Study area and population
Study area
This study was conducted in Fentalle (Kereyu pastoralist) Wereda (equivalent to district), located in the east Shoa zone of Oromia, in the southern part of the northern rift valley of Ethiopia. The area falls within an altitude range of 800–1100 masl. The total land area is 1170 km2 with a total population of 76,367; it is located 200 km east of the capital city, Addis Ababa. Metehara is the capital and administrative centre of the district. A detailed description of the study area is given elsewhere [10].
There are four health centres and 18 health posts for the entire population of the Kereyu District. There is one referral hospital in Metehara Sugar Corporation called Merti Hospital. There was a total of 65 HEWs and 46 other health professionals (clinical nurses, midwifes, environmental health workers and pharmacy technicians and laboratory technicians) in the district, excluding the HCPs working at Merti Hospital, which is not the administrative unit of the district health office but does provide services to the pastoralist community. The number of HEWs reported at the district health office includes those training for upskilling, travelling for workshops, not available due to maternity or sick leave as well as those in the process of transfer to other districts. We included all available HEWs at the health posts during the study period.
Study design and sampling
We conducted a facility-based cross sectional study from September 2014 to January 2015.
Before the actual data collection, we identified all the health facilities in the district with the help of HEWs and identified the TB focal person responsible for coordinating the district’s TB prevention and care activities. We included all HCPs and HEWs in the district in the study. Those who were available during the study period were given a self-administered questionnaire to complete and return. If the HCPs reported insufficient time to return the questionnaire the same day, they were allowed to return it the next day.
Measurements
Data were collected using semi-structured and self-administered questionnaires prepared in English as in previous similar studies [20, 21, 25] and translated to Amharic (the federal working language). We gave training to three experienced local coordinators and involved them in the facilitation of the data collection process with the principal investigator.
The survey contained 71 questions, with sections on sociodemographic characteristics, TB knowledge, attitude, perceived stigma regarding TB and perception of collaboration with traditional healers. The tools were pre-tested before the actual data collection to assess the comprehensibility of the questionnaire.
Knowledge
Knowledge is defined as the fact or condition of knowing something with familiarity gained through experience or association [22]. The knowledge section had 24 questions and was divided into three sections: TB diagnosis (10 questions about signs and symptoms of susceptive TB, active TB and relapse TB), nature of the disease (12 questions about transmission, cause, factors in the spread of TB) and treatment duration (2 questions, one for intensive phase and one for the whole duration of TB treatment required). The correct (yes) response to each question was scored as one for a positive response, and incorrect (no/I don’t know) response was scored as zero for a negative response. The scores were added together to generate a knowledge score from 0 to 24 (including each sign and symptom mentioned and factors for the exposure of TB), and the overall score was dichotomized using a median of 18 as a cut-off value. Those who scored 18 and above were coded “1” for good overall TB knowledge, and those below 18 coded “0” for poor overall TB knowledge. Likewise, scores were generated for the two sub-scales of knowledge regarding TB (TB diagnosis and nature of the disease), and the sub-scale of knowledge was categorized as poor and good levels of knowledge.
The term ‘nature of the disease’ is used to summarize the response of the HCPs about the organ affected most by TB, transmission, cause, factors for the spread of TB and the public importance of TB in the community.
Attitude
Attitude is defined as how people feel about certain subjects or issues [21]. The attitude section contained 10 questions addressing two sub sections: attitude towards TB patients (3 questions) and attitude towards TB control (7 questions). A 5-point Likert scale was used to obtain responses to these questions and was treated as a continuous interval variable for analysis. The overall score for attitude regarding TB, attitude towards TB patients and TB control system was obtained by computing the included items using the SPSS syntax Compute by summing included items and multiplying the sum by 5 (number of Likert points). The attitude score was not normally distributed, therefore, overall attitude score was dichotomized using the median score of 37 as a cut-off value; those who had a mean score of 37 and above were coded “1” for a favourable attitude regarding TB and below 37 coded “0”, indicating an unfavourable attitude regarding TB.
Perceived stigma
Perceived stigma refers to the fear of discrimination or, in general, to the awareness of negative attitude and/or practices related to a particular condition [26]. The perceived stigma section had three sections: feeling about a person with TB, perceived community feelings towards TB patients and feelings about being near a person with TB. The first section had five items and was summed to create a perceived stigma score towards TB patients for analysis. Each item was coded as a “yes” or “no” response where ‘yes’ indicated the absence of perceived stigma and ‘no’ indicated the presence of perceived stigma. Negatively stated questions were reverse coded to obtain the correct scoring. The responses consistent with “lack of stigma” were scored one and the rest scored zero. The sum of the responses to (1) I feel compassion and desire to help; (2) I feel compassion but tend to stay away from TB patients; (3) It is their problem, and I cannot get TB; (4) I feel fear because they might infect me; and (5) I have no particular feelings were used to generate stigma score from 0 to 5. The overall score was dichotomized using the median as a cut-off point. Since the stigma scores were not normally distributed, the median score (median 1, IQR = 1) was used to classify the HCPs as having high or low perceived stigma towards TB patients. Those who have a score above 1 were coded as one, showing high perceived stigma towards TB patients, and those who scored 1 or lower were coded as zero, showing low perceived stigma towards TB patients (Additional file 1).
Perception about THs
Perception is man’s primary form of cognitive contact with the world around him [27]. In this study, we assessed the HCPs’ perception for possible future collaboration of the conventional health system with THs on TB control. Perception of the HCPs regarding collaboration with THs and their willingness to collaborate on TB prevention and care was assessed using 17 items. The questions had multiple choice and ‘yes’ or ‘no’ responses. The proportion of responses to some of the items used and relationship to the conclusion of this paper is reported.
Data analysis
Quantitative data
The data were entered and analysed using statistical software for social science (SPSS) version 22 and STATA version 14. We applied descriptive statistics to summarize the socio-demographic status of the HCPs, describe their knowledge and attitude and perceived stigma regarding TB and their perception about collaboration with THs on TB diagnosis and treatment. A Chi-square test was used in bivariate analysis to determine the association between the outcome variables and selected covariates. Univariate logistic regression was used to assess the strength of the association. The statistical significance of the differences was evaluated using p value < 0.05 and a 95% confidence interval.