The study is reported along with the STROBE Statement for reporting observational studies(46).
Study design and settings
A cross-sectional survey was conducted among TB suspects in Kassala State, which is one of the 18 states of Sudan. It is located in the Eastern part of the country, extended over an area of 42,282 kilometres divided into 11 localities. The total estimated population in 2018 was 2.5 million and consisted of multi-ethnic communities. At least 70% of the state's population is living in rural areas, and the TB disease is endemic in the state. There are 22 TBMUs that receive all TB suggestive cases in the state who require sputum smear microscopic examination for Mycobacterium bacilli(47) and TB treatment.
The ethical approval for the study was obtained from the Research Ethics Committee in the Kassala State Ministry of Health. Also, selected study sites provided permission to access participants' information.
Data Collection And Participants
Trained data collectors gathered the study-related data using a structured questionnaire. The questionnaire was formulated in line with the RAA-model, previous studies(48–52) and results of an elicitation study(53). The elicitation study included 20 participants from the study population, and they were asked to list the advantages, disadvantages, the people who would approve or disapprove of testing, and the possible factors that might facilitate or impede their HIV testing in the next three months. Seventeen beliefs were included in the questionnaire.
The study was carried out from July 2017 to February 2018 in five randomly selected TBMUs. Eligibility criteria were TB suspects referred to TBMUs for TB diagnosis and who were aged 18 years and above. The survey instrument was piloted before the start of the study.
An in-site systematic sampling approach(54) was employed to recruit potential participants. Potential participants were randomly chosen from the clinic registers. Any participant that refused to participate was replaced by the immediate next one until reaching the required sample size. Data collectors informed participants about the study objectives and data confidentiality before they were enrolled in the study. All participants provided informed consent before they were asked to answer the questionnaire in a private room.
Variables And Measurements
In this study, the outcome variable was the intention to use HTC services in TBMUs in Kassala State in the next three months. The explanatory variables were behavioural beliefs, normative beliefs, control beliefs, risk perception regarding HIV testing in the same context and time. Also, past exposure to HIV testing information was measured.
Sociodemographic characteristics were assessed by six questions on age, gender, residence, marital status, level of education, and working status. Five structured questions about HIV and HTC related-knowledge were used to determine the level of knowledge among the participants. The response options for each question were yes, no, and I don't know. The answers were coded as 1 for 'yes' and 0 for 'no' or 'don't know'.
The intention to use HTC services was measured by three items which were: I intent, I expect, and I want to use HTC services in the next three months and were measured on a 7–point scale. The higher score indicated a more positive intention.
The behavioural beliefs were measured by asking the participants to rate five statements regarding using HTC services in the next three months. The statements were: "I will Know my HIV status", "Facilitates my treatment if I have positive test result", "I could prevent infecting my family from HIV infection", "I would feel worries about HIV test result", and "I would have information about HIV infection". The participants replied on a seven-point Likert scale ranging from 1(unlikely) to 7, likely). The normative beliefs were evaluated by asking the respondents to indicate the extent to which their friends, doctor, partner, and parent thought they should use HTC services in the next three months. The participants answered on a seven-point Likert scale (with disagree (1) and agree (7) as anchors).
The control beliefs, assessed by five items: "Health care providers will keep my HIV test result confidential", "I would fear to be stigmatized if I get HIV positive result", "I would fear losing my partner if I infected with HIV infection", "I would find it difficult to disclose my HIV positive result to others", and "I could have the cost to reach HTC services". Responses were on a seven-point unipolar (unlikely (1) -likely (7)) scale.
Two items assessed the perceived HIV risk: "I think I myself may be infected with HIV infection", and "I think my partner may be infected with HIV infection" scored on a seven points (disagree(1)-agree(7)) scale.
Previous exposure to HIV testing information was measured by one item:" In the last year I had experience with HTC services" In the last year I had experience with HTC services", responses ranged from disagree (1) to agree (7).
Social desirability bias is a potential problem in a self-reported questionnaire, particularly when examining a sensitive behaviour(55). To reduce the impact of the social desirability bias, we used well-trained data collectors and each participant was assured about the confidentiality of the information he would provide. Participants were asked to sit in a private room to complete the questionnaire.
The sample size for the study was planned for correlation analysis, with a 95% confidence interval (56). For a correlation of .05 with 95% confidence interval width of .10, the required sample size was 383 participants.
For descriptive statistics, frequencies and proportions were computed for the sociodemographic variables. Averages were calculated for age and knowledge about HTC and HIV
A Confidence Interval Based Estimation of Relevance (CIBER) analytical approach was used to assess whether beliefs were relevant to the dependent variable in terms of intervention development(57). The CIBER is a useful approach that helps in visualizing a univariate distribution of each sub-determinant, their association with the outcome variable and confidence interval of both in one plot. It uses a diamond shape with fills colour. The left-hand panel diamonds represent each sub-determinants' mean and its 99.99% confidence interval. The diamond fill colour provides more information about the item means: redder colour stands for lower means, blue colour for middle scale means and green colour for a higher item means. The right-hand diamonds are indicative of the association with a 95% confidence interval. The diamonds fill colours indicate association strength and direction: the redder colour of the diamond indicates a stronger and more negative association, greyer colour shows a weaker association and greener colour indicates a strong and more positive association. Also, the CIBER provides the confidence interval of the explained variance (R2) in the outcome variable based on sub-determinants (beliefs) that were included in the analysis. Data analyses were conducted using R version 3.6.1(58) and Statistical Package for Social Sciences (SPSS) version 21 for analysis.