Predictors of tuberculosis knowledge, attitudes and practices in urban slums in Nigeria: a cross-sectional study

Introduction Nigeria is among six countries responsible for the majority of tuberculosis (TB) cases in the world. The Nigerian government has emphasized community-based case finding to increase detection of TB. This process requires efforts to improve knowledge, attitudes and practices (KAP) of TB, particularly in the poorest of communities. This study presents data from a KAP survey administered in two underserved Nigerian communities. Methods a structured survey was administered by trained interviewers among adult residents in two slum communities in Lagos, Nigeria. Participants were selected through multistage random sampling. KAP scores were computed and the predictors of higher scores were assessed. Results a total of 504 respondents were surveyed. The mean KAP scores were relatively low: 9.8 ± 7.1 for knowledge (out of a maximum 34), 5.3 ± 3.4 for attitude (maximum = 10), and 5.2 ± 1.5 for practice (maximum = 7). The predictors of good knowledge were increasing age, post secondary education and professional occupation. The predictors of positive attitude were post secondary education and good TB knowledge. Good knowledge was a predictor of good practice Conclusion our findings underscore the need to improve the education about TB in underserved communities. Improving KAP scores will ultimately lead to higher rates of TB detection and treatment.


Introduction
Tuberculosis (TB) is one of the top 10 causes of death worldwide with an estimated 10.4 million new cases and 1.8 million deaths from the disease in 2015; over 95% of these deaths occurring in low-and middle-income countries (LMIC) [1]. In the era of the Sustainable Development Goals, the World Health Organization (WHO) produced "The End TB Strategy" with the aim of a 90% reduction in TB deaths and 80% reduction in TB incidence rate by 2030, compared with 2015 [2,3]. Africa is home to 16 of the 30 highest TB burden countries in the world [4]. This high burden can be, in part, attributed to issues associated with population growth, the HIV/AIDS epidemic, multidrug resistant TB, inadequate health systems and poverty [5]. In Nigeria, TB diagnosis and treatment is free under the Stop TB initiative [6]. Despite existence of the program, the nation is among six countries responsible for 60% of the total TB burden in the world with an incidence rate of 322 per 100 000 population in 2015 [1,2].
The strategies to control TB in the country include: expansion of the directly observed treatment, short-course (DOTS) program; TB/HIV collaborative activities; multi-drug resistant (MDR) TB services; engagement of all care providers; engagement of communities and patients in TB activities; and, advocacy, communication and social mobilization (ACSM) to improve community awareness on TB [6]. To achieve significant progress in the control efforts, community members need to be knowledgeable about TB and have positive attitudes and satisfactory preventive practices. DOTS relies on suspect TB cases presenting themselves for care at health facilities, which can be facilitated when community knowledge is high and stigmatizing attitudes are low [7]. Poor communities are more vulnerable to TB because of lack of awareness, overcrowded and substandard living conditions, poor nutrition, intercurrent disease, and economic, geographical, social and cultural barriers to accessing TB services [8,9]. To address these issues, the Nigerian government has emphasized community-based case finding with a goal of increasing detection of presumptive TB cases referred by community volunteers between 2013 and 2014 from 11% to 23% [6]. As passive case finding remains the mainstay in routine control programs, improving community knowledge and attitudes, from what was previously determined to be low levels, will be essential in this community-based case finding process [7,10]. To examine the current level of knowledge and attitude in the "End TB Strategy" era with the goal of guiding future interventions, we evaluated the knowledge, attitudes and practices in two vulnerable communities in Nigeria.

Methods
Study areas: this cross-sectional study was carried out in the Idi-Araba and Okokomaiko communities, which fall under the Mushin and Ojo (urban), respectively, local government areas (LGAs) in Lagos, Nigeria. Both communities are large, congested slums with poor sanitation and low-quality housing. The study population was comprised of adults living in either of the communities.
Sampling: a calculated sample size of 504 respondents (252 in each community) were selected by a multi-stage sampling technique. A list of streets was obtained from the Community Development Association (CDA) of each community. Sixteen streets were selected from each community by simple random sampling and the habitable houses on each street were enumerated. Systematic random sampling was then used to select houses on the streets, while selection of households in the houses was chosen by simple random sampling (in instances where there were multiple households per house, one household was selected from each house). One eligible respondent was then selected per household by simple random sampling (where there was more than one). If the occupants of the house were unavailable, or if eligible household members were not willing to participate in the study, the next consecutive house was used.
Data collection methods: data were collected using a structured interviewer-administered questionnaire, which was adapted from the WHO sample ACSM KAP survey questionnaire [11]. The tool was pretested in a similar community outside of the study sites and appropriate amendments were made. The survey collected information on personal characteristics, TB knowledge and awareness, TB attitudes and care-seeking behavior, TB attitudes and stigma and TB information and preventive practices. Face-to-face interviews were conducted by six trained research assistants with post-secondary education and fluent in the local languages. The research assistants were supervised by some members of the research team. Interviews were conducted in the homes of respondents and took between 15 to 25 minutes to complete.
Data analysis: data were analyzed using Statistical Package for Social Sciences (SPSS) version 15 (SPSS Inc, Chicago, IL). The overall KAP of respondents were assessed. For knowledge, 11 questions were scored and some allowed multiple responses; a score of one was given to correct responses and the possible scores ranged from 0-34. Ten questions on attitude were scored and the positive responses were each given a score of one. For preventive practices, 4 questions were scored, 3 of which were on a three-point scale graded 0-2; the maximum obtainable score was 7. The cut-off for what was considered to be "good" knowledge, "positive" attitude and "good" practice were the mean values for each of these, respectively.
Statistical differences within interview sites for categorical variables were evaluated using the Chi-squared or Fisher's exact tests, as relevant and for continuous variables using the Wilcoxon rank sum test. Bivariate associations between sociodemographic variables and KAP levels were evaluated using Chi-squared test. Logistic regression was conducted to examine multivariate associations between respondents' characteristics and TB knowledge, attitude and practices. Variables with statistical differences at p<0.2 in bivariate analyses were included in logistic regression analyses using a block entry approach. Odds ratios (OR) and 95% confidence intervals (CI) were computed for each predictor variable. Level of significance was set at 0.05.
Ethical Approval: the study proposal was approved by the Health

Research and Ethics Committee of Lagos University Teaching
Hospital. Written informed consent was obtained from the respondents prior to the administration of questionnaires and confidentiality was maintained by not using identifiers.

Results
Socio-demographic characteristics: the respondents ranged in age from 18 to 89 years with a median age of 34 years (IQR: 27-43).
Most respondents were female (60.7%), married (65.5%), had at least a secondary-level education (62.5%) and were employed (78.0%). Over a third (45%) of them were semi-skilled. The two communities differed in ethnicity, employment status and occupation, with Idi-Araba having higher proportions of unemployed and unskilled respondents and respondents of Hausa ethnicity ( Table 1). Table 2 displays the correct responses to knowledge questions asked respondents. Three-quarters of respondents in the communities had heard about TB prior to the study. However, only 15.3% knew that TB is caused by a germ, 30.8% correctly responded that TB could be transmitted by air via a cough or sneeze and very few could properly identify the various signs and symptoms of TB (other than basic cough). Upon direct prompting, the least proportion (14.5%) of respondents knew that diabetics are at greater risk of TB followed by 34.3% who knew that people living with HIV were at greater risk. A quarter were aware of a health facility for TB diagnosis and treatment in their LGA while 11.1% and 9.1% knew of free diagnosis and free treatment for TB respectively. A significantly higher proportion of respondents in Idi-Araba (79.8%) had heard of TB than respondents in Okokomaiko (70.2%; p=0.01).

Knowledge of TB:
Significantly higher proportions of respondents in Idi-Araba than in Okokomaiko mentioned cough and weight loss as symptoms of TB; knew that TB is transmissible; mentioned cross ventilation as a mode of prevention; knew that adults as well as children can be affected; mentioned that both genders can be affected and knew that people living with HIV are at greater risk of TB (p<0.05). Respondents reported the most common source of TB information to be family and friends (29.8%) followed by radio or television (26.2%) while only 10.1% got their information from health workers, 4% from newspapers or magazines, 3.4% from teachers, 0.8% from religious leaders, 0.6% from brochures or posters and 0.2% from other sources. More respondents in Idi-Araba (36.1%) got their TB information from family and friends while more respondents in Okokomaiko (26.2%) got theirs from radio or television; this difference was not statistically significant. The mean knowledge score was 9.8 ± 7.1, where 56.2% were designated as having good knowledge and 43.8% with relatively poor knowledge. The respondents in Idi-Araba had a significantly higher mean knowledge score than those in Okokomaiko (p=0.01).
Attitudes towards TB: of the respondents, 22% felt they could get TB -66 from Idi-Araba and 44 from Okokomaiko. Eighty-eight (17.5%) respondents felt well-informed about TB; 53 from Idi-Araba and 35 from Okokomaiko. There were no statistically significant differences between the two groups. Data on attitude toward TB are shown in Table 3. Overall, a majority of patients understood TB to be a serious disease, including in Nigeria. A relatively small proportion (31.9%) of respondents were aware that HIV-infected patients should be concerned about TB. Most surprisingly, very few people (10.1%) perceived TB diagnosis and treatment to be free in Nigeria.

Significantly higher proportions of respondents in Idi-Araba than in
Okokomaiko correctly perceived TB to be a serious disease, considered TB to be a serious problem in Nigeria, would talk to a health personnel if they had TB, would go to a health facility with symptoms of TB, would support BCG immunization for children, and would like more information about TB (p<0.05). The mean attitude score was 5.3 ± 3.4, where 63.3% of respondents were designated as having a positive attitude and 36.7% with a negative attitude. The respondents in Idi-Araba had a significantly higher mean attitude score than those in Okokomaiko (p<0.01).
Preventive practices: Table 4 shows the preventive practices of respondents. Over half of the respondents reported the practices of always covering mouth when coughing (65.1%), always covering nose/mouth when sneezing (57.7%) and always ensuring crossventilation at home (53%). Okokomaiko had significantly higher proportions of respondents that used hands and handkerchiefs to cover their mouths while Idi-Araba had a significantly higher proportion of respondents that always ensure cross ventilation at home (p<0.05). The mean practice score was 5.2 ± 1.5, where 48.8% of respondents were designated as engaging in good practices and 51.2% in poor practices. The respondents in Idi-Araba had a significantly higher mean practice score than those in Okokomaiko (p=0.04).

Predictors of good knowledge, positive attitude and good
practice: bivariate analyses showed that there were statistically significant associations between age group, sex, education and level of knowledge; between education, occupation, knowledge of TB and attitude; between sex, occupation, knowledge of TB, attitude towards TB and practice. In multivariate analyses, predictors of good knowledge were increasing age, post secondary education and professional occupation. The predictors of positive attitude were post secondary education and good TB knowledge. Good knowledge was a predictor of good practice (Table 5).

Discussion
To date, few studies have been published that assessed community KAP regarding TB in Nigeria and fewer still have been conducted in urban slum communities, despite the high risk for TB in these settings [12]. The two communities included in this study were comparable in most of the socio-demographic variables examined. A large majority of respondents had heard of TB prior to the survey, similar to the reported rate of 74.7% from the Nigerian 2008 Demographic and Health Survey [13]. However, a more recent study demonstrated higher TB awareness (97.3%) within a rural community in South-East Nigeria possibly because of TB workshops, seminars and public lectures given within that community coupled with health talks on TB routinely given at primary health care centers [10]. Very few respondents got their TB information from health workers, underscoring the poor penetration of TB control activities within communities. TB education strategies proposed by the government include decentralization of TB care and control services into the community [14]; additionally, we recommend closer interactions between health workers and communities to ensure that more factual information is being passed on than would received from the more common sources of family, friends and radio or television. The mean knowledge score of respondents in this study was low. Using the mean score as a cut-off, just over half of respondents had good knowledge. Particularly worrisome were the very low proportions that

Conclusion
The identified gaps in TB knowledge, attitudes and practice in this study underscore the need to improve the education about TB in underserved communities. The findings may be of benefit in the planning of public health interventions to improve KAP scores, which will ultimately lead to higher rates of TB detection and treatment.  This study provides useful information about the predictors of good knowledge, positive attitudes and good practices.

Competing interests
The authors declare no competing interests.  Table 1: Socio-demographic characteristics of respondents