This study was a cross sectional descriptive research design using quantitative approach. A cross-sectional study design was chosen because; the present study compared a single point in time variables. Quantitative method was selected because these methods allowed researcher assess more variables at the same time.

This research targeted all TB patients that visited different health facilities in Gakenke District Northern Province, Rwanda from April to December 2019. Randomly, three health centers from Gakeneke District were selected.

Gakenke District is one of five districts of the Northern Province of Rwanda, with 348,334 inhabitants (NISR, 2012). People survive mainly by agricultural, farming and commercial activities (selling of goods and services). Commonly, diseases like intestinal worms, diarrhea, malaria, HIV and AIDS, tuberculosis, as well as upper and lower respiratory tract infections are seen in the outpatient departments of health facilities of Gakenke District (Gakenke Annual Report, 2017).

The data was collected among TB patients that frequently utilize the different health facilities in the three sectors of Gakenke District. The sample size of 384 TB patients was dram by using Cochran formula. Cochran (1963:75) formula used to determine proportions in single cross-sectional surveys was used for sample size determination (Cochran, 1963).

As there is no study in Rwanda showing the level of TB knowledge among TB patient, based on the above explanations the following assumptions were applied: p, level of knowledge 50%, d is the expected margin of error (5%), Z, the standard score corresponding to a 95% confidence interval (1.96).

However, 376 TB patients completed the questionnaire. This shows that the response rate was 97.9%.

Two-stage sampling method was used, in the following way:

Three health centres were randomly selected among seven sectors of Gakenke District by a simple random sampling. The selected health centres are Rutake, Rutenderi and Karambo.

Selection of clients: all adult clients frequenting the health center during the study period using a systematic sampling method. All new cases ranking second, fourth, and….ordered number in the exhaustive list of OPD service client register during open hours of the day.

Data were collected by using a questionnaire adopted from previous study conducted in Uganda. Eligible participants were interviewed face-to-face using a structured questionnaire with both open and closed questions to collect information on socio-demographics, socio-economic, TB related knowledge including cause of TB, TB symptoms, and TB prevention measures. The researcher provided a verbal introduction about the purpose of study.

Data was analyzed using SPSS version 22. Demographic and outcome data was summarized into frequencies, percentages and measures of central tendency. Responses to the 36 (including multiple response questions) questions about TB were scored (one point for correct and zero for incorrect) and categorized into 0-50, 51-75 and 76-100 percent corresponding to poor, moderate and high TB knowledge respectively as a composite outcome. The association between exposure variables and TB knowledge was explored by univariate analysis and Multivariate analysis. Tests were two-sided and considered significant if P<0.05. Potential determinants of TB knowledge with P < 0.05 at univariate level were included in a multivariable ordinal logistic regression model to estimate their adjusted odds ratios (aOR).

Ethical clearance certificate was obtained from the Research and Ethical Clearance Board at Mount Kenya University Rwanda. A permission to collect data was obtained from Director of Health unit at Gakenke District. The researcher informed the participants that then participation was voluntarily and that they had to sign consent form for confidentiality of data. No names of participants were put down but only codes were used and all information about this subject was kept confidentially.

The study is carried out in three health centres of Gakenke District (Rutake, Rutenderi and Karambo).

The table 1 contains the demographic characteristics of the respondents which contain sex, age and education level. The findings shows that out of 376 TB patients who participated in the study, 71.0% were male, 51.6% were aged 45 years and above, 81.9% of respondents were married, and 15.2% of respondents had no formal education. The results further showed that 30.9% of respondents were farmers, 72.9% were Christian, 74.5% never smoke, and 61.2% had ever drink, 34.9% had more than 5 family members, 32.4% had less than 3 73.4% had the income. The majority of TB patients who participated in the study 326(86.7%) have tested for HIV, near a half of respondents 170(45.2%) live between 2-5 km from a nearest health facility.

The individual questions about the TB knowledge including the knowledge of groups at risk to be contaminated by TB, the fact of being informed on how TB is transmitted, knowledge on mechanisms to prevent and manage TB by different independent variables. (Table 2)

The findings on TB knowledge among TB patients from Gakenke District shows that 93(24.7%) knew the real cause of TB (germs), 127(33.8%) knew that coughing over 15 days is one of TB symptom, 168 (44.7%) respondents knew that TB is transmitted through airborne, 178(47.3%) knew that not sharing the same eats and drink with a person who has TB is one way of TB prevention, the majority 162(56.9%) TB patients knew that TB is a curable disease, A total of 128 (34.0%) of TB patients knew that TB treatment can last between 6-8 months, 17(18.9%) knew that TB treatment can last less than one month while some of respondents 45(12.0%) don’t know the duration of TB treatment. (Figure 1)

To determine the level of knowledge, total scores were estimated from 36 questions ( considering multiple responses) related to TB. The mean score was 23, considering the mean score; respondents with a score of above mean score were considered as having good knowledge, respondents with a score between 18-22 were considered to have moderate knowledge, and respondents with a score less than 18 were considered to have poor knowledge. The findings presented in Figure 1 shows that the majority 204(54.3%) of respondents demonstrated a good knowledge about TB, 20.2% demonstrated a moderate knowledge while 25.5% of respondents had a poor knowledge.

Both bivariate and multivariate analyses were performed to estimate socio-demographic factors associated with TB knowledge. Socio-demographic factors associated with TB knowledge are presented in table 3. For analysis purpose, in both bivariate and multivariate analysis moderate and poor knowledge were combined.

Findings on the socio-demographic factors associated with TB knowledge show that out of 204 respondents who demonstrated good knowledge about TB, 163(79.9%) were male, and out of 172 respondents with poor/moderate knowledge, 68(39.5%) were female. The findings show that gender was associated with TB knowledge (p<0.001). The majority of respondents with good knowledge were aged 45 years and above 90(44.1%), with this aged group 104(60.5%) demonstrated poor/moderate knowledge. Age group of respondents were significantly related to TB knowledge (p<0.001). The findings shows that all respondents (204) with good knowledge were married, 68(39.5%) of respondents with poor/knowledge were single; marital status was significantly related to TB knowledge (p<0.001). Religion, smoking, drinking, and distance to health facility were significantly associated with TB knowledge (p<0.001). Factors such as ever tested for HIV and number of family member were not significantly associated with TB knowledge.

The relationship between factors such as education level, occupation and family income and TB knowledge were investigated. (Table 4)

Regarding socio-economic factors (education, occupation and income), the findings shows that 147 (72.1%) respondents with good knowledge had primary education level, 98(57.0%) of respondents with poor/knowledge had primary education; education level was significantly related to TB knowledge (p<0.001). The findings shows that 149(73.0) of respondents with good knowledge had the income of 20,000-40,000 Rfws, 127(73.8%) of respondents with poor knowledge had the income between 20,000-40,000 Rfws. But no significant relationship observed between family income and TB knowledge (P=0.872). A total of 108 (52.9%) respondents with good knowledge were farmers, 86(50%) of respondents with poor TB knowledge were self-employed. (Table 5)

None of socio-economic factors were taken into multivariate analysis. Socio-demographic factors such as gender, age group, drinking, distance to health facility were included in multivariate analysis. The findings show that male were three times more likely to have good knowledge about TB compared to female (AOR=3.31, 95%CI: 1.98-5.53, p<0.001). Compared to TB patients aged 45 years and above, respondents aged 25-34 years old were more likely to have good knowledge about TB (AOR=38.71, 95%CI: 9.22-162.48, p<0.001). Respondents who ever drink alcohol were less likely to have good knowledge about TB (AOR=0.148, 95%CI: 0.091-0.239, p<0.001). TB patients who live between 2-5 km from nearest health facility were more likely to have good knowledge about TB compared to those who live at more than 5 km (AOR=33.58, 95%CI: 14. 95-74.40, p<0.001).

Knowledge of tuberculosis has been shown to influence health seeking behavior. In this study we report sub-optimal knowledge about TB in the general population of Gakenke District. Researcher also identified that younger age, having primary education, ever drinking, marital status, and distance to health facility, religion and occupation; were associated with a good level of TB knowledge in the bivariate model. In multivariate analysis, being male, aged 25-34 years old, and living at distance of 2-5 km, were significantly associated with good TB knowledge.

In contrast to the findings from this study where the researcher observed that 54.3% of TB patients had good TB knowledge, a study conducted in Uganda found that only 208(18.3%) respondents demonstrated good knowledge, the majority 358(31.4%) demonstrated moderate knowledge 5. Comparing individual variables on TB knowledge, this study found that 24.7%, in line with this finding a study conducted in Uganda found that few respondents identified a germ as the cause of TB (26.7%).

A study conducted in Ethiopia found that 74.4% of respondents were found to have good knowledge 6, compared to this study which found 54.3% of respondents with good TB knowledge. The difference between TB knowledge reported in Ethiopia and in Rwanda may due to the study area and study participants, in Ethiopia the study was conducted in urban area and among health care professionals while this study was conducted in rural area (Gakenke) and among TB patients.

In contrast to our findings, a study conducted in Brazil found 67% knew how tuberculosis is transmitted, 87% knew its key symptoms and 81% declared they would take preventive therapy if prescribed. Among KAP-interviewed index cases, 67% knew they could spread tuberculosis, 70% feared for the health of their families and 88% would like their family to be evaluated in the same services 7. In contrast to our findings a study conducted in Pakistan found overall lower knowledge on TB where cough was the most commonly stated symptom in both the urban (66.2%) and rural (63.4%) areas. Other symptoms cited were: coughing up blood (urban 56.5%, rural 50.0%), weight loss (urban 57.4%, rural 38.0%). Possible modes of transmission indicated by the respondents were sharing food utensils (urban 35.6%, rural 29.6%), eating in the same plate (urban 44.7%, rural 44.1%), shaking hands (urban 3.0%, rural 5.4%) and touching items in the public places (urban 0.0%, rural 1.7%) 8.

Inconstancy between these previous studies and the present study regarding the factors associated with TB knowledge may due to the study design, study population and geographical location.

Regarding the factors associated with TB knowledge, this study found that male are more likely (AOR=3.31, 95%CI: 1.98-5.53, p<0.001) to have good knowledge compared to female respondents, younger respondents were more likely to have good knowledge about TB (AOR=38.71, 95%CI: 9.22-162.48, p<0.001). Respondents who live between 2-5 km from nearest health facility were more likely (AOR=33.58, 95%CI: 14.95-74.40, p<0.001) to have good knowledge about TB.

A study conducted in Uganda found that independent determinants of poor knowledge of TB in the multivariable analysis included (aOR, 95% CI) lack of formal education (0.56; 0.38 - 0.83, P = 0.004), unemployment (0.67; 0.49 - 0.90, P = 0.010) and never testing for HIV (0.69; 0.51 - 0.92, P<0.012). Whilst, older age (1.73; 1.30 - 2.29, P<0.001) and residing in Lira (2.02; 1.50 - 2.72, P<0.001) were independent determinants of higher knowledge of TB 5.

A study conducted in Ethiopian revealed that training is the strongest determinant of knowledge, AOR 3.386 and 95% CI (1.377, 8.330). On the other hand, job location and age category, AOR 0.592 and 95% CI (0.286, 1.223) and 0.913 95% CI (0.649, 1.284), respectively, were not found to be associated with TBIC knowledge in the multivariate models 6.

A study conducted in Brazil found that illiterate relatives (adjusted OR = 4.39; 95%CI 1.11 - 17.36), those who do not watch or watch little television (adjusted OR = 3.99; 95%CI 1.2 - 13.26), and also those who do not have the Internet access (adjusted OR = 5.01; 95% CI 1.29 - 19.38) were more likely to have low knowledge regarding tuberculosis 9. A study conducted in Philippine revealed that a wide lack of knowledge was found about the link between HIV/AIDS infection and TB among individuals with medium and low socio-economic status and non-college educational background. Other several studies found that Individuals who could read and write were more likely to know the cause of TB which is consistent with our findings with previous reports 10,11. Males were more likely to know the cause of TB compared with females. The poor knowledge among women and non-educated individuals concerning the cause of TB will result in inappropriate health care seeking behavior 12.