcan address all mothers according to their favourable place and timeto increase knowledge of breast cancer through information andhealth education regarding its sign and symptoms, risk factors, earlydetection, and management systems to reduce the burden of BC.

Ethics approval and informed consent

The study was performed by interviewing reproductive age groupwomen after an ethical consent was obtained from Madda WalabuUniversity ethical clearance committee and individual verbal consent isobtained from the study participants. This manuscript has never beensubmitted and considered for publication to any other journal.

Not applicable.

The data will be available upon request.

This study was supported by Madda Walabu University.

The authors declare that they have no any competing interests.

This work was carried out in collaboration between all authors.‘Author AH’ and ‘Author MK’ developed the concept, designed thestudy, wrote the protocol, performed data collection, analysis andinterpretation, and wrote the first draft of the manuscript. ‘Author AL’,and ‘Author SN’ developed the concept, designed the study, performeddata collection, performed data analysis and interpretation, and wrotethe first draft of the manuscript. All authors read and approved the finalmanuscript.

We are grateful to Madda Walabu University for supporting this study. We arealso very grateful to data collectors to undertake this study.

MK is a lecturer and an academic and research coordinator at Madda WalabuUniversity, AH is a lecturer and head of public health department at Madda WalabuUniversity, AL is a lecturer at Madda Walabu University, and SN is a lecturer atWolayta Sodo University.